Have a language expert improve your writing

Run a free plagiarism check in 10 minutes, generate accurate citations for free.

  • Knowledge Base

Methodology

  • What Is a Case-Control Study? | Definition & Examples

What Is a Case-Control Study? | Definition & Examples

Published on February 4, 2023 by Tegan George . Revised on June 22, 2023.

A case-control study is an experimental design that compares a group of participants possessing a condition of interest to a very similar group lacking that condition. Here, the participants possessing the attribute of study, such as a disease, are called the “case,” and those without it are the “control.”

It’s important to remember that the case group is chosen because they already possess the attribute of interest. The point of the control group is to facilitate investigation, e.g., studying whether the case group systematically exhibits that attribute more than the control group does.

Table of contents

When to use a case-control study, examples of case-control studies, advantages and disadvantages of case-control studies, other interesting articles, frequently asked questions.

Case-control studies are a type of observational study often used in fields like medical research, environmental health, or epidemiology. While most observational studies are qualitative in nature, case-control studies can also be quantitative , and they often are in healthcare settings. Case-control studies can be used for both exploratory and explanatory research , and they are a good choice for studying research topics like disease exposure and health outcomes.

A case-control study may be a good fit for your research if it meets the following criteria.

  • Data on exposure (e.g., to a chemical or a pesticide) are difficult to obtain or expensive.
  • The disease associated with the exposure you’re studying has a long incubation period or is rare or under-studied (e.g., AIDS in the early 1980s).
  • The population you are studying is difficult to contact for follow-up questions (e.g., asylum seekers).

Retrospective cohort studies use existing secondary research data, such as medical records or databases, to identify a group of people with a common exposure or risk factor and to observe their outcomes over time. Case-control studies conduct primary research , comparing a group of participants possessing a condition of interest to a very similar group lacking that condition in real time.

Receive feedback on language, structure, and formatting

Professional editors proofread and edit your paper by focusing on:

  • Academic style
  • Vague sentences
  • Style consistency

See an example

case control study is a type of

Case-control studies are common in fields like epidemiology, healthcare, and psychology.

You would then collect data on your participants’ exposure to contaminated drinking water, focusing on variables such as the source of said water and the duration of exposure, for both groups. You could then compare the two to determine if there is a relationship between drinking water contamination and the risk of developing a gastrointestinal illness. Example: Healthcare case-control study You are interested in the relationship between the dietary intake of a particular vitamin (e.g., vitamin D) and the risk of developing osteoporosis later in life. Here, the case group would be individuals who have been diagnosed with osteoporosis, while the control group would be individuals without osteoporosis.

You would then collect information on dietary intake of vitamin D for both the cases and controls and compare the two groups to determine if there is a relationship between vitamin D intake and the risk of developing osteoporosis. Example: Psychology case-control study You are studying the relationship between early-childhood stress and the likelihood of later developing post-traumatic stress disorder (PTSD). Here, the case group would be individuals who have been diagnosed with PTSD, while the control group would be individuals without PTSD.

Case-control studies are a solid research method choice, but they come with distinct advantages and disadvantages.

Advantages of case-control studies

  • Case-control studies are a great choice if you have any ethical considerations about your participants that could preclude you from using a traditional experimental design .
  • Case-control studies are time efficient and fairly inexpensive to conduct because they require fewer subjects than other research methods .
  • If there were multiple exposures leading to a single outcome, case-control studies can incorporate that. As such, they truly shine when used to study rare outcomes or outbreaks of a particular disease .

Disadvantages of case-control studies

  • Case-control studies, similarly to observational studies, run a high risk of research biases . They are particularly susceptible to observer bias , recall bias , and interviewer bias.
  • In the case of very rare exposures of the outcome studied, attempting to conduct a case-control study can be very time consuming and inefficient .
  • Case-control studies in general have low internal validity  and are not always credible.

Case-control studies by design focus on one singular outcome. This makes them very rigid and not generalizable , as no extrapolation can be made about other outcomes like risk recurrence or future exposure threat. This leads to less satisfying results than other methodological choices.

If you want to know more about statistics , methodology , or research bias , make sure to check out some of our other articles with explanations and examples.

  • Student’s  t -distribution
  • Normal distribution
  • Null and Alternative Hypotheses
  • Chi square tests
  • Confidence interval
  • Quartiles & Quantiles
  • Cluster sampling
  • Stratified sampling
  • Data cleansing
  • Reproducibility vs Replicability
  • Peer review
  • Prospective cohort study

Research bias

  • Implicit bias
  • Cognitive bias
  • Placebo effect
  • Hawthorne effect
  • Hindsight bias
  • Affect heuristic
  • Social desirability bias

Prevent plagiarism. Run a free check.

A case-control study differs from a cohort study because cohort studies are more longitudinal in nature and do not necessarily require a control group .

While one may be added if the investigator so chooses, members of the cohort are primarily selected because of a shared characteristic among them. In particular, retrospective cohort studies are designed to follow a group of people with a common exposure or risk factor over time and observe their outcomes.

Case-control studies, in contrast, require both a case group and a control group, as suggested by their name, and usually are used to identify risk factors for a disease by comparing cases and controls.

A case-control study differs from a cross-sectional study because case-control studies are naturally retrospective in nature, looking backward in time to identify exposures that may have occurred before the development of the disease.

On the other hand, cross-sectional studies collect data on a population at a single point in time. The goal here is to describe the characteristics of the population, such as their age, gender identity, or health status, and understand the distribution and relationships of these characteristics.

Cases and controls are selected for a case-control study based on their inherent characteristics. Participants already possessing the condition of interest form the “case,” while those without form the “control.”

Keep in mind that by definition the case group is chosen because they already possess the attribute of interest. The point of the control group is to facilitate investigation, e.g., studying whether the case group systematically exhibits that attribute more than the control group does.

The strength of the association between an exposure and a disease in a case-control study can be measured using a few different statistical measures , such as odds ratios (ORs) and relative risk (RR).

No, case-control studies cannot establish causality as a standalone measure.

As observational studies , they can suggest associations between an exposure and a disease, but they cannot prove without a doubt that the exposure causes the disease. In particular, issues arising from timing, research biases like recall bias , and the selection of variables lead to low internal validity and the inability to determine causality.

Sources in this article

We strongly encourage students to use sources in their work. You can cite our article (APA Style) or take a deep dive into the articles below.

George, T. (2023, June 22). What Is a Case-Control Study? | Definition & Examples. Scribbr. Retrieved February 22, 2024, from https://www.scribbr.com/methodology/case-control-study/
Schlesselman, J. J. (1982). Case-Control Studies: Design, Conduct, Analysis (Monographs in Epidemiology and Biostatistics, 2) (Illustrated). Oxford University Press.

Is this article helpful?

Tegan George

Tegan George

Other students also liked, what is an observational study | guide & examples, control groups and treatment groups | uses & examples, cross-sectional study | definition, uses & examples, what is your plagiarism score.

What Is A Case Control Study?

Julia Simkus

Editor at Simply Psychology

BA (Hons) Psychology, Princeton University

Julia Simkus is a graduate of Princeton University with a Bachelor of Arts in Psychology. She is currently studying for a Master's Degree in Counseling for Mental Health and Wellness in September 2023. Julia's research has been published in peer reviewed journals.

Learn about our Editorial Process

Saul Mcleod, PhD

Editor-in-Chief for Simply Psychology

BSc (Hons) Psychology, MRes, PhD, University of Manchester

Saul Mcleod, Ph.D., is a qualified psychology teacher with over 18 years experience of working in further and higher education. He has been published in peer-reviewed journals, including the Journal of Clinical Psychology.

Olivia Guy-Evans, MSc

Associate Editor for Simply Psychology

BSc (Hons) Psychology, MSc Psychology of Education

Olivia Guy-Evans is a writer and associate editor for Simply Psychology. She has previously worked in healthcare and educational sectors.

On This Page:

A case-control study is a research method where two groups of people are compared – those with the condition (cases) and those without (controls). By looking at their past, researchers try to identify what factors might have contributed to the condition in the ‘case’ group.

Explanation

A case-control study looks at people who already have a certain condition (cases) and people who don’t (controls). By comparing these two groups, researchers try to figure out what might have caused the condition. They look into the past to find clues, like habits or experiences, that are different between the two groups.

The “cases” are the individuals with the disease or condition under study, and the “controls” are similar individuals without the disease or condition of interest.

The controls should have similar characteristics (i.e., age, sex, demographic, health status) to the cases to mitigate the effects of confounding variables .

Case-control studies identify any associations between an exposure and an outcome and help researchers form hypotheses about a particular population.

Researchers will first identify the two groups, and then look back in time to investigate which subjects in each group were exposed to the condition.

If the exposure is found more commonly in the cases than the controls, the researcher can hypothesize that the exposure may be linked to the outcome of interest.

Case Control Study

Figure: Schematic diagram of case-control study design. Kenneth F. Schulz and David A. Grimes (2002) Case-control studies: research in reverse . The Lancet Volume 359, Issue 9304, 431 – 434

Quick, inexpensive, and simple

Because these studies use already existing data and do not require any follow-up with subjects, they tend to be quicker and cheaper than other types of research. Case-control studies also do not require large sample sizes.

Beneficial for studying rare diseases

Researchers in case-control studies start with a population of people known to have the target disease instead of following a population and waiting to see who develops it. This enables researchers to identify current cases and enroll a sufficient number of patients with a particular rare disease.

Useful for preliminary research

Case-control studies are beneficial for an initial investigation of a suspected risk factor for a condition. The information obtained from cross-sectional studies then enables researchers to conduct further data analyses to explore any relationships in more depth.

Limitations

Subject to recall bias.

Participants might be unable to remember when they were exposed or omit other details that are important for the study. In addition, those with the outcome are more likely to recall and report exposures more clearly than those without the outcome.

Difficulty finding a suitable control group

It is important that the case group and the control group have almost the same characteristics, such as age, gender, demographics, and health status.

Forming an accurate control group can be challenging, so sometimes researchers enroll multiple control groups to bolster the strength of the case-control study.

Do not demonstrate causation

Case-control studies may prove an association between exposures and outcomes, but they can not demonstrate causation.

A case-control study is an observational study where researchers analyzed two groups of people (cases and controls) to look at factors associated with particular diseases or outcomes.

Below are some examples of case-control studies:
  • Investigating the impact of exposure to daylight on the health of office workers (Boubekri et al., 2014).
  • Comparing serum vitamin D levels in individuals who experience migraine headaches with their matched controls (Togha et al., 2018).
  • Analyzing correlations between parental smoking and childhood asthma (Strachan and Cook, 1998).
  • Studying the relationship between elevated concentrations of homocysteine and an increased risk of vascular diseases (Ford et al., 2002).
  • Assessing the magnitude of the association between Helicobacter pylori and the incidence of gastric cancer (Helicobacter and Cancer Collaborative Group, 2001).
  • Evaluating the association between breast cancer risk and saturated fat intake in postmenopausal women (Howe et al., 1990).

Frequently asked questions

1. what’s the difference between a case-control study and a cross-sectional study.

Case-control studies are different from cross-sectional studies in that case-control studies compare groups retrospectively while cross-sectional studies analyze information about a population at a specific point in time.

In  cross-sectional studies , researchers are simply examining a group of participants and depicting what already exists in the population.

2. What’s the difference between a case-control study and a longitudinal study?

Case-control studies compare groups retrospectively, while longitudinal studies can compare groups either retrospectively or prospectively.

In a  longitudinal study , researchers monitor a population over an extended period of time, and they can be used to study developmental shifts and understand how certain things change as we age.

In addition, case-control studies look at a single subject or a single case, whereas longitudinal studies can be conducted on a large group of subjects.

3. What’s the difference between a case-control study and a retrospective cohort study?

Case-control studies are retrospective as researchers begin with an outcome and trace backward to investigate exposure; however, they differ from retrospective cohort studies.

In a  retrospective cohort study , researchers examine a group before any of the subjects have developed the disease, then examine any factors that differed between the individuals who developed the condition and those who did not.

Thus, the outcome is measured after exposure in retrospective cohort studies, whereas the outcome is measured before the exposure in case-control studies.

Boubekri, M., Cheung, I., Reid, K., Wang, C., & Zee, P. (2014). Impact of windows and daylight exposure on overall health and sleep quality of office workers: a case-control pilot study. Journal of Clinical Sleep Medicine: JCSM: Official Publication of the American Academy of Sleep Medicine, 10 (6), 603-611.

Ford, E. S., Smith, S. J., Stroup, D. F., Steinberg, K. K., Mueller, P. W., & Thacker, S. B. (2002). Homocyst (e) ine and cardiovascular disease: a systematic review of the evidence with special emphasis on case-control studies and nested case-control studies. International journal of epidemiology, 31 (1), 59-70.

Helicobacter and Cancer Collaborative Group. (2001). Gastric cancer and Helicobacter pylori: a combined analysis of 12 case control studies nested within prospective cohorts. Gut, 49 (3), 347-353.

Howe, G. R., Hirohata, T., Hislop, T. G., Iscovich, J. M., Yuan, J. M., Katsouyanni, K., … & Shunzhang, Y. (1990). Dietary factors and risk of breast cancer: combined analysis of 12 case—control studies. JNCI: Journal of the National Cancer Institute, 82 (7), 561-569.

Lewallen, S., & Courtright, P. (1998). Epidemiology in practice: case-control studies. Community eye health, 11 (28), 57–58.

Strachan, D. P., & Cook, D. G. (1998). Parental smoking and childhood asthma: longitudinal and case-control studies. Thorax, 53 (3), 204-212.

Tenny, S., Kerndt, C. C., & Hoffman, M. R. (2021). Case Control Studies. In StatPearls . StatPearls Publishing.

Togha, M., Razeghi Jahromi, S., Ghorbani, Z., Martami, F., & Seifishahpar, M. (2018). Serum Vitamin D Status in a Group of Migraine Patients Compared With Healthy Controls: A Case-Control Study. Headache, 58 (10), 1530-1540.

Further Information

  • Schulz, K. F., & Grimes, D. A. (2002). Case-control studies: research in reverse. The Lancet, 359(9304), 431-434.
  • What is a case-control study?

Print Friendly, PDF & Email

Leave a Comment Cancel reply

You must be logged in to post a comment.

Case-Control Studies

case control study is a type of

Introduction

Cohort studies have an intuitive logic to them, but they can be very problematic when one is investigating outcomes that only occur in a small fraction of exposed and unexposed individuals. They can also be problematic when it is expensive or very difficult to obtain exposure information from a cohort. In these situations a case-control design offers an alternative that is much more efficient. The goal of a case-control study is the same as that of cohort studies, i.e., to estimate the magnitude of association between an exposure and an outcome. However, case-control studies employ a different sampling strategy that gives them greater efficiency.

Learning Objectives

After completing this module, the student will be able to:

  • Define and explain the distinguishing features of a case-control study
  • Describe  and identify the types of epidemiologic questions that can be addressed by case-control studies
  • Define what is meant by the term "source population"
  • Describe the purpose of controls in a case-control study
  • Describe differences between hospital-based and population-based case-control studies
  • Describe the principles of valid control selection
  • Explain the importance of using specific diagnostic criteria and explicit case definitions in case-control studies
  • Estimate and interpret the odds ratio from a case-control study
  • Identify the potential strengths and limitations of case-control studies

Overview of Case-Control Design

In the module entitled Overview of Analytic Studies it was noted that Rothman describes the case-control strategy as follows:

"Case-control studies are best understood by considering as the starting point a source population , which represents a hypothetical study population in which a cohort study might have been conducted. The source population is the population that gives rise to the cases included in the study. If a cohort study were undertaken, we would define the exposed and unexposed cohorts (or several cohorts) and from these populations obtain denominators for the incidence rates or risks that would be calculated for each cohort. We would then identify the number of cases occurring in each cohort and calculate the risk or incidence rate for each. In a case-control study the same cases are identified and classified as to whether they belong to the exposed or unexposed cohort. Instead of obtaining the denominators for the rates or risks, however, a control group is sampled from the entire source population that gives rise to the cases. Individuals in the control group are then classified into exposed and unexposed categories. The purpose of the control group is to determine the relative size of the exposed and unexposed components of the source population. Because the control group is used to estimate the distribution of exposure in the source population, the cardinal requirement of control selection is that the controls be sampled independently of exposure status."

To illustrate this consider the following hypothetical scenario in which the source population is the state of Massachusetts. Diseased individuals are red, and non-diseased individuals are blue. Exposed individuals are indicated by a whitish midsection. Note the following aspects of the depicted scenario:

  • The disease is rare.
  • There is a fairly large number of exposed individuals in the state, but most of these are not diseased.

Map of Massachusetts with thousands of icon people overlaid. A very small percentage of them are identified as having a rare disease.

If we somehow had exposure and outcome information on all of the subjects in the source population and looked at the association using a cohort design, we might find the data summarized in the contingency table below.

In this hypothetical example, we have data on all 6,000,000 people in the source population, and we could compute the probability of disease (i.e., the risk or incidence) in both the exposed group and the non-exposed group, because we have the denominators for both the exposed and non-exposed groups.

The table above summarizes all of the necessary information regarding exposure and outcome status for the population and enables us to compute a risk ratio as a measure of the strength of the association. Intuitively, we compute the probability of disease (the risk) in each exposure group and then compute the risk ratio as follows:

The problem , of course, is that we usually don't have the resources to get the data on all subjects in the population. If we took a random sample of even 5-10% of the population, we would have few diseased people in our sample, certainly not enough to produce a reasonably precise measure of association. Moreover, we would expend an inordinate amount of effort and money collecting exposure and outcome data on a large number of people who would not develop the outcome.

We need a method that allows us to retain all the people in the numerator of disease frequency (diseased people or "cases") but allows us to collect information from only a small proportion of the people that make up the denominator (population, or "controls"), most of whom do not have the disease of interest. The case-control design allows us to accomplish this. We identify and collect exposure information on all the cases, but identify and collect exposure information on only a sample of the population. Once we have the exposure information, we can assign subjects to the numerator and denominator of the exposed and unexposed groups. This is what Rothman means when he says,

"The purpose of the control group is to determine the relative size of the exposed and unexposed components of the source population."

In the above example, we would have identified all 1,300 cases, determined their exposure status, and ended up categorizing 700 as exposed and 600 as unexposed. We might have ransomly sampled 6,000 members of the population (instead of 6 million) in order to determine the exposure distribution in the total population. If our sampling method was random, we would expect that about 1,000 would be exposed and 5,000 unexposed (the same ratio as in the overall population). We calculate a similar measure as the risk ratio above, but substituting in the denominator a sample of the population ("controls") instead of the whole population:

Note that when we take a sample of the population, we no longer have a measure of disease frequency, because the denominator no longer represents the population. Therefore, we can no longer compute the probability or rate of disease incidence in each exposure group. We also can't calculate a risk or rate difference measure for the same reason. However, as we have seen, we can compute the relative probability of disease in the exposed vs. unexposed group. The term generally used for this measure is an odds ratio , described in more detail later in the module.

Consequently, when the outcome is uncommon, as in this case, the risk ratio can be estimated much more efficiently by using a case-control design. One would focus first on finding an adequate number of cases in order to determine the ratio of exposed to unexposed cases. Then, one only needs to take a sample of the population in order to estimate the relative size of the exposed and unexposed components of the source population. Note that if one can identify all of the cases that were reported to a registry or other database within a defined period of time, then it is possible to compute an estimate of the incidence of disease if the size of the population is known from census data.   While this is conceptually possible, it is rarely done, and we will not discuss it further in this course.

Toggle open/close quiz question

A Nested Case-Control Study

Suppose a prospective cohort study were conducted among almost 90,000 women for the purpose of studying the determinants of cancer and cardiovascular disease. After enrollment, the women provide baseline information on a host of exposures, and they also provide baseline blood and urine samples that are frozen for possible future use. The women are then followed, and, after about eight years, the investigators want to test the hypothesis that past exposure to pesticides such as DDT is a risk factor for breast cancer. Eight years have passed since the beginning of the study, and 1.439 women in the cohort have developed breast cancer. Since they froze blood samples at baseline, they have the option of analyzing all of the blood samples in order to ascertain exposure to DDT at the beginning of the study before any cancers occurred. The problem is that there are almost 90,000 women and it would cost $20 to analyze each of the blood samples. If the investigators could have analyzed all 90,000 samples this is what they would have found the results in the table below.

Table of Breast Cancer Occurrence Among Women With or Without DDT Exposure

If they had been able to afford analyzing all of the baseline blood specimens in order to categorize the women as having had DDT exposure or not, they would have found a risk ratio = 1.87 (95% confidence interval: 1.66-2.10). The problem is that this would have cost almost $1.8 million, and the investigators did not have the funding to do this.

While 1,439 breast cancers is a disturbing number, it is only 1.6% of the entire cohort, so the outcome is relatively rare, and it is costing a lot of money to analyze the blood specimens obtained from all of the non-diseased women. There is, however, another more efficient alternative, i.e., to use a case-control sampling strategy. One could analyze all of the blood samples from women who had developed breast cancer, but only a sample of the whole cohort in order to estimate the exposure distribution in the population that produced the cases.

If one were to analyze the blood samples of 2,878 of the non-diseased women (twice as many as the number of cases), one would obtain results that would look something like those in the next table.

Odds of Exposure: 360/1079 in the cases versus 432/2,446 in the non-diseased controls.

Totals Samples analyzed = 1,438+2,878 = 4,316

Total Cost = 4,316 x $20 = $86,320

With this approach a similar estimate of risk was obtained after analyzing blood samples from only a small sample of the entire population at a fraction of the cost with hardly any loss in precision. In essence, a case-control strategy was used, but it was conducted within the context of a prospective cohort study. This is referred to as a case-control study "nested" within a cohort study.

Rothman states that one should look upon all case-control studies as being "nested" within a cohort. In other words the cohort represents the source population that gave rise to the cases. With a case-control sampling strategy one simply takes a sample of the population in order to obtain an estimate of the exposure distribution within the population that gave rise to the cases. Obviously, this is a much more efficient design.

It is important to note that, unlike cohort studies, case-control studies do not follow subjects through time. Cases are enrolled at the time they develop disease and controls are enrolled at the same time. The exposure status of each is determined, but they are not followed into the future for further development of disease.

As with cohort studies, case-control studies can be prospective or retrospective. At the start of the study, all cases might have already occurred and then this would be a retrospective case-control study. Alternatively, none of the cases might have already occurred, and new cases will be enrolled prospectively. Epidemiologists generally prefer the prospective approach because it has fewer biases, but it is more expensive and sometimes not possible. When conducted prospectively, or when nested in a prospective cohort study, it is straightforward to select controls from the population at risk. However, in retrospective case-control studies, it can be difficult to select from the population at risk, and controls are then selected from those in the population who didn't develop disease. Using only the non-diseased to select controls as opposed to the whole population means the denominator is not really a measure of disease frequency, but when the disease is rare , the odds ratio using the non-diseased will be very similar to the estimate obtained when the entire population is used to sample for controls. This phenomenon is known as the r are-disease assumption . When case-control studies were first developed, most were conducted retrospectively, and it is sometimes assumed that the rare-disease assumption applies to all case-control studies. However, it actually only applies to those case-control studies in which controls are sampled only from the non-diseased rather than the whole population.  

The difference between sampling from the whole population and only the non-diseased is that the whole population contains people both with and without the disease of interest. This means that a sampling strategy that uses the whole population as its source must allow for the fact that people who develop the disease of interest can be selected as controls. Students often have a difficult time with this concept. It is helpful to remember that it seems natural that the population denominator includes people who develop the disease in a cohort study. If a case-control study is a more efficient way to obtain the information from a cohort study, then perhaps it is not so strange that the denominator in a case-control study also can include people who develop the disease. This topic is covered in more detail in EP813 Intermediate Epidemiology.

Retrospective and Prospective Case-Control Studies

Students usually think of case-control studies as being only retrospective, since the investigators enroll subjects who have developed the outcome of interest. However, case-control studies, like cohort studies, can be either retrospective or prospective. In a prospective case-control study, the investigator still enrolls based on outcome status, but the investigator must wait to the cases to occur.

When is a Case-Control Study Desirable?

Given the greater efficiency of case-control studies, they are particularly advantageous in the following situations:

  • When the disease or outcome being studied is rare.
  • When the disease or outcome has a long induction and latent period (i.e., a long time between exposure and the eventual causal manifestation of disease).
  • When exposure data is difficult or expensive to obtain.
  • When the study population is dynamic.
  • When little is known about the risk factors for the disease, case-control studies provide a way of testing associations with multiple potential risk factors. (This isn't really a unique advantage to case-control studies, however, since cohort studies can also assess multiple exposures.)

Another advantage of their greater efficiency, of course, is that they are less time-consuming and much less costly than prospective cohort studies.

The DES Case-Control Study

A classic example of the efficiency of the case-control approach is the study (Herbst et al.: N. Engl. J. Med. Herbst et al. (1971;284:878-81) that linked in-utero exposure to diethylstilbesterol (DES) with subsequent development of vaginal cancer 15-22 years later. In the late 1960s, physicians at MGH identified a very unusual cancer cluster. Eight young woman between the ages of 15-22 were found to have cancer of the vagina, an uncommon cancer even in elderly women. The cluster of cases in young women was initially reported as a case series, but there were no strong hypotheses about the cause.

In retrospect, the cause was in-utero exposure to DES. After World War II, DES started being prescribed for women who were having troubles with a pregnancy -- if there were signs suggesting the possibility of a miscarriage, DES was frequently prescribed. It has been estimated that between 1945-1950 DES was prescribed for about 20% of all pregnancies in the Boston area. Thus, the unborn fetus was exposed to DES in utero, and in a very small percentage of cases this resulted in development of vaginal cancer when the child was 15-22 years old (a very long latent period). There were several reasons why a case-control study was the only feasible way to identify this association: the disease was extremely rare (even in subjects who had been exposed to DES), there was a very long latent period between exposure and development of disease, and initially they had no idea what was responsible, so there were many possible exposures to consider.

In this situation, a case-control study was the only reasonable approach to identify the causative agent. Given how uncommon the outcome was, even a large prospective study would have been unlikely to have more than one or two cases, even after 15-20 years of follow-up. Similarly, a retrospective cohort study might have been successful in enrolling a large number of subjects, but the outcome of interest was so uncommon that few, if any, subjects would have had it. In contrast, a case-control study was conducted in which eight known cases and 32 age-matched controls provided information on many potential exposures. This strategy ultimately allowed the investigators to identify a highly significant association between the mother's treatment with DES during pregnancy and the eventual development of adenocarcinoma of the vagina in their daughters (in-utero at the time of exposure) 15 to 22 years later.

For more information see the DES Fact Sheet from the National Cancer Institute.

An excellent summary of this landmark study and the long-range effects of DES can be found in a Perspective article in the New England Journal of Medicine. A cohort of both mothers who took DES and their children (daughters and sons) was later formed to look for more common outcomes. Members of the faculty at BUSPH are on the team of investigators that follow this cohort for a variety of outcomes, particularly reproductive consequences and other cancers.

Selecting & Defining Cases and Controls

The "case" definition.

Careful thought should be given to the case definition to be used. If the definition is too broad or vague, it is easier to capture people with the outcome of interest, but a loose case definition will also capture people who do not have the disease. On the other hand, an overly restrictive case definition is employed, fewer cases will be captured, and the sample size may be limited. Investigators frequently wrestle with this problem during outbreak investigations. Initially, they will often use a somewhat broad definition in order to identify potential cases. However, as an outbreak investigation progresses, there is a tendency to narrow the case definition to make it more precise and specific, for example by requiring confirmation of the diagnosis by laboratory testing. In general, investigators conducting case-control studies should thoughtfully construct a definition that is as clear and specific as possible without being overly restrictive.

Investigators studying chronic diseases generally prefer newly diagnosed cases, because they tend to be more motivated to participate, may remember relevant exposures more accurately, and because it avoids complicating factors related to selection of longer duration (i.e., prevalent) cases. However, it is sometimes impossible to have an adequate sample size if only recent cases are enrolled.

Sources of Cases

Typical sources for cases include:

  • Patient rosters at medical facilities
  • Death certificates
  • Disease registries (e.g., cancer or birth defect registries; the SEER Program [Surveillance, Epidemiology and End Results] is a federally funded program that identifies newly diagnosed cases of cancer in population-based registries across the US )
  • Cross-sectional surveys (e.g., NHANES, the National Health and Nutrition Examination Survey)

Selection of the Controls

As noted above, it is always useful to think of a case-control study as being nested within some sort of a cohort, i.e., a source population that produced the cases that were identified and enrolled. In view of this there are two key principles that should be followed in selecting controls:

  • The comparison group ("controls") should be representative of the source population that produced the cases.
  • The "controls" must be sampled in a way that is independent of the exposure, meaning that their selection should not be more (or less) likely if they have the exposure of interest.

If either of these principles are not adhered to, selection bias can result (as discussed in detail in the module on Bias).

case control study is a type of

Note that in the earlier example of a case-control study conducted in the Massachusetts population, we specified that our sampling method was random so that exposed and unexposed members of the population had an equal chance of being selected. Therefore, we would expect that about 1,000 would be exposed and 5,000 unexposed (the same ratio as in the whole population), and came up with an odds ratio that was same as the hypothetical risk ratio we would have had if we had collected exposure information from the whole population of six million:

What if we had instead been more likely to sample those who were exposed, so that we instead found 1,500 exposed and 4,500 unexposed among the 6,000 controls?   Then the odds ratio would have been:

This odds ratio is biased because it differs from the true odds ratio.   In this case, the bias stemmed from the fact that we violated the second principle in selection of controls. Depending on which category is over or under-sampled, this type of bias can result in either an underestimate or an overestimate of the true association.

A hypothetical case-control study was conducted to determine whether lower socioeconomic status (the exposure) is associated with a higher risk of cervical cancer (the outcome). The "cases" consisted of 250 women with cervical cancer who were referred to Massachusetts General Hospital for treatment for cervical cancer. They were referred from all over the state. The cases were asked a series of questions relating to socioeconomic status (household income, employment, education, etc.). The investigators identified control subjects by going door-to-door in the community around MGH from 9:00 AM to 5:00  PM. Many residents are not home, but they persist and eventually enroll enough controls. The problem is that the controls were selected by a different mechanism than the cases, AND the selection mechanism may have tended to select individuals of different socioeconomic status, since women who were at home may have been somewhat more likely to be unemployed. In other words, the controls were more likely to be enrolled (selected) if they had the exposure of interest (lower socioeconomic status). 

Toggle open/close quiz question

Sources for "Controls"

Population controls:.

A population-based case-control study is one in which the cases come from a precisely defined population, such as a fixed geographic area, and the controls are sampled directly from the same population. In this situation cases might be identified from a state cancer registry, for example, and the comparison group would logically be selected at random from the same source population. Population controls can be identified from voter registration lists, tax rolls, drivers license lists, and telephone directories or by "random digit dialing". Population controls may also be more difficult to obtain, however, because of lack of interest in participating, and there may be recall bias, since population controls are generally healthy and may remember past exposures less accurately.

Example of a Population-based Case-Control Study: Rollison et al. reported on a "Population-based Case-Control Study of Diabetes and Breast Cancer Risk in Hispanic and Non-Hispanic White Women Living in US Southwestern States". (ALink to the article - Citation: Am J Epidemiol 2008;167:447–456).

"Briefly, a population-based case-control study of breast cancer was conducted in Colorado, New Mexico, Utah, and selected counties of Arizona. For investigation of differences in the breast cancer risk profiles of non-Hispanic Whites and Hispanics, sampling was stratified by race/ethnicity, and only women who self-reported their race as non-Hispanic White, Hispanic, or American Indian were eligible, with the exception of American Indian women living on reservations. Women diagnosed with histologically confirmed breast cancer between October 1999 and May 2004 (International Classification of Diseases for Oncology codes C50.0–C50.6 and C50.8–C50.9) were identified as cases through population-based cancer registries in each state."

"Population-based controls were frequency-matched to cases in 5-year age groups. In New Mexico and Utah, control participants under age 65 years were randomly selected from driver's license lists; in Arizona and Colorado, controls were randomly selected from commercial mailing lists, since driver's license lists were unavailable. In all states, women aged 65 years or older were randomly selected from the lists of the Centers for Medicare and Medicaid Services (Social Security lists). Of all women contacted, 68 percent of cases and 42 percent of controls participated in the study."

"Odds ratios and 95% confidence intervals were calculated using logistic regression, adjusting for age, body mass index at age 15 years, and parity. Having any type of diabetes was not associated with breast cancer overall (odds ratio = 0.94, 95% confidence interval: 0.78, 1.12). Type 2 diabetes was observed among 19% of Hispanics and 9% of non-Hispanic Whites but was not associated with breast cancer in either group."

In this example, it is clear that the controls were selected from the source population (principle 1), but less clear that they were enrolled independent of exposure status (principle 2), both because drivers' licenses were used for selection and because the participation rate among controls was low. These factors would only matter if they impacted on the estimate of the proportion of the population who had diabetes.

Hospital or Clinic Controls:

case control study is a type of

  • They have diseases that are unrelated to the exposure being studied. For example, for a study examining the association between smoking and lung cancer, it would not be appropriate to include patients with cardiovascular disease as control, since smoking is a risk factor for cardiovascular disease. To include such patients as controls would result in an underestimate of the true association.
  • Second, control patients in the comparison should have diseases with similar referral patterns as the cases, in order to minimize selection bias. For example, if the cases are women with cervical cancer who have been referred from all over the state, it would be inappropriate to use controls consisting of women with diabetes who had been referred primarily from local health centers in the immediate vicinity of the hospital. Similarly, it would be inappropriate to use patients from the emergency room, because the selection of a hospital for an emergency is different than for cancer, and this difference might be related to the exposure of interest.

The advantages of using controls who are patients from the same facility are:

  • They are easier to identify
  • They are more likely to participate than general population controls.
  • They minimize selection bias because they generally come from the same source population (provided referral patterns are similar).
  • Recall bias would be minimized, because they are sick, but with a different diagnosis.

Example: Several years ago the vascular surgeons at Boston Medical Center wanted to study risk factors for severe atherosclerosis of the lower extremities. The cases were patients who were referred to the hospital for elective surgery to bypass severe atherosclerotic blockages in the arteries to the legs. The controls consisted of patients who were admitted to the same hospital for elective joint replacement of the hip or knee. The patients undergoing joint replacement were similar in age and they also were following the same referral pathways. In other words, they met the "would" criterion: if one of the joint replacement surgery patients had developed severe atherosclerosis in their leg arteries, they would have been referred to the same hospital.

Friend, Neighbor, Spouse, and Relative Controls:

Occasionally investigators will ask cases to nominate controls who are in one of these categories, because they have similar characteristics, such as genotype, socioeconomic status, or environment, i.e., factors that can cause confounding, but are hard to measure and adjust for. By matching cases and controls on these factors, confounding by these factors will be controlled.   However, one must be careful that the controls satisfy the two fundamental principles. Often, they do not.

How Many Controls?

Since case-control studies are often used for uncommon outcomes, investigators often have a limited number of cases but a plentiful supply of potential controls. In this situation the statistical power of the study can be increased somewhat by enrolling more controls than cases. However, the additional power that is achieved diminishes as the ratio of controls to cases increases, and ratios greater than 4:1 have little additional impact on power. Consequently, if it is time-consuming or expensive to collect data on controls, the ratio of controls to cases should be no more than 4:1. However, if the data on controls is easily obtained, there is no reason to limit the number of controls.

Methods of Control Sampling

There are three strategies for selecting controls that are best explained by considering the nested case-control study described on page 3 of this module:

  • Survivor sampling: This is the most common method. Controls consist of individuals from the source population who do not have the outcome of interest.
  • Case-base sampling (also known as "case-cohort" sampling): Controls are selected from the population at risk at the beginning of the follow-up period in the cohort study within which the case-control study was nested.
  • Risk Set Sampling: In the nested case-control study a control would be selected from the population at risk at the point in time when a case was diagnosed.

The Rare Outcome Assumption

It is often said that an odds ratio provides a good estimate of the risk ratio only when the outcome of interest is rare, but this is only true when survivor sampling is used. With case-base sampling or risk set sampling, the odds ratio will provide a good estimate of the risk ratio regardless of the frequency of the outcome, because the controls will provide an accurate estimate of the distribution in the source population (i.e., not just in non-diseased people).

More on Selection Bias

Always consider the source population for case-control studies, i.e. the "population" that generated the cases. The cases are always identified and enrolled by some method or a set of procedures or circumstances. For example, cases with a certain disease might be referred to a particular tertiary hospital for specialized treatment. Alternatively, if there is a database or a disease registry for a geographic area, cases might be selected at random from the database. The key to avoiding selection bias is to select the controls by a similar, if not identical, mechanism in order to ensure that the controls provide an accurate representation of the exposure status of the source population.

Example 1: In the first example above, in which cases were randomly selected from a geographically defined database, the source population is also defined geographically, so it would make sense to select population controls by some random method. In contrast, if one enrolled controls from a particular hospital within the geographic area, one would have to at least consider whether the controls were inherently more or less likely to have the exposure of interest. If so, they would not provide an accurate estimate of the exposure distribution of the source population, and selection bias would result.

Example 2: In the second example above, the source population was defined by the patterns of referral to a particular hospital for a particular disease. In order for the controls to be representative of the "population" that produced those cases, the controls should be selected by a similar mechanism, e.g., by contacting the referring health care providers and asking them to provide the names of potential controls. By this mechanism, one can ensure that the controls are representative of the source population, because if they had had the disease of interest they would have been just as likely as the cases to have been included in the case group (thus fulfilling the "would" criterion).

Example 3: A food handler at a delicatessen who is infected with hepatitis A virus is responsible for an outbreak of hepatitis which is largely confined to the surrounding community from which most of the customers come. Many (but not all) of the infected cases are identified by passive and active surveillance. How should controls be selected? In this situation, one might guess that the likelihood of people going to the delicatessen would be heavily influenced by their proximity to it, and this would to a large extent define the source population. In a case-control study undertaken to identify the source, the delicatessen is one of the exposures being tested. Consequently, even if the cases were reported to the state-wide surveillance system, it would not be appropriate to randomly select controls from the state, the county, or even the town where the delicatessen is located. In other words, the "would" criterion doesn't work here, because anyone in the state with clinical hepatitis would end up in the surveillance system, but someone who lived far from the deli would have a much lower likelihood of having the exposure. A better approach would be to select controls who were matched to the cases by neighborhood, age, and gender. These controls would have similar access to go to the deli if they chose to, and they would therefore be more representative of the source population.

Analysis of Case-Control Studies

The computation and interpretation of the odds ratio in a case-control study has already been discussed in the modules on Overview of Analytic Studies and Measures of Association. Additionally, one can compute the confidence interval for the odds ratio, and statistical significance can also be evaluated by using a chi-square test (or a Fisher's Exact Test if the sample size is small) to compute a p-value. These calculations can be done using the Case-Control worksheet in the Excel file called EpiTools.XLS.

Image of the Case-Control worksheet in the Epi_Tools file

Advantages and Disadvantages of Case-Control Studies

Advantages:

  • They are efficient for rare diseases or diseases with a long latency period between exposure and disease manifestation.
  • They are less costly and less time-consuming; they are advantageous when exposure data is expensive or hard to obtain.
  • They are advantageous when studying dynamic populations in which follow-up is difficult.

Disadvantages:

  • They are subject to selection bias.
  • They are inefficient for rare exposures.
  • Information on exposure is subject to observation bias.
  • They generally do not allow calculation of incidence (absolute risk).
  • En español – ExME
  • Em português – EME

Case-control and Cohort studies: A brief overview

Posted on 6th December 2017 by Saul Crandon

Man in suit with binoculars

Introduction

Case-control and cohort studies are observational studies that lie near the middle of the hierarchy of evidence . These types of studies, along with randomised controlled trials, constitute analytical studies, whereas case reports and case series define descriptive studies (1). Although these studies are not ranked as highly as randomised controlled trials, they can provide strong evidence if designed appropriately.

Case-control studies

Case-control studies are retrospective. They clearly define two groups at the start: one with the outcome/disease and one without the outcome/disease. They look back to assess whether there is a statistically significant difference in the rates of exposure to a defined risk factor between the groups. See Figure 1 for a pictorial representation of a case-control study design. This can suggest associations between the risk factor and development of the disease in question, although no definitive causality can be drawn. The main outcome measure in case-control studies is odds ratio (OR) .

case control study is a type of

Figure 1. Case-control study design.

Cases should be selected based on objective inclusion and exclusion criteria from a reliable source such as a disease registry. An inherent issue with selecting cases is that a certain proportion of those with the disease would not have a formal diagnosis, may not present for medical care, may be misdiagnosed or may have died before getting a diagnosis. Regardless of how the cases are selected, they should be representative of the broader disease population that you are investigating to ensure generalisability.

Case-control studies should include two groups that are identical EXCEPT for their outcome / disease status.

As such, controls should also be selected carefully. It is possible to match controls to the cases selected on the basis of various factors (e.g. age, sex) to ensure these do not confound the study results. It may even increase statistical power and study precision by choosing up to three or four controls per case (2).

Case-controls can provide fast results and they are cheaper to perform than most other studies. The fact that the analysis is retrospective, allows rare diseases or diseases with long latency periods to be investigated. Furthermore, you can assess multiple exposures to get a better understanding of possible risk factors for the defined outcome / disease.

Nevertheless, as case-controls are retrospective, they are more prone to bias. One of the main examples is recall bias. Often case-control studies require the participants to self-report their exposure to a certain factor. Recall bias is the systematic difference in how the two groups may recall past events e.g. in a study investigating stillbirth, a mother who experienced this may recall the possible contributing factors a lot more vividly than a mother who had a healthy birth.

A summary of the pros and cons of case-control studies are provided in Table 1.

case control study is a type of

Table 1. Advantages and disadvantages of case-control studies.

Cohort studies

Cohort studies can be retrospective or prospective. Retrospective cohort studies are NOT the same as case-control studies.

In retrospective cohort studies, the exposure and outcomes have already happened. They are usually conducted on data that already exists (from prospective studies) and the exposures are defined before looking at the existing outcome data to see whether exposure to a risk factor is associated with a statistically significant difference in the outcome development rate.

Prospective cohort studies are more common. People are recruited into cohort studies regardless of their exposure or outcome status. This is one of their important strengths. People are often recruited because of their geographical area or occupation, for example, and researchers can then measure and analyse a range of exposures and outcomes.

The study then follows these participants for a defined period to assess the proportion that develop the outcome/disease of interest. See Figure 2 for a pictorial representation of a cohort study design. Therefore, cohort studies are good for assessing prognosis, risk factors and harm. The outcome measure in cohort studies is usually a risk ratio / relative risk (RR).

case control study is a type of

Figure 2. Cohort study design.

Cohort studies should include two groups that are identical EXCEPT for their exposure status.

As a result, both exposed and unexposed groups should be recruited from the same source population. Another important consideration is attrition. If a significant number of participants are not followed up (lost, death, dropped out) then this may impact the validity of the study. Not only does it decrease the study’s power, but there may be attrition bias – a significant difference between the groups of those that did not complete the study.

Cohort studies can assess a range of outcomes allowing an exposure to be rigorously assessed for its impact in developing disease. Additionally, they are good for rare exposures, e.g. contact with a chemical radiation blast.

Whilst cohort studies are useful, they can be expensive and time-consuming, especially if a long follow-up period is chosen or the disease itself is rare or has a long latency.

A summary of the pros and cons of cohort studies are provided in Table 2.

case control study is a type of

The Strengthening of Reporting of Observational Studies in Epidemiology Statement (STROBE)

STROBE provides a checklist of important steps for conducting these types of studies, as well as acting as best-practice reporting guidelines (3). Both case-control and cohort studies are observational, with varying advantages and disadvantages. However, the most important factor to the quality of evidence these studies provide, is their methodological quality.

  • Song, J. and Chung, K. Observational Studies: Cohort and Case-Control Studies .  Plastic and Reconstructive Surgery.  2010 Dec;126(6):2234-2242.
  • Ury HK. Efficiency of case-control studies with multiple controls per case: Continuous or dichotomous data .  Biometrics . 1975 Sep;31(3):643–649.
  • von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative.  The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies.   Lancet 2007 Oct;370(9596):1453-14577. PMID: 18064739.

' src=

Saul Crandon

Leave a reply cancel reply.

Your email address will not be published. Required fields are marked *

Save my name, email, and website in this browser for the next time I comment.

No Comments on Case-control and Cohort studies: A brief overview

' src=

Very well presented, excellent clarifications. Has put me right back into class, literally!

' src=

Very clear and informative! Thank you.

' src=

very informative article.

' src=

Thank you for the easy to understand blog in cohort studies. I want to follow a group of people with and without a disease to see what health outcomes occurs to them in future such as hospitalisations, diagnoses, procedures etc, as I have many health outcomes to consider, my questions is how to make sure these outcomes has not occurred before the “exposure disease”. As, in cohort studies we are looking at incidence (new) cases, so if an outcome have occurred before the exposure, I can leave them out of the analysis. But because I am not looking at a single outcome which can be checked easily and if happened before exposure can be left out. I have EHR data, so all the exposure and outcome have occurred. my aim is to check the rates of different health outcomes between the exposed)dementia) and unexposed(non-dementia) individuals.

' src=

Very helpful information

' src=

Thanks for making this subject student friendly and easier to understand. A great help.

' src=

Thanks a lot. It really helped me to understand the topic. I am taking epidemiology class this winter, and your paper really saved me.

Happy new year.

' src=

Wow its amazing n simple way of briefing ,which i was enjoyed to learn this.its very easy n quick to pick ideas .. Thanks n stay connected

' src=

Saul you absolute melt! Really good work man

' src=

am a student of public health. This information is simple and well presented to the point. Thank you so much.

' src=

very helpful information provided here

' src=

really thanks for wonderful information because i doing my bachelor degree research by survival model

' src=

Quite informative thank you so much for the info please continue posting. An mph student with Africa university Zimbabwe.

' src=

Thank you this was so helpful amazing

' src=

Apreciated the information provided above.

' src=

So clear and perfect. The language is simple and superb.I am recommending this to all budding epidemiology students. Thanks a lot.

' src=

Great to hear, thank you AJ!

' src=

I have recently completed an investigational study where evidence of phlebitis was determined in a control cohort by data mining from electronic medical records. We then introduced an intervention in an attempt to reduce incidence of phlebitis in a second cohort. Again, results were determined by data mining. This was an expedited study, so there subjects were enrolled in a specific cohort based on date(s) of the drug infused. How do I define this study? Thanks so much.

' src=

thanks for the information and knowledge about observational studies. am a masters student in public health/epidemilogy of the faculty of medicines and pharmaceutical sciences , University of Dschang. this information is very explicit and straight to the point

' src=

Very much helpful

Subscribe to our newsletter

You will receive our monthly newsletter and free access to Trip Premium.

Related Articles

""

Cluster Randomized Trials: Concepts

This blog summarizes the concepts of cluster randomization, and the logistical and statistical considerations while designing a cluster randomized controlled trial.

""

Expertise-based Randomized Controlled Trials

This blog summarizes the concepts of Expertise-based randomized controlled trials with a focus on the advantages and challenges associated with this type of study.

""

An introduction to different types of study design

Conducting successful research requires choosing the appropriate study design. This article describes the most common types of designs conducted by researchers.

Study Design 101: Case Control Study

  • Case Report
  • Case Control Study
  • Cohort Study
  • Randomized Controlled Trial
  • Practice Guideline
  • Systematic Review
  • Meta-Analysis
  • Helpful Formulas
  • Finding Specific Study Types

A study that compares patients who have a disease or outcome of interest (cases) with patients who do not have the disease or outcome (controls), and looks back retrospectively to compare how frequently the exposure to a risk factor is present in each group to determine the relationship between the risk factor and the disease.

Case control studies are observational because no intervention is attempted and no attempt is made to alter the course of the disease. The goal is to retrospectively determine the exposure to the risk factor of interest from each of the two groups of individuals: cases and controls. These studies are designed to estimate odds.

Case control studies are also known as "retrospective studies" and "case-referent studies."

  • Good for studying rare conditions or diseases
  • Less time needed to conduct the study because the condition or disease has already occurred
  • Lets you simultaneously look at multiple risk factors
  • Useful as initial studies to establish an association
  • Can answer questions that could not be answered through other study designs

Disadvantages

  • Retrospective studies have more problems with data quality because they rely on memory and people with a condition will be more motivated to recall risk factors (also called recall bias).
  • Not good for evaluating diagnostic tests because it's already clear that the cases have the condition and the controls do not
  • It can be difficult to find a suitable control group

Design pitfalls to look out for

Care should be taken to avoid confounding, which arises when an exposure and an outcome are both strongly associated with a third variable. Controls should be subjects who might have been cases in the study but are selected independent of the exposure. Cases and controls should also not be "over-matched."

Is the control group appropriate for the population? Does the study use matching or pairing appropriately to avoid the effects of a confounding variable? Does it use appropriate inclusion and exclusion criteria?

Fictitious Example

There is a suspicion that zinc oxide, the white non-absorbent sunscreen traditionally worn by lifeguards is more effective at preventing sunburns that lead to skin cancer than absorbent sunscreen lotions. A case-control study was conducted to investigate if exposure to zinc oxide is a more effective skin cancer prevention measure. The study involved comparing a group of former lifeguards that had developed cancer on their cheeks and noses (cases) to a group of lifeguards without this type of cancer (controls) and assess their prior exposure to zinc oxide or absorbent sunscreen lotions.

This study would be retrospective in that the former lifeguards would be asked to recall which type of sunscreen they used on their face and approximately how often. This could be either a matched or unmatched study, but efforts would need to be made to ensure that the former lifeguards are of the same average age, and lifeguarded for a similar number of seasons and amount of time per season.

Real-life Examples

Boubekri, M., Cheung, I., Reid, K., Wang, C., & Zee, P. (2014). Impact of windows and daylight exposure on overall health and sleep quality of office workers: a case-control pilot study. Journal of Clinical Sleep Medicine : JCSM : Official Publication of the American Academy of Sleep Medicine, 10 (6), 603-611. https://doi.org/10.5664/jcsm.3780

This pilot study explored the impact of exposure to daylight on the health of office workers (measuring well-being and sleep quality subjectively, and light exposure, activity level and sleep-wake patterns via actigraphy). Individuals with windows in their workplaces had more light exposure, longer sleep duration, and more physical activity. They also reported a better scores in the areas of vitality and role limitations due to physical problems, better sleep quality and less sleep disturbances.

Togha, M., Razeghi Jahromi, S., Ghorbani, Z., Martami, F., & Seifishahpar, M. (2018). Serum Vitamin D Status in a Group of Migraine Patients Compared With Healthy Controls: A Case-Control Study. Headache, 58 (10), 1530-1540. https://doi.org/10.1111/head.13423

This case-control study compared serum vitamin D levels in individuals who experience migraine headaches with their matched controls. Studied over a period of thirty days, individuals with higher levels of serum Vitamin D was associated with lower odds of migraine headache.

Related Formulas

  • Odds ratio in an unmatched study
  • Odds ratio in a matched study

Related Terms

A patient with the disease or outcome of interest.

Confounding

When an exposure and an outcome are both strongly associated with a third variable.

A patient who does not have the disease or outcome.

Matched Design

Each case is matched individually with a control according to certain characteristics such as age and gender. It is important to remember that the concordant pairs (pairs in which the case and control are either both exposed or both not exposed) tell us nothing about the risk of exposure separately for cases or controls.

Observed Assignment

The method of assignment of individuals to study and control groups in observational studies when the investigator does not intervene to perform the assignment.

Unmatched Design

The controls are a sample from a suitable non-affected population.

Now test yourself!

1. Case Control Studies are prospective in that they follow the cases and controls over time and observe what occurs.

a) True b) False

2. Which of the following is an advantage of Case Control Studies?

a) They can simultaneously look at multiple risk factors. b) They are useful to initially establish an association between a risk factor and a disease or outcome. c) They take less time to complete because the condition or disease has already occurred. d) b and c only e) a, b, and c

Evidence Pyramid - Navigation

  • Meta- Analysis
  • Case Reports
  • << Previous: Case Report
  • Next: Cohort Study >>

Creative Commons License

  • Last Updated: Sep 25, 2023 10:59 AM
  • URL: https://guides.himmelfarb.gwu.edu/studydesign101

GW logo

  • Himmelfarb Intranet
  • Privacy Notice
  • Terms of Use
  • GW is committed to digital accessibility. If you experience a barrier that affects your ability to access content on this page, let us know via the Accessibility Feedback Form .
  • Himmelfarb Health Sciences Library
  • 2300 Eye St., NW, Washington, DC 20037
  • Phone: (202) 994-2850
  • [email protected]
  • https://himmelfarb.gwu.edu

Quantitative study designs: Case Control

Quantitative study designs.

  • Introduction
  • Cohort Studies
  • Randomised Controlled Trial

Case Control

  • Cross-Sectional Studies
  • Study Designs Home

In a Case-Control study there are two groups of people: one has a health issue (Case group), and this group is “matched” to a Control group without the health issue based on characteristics like age, gender, occupation. In this study type, we can look back in the patient’s histories to look for exposure to risk factors that are common to the Case group, but not the Control group. It was a case-control study that demonstrated a link between carcinoma of the lung and smoking tobacco . These studies estimate the odds between the exposure and the health outcome, however they cannot prove causality. Case-Control studies might also be referred to as retrospective or case-referent studies. 

Stages of a Case-Control study

This diagram represents taking both the case (disease) and the control (no disease) groups and looking back at their histories to determine their exposure to possible contributing factors.  The researchers then determine the likelihood of those factors contributing to the disease.

case control study is a type of

(FOR ACCESSIBILITY: A case control study is likely to show that most, but not all exposed people end up with the health issue, and some unexposed people may also develop the health issue)

Which Clinical Questions does Case-Control best answer?

Case-Control studies are best used for Prognosis questions.

For example: Do anticholinergic drugs increase the risk of dementia in later life? (See BMJ Case-Control study Anticholinergic drugs and risk of dementia: case-control study )

What are the advantages and disadvantages to consider when using Case-Control?

* Confounding occurs when the elements of the study design invalidate the result. It is usually unintentional. It is important to avoid confounding, which can happen in a few ways within Case-Control studies. This explains why it is lower in the hierarchy of evidence, superior only to Case Studies.

What does a strong Case-Control study look like?

A strong study will have:

  • Well-matched controls, similar background without being so similar that they are likely to end up with the same health issue (this can be easier said than done since the risk factors are unknown). 
  • Detailed medical histories are available, reducing the emphasis on a patient’s unreliable recall of their potential exposures. 

What are the pitfalls to look for?

  • Poorly matched or over-matched controls.  Poorly matched means that not enough factors are similar between the Case and Control. E.g. age, gender, geography. Over-matched conversely means that so many things match (age, occupation, geography, health habits) that in all likelihood the Control group will also end up with the same health issue! Either of these situations could cause the study to become ineffective. 
  • Selection bias: Selection of Controls is biased. E.g. All Controls are in the hospital, so they’re likely already sick, they’re not a true sample of the wider population. 
  • Cases include persons showing early symptoms who never ended up having the illness. 

Critical appraisal tools 

To assist with critically appraising case control studies there are some tools / checklists you can use.

CASP - Case Control Checklist

JBI – Critical appraisal checklist for case control studies

CEBMA – Centre for Evidence Based Management  – Critical appraisal questions (focus on leadership and management)

STROBE - Observational Studies checklists includes Case control

SIGN - Case-Control Studies Checklist

NCCEH - Critical Appraisal of a Case Control Study for environmental health

Real World Examples

Smoking and carcinoma of the lung; preliminary report

  • Doll, R., & Hill, A. B. (1950). Smoking and carcinoma of the lung; preliminary report.  British Medical Journal ,  2 (4682), 739–748. Retrieved from  https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2038856/
  • Key Case-Control study linking tobacco smoking with lung cancer
  • Notes a marked increase in incidence of Lung Cancer disproportionate to population growth.
  • 20 London Hospitals contributed current Cases of lung, stomach, colon and rectum cancer via admissions, house-physician and radiotherapy diagnosis, non-cancer Controls were selected at each hospital of the same-sex and within 5 year age group of each.
  • 1732 Cases and 743 Controls were interviewed for social class, gender, age, exposure to urban pollution, occupation and smoking habits.
  • It was found that continued smoking from a younger age and smoking a greater number of cigarettes correlated with incidence of lung cancer.

Anticholinergic drugs and risk of dementia: case-control study

  • Richardson, K., Fox, C., Maidment, I., Steel, N., Loke, Y. K., Arthur, A., . . . Savva, G. M. (2018). Anticholinergic drugs and risk of dementia: case-control study. BMJ , 361, k1315. Retrieved from  http://www.bmj.com/content/361/bmj.k1315.abstract .
  • A recent study linking the duration and level of exposure to Anticholinergic drugs and subsequent onset of dementia.
  • Anticholinergic Cognitive Burden (ACB) was estimated in various drugs, the higher the exposure (measured as the ACB score) the greater likeliness of onset of dementia later in life.
  • Antidepressant, urological, and antiparkinson drugs with an ACB score of 3 increased the risk of dementia. Gastrointestinal drugs with an ACB score of 3 were not strongly linked with onset of dementia.
  • Tricyclic antidepressants such as Amitriptyline have an ACB score of 3 and are an example of a common area of concern.

Omega-3 deficiency associated with perinatal depression: Case-Control study 

  • Rees, A.-M., Austin, M.-P., Owen, C., & Parker, G. (2009). Omega-3 deficiency associated with perinatal depression: Case control study. Psychiatry Research , 166(2), 254-259. Retrieved from  http://www.sciencedirect.com/science/article/pii/S0165178107004398 .
  • During pregnancy women lose Omega-3 polyunsaturated fatty acids to the developing foetus.
  • There is a known link between Omgea-3 depletion and depression
  • Sixteen depressed and 22 non-depressed women were recruited during their third trimester
  • High levels of Omega-3 were associated with significantly lower levels of depression.
  • Women with low levels of Omega-3 were six times more likely to be depressed during pregnancy.

References and Further Reading

Doll, R., & Hill, A. B. (1950). Smoking and carcinoma of the lung; preliminary report. British Medical Journal, 2(4682), 739–748. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2038856/

Greenhalgh, Trisha. How to Read a Paper: the Basics of Evidence-Based Medicine, John Wiley & Sons, Incorporated, 2014. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/deakin/detail.action?docID=1642418 .

Himmelfarb Health Sciences Library. (2019). Study Design 101: Case-Control Study. Retrieved from https://himmelfarb.gwu.edu/tutorials/studydesign101/casecontrols.cfm   

Hoffmann, T., Bennett, S., & Del Mar, C. (2017). Evidence-Based Practice Across the Health Professions (Third edition. ed.): Elsevier. 

Lewallen, S., & Courtright, P. (1998). Epidemiology in practice: case-control studies. Community Eye Health, 11(28), 57.  https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1706071/  

Pelham, B. W. a., & Blanton, H. (2013). Conducting research in psychology : measuring the weight of smoke /Brett W. Pelham, Hart Blanton (Fourth edition. ed.): Wadsworth Cengage Learning. 

Rees, A.-M., Austin, M.-P., Owen, C., & Parker, G. (2009). Omega-3 deficiency associated with perinatal depression: Case control study. Psychiatry Research, 166(2), 254-259. Retrieved from http://www.sciencedirect.com/science/article/pii/S0165178107004398

Richardson, K., Fox, C., Maidment, I., Steel, N., Loke, Y. K., Arthur, A., … Savva, G. M. (2018). Anticholinergic drugs and risk of dementia: case-control study. BMJ, 361, k1315. Retrieved from http://www.bmj.com/content/361/bmj.k1315.abstract

Statistics How To. (2019). Case-Control Study: Definition, Real Life Examples. Retrieved from https://www.statisticshowto.com/case-control-study/  

  • << Previous: Randomised Controlled Trial
  • Next: Cross-Sectional Studies >>
  • Last Updated: Jan 17, 2024 11:49 AM
  • URL: https://deakin.libguides.com/quantitative-study-designs

Cookies on GOV.UK

We use some essential cookies to make this website work.

We’d like to set additional cookies to understand how you use GOV.UK, remember your settings and improve government services.

We also use cookies set by other sites to help us deliver content from their services.

You have accepted additional cookies. You can change your cookie settings at any time.

You have rejected additional cookies. You can change your cookie settings at any time.

case control study is a type of

  • Health and social care
  • Public health
  • Health improvement

Case-control study: comparative studies

How to use a case-control study to evaluate your digital health product.

This page is part of a collection of guidance on evaluating digital health products .

A case-control study is a type of observational study. It looks at 2 sets of participants. One group has the condition you are interested in (the cases) and one group does not have it (the controls).

In other respects, the participants in both groups are similar. You can then look at a particular factor that might have caused the condition, such as your digital product, and compare participants from the 2 groups in relation to that.

A case-control study is an observational study because you observe the effects on existing groups rather than designing an experiment where participants are allocated into different groups.

What to use it for

A case-control study can help you to find out if your digital product or service achieves its aims, so it can be useful when you have developed your product (summative evaluation).

It can be a useful method when it would be difficult or impossible to randomise participants, for example, if your product aims to help people with rare health conditions.

Case-control studies have many benefits.

  • help to estimate the effects of your digital product when randomisation is not possible
  • use existing data, which could be cheaper and easier
  • operate with fewer participants compared to other designs

There can also be drawbacks of a case-control study.

For example:

  • you need to pay careful attention to factors that may influence your results, confounding factors and biases – see explanation in ‘How to carry out a case-control study’ below
  • there may be challenges when accessing pre-existing data
  • you cannot draw definitive answers about the effects of your product as you haven’t randomly selected participants for your evaluation

How to carry out a case-control study

In a traditional case-control design, cases and controls are looked at retrospectively – that is, the health condition and the factor that might have caused it have already occurred when you start the study.

Sources of cases and controls typically include:

  • routinely collected data at medical facilities
  • disease registries
  • cross-sectional surveys

Some researchers use the term prospective case-control study when, for example, a prospective group exposed to an intervention is compared to a retrospective control.

Choosing your control

Selecting an appropriate control is an important part of a case-control study. The comparison group should be as similar as possible to the source population that produced the cases. This means the participants will be similar to each other in terms of factors that may influence the outcomes you’re looking at. Ideally, they will only differ in whether they received your digital product (cases) or not (controls).

There are 2 main types of case-control design: matched and unmatched.

Essentially, in an unmatched case-control design, a shared control group is selected for all cases at random given certain attributes. In a matched case-control design, controls are selected case-by-case based on specified characteristics. You should pick characteristics that have an effect on the usage of digital devices and services.

Commonly used matching factors include:

  • socio-economic status

However, think about other characteristics and attributes that might influence the use of your product, and the subsequent outcomes.

Confounding variables and biases

Confounding variables (variables other than the one you are interested in that may influence the results) and biases (errors that influence the sample selected and results observed) are important to consider when conducting any research. This is especially important in designs that are non-randomised.

  • selection bias can happen when participants are assigned without randomisation
  • attribution bias may occur when patients with unfavourable outcomes are less likely to attend follow-ups

Analysing your data

The analysis most commonly used in case-control studies is an odds ratio, which is the chance (odds) of the outcomes occurring in the case group versus the control group.

Example: Can telemedicine help with post-bariatric surgery care? A case-control design

In 2019, Wang and colleagues published a paper entitled Exploring the Effects of Telemedicine on Bariatric Surgery Follow-up: a Matched Case Control Study .

The study showed that people who go through bariatric surgery have better outcomes if they attend their follow-up appointments after surgery in comparison to those who do not. However, attending appointments can be challenging for people who live in remote areas. In Ontario, Canada, telemedicine suites were set up to enable healthcare provider-patient videoconferencing.

The researchers used a matched case-control study to investigate if telemedicine videoconferencing can support post-surgery appointment attendance rates in people who live further away from the hospital sites. They used the existing data from the bariatric surgery hospital programme to identify eligible patients.

All patients attending the bariatric surgery were offered telemedicine services. The cases were the participants who used telemedicine services; they were compared to those who did not (the controls).

Cases and controls were matched on various characteristics, specifically:

  • time since bariatric surgery
  • body mass index ( BMI )
  • travel distance from the hospital site

Researchers measured:

  • the percentage of appointments attended
  • rates of dropout
  • pre-and post-surgery weight and BMI
  • various physical and psychological outcomes

They also calculated rurality index to classify patients into urban, non-urban and rural areas. These variables were used to compare cases (those who used telemedicine) and controls (those who did not).

During the study period, they identified that 487 patients of 1,262 who received bariatric surgery used telemedicine services. Of those, 192 agreed to participate in the study.

They found that patients who used telemedicine did as well as patients who attended in person, both in terms of appointment attendance rates and in terms of physical and psychological outcomes.

Moreover, the researchers found that the cases (telemedicine users) came from more rural areas than the controls. The authors argued that this demonstrated that telemedicine can help overcome the known challenges for patients in more rural areas to attend appointments.

Randomising patients to telemedicine or withdrawing the telemedicine would be difficult, undesirable and possibly unethical. Case-control was a good alternative to assess the potential impact on patient outcomes in a service that is already up and running.

More information and resources

A 2003 study by Mann provides an accessible overview of observational research methods, including an explanation of biases and confounding variables.

On the website for Strengthening the Reporting of Observational Studies in Epidemiology ( STROBE ), there is a checklist of items that should be included in reports of case-control studies .

A 2016 study by Pearce offers considerations for the analysis of a matched case-control study.

Examples of case-control studies in digital health

In a 2020 study by Heuvel and others , researchers assessed a new digital health tool to monitor women at increased risk of preeclampsia at home. They investigated if the digital tool allows for fewer antenatal visits without compromising women’s safety, and whether it positively affects pregnancy outcomes. This study used a prospective case group compared to a retrospective control group.

In a 2019 study by Depp and others , the research team examined whether schizophrenia symptoms were associated with mobility (measured using GPS sensors). They compared participants with schizophrenia to healthy controls and they found that less mobility was associated with greater symptoms of schizophrenia.

Related content

Is this page useful.

  • Yes this page is useful
  • No this page is not useful

Help us improve GOV.UK

Don’t include personal or financial information like your National Insurance number or credit card details.

To help us improve GOV.UK, we’d like to know more about your visit today. We’ll send you a link to a feedback form. It will take only 2 minutes to fill in. Don’t worry we won’t send you spam or share your email address with anyone.

case control study is a type of

  • Free Case Studies
  • Business Essays

Write My Case Study

Buy Case Study

Case Study Help

  • Case Study For Sale
  • Case Study Service
  • Hire Writer

Case-Control Study: Definition, Types and Examples

During forming a group of cases, it is necessary to apply strict, objective criteria for the result. You should be sure of the homogeneity of the result because similar illness or effects may have different risk factors, for example not all diseases of the intestinal infections that are detected in the population under study only in the presence of diabetic syndrome can be selected for the study. It should be sought to use, if possible, “incident” cases (again diagnosed) than “prevalent” (already existing at the given time).During using “superior” cases, the effect of the disease on potential risk factors can lead to complications in the interpretation of data. For example, in a study on the impact of coffee consumption on the risk of peptic ulcer disease, “prevalent” cases (long suffering from ulcers and precautionary coffee drinkers) will differ in relation to the exposure from “incident” cases, in which the disease has occurred relatively recently and what else have not had time to change their attitude to drinking this soda.Observational studies proved, that case-control studies have less reliability than cohort studies.

This is not entirely true: a well-organized case-control study in a number of situations can provide much more reliable results than cohort studies. The main stages are:

  • Formation of a sample (cohort) from the general population, taking into account the features of inclusion and exclusion.
  • Collecting information on the prevalence of risk factors and illnesses.

What is a Case-Control Study?

The case-control study is one of the very important types of studies, that has a number of obvious benefits. First of all, this scheme of analytical research is excellent for rare diseases (co-study in such a situation, the population of the study may be excessively high). The case-control study allows you to get an answer quickly and, therefore,it is the method of choice during investigating flashes.

In the case-control study, you can study simultaneously (and quickly) a multitude of factors for studying one result. However, only one output can be studied.Problems, encountered in the case-control study vs. cohort study , are related to the fact that interest in the data on the impact of the factor may be inaccessible or inaccurate. Sometimes, it’s just not possible to choose a sufficient number of controls that satisfy the requirements set.

The choice of a scheme for analytical research depends, first of all, on specific tasks and main steps, but largely determined by the available resources and timing for it. Knowing the possibilities of different approaches, their advantages and disadvantages allow the epidemiologist to plan research optimally.In the case-control study, it is virtually impossible to identify the rare causes of the disease. In such cases, scanty data does not allow us to assess the validity of the differences in the incidence of risk factors in the comparison groups and, therefore, to draw conclusions about the presence or absence of a causal relationship.

Advantages and Disadvantages of a Case-Control Study, Types of Case-Control Studies

In the case-control study, the search for causal relationships goes in the direction of the investigation to the foreseeable cause.

Case-control study examples can only be retrospective, as it is conducted on the basis of archival data. Often, the source of information in the case-control studies is the history of the disease, which is in the archives of medical institutions, the memories of patients or their relatives in the context of an interview or by the results of the questionnaire.This retrospective study can be done as a preliminary study of the causal relationship between the predicted risk factor and the specific disease. In the future, this problem can be studied in cohort studies.Positive aspects of the case-control study are the possibility of conducting them regardless of the prevalence of the disease under study.

Relatively small expenditures of time, forces and means are needed to create a basic group of patients (even rarely encountered diseases), to pick up a control group for them, to question and make at least indicative conclusions. In the study of such diseases, you have to pick up a cohort of hundreds of thousands of people, to watch them for a long time. This would entail considerable time, material and moral costs.Case-control studies have a relatively short duration. The duration of the research depends directly on the productivity of the personnel involved in the study. In order to obtain conclusions, it is not necessary, as in the cohort study, to conduct observations for a period that exceeds the latent period of disease development.

There is a possibility to identify several risk factors for one disease simultaneously.The main disadvantage is the inability to quantify the risk of a disease (death) from an alleged cause.The case-control study is characterized by relatively small economic costs. This makes them attractive when the researcher is limited in funding. However, you should not forget that each study has its own indications and limitations.

Related posts:

  • Research Design: Definition, Types and How to Write
  • Quantitative Research Design: Definition, Methods and Types
  • Exploratory Research Design: Definition, Types and Ways to Implement
  • Case Study: Definition, How to Write, Format and Examples
  • What Is Cohort Study: Types, Study Design and Examples
  • Explanation of three types Examples
  • Descriptive Research Design: Definition, Methods and Examples

' src=

Quick Links

Privacy Policy

Terms and Conditions

Testimonials

Our Services

Case Study Writing Service

Case Studies For Sale

Our Company

Welcome to the world of case studies that can bring you high grades! Here, at ACaseStudy.com, we deliver professionally written papers, and the best grades for you from your professors are guaranteed!

[email protected] 804-506-0782 350 5th Ave, New York, NY 10118, USA

Acasestudy.com © 2007-2019 All rights reserved.

case control study is a type of

Hi! I'm Anna

Would you like to get a custom case study? How about receiving a customized one?

Haven't Found The Case Study You Want?

For Only $13.90/page

This paper is in the following e-collection/theme issue:

Published on 22.2.2024 in Vol 10 (2024)

Mediating Effect of Tobacco Dependence on the Association Between Maternal Smoking During Pregnancy and Chronic Obstructive Pulmonary Disease: Case-Control Study

Authors of this article:

Author Orcid Image

Original Paper

  • Jinxuan Li 1, 2, 3, 4, 5, 6, 7 , MD   ; 
  • Jianying Xu 8 , MD   ; 
  • Lan Yang 9 , MD   ; 
  • Yongjian Xu 10 , MD   ; 
  • Xiangyan Zhang 11 , MD   ; 
  • Chunxue Bai 12 , MD   ; 
  • Jian Kang 13 , MD   ; 
  • Pixin Ran 14 , MD   ; 
  • Huahao Shen 15 , MD   ; 
  • Fuqiang Wen 16 , MD   ; 
  • Kewu Huang 17 , MD   ; 
  • Wanzhen Yao 18 , MD   ; 
  • Tieying Sun 19, 20 , MD   ; 
  • Guangliang Shan 21 , MD   ; 
  • Ting Yang 2, 3, 4, 5, 22 , MD   ; 
  • Yingxiang Lin 17 , MD   ; 
  • Jianguo Zhu 20 , MD   ; 
  • Ruiying Wang 8 , MD   ; 
  • Zhihong Shi 9 , MD   ; 
  • Jianping Zhao 10 , MD   ; 
  • Xianwei Ye 11 , MD   ; 
  • Yuanlin Song 12 , MD   ; 
  • Qiuyue Wang 13 , MD   ; 
  • Gang Hou 2, 3, 4, 5, 22 , MD   ; 
  • Yumin Zhou 14 , MD   ; 
  • Wen Li 15 , MD   ; 
  • Liren Ding 15 , MD   ; 
  • Hao Wang 16 , MD   ; 
  • Yahong Chen 18 , MD   ; 
  • Yanfei Guo 19, 20 , MD   ; 
  • Fei Xiao 20 , MD   ; 
  • Yong Lu 17 , MD   ; 
  • Xiaoxia Peng 23, 24 , MD   ; 
  • Biao Zhang 21 , MD   ; 
  • Zuomin Wang 25 , MD   ; 
  • Hong Zhang 17 , MD   ; 
  • Xiaoning Bu 17 , MD   ; 
  • Xiaolei Zhang 2, 3, 4, 5, 22 , MD   ; 
  • Li An 17 , MD   ; 
  • Shu Zhang 17 , MD   ; 
  • Zhixin Cao 17 , MD   ; 
  • Qingyuan Zhan 2, 3, 4, 5, 22 , MD   ; 
  • Yuanhua Yang 17 , MD   ; 
  • Lirong Liang 26, 27 , MD   ; 
  • Bin Cao 2, 3, 4, 5, 22 , MD   ; 
  • Huaping Dai 2, 3, 4, 5, 22 , MD   ; 
  • Kian Fan Chung 28 , MD   ; 
  • Zhengming Chen 29 , PhD   ; 
  • Jiang He 30 , MD   ; 
  • Sinan Wu 2, 3, 4, 5, 31 , MD   ; 
  • Dan Xiao 2, 3, 4, 5, 6, 7 * , MD   ; 
  • Chen Wang 2, 3, 4, 5, 7, 22 * , MD   ; 
  • China Pulmonary Health Study Group 32

1 China-Japan Friendship School of Clinical Medicine, Capital Medical University, Beijing, China

2 National Center for Respiratory Medicine, Beijing, China

3 State Key Laboratory of Respiratory Health and Multimorbidity, Beijing, China

4 National Clinical Research Center for Respiratory Diseases, Beijing, China

5 Institute of Respiratory Medicine, Chinese Academy of Medical Sciences, Beijing, China

6 Department of Tobacco Control and Prevention of Respiratory Diseases, Center of Respiratory Medicine, China-Japan Friendship Hospital, Beijing, China

7 WHO Collaborating Centre for Tobacco Cessation and Respiratory Diseases Prevention, Beijing, China

8 Department of Pulmonary and Critical Care Medicine, Shanxi Bethune Hospital, Shanxi Academy of Medical Sciences, Taiyuan, China

9 Department of Pulmonary and Critical Care Medicine, First Affiliated Hospital of Xi’an Jiaotong University, Xi’an, China

10 Department of Pulmonary and Critical Care Medicine, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China

11 Department of Pulmonary and Critical Care Medicine, Guizhou Provincial People’s Hospital, Guiyang, China

12 Department of Pulmonary and Critical Care Medicine, Zhongshan Hospital, Fudan University, Shanghai, China

13 Department of Pulmonary and Critical Care Medicine, First Hospital of China Medical University, Shenyang, China

14 State Key Laboratory of Respiratory Disease, National Clinical Research Center for Respiratory Diseases, Guangzhou Institute of Respiratory Diseases, First Affiliated Hospital, Guangzhou Medical University, Guangzhou, China

15 Department of Pulmonary and Critical Care Medicine, Second Affiliated Hospital of Zhejiang University School of Medicine, Hangzhou, China

16 State Key Laboratory of Biotherapy of China and Department of Respiratory and Critical Care Medicine, West China Hospital of Sichuan University, Chengdu, China

17 Beijing Key Laboratory of Respiratory and Pulmonary Circulation Disorders, Department of Pulmonary and Critical Care Medicine, Beijing Chao-Yang Hospital, Capital Medical University, Beijing, China

18 Department of Pulmonary and Critical Care Medicine, Peking University Third Hospital, Beijing, China

19 Department of Respiratory and Critical Care Medicine, Beijing Hospital, Beijing, China

20 National Center of Gerontology, Beijing, China

21 Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences, School of Basic Medicine, Peking Union Medical College, Beijing, China

22 Department of Pulmonary and Critical Care Medicine, Center of Respiratory Medicine, China-Japan Friendship Hospital, Beijing, China

23 Center for Clinical Epidemiology and Evidence-Based Medicine, Beijing Children's Hospital, Capital Medical University, Beijing, China

24 National Center for Children’s Health, Beijing, China

25 Department of Stomatology, Beijing Chao-Yang Hospital, Capital Medical University, Beijing, China

26 Department of Epidemiology, Beijing Chao-Yang Hospital, Capital Medical University, Beijing, China

27 Beijing Institute of Respiratory Medicine, Beijing, China

28 National Heart and Lung Institute, Imperial College London and Royal Brompton and Harefield NHS Trust, London, United Kingdom

29 Clinical Trial Service Unit and Epidemiological Studies Unit, Nuffield Department of Population Health, University of Oxford, Oxford, United Kingdom

30 Department of Epidemiology, Tulane University School of Public Health and Tropical Medicine, New Orleans, LA, United States

31 Data and Project Management Unit, Center of Respiratory Medicine, China-Japan Friendship Hospital, Beijing, China

32 See Acknowledgements

*these authors contributed equally

Corresponding Author:

Dan Xiao, MD

Department of Tobacco Control and Prevention of Respiratory Diseases

Center of Respiratory Medicine

China-Japan Friendship Hospital

Yinghua Street

Beijing, 100029

Phone: 86 010 8420 5425

Email: [email protected]

Background: Maternal smoking during pregnancy (MSDP) is a known risk factor for offspring developing chronic obstructive pulmonary disease (COPD), but the underlying mechanism remains unclear.

Objective: This study aimed to explore whether the increased COPD risk associated with MSDP could be attributed to tobacco dependence (TD).

Methods: This case-control study used data from the nationwide cross-sectional China Pulmonary Health study, with controls matched for age, sex, and smoking status. TD was defined as smoking within 30 minutes of waking, and the severity of TD was assessed using the Fagerstrom Test for Nicotine Dependence. COPD was diagnosed when the ratio of forced expiratory volume in 1 second to forced vital capacity was <0.7 in a postbronchodilator pulmonary function test according to the 2017 Global Initiative for Chronic Obstructive Lung Disease criteria. Logistic regression was used to examine the correlation between MSDP and COPD, adjusting for age, sex, BMI, educational attainment, place of residence, ethnic background, occupation, childhood passive smoking, residential fine particulate matter, history of childhood pneumonia or bronchitis, average annual household income, and medical history (coronary heart disease, hypertension, and diabetes). Mediation analysis examined TD as a potential mediator in the link between MSDP and COPD risk. The significance of the indirect effect was assessed through 1000 iterations of the “bootstrap” method.

Results: The study included 5943 participants (2991 with COPD and 2952 controls). Mothers of the COPD group had higher pregnancy smoking rates (COPD: n=305, 10.20%; controls: n=211, 7.10%; P <.001). TD was more prevalent in the COPD group (COPD: n=582, 40.40%; controls: n=478, 33.90%; P <.001). After adjusting for covariates, MSDP had a significant effect on COPD (β=.097; P <.001). There was an association between MSDP and TD (β=.074; P <.001) as well as between TD and COPD (β=.048; P =.007). Mediation analysis of TD in the MSDP-COPD association showed significant direct and indirect effects (direct: β=.094; P <.001 and indirect: β=.004; P =.03). The indirect effect remains present in the smoking population (direct: β=.120; P <.001 and indirect: β=.002; P =.03).

Conclusions: This study highlighted the potential association between MSDP and the risk of COPD in offspring, revealing the mediating role of TD in this association. These findings contribute to a deeper understanding of the impact of prenatal tobacco exposure on lung health, laying the groundwork for the development of relevant prevention and treatment strategies.

Introduction

Chronic obstructive pulmonary disease (COPD) is a global health challenge, ranking among the leading causes of both morbidity and mortality worldwide [ 1 ]. It was estimated that COPD affected a substantial portion of the global population with a prevalence rate of 10.3% among individuals, and the number of individuals afflicted with COPD has been on the rise [ 2 ]. Notably, China accounts for nearly a quarter of global COPD cases, with a 67% increase in prevalence among individuals aged 40 years and older between 2012 and 2015 [ 3 ].

COPD is influenced by various factors, including genetics, lifestyle, environmental factors, and other influencing factors [ 2 - 5 ]. Smoking, recognized as the most significant cause of COPD, leads to its development in approximately half of all smokers [ 6 , 7 ]. The prenatal and perinatal phases are crucial for lung development, with maternal smoking linked to offspring susceptibility to various health issues [ 8 - 10 ]. A systematic review suggested that maternal smoking during pregnancy (MSDP) is associated with an increased risk of COPD [ 11 ]. MSDP may directly impact fetal development, influencing the formation of the central nervous system and respiratory system [ 12 ]. Given the vulnerability of embryos and fetuses to the external environment, maternal smoking may exert a more pronounced effect on fetal development [ 13 ].

Despite widespread recognition of the adverse effects of MSDP on offspring COPD [ 11 ], the underlying mechanisms remain incompletely elucidated. Tobacco dependence (TD) is a complex condition influenced by genetic and environmental factors, classified as a mental disorder according to the International Classification of Diseases, and TD should be recognized as a lethal noncommunicable disease [ 14 ]. Previous studies have identified TD as an independent risk factor for atherosclerosis [ 15 ]. Notably, TD is prevalent among smokers, often associated with heightened smoking intensity and lower cessation rates [ 16 ], and approximately 40% of smokers experience impaired lung function and develop COPD [ 17 ]. A meta-analysis indicates an elevated risk of smoking and TD in offspring associated with MSDP [ 18 ]. Given these associations, our hypothesis posits that TD plays a pivotal mediating role in the association between MSDP and COPD.

Therefore, this study aims to investigate whether TD serves as a mediator in the association between MSDP and the risk of COPD in offspring. To achieve this, we used data from the national cross-sectional China Pulmonary Health (CPH) study [ 3 ] and conducted a case-control study to explore the role of TD in the association between MSDP and COPD. The findings will contribute to the development of targeted interventions focused on TD.

Participants and Study Design

The study population was drawn from the national cross-sectional CPH study [ 3 , 19 , 20 ], the largest study assessing the burden of COPD in China. The CPH study, encompassing a vast nationwide cross-sectional examination, encompassed 57,779 Chinese adults aged 20 years and older across 10 provinces in China. This extensive study incorporated 80 urban and 80 rural areas, using a multistage stratified cluster sampling approach, The details of recruitment for the CPH can be found elsewhere [ 3 ]. Trained health workers in local community health centers administered standardized questionnaire surveys to all participants, capturing essential information on sociodemographic status, medical history, lifestyle, and more. The questionnaire information involved in this study can be obtained in Table S1 in Multimedia Appendix 1 .

In this study, a case-control design was used, and the workflow is illustrated in Figure 1 . Inclusion criteria for the case group were as follows: (1) provision of signed informed consent, (2) age of 18 years or older, and (3) availability of complete questionnaire information. Individuals with other conditions such as asthma, tuberculosis, lung cancer, lung inflammation, and missing information were excluded. Control participants were matched for age, sex, and smoking from cohort. The matching tolerance can be found in Table S2 in Multimedia Appendix 1 . Ultimately, a total of 2991 patients in the COPD group and 2952 controls were included in this study.

case control study is a type of

Definition of MSDP and Evaluation of TD

MSDP was self-reported through a questionnaire, with all participants asked to respond to the question, “Did your mother smoke during pregnancy? (yes or no).” Upon awakening, smoking within the first 30 minutes was defined as indicative of TD, based on prior study [ 21 ]. The severity of TD was assessed using the Fagerstrom Test for Nicotine Dependence (FTND) [ 22 ]. FTND is a standardized tool for evaluating the degree of physical nicotine addiction. It consists of six items: (1) How soon after waking do you smoke your first cigarette (within 5 minutes, 6 to 30 minutes, 31 to 60 minutes, and after 60 minutes)? (2) Do you find it difficult to refrain from smoking in places where it is prohibited (yes or no)? (3) Which cigarette would you find most difficult to give up (the first one in the morning or any other)? (4) How many cigarettes do you smoke per day (10 or less, 11 to 20, 21 to 30, and 31 or more)? (5) Do you smoke more frequently during the first hour after waking compared to the rest of the day (yes or no)? and (6) Do you smoke when you are so ill that you are confined to bed most of the day (yes or no)? The FTND generates a total score ranging from 0 to 10, with scores calculated as the sum of individual items. Yes or no items are scored as 0 or 1, and multiple-choice items are scored as 0 to 3. The questionnaire provided a comprehensive assessment of cigarette consumption, compulsion, and dependence. A higher total FTND score indicates a greater degree of TD. All participants reported their smoking status (yes or no). Smoking was defined as having consumed 100 cigarettes in one’s lifetime, encompassing both current and former smokers.

To address potential recall bias in self-reported MSDP, we implemented rigorous quality control measures. Our questionnaire design emphasized clarity and comprehensibility, and interviewers received detailed training to ensure precise definitions of smoking behavior. Memory aids were provided to aid accurate recall, and privacy protection measures were emphasized to minimize bias.

Definition of COPD

All study participants underwent postbronchodilator pulmonary function tests, including measurements of forced expiratory volume in 1 second (FEV 1 ) and forced vital capacity (FVC) and FEV 1 /FVC ratio. COPD was diagnosed in accordance with the criteria set forth by the 2017 Global Initiative for Chronic Obstructive Lung Disease [ 23 ] if the participant’s FEV 1 /FVC ratio was <0.7.

This study incorporated several covariates, including age (years), sex (male or female), BMI (calculated as weight in kilograms divided by the square of height in meters), educational attainment (primary and below, junior middle school, senior high school, or bachelor and above), place of residence (urban or rural), ethnic background (Han Chinese or other), occupation (farmer, worker, or other), childhood passive smoking, residential fine particulate matter (PM 2.5 ) exposure, history of childhood pneumonia or bronchitis (yes or no), average annual household income, and medical history (coronary heart disease, hypertension, and diabetes).

Statistical Analysis

For normally distributed continuous variables, descriptive statistics are presented as mean (SD). Categorical variables are expressed as numbers and percentages. The Kolmogorov-Smirnov test was used to assess the normality of continuous variables. Comparisons of continuous variables were conducted using the Mann-Whitney U test, while categorical variables were compared using the chi-square test.

Mediation analysis, as outlined by Baron and Kenny [ 24 ], was used to investigate the potential mediating role of TD in the association between MSDP and the risk of COPD. The association between MSDP and TD was modeled using logistic regression, whereas the association with FTND scores was assessed using linear regression. The total effect of the initial variable on the outcome was defined as the sum of its direct and indirect effects. The indirect effect was further defined as the product of the initial variable’s impact on the intermediate variable and the intermediate variable’s impact on the outcome, with adjustments made for the initial variable. The statistical significance of the indirect effect was assessed through 1000 repetitions of a bootstrap procedure. Regarding the sensitivity analysis, we conducted a similar analysis by defining TD as a total FTND score ≥4. The sample size for this study was determined using the Monte Carlo method [ 25 ]. The details of the sample size calculation process are provided in Table S3 in Multimedia Appendix 1 . All statistical analyses were conducted using R (version 4.2.1; R Foundation for Statistical Computing), with the use of packages such as bruceR , interactions , and mediation to estimate mediating effects. A significance level of P =.05 (2-tailed) was considered statistically significant.

Ethical Considerations

The study received approval from the ethics review committee of Beijing Capital Medical University (11-KE-42) and other collaborating institutes. Written informed consent was obtained from all study participants in accordance with the principles outlined in the Declaration of Helsinki. To ensure privacy and confidentiality, the data used in this study underwent anonymization and deidentification processes.

Participants

A total of 5943 participants were enrolled in this study, comprising 2991 individuals with COPD and 2952 matched controls. The demographic characteristics of all participants are elaborated in Table 1 . In the control group, the mean age was 59.17 (SD 11.44) years, while in the COPD group, it was 59.50 (SD 11.66) years. Among controls, there were 1783 (60.40%) male participants, and in the COPD group, 1816 (60.70%) were male participants. Smoking was observed in 1442 (48.20%) individuals within the COPD group and 1412 (47.80%) individuals in the control group. For matching purposes, no statistically significant differences were detected between the 2 groups in terms of participants’ age, sex, and smoking. However, a higher proportion of mothers in the COPD group reported MSDP compared to the control group (COPD: n=305, 10.20% and controls: n=211, 7.10%; P <.001).

a COPD: chronic obstructive pulmonary disease.

b P value for 2-tailed t test.

c P value for chi-square test.

d MSDP: maternal smoking during pregnancy.

e GOLD: Global Initiative for Chronic Obstructive Lung Disease.

f Not available.

g FEV 1 : forced expiratory volume in 1 second.

h FVC: forced vital capacity.

i AAHI (average annual household income) is measured in 10,000 Yuan (a currency exchange rate of 1 Chinese Yuan (CNY)=US $0.16147 is applicable).

j PM 2.5 : fine particulate matter.

Differences of TD in Smokers Between the 2 Groups

As depicted in Table 2 , the prevalence of TD was notably higher in the COPD group in comparison to the control group (COPD: n=582, 40.40%; controls: n=478, 33.90%; P =.001). Patients with COPD, when contrasted with the control group, exhibited a shorter time interval between waking up and the desire to smoke their first cigarette (Q1: P =.004) and encountered greater difficulty refraining from smoking in public places (Q2: P =.006). Among the COPD group, 19.40% (n=280) reported a tendency to smoke more frequently during the early hours after waking, whereas in the control group, this behavior was observed in 16.50% (n=233; P =.048). Additionally, the proportion of individuals smoking when unwell was higher in the COPD group than in the control group (Q6: P <.001). Q3 showed no statistically significant difference between the 2 groups, but with a P value <0.1, suggesting the potential presence of marginal effects.

b FTND: Fagerstrom Test for Nicotine Dependence.

d Q1: How soon after you wake up do you smoke your first cigarette?

e Q2: Do you find it difficult to refrain from smoking in places where it is forbidden?

f Q3: Which cigarette would you find most difficult to give up?

g Q4: How many cigarettes do you smoke per day?

h Q5: Do you smoke more frequently during the first hour after waking compared to the rest of the day?

i Q6: Do you smoke when you are so ill that you are confined to bed most of the day?

j TD: tobacco dependence.

Direct and Indirect Effects of MSDP on COPD in All Participants

The results of the mediation analysis for all participants are presented in Table 3 . After adjusting for age, sex, BMI, educational attainment, place of residence, ethnicity, occupation, childhood passive smoking, residential PM 2.5 exposure, history of childhood pneumonia or bronchitis, average annual household income, and medical history (coronary heart disease, hypertension, and diabetes), the total effect of MSDP on COPD was statistically significant (β=.097; P <.001). Furthermore, the direct effect of MSDP on COPD remained significant after adjusting for TD and other covariates (β=.094; P <.001). The indirect effect of TD was also found to be statistically significant (β=.004; P =.03). The sensitivity analysis revealed similar results (Table S4 in Multimedia Appendix 1 ).

a TD: tobacco dependence.

b MSDP: maternal smoking during pregnancy.

c COPD: chronic obstructive pulmonary disease.

d All analyses adjusted for age, sex, BMI, educational attainment, place of residence, ethnicity, occupation, childhood passive smoking, residential PM 2.5 exposure, history of childhood pneumonia or bronchitis, average annual household income, and medical history (coronary heart disease, hypertension, and diabetes).

e Total effect of independent variables on dependent variables.

f Coefficients of independent variables on mediating variables after adjustment for covariates.

g Coefficients of mediating variables on dependent variables after adjusted for covariates and IVs.

h Indirect effect of independent variables on dependent variables.

i Direct effect of independent variables on dependent variables.

Direct and Indirect Effects of MSDP on COPD in Smokers

The results of the mediation analysis for smokers are presented in Table 4 . In this segment, all analyses were adjusted for age, sex, BMI, educational attainment, place of residence, ethnicity, occupation, childhood passive smoking, residential PM 2.5 exposure, history of childhood pneumonia or bronchitis, average annual household income, and medical history (coronary heart disease, hypertension, and diabetes). Among smokers, both the total effect and direct effect (β) of MSDP on COPD were .123 and .120, respectively. Additionally, the indirect effect of TD was found to be statistically significant (β=.003; P =.03). The sensitivity analysis revealed similar results (Table S5 in Multimedia Appendix 1 ).

e Coefficients of independent variables on dependent variables after correction for covariates.

g Coefficients of mediating variables on dependent variables after correction for covariates and IVs.

Principal Findings

The aim of this study was to investigate whether TD plays a role in the association between MSDP and COPD in offspring. A case-control study was conducted within the Chinese population. It was the first study to observe that MSDP has both a direct impact on COPD and an indirect influence mediated through TD. This mediation effect persists among the smoking population, and sensitivity analysis confirms the stability of these results.

It is well-established that smoking represents the primary risk factor for COPD, and an increasing body of evidence suggests that early life tobacco exposure plays a role in the onset and progression of COPD [ 26 ]. MSDP, as highlighted by Jaakkola et al [ 27 ], exerts substantial adverse effects on conditions such as asthma, chronic bronchitis, and chronic respiratory symptoms. Nevertheless, limited research has delved into the impact of prenatal tobacco smoke exposure on the development of COPD in later adulthood. This study, however, reveals an association between MSDP and the risk of COPD in offspring. Early life tobacco exposure has enduring consequences on lung function in offspring, with suboptimal intrauterine conditions leading to disturbances in lung development. As a result, affected individuals exhibit diminished lung function at birth, which often persists throughout their lifetime. This, in turn, elevates the risk of subsequent COPD, particularly in childhood wheeze disorders and genetically susceptible individuals [ 28 ]. To curb the current upward trend in COPD incidence, it is imperative to mitigate lung development risks by enhancing antenatal and neonatal care and reducing exposure to environmental pollutants, including passive tobacco smoke, both prenatally and postnatally.

In this study, MSDP emerges as a risk factor for TD in offspring, a finding consistent with prior research. Previous studies have established that individuals with a history of MSDP face an elevated risk of TD in adolescence and adulthood [ 18 ]. The precise mechanism underlying the connection between MSDP and offspring TD remains elusive. One plausible mechanism could involve the neurotoxic effects of harmful compounds present in tobacco smoke, which can readily traverse the placenta, potentially leading to smoking behavior and dependence in offspring [ 29 ]. Another conceivable mechanism is that MSDP is linked to preterm birth [ 30 ] and low birth weight [ 31 ], both of which could predispose offspring to various health challenges. Furthermore, it is essential for health care professionals to recognize that TD may be linked to early life tobacco exposure. Therefore, children exposed to MSDP should receive comprehensive education on topics such as diet, exercise, and smoking avoidance to mitigate their risk of developing COPD in adulthood.

This study revealed a direct association between TD and COPD. Previous research has identified smoking as one of the primary risk factors for COPD, causing inflammation and damage to the airways and lung tissue, ultimately leading to the development of COPD [ 32 ]. Recent studies have further suggested the correlation between TD and COPD [ 33 ]. Additionally, other studies suggest that TD may exacerbate the severity and progression of COPD [ 34 ]. Among individuals diagnosed with COPD, the presence of TD makes quitting smoking more challenging [ 35 ]. Thus, this study underscores the critical importance of addressing TD in the management and mitigation of COPD.

To our knowledge, we have discovered for the first time in this study that TD plays a mediating role in the association between MSDP and COPD. However, the specific reasons underlying this mediating mechanism remain unclear. There are several possible mechanisms to consider. First, genetic factors may be implicated, as MSDP has the potential to induce persistent epigenetic changes until adolescence [ 36 ]. These genetic alterations may contribute to the development of TD [ 37 ]. Consequently, TD can exacerbate challenges in smoking cessation and the progression of COPD [ 38 ]. Second, environmental factors may also be involved. Mothers who smoke during pregnancy may continue smoking during infancy, creating an early growth environment that promotes the risk of TD in offspring during lung development [ 39 , 40 ], subsequently leading to COPD. Our results highlight that TD serves as a partial mediator in the connection between MSDP and offspring COPD. Consequently, it becomes imperative to consider TD when addressing MSDP and smoking cessation efforts in offspring. Prior studies have advocated smoking cessation strategies for TD smokers, incorporating the use of medications such as varenicline, nicotine replacement therapy products, and bupropion [ 41 ]. Hence, clinicians should be cognizant of the necessity for appropriate smoking cessation interventions among individuals displaying signs of TD, ultimately contributing to the reduction of COPD risk.

Limitations

This study bears significant implications as it substantiates the partial mediating role of TD in the link between MSDP and COPD, underlining the importance of developing interventions targeting TD to mitigate the incidence of COPD in offspring. Nonetheless, it is essential to acknowledge the study’s limitations. First, the data concerning MSDP relied on questionnaire responses, introducing a potential source of recall bias. Due to the presence of recall bias, we cannot entirely rule out the possibility of uncertainty or erroneous memory in participants when reporting maternal smoking behavior during pregnancy. To enhance the accuracy of capturing MSDP information, future studies may consider using prospective cohort designs. Second, the absence of genetic data represents a significant limitation in this study. Despite the inclusion of various participant covariates, the intricate role of genetic factors in influencing COPD was not considered due to data constraints. Future research endeavors could overcome this limitation by integrating comprehensive genetic analyses into the study design. Finally, our classification of nonsmokers as nonnicotine-dependent during the overall analysis might introduce some bias. To fortify the robustness of our findings, we conducted a parallel subgroup analysis among the smoking population that consistently demonstrated the robustness of our results.

Conclusions

MSDP exerts an adverse influence on the risk of COPD in offspring, and TD plays a partially mediating role in this association. These findings underscore the potential for clinicians to mitigate the impact of MSDP on COPD in offspring by addressing TD. Consequently, early intervention strategies aimed at reducing TD become imperative in the endeavor to mitigate the risk of COPD.

Acknowledgments

The authors thank the participants of the China Pulmonary Health study. For continuous support, assistance, and cooperation, the authors thank Xinran Zhang and Xiaoying Gu (China-Japan Friendship Hospital); Ping Li, Mengyu Cheng, Wen Han, Hu Liu, Pengfei Wang, Jing Li, and Jing Wu (Shanxi Bethune Hospital Shanxi Academy of Medical Sciences); Yang Li and Ting Liu (First Affiliated Hospital of Xi’an Jiaotong University); Liekou Ma (Fufeng People’s Hospital); Luoping Yuan (Luonan Hospital); Jianbo Liu (Jingyang Hospital); Hubin Xi (Shanyang People’s Hospital); Baoping Wu (Changning Township Hospital of Wugong); Zhifang Liu (Meixian People’s Hospital); Cheng Zhang, Hong Yu, Weijia Liu, Ruiming Wu, Li Zhao, Yankun Jin, Lu Zhang, Mengning Zhen, Ping Lu, and Ling Li (Guizhou Provincial People’s Hospital); Bin Wu and Weimin Yao (Affiliated Hospital of Guangdong Medical University); Yanjie Yang, Li Manning, Qi Chen, Ying Gong, Mingfei Zhang, and Sulan Wei (Shanghai Zhongshan Hospital); Zhihua Chen, Gang Huang, Niya Zhou, Bin Shen, Wen Hua, Bin Zhang, Youlian Yu, and Juan Xiong (Second Affiliated Hospital of Zhejiang University); Ting Yang, Yongchun Shen, Diandian Li, Hongyu Long, Zenglin Liao, Xiaoou Li, Yanqiu Wu, Xiang Tong, Xiaying Peng, Bo Wang, Zhixin Qiu, Jian Luo, Lanlan Zhang, Shuang Zhao, Xingyu Xiong, Yinyin Yang, Yalun Li, Yanqqi He, Faming Jiang, Ting Wang, Jiajia Dong, Jing An, Linwei Li, Lian Liu, and Yonggang Zhang (West China Hospital); Liuqun Jia (First Affiliated Hospital of Zhengzhou University); Caishuang Pang (Chongqing Cancer Hospital); Qianjing Hu (People’s Hospital of Yubei District of Chonqing); Shujin Guo and Xiaying Peng (Sichuan Province People’s Hospital); Min Li (First Affiliated Hospital of Kunming Medical University); Lingli Guo (North China University of Science and Technology); Xue Zhang (Luoyang Orthopedic Hospital of Henan Province); Wen Du (West China Teaching Hospital); Yinyin Yang (Chengdu Second People’s Hospital); Lin Li (Mianyang Central Hospital); Jingyu Quan, Baosen Pang, Min Zhu, Xiaohong Chang, Jun Zhang, Baomei Wu, Ping Xin, Xiuxia Huang, Zhiyuan An, Shuilian Chu, Qiuyun Liu, Yanrui Jia, Jie Xia, Ying Cui, Jing Zhao, Chunyan Zhang, Jingyu Yang, and Xu Wu (Beijing Institute of Respiratory Medicine and Beijing Chao-Yang Hospital); Ruiyan Lin, Jie Song, Zhenyang Xu, and Xiaohui Li (Beijing Luhe Hospital); Rui Wu, Yanling Ding, Ming Lu, Jing Zhang, Lina Sun, Chengcheng Liao, Yun Sun, Yixuan Liao, Fan Lin, Yu Bai, and Meng Wang (Peking University Third Hospital); Lei Wang and Rong Gao (Beixiaguan Primary Care Center); Lingbo Sun and Xiaoliang Xie (Peking University Health Science Center Primary Care Center); and Pengjun Zhang, Hongsheng Zhang, Di Chai, and Xiaomeng Li (Beijing Hospital). The study was funded by the Chinese Academy of Medical Sciences (CAMS) Initiative for Innovative Medicine (CAMS 2021-I2M-1-010) and the Special Research Foundation for Public Welfare of Health, Ministry of Health of China (grant 201002008).

Data Availability

The data sets generated and analyzed during this study are available from the corresponding author on reasonable request.

Authors' Contributions

JL, DX, KH, TY, JX, LY, J Zhao, Xiangyan Zhang, CB, JK, PR, HS, FW, YC, TS, GS, Y Lin, GH, SW, J Zhu, JH, and CW conceived and designed the study. CW supervised the study. JL did the statistical analysis. All authors contributed to data collection and interpretation. JL, DX, and CW drafted the report. All authors revised the report and approved the final version before submission.

Conflicts of Interest

None declared.

Supplementary tables.

  • Varmaghani M, Dehghani M, Heidari E, Sharifi F, Moghaddam SS, Farzadfar F. Global prevalence of chronic obstructive pulmonary disease: systematic review and meta-analysis. East Mediterr Health J. 2019;25 (1):47-57. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Adeloye D, Song P, Zhu Y, Campbell H, Sheikh A, Rudan I, et al. NIHR RESPIRE Global Respiratory Health Unit. Global, regional, and national prevalence of, and risk factors for, chronic obstructive pulmonary disease (COPD) in 2019: a systematic review and modelling analysis. Lancet Respir Med. 2022;10 (5):447-458. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Wang C, Xu J, Yang L, Xu Y, Zhang X, Bai C, et al. Prevalence and risk factors of chronic obstructive pulmonary disease in China (the China Pulmonary Health [CPH] study): a national cross-sectional study. Lancet. 2018;391 (10131):1706-1717. [ CrossRef ] [ Medline ]
  • Silverman EK. Genetics of COPD. Annu Rev Physiol. 2020;82:413-431. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Niu Y, Zhou Y, Chen R, Yin P, Meng X, Wang W, et al. Long-term exposure to ozone and cardiovascular mortality in China: a nationwide cohort study. Lancet Planet Health. 2022;6 (6):e496-e503. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Wheaton AG, Liu Y, Croft JB, VanFrank B, Croxton TL, Punturieri A, et al. Chronic obstructive pulmonary disease and smoking status—United States, 2017. MMWR Morb Mortal Wkly Rep. 2019;68 (24):533-538. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Sin DD. The importance of early chronic obstructive pulmonary disease: a lecture from 2022 Asian Pacific Society of respirology. Tuberc Respir Dis (Seoul). 2023;86 (2):71-81. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Chu X, Ye J, Wen Y, Li P, Cheng B, Cheng S, et al. Maternal smoking during pregnancy and risks to depression and anxiety in offspring: an observational study and genome-wide gene-environment interaction analysis in UK biobank cohort. J Psychiatr Res. 2021;140:149-158. [ CrossRef ] [ Medline ]
  • Liu L, Cheng S, Wen Y, Jia Y, Cheng B, Meng P, et al. Maternal smoking around birth may lower the protective effects of breastfeeding on anxiety, depression and neuroticism in adult offspring: a UK biobank study. Eur Arch Psychiatry Clin Neurosci. 2023;273 (2):481-492. [ CrossRef ] [ Medline ]
  • Ayubi E, Safiri S, Mansori K. Association between maternal smoking during pregnancy and risk of bone fractures in offspring: a systematic review and meta-analysis. Clin Exp Pediatr. 2021;64 (3):96-102. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Duan P, Wang Y, Lin R, Zeng Y, Chen C, Yang L, et al. Impact of early life exposures on COPD in adulthood: a systematic review and meta-analysis. Respirology. 2021;26 (12):1131-1151. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Hajdusianek W, Żórawik A, Waliszewska-Prosół M, Poręba R, Gać P. Tobacco and nervous system development and function—new findings 2015-2020. Brain Sci. 2021;11 (6):797. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Gómez-Roig MD, Pascal R, Cahuana MJ, García-Algar O, Sebastiani G, Andreu-Fernández V, et al. Environmental exposure during pregnancy: influence on prenatal development and early life: a comprehensive review. Fetal Diagn Ther. 2021;48 (4):245-257. [ CrossRef ] [ Medline ]
  • Xiao D, Wang C. Tobacco dependence should be recognised as a lethal non-communicable disease. BMJ. 2019;365:l2204. [ CrossRef ] [ Medline ]
  • Zhu J, Nelson K, Toth J, Muscat JE. Nicotine dependence as an independent risk factor for atherosclerosis in the National Lung Screening Trial. BMC Public Health. 2019;19 (1):103. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Hsieh MT, Tseng PT, Wu YC, Tu YK, Wu HC, Hsu CW, et al. Effects of different pharmacologic smoking cessation treatments on body weight changes and success rates in patients with nicotine dependence: a network meta-analysis. Obes Rev. 2019;20 (6):895-905. [ CrossRef ] [ Medline ]
  • Pérez-Rubio G, Córdoba-Lanús E, Cupertino P, Cartujano-Barrera F, Campos MA, Falfán-Valencia R. Role of genetic susceptibility in nicotine addiction and chronic obstructive pulmonary disease. Rev Invest Clin. 2019;71 (1):36-54. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Duko B, Pereira G, Tait RJ, Nyadanu SD, Betts K, Alati R. Prenatal tobacco exposure and the risk of tobacco smoking and dependence in offspring: a systematic review and meta-analysis. Drug Alcohol Depend. 2021;227:108993. [ CrossRef ] [ Medline ]
  • Huang K, Yang T, Xu J, Yang L, Zhao J, Zhang X, et al. Prevalence, risk factors, and management of asthma in China: a national cross-sectional study. Lancet. 2019;394 (10196):407-418. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Niu Y, Yang T, Gu X, Chen R, Meng X, Xu J, et al. Long-term ozone exposure and small airway dysfunction: the China Pulmonary Health (CPH) study. Am J Respir Crit Care Med. 2022;205 (4):450-458. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Liu Z, Li YH, Cui ZY, Li L, Nie XQ, Yu CD, et al. Prevalence of tobacco dependence and associated factors in China: findings from nationwide China Health Literacy Survey during 2018-19. Lancet Reg Health West Pac. 2022;24:100464. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Heatherton TF, Kozlowski LT, Frecker RC, Fagerström KO. The Fagerström Test for Nicotine Dependence: a revision of the Fagerström Tolerance Questionnaire. Br J Addict. 1991;86 (9):1119-1127. [ CrossRef ] [ Medline ]
  • Vogelmeier CF, Criner GJ, Martinez FJ, Anzueto A, Barnes PJ, Bourbeau J, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive lung disease 2017 report. GOLD executive summary. Am J Respir Crit Care Med. 2017;195 (5):557-582. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986;51 (6):1173-1182. [ CrossRef ] [ Medline ]
  • Schoemann AM, Boulton AJ, Short SD. Determining power and sample size for simple and complex mediation models. Soc Psychol Pers Sci. 2017;8 (4):379-386. [ CrossRef ]
  • Savran O, Ulrik CS. Early life insults as determinants of chronic obstructive pulmonary disease in adult life. Int J Chron Obstruct Pulmon Dis. 2018;13:683-693. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Jaakkola JJK, Kosheleva AA, Katsnelson BA, Kuzmin SV, Privalova LI, Spengler JD. Prenatal and postnatal tobacco smoke exposure and respiratory health in Russian children. Respir Res. 2006;7 (1):48. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Stocks J, Sonnappa S. Early life influences on the development of chronic obstructive pulmonary disease. Ther Adv Respir Dis. 2013;7 (3):161-173. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Hyman SE. Addiction: a disease of learning and memory. Am J Psychiatry. 2005;162 (8):1414-1422. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Soneji S, Beltrán-Sánchez H. Association of maternal cigarette smoking and smoking cessation with preterm birth. JAMA Netw Open. 2019;2 (4):e192514. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Di HK, Gan Y, Lu K, Wang C, Zhu Y, Meng X, et al. Maternal smoking status during pregnancy and low birth weight in offspring: systematic review and meta-analysis of 55 cohort studies published from 1986 to 2020. World J Pediatr. 2022;18 (3):176-185. [ CrossRef ] [ Medline ]
  • Hikichi M, Mizumura K, Maruoka S, Gon Y. Pathogenesis of chronic obstructive pulmonary disease (COPD) induced by cigarette smoke. J Thorac Dis. 2019;11 (Suppl 17):S2129-S2140. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Galiatsatos P, Kaplan B, Lansey DG, Ellison-Barnes A. Tobacco use and tobacco dependence management. Clin Chest Med. 2023;44 (3):479-488. [ CrossRef ] [ Medline ]
  • Tashkin DP. Smoking cessation in COPD: confronting the challenge. Intern Emerg Med. 2021;16 (3):545-547. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Montes de Oca M, Laucho-Contreras ME. Smoking cessation and vaccination. Eur Respir Rev. 2023;32 (167):220187. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Rauschert S, Melton PE, Burdge G, Craig JM, Godfrey KM, Holbrook JD, et al. Maternal smoking during pregnancy induces persistent epigenetic changes into adolescence, independent of postnatal smoke exposure and is associated with cardiometabolic risk. Front Genet. 2019;10:770. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Hutchison KE, Allen DL, Filbey FM, Jepson C, Lerman C, Benowitz NL, et al. CHRNA4 and tobacco dependence: from gene regulation to treatment outcome. Arch Gen Psychiatry. 2007;64 (9):1078-1086. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Falomir-Pichastor JM, Blondé J, Desrichard O, Felder M, Riedo G, Folly L. Tobacco dependence and smoking cessation: the mediating role of smoker and ex-smoker self-concepts. Addict Behav. 2020;102:106200. [ CrossRef ] [ Medline ]
  • Huggett SB, Keyes M, Iacono WG, McGue M, Corley RP, Hewitt JK, et al. Age of initiation and transition times to tobacco dependence: early onset and rapid escalated use increase risk for dependence severity. Drug Alcohol Depend. 2019;202:104-110. [ CrossRef ] [ Medline ]
  • Rydell M, Cnattingius S, Granath F, Magnusson C, Galanti MR. Prenatal exposure to tobacco and future nicotine dependence: population-based cohort study. Br J Psychiatry. 2012;200 (3):202-209. [ FREE Full text ] [ CrossRef ] [ Medline ]
  • Rigotti NA, Kruse GR, Livingstone-Banks J, Hartmann-Boyce J. Treatment of tobacco smoking: a review. JAMA. 2022;327 (6):566-577. [ CrossRef ] [ Medline ]

Abbreviations

Edited by A Mavragani, T Sanchez; submitted 28.09.23; peer-reviewed by X Qi, K Zhang, Y Yang, H Guo, Y Fan; comments to author 09.11.23; revised version received 29.11.23; accepted 07.01.24; published 22.02.24.

©Jinxuan Li, Jianying Xu, Lan Yang, Yongjian Xu, Xiangyan Zhang, Chunxue Bai, Jian Kang, Pixin Ran, Huahao Shen, Fuqiang Wen, Kewu Huang, Wanzhen Yao, Tieying Sun, Guangliang Shan, Ting Yang, Yingxiang Lin, Jianguo Zhu, Ruiying Wang, Zhihong Shi, Jianping Zhao, Xianwei Ye, Yuanlin Song, Qiuyue Wang, Gang Hou, Yumin Zhou, Wen Li, Liren Ding, Hao Wang, Yahong Chen, Yanfei Guo, Fei Xiao, Yong Lu, Xiaoxia Peng, Biao Zhang, Zuomin Wang, Hong Zhang, Xiaoning Bu, Xiaolei Zhang, Li An, Shu Zhang, Zhixin Cao, Qingyuan Zhan, Yuanhua Yang, Lirong Liang, Bin Cao, Huaping Dai, Kian Fan Chung, Zhengming Chen, Jiang He, Sinan Wu, Dan Xiao, Chen Wang, China Pulmonary Health Study Group. Originally published in JMIR Public Health and Surveillance (https://publichealth.jmir.org), 22.02.2024.

This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Public Health and Surveillance, is properly cited. The complete bibliographic information, a link to the original publication on https://publichealth.jmir.org, as well as this copyright and license information must be included.

Association of Obesity With Type 2 Diabetes Mellitus: A Hospital-Based Unmatched Case-Control Study

Affiliations.

  • 1 Medicine, Ayub Medical College, Abbottabad, PAK.
  • 2 Community Medicine, Ayub Medical College, Abbottabad, PAK.
  • PMID: 38384596
  • PMCID: PMC10880576
  • DOI: 10.7759/cureus.52728

Background The prevalence of type 2 diabetes mellitus (T2DM) and obesity is alarmingly increasing with the accessibility of the modern lifestyle. This study aimed to assess the association of obesity with T2DM among the patients visiting the Medicine Department of Ayub Teaching Hospital, Abbottabad, Pakistan. Method This hospital-based, unmatched case-control study was conducted from March 2022 to September 2022. A total of 200 patients (age ≥ 18) (100 cases and 100 controls) were recruited. Those patients with a history of T2DM were selected as cases, and those without diabetes were selected as controls after taking informed written consent. Patients with BMI ≥ 25 were considered obese. Data were collected through a non-probability convenience sampling technique using a self-structured non-validated questionnaire. Data were organized and analyzed through IBM SPSS Statistics for Windows, version 26.0 (IBM Corp., Armonk, NY). Results We found a significant positive association of obesity with T2DM with a crude odds ratio of 3.6 (95% CI: 2.0-6.6), a p-value of 0.000, and an adjusted odd ratio of 3.7 (95% CI: 1.9 - 7.1), with a p-value of 0.004 (adjusted for potential confounders, including gender, age group, stress, and status of physical activeness) using a logistic regression model. Conclusion It is concluded that obesity is strongly associated with developing T2DM and lack of physical activity, people over 45 years, and males with obesity have a higher chance of developing T2DM.

Keywords: body mass index; case-control study; diabetes mellitus; obesity; type 2 diabetes mellitus.

Copyright © 2024, Ali et al.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • View all journals
  • My Account Login
  • Explore content
  • About the journal
  • Publish with us
  • Sign up for alerts
  • Open access
  • Published: 14 February 2024

The social anatomy of climate change denial in the United States

  • Dimitrios Gounaridis 1 &
  • Joshua P. Newell 1  

Scientific Reports volume  14 , Article number:  2097 ( 2024 ) Cite this article

19k Accesses

574 Altmetric

Metrics details

  • Climate-change policy
  • Environmental social sciences

Using data from Twitter (now X), this study deploys artificial intelligence (AI) and network analysis to map and profile climate change denialism across the United States. We estimate that 14.8% of Americans do not believe in climate change. This denialism is highest in the central and southern U.S. However, it also persists in clusters within states (e.g., California) where belief in climate change is high. Political affiliation has the strongest correlation, followed by level of education, COVID-19 vaccination rates, carbon intensity of the regional economy, and income. The analysis reveals how a coordinated social media network uses periodic events, such as cold weather and climate conferences, to sow disbelief about climate change and science, in general. Donald Trump was the strongest influencer in this network, followed by conservative media outlets and right-wing activists. As a form of knowledge vulnerability, climate denialism renders communities unprepared to take steps to increase resilience. As with other forms of misinformation, social media companies (e.g., X, Facebook, YouTube, TikTok) should flag accounts that spread falsehoods about climate change and collaborate on targeted educational campaigns.

Introduction

Climate change denialism persists in the United States, with estimates ranging from 12% to 26% of the U.S. population 1 , 2 . It is more pronounced in some states and regions 3 . Reasons for this denialism are multifaceted: Political affiliation and ideology, income, education, and exposure to extreme weather events are all important factors 4 , 5 , 6 . Denialism is more prevalent where local economies are highly dependent on fossil fuels 7 , in rural communities, and in populations where mistrust in science is pronounced 8 , 9 . Social media reaches millions of users, providing a key mechanism for influencers to spread misinformation 10 . The ability of social media to influence and harden attitudes was apparent in the response to COVID-19 vaccines 11 .

Understanding how and why climate change opinion varies geographically and documenting it at an actionable scale is crucial for communication campaigns, outreach, and other interventions 12 , 13 . Most estimates of the extent and geographic configuration of climate change denialism rely primarily on national surveys, with the Yale Climate Opinion Survey being the only dataset that provides estimates at the state and county levels for the entire U.S. 3 . These survey efforts, however, are time-intensive and expensive and are therefore destined to cover short time spans and, often, limited geographic extent. The Yale Survey combines data from more than 2500 national surveys and uses multinomial regression modeling to downscale estimates to subnational levels. Independent representative surveys conducted in states and metropolitan areas validate the predictions from the Yale Survey models 3 .

Mining social media data (e.g., Facebook, YouTube, and X, formerly Twitter) is a tantalizing alternative to survey-based approaches 14 , 15 . X is a social media platform with an extensive data repository. By adjusting for the skew toward certain demographic groups in users, data from this platform is useful for estimating public views on an array of topics, such as politics, social issues, and COVID-19 vaccination rates 16 , 17 . Data from Twitter has also been used in predictive modeling of election outcomes 18 . Account holders can misuse it to oppose scientific knowledge and spread misinformation 19 .

This study used Twitter data (2017–2019) to: (i) estimate the prevalence of climate change denialism at the state and county levels; (ii) identify typical profiles of climate change deniers; (iii) understand how social media promulgates climate change denialism through key influencers; and (iv) determine how world events are leveraged to promulgate attitudes about climate change.

We used a Deep Learning text recognition model to classify 7.4 million geocoded tweets containing keywords related to climate change. Posted by 1.3 million unique users in the U.S., these tweets were collected between September 2017 and May 2019 (see Online Methods S1 ). We classified these tweets about climate change into ‘for’ (belief) and ‘against’ (denial). Our analysis resulted in a profile of climate change deniers at the county level, provided insight into the networks of social media figures influential in promoting climate change denial, and generated insight into how these influencers use current events to foster this denial.

After confirming the validity of using social media data instead of information collected through surveys to capture public opinion on climate change at policy-relevant geographical scales, we found that denialism clusters in particular regions (and counties) of the country and amongst certain socio-demographic groups. Our analysis reveals how politicians, media figures, and conservative activists promulgated misinformation in the Twittersphere. It maps out how denialists and climate change believers have formed mostly separate Twitter communities, creating echo chambers. Such information provides a basis for developing strategies to counter this knowledge vulnerability and reduce the spread of mis- or disinformation by targeting the communities most at risk of not adopting measaures to increase resilience to the effects of climate change.

Where in the U.S. is climate change denial prevalent?

Our study found that 14.8% of Americans deny that climate change is real (Fig.  1 A), a percentage consistent with previous national studies (Fig. S4 ). Using geolocation information, we determined that denialism is highest in the Central part of the U.S. and in the South, with more than 20% of the populations of OK, MS, AL, and ND consisting of deniers. Along the West and East Coasts and New England, belief in climate change is highest. However, climate change denial varies substantially within states, often clustering in geographic swaths across multiple counties (Fig.  1 B). For example, in Shasta County, California climate change denial is as high as 52%; yet overall less than 12% of the population of California does not believe in climate change. Similarly, the average percentage of deniers is 21% in Texas, but at the county-level this ranges from 13% in Travis County to 67% in Hockley County.

figure 1

Climate change denialism in the United States, by state ( A ) and county ( B ). Note: Figure created using QGIS 3.30 ( https://www.qgis.org/ ).

To validate these results, we compared them to the Yale Climate Opinion Surveys at the national, state, and county levels (Fig. S5 ). The mean absolute difference between the two models was three percentage points (S.D. = 2.7), with the Twitter data yielding a higher percentage of deniers (Fig. S5 A). Compared to the Yale Survey, our model showed higher proportions of deniers in Southern states (for example, MS, AL, TN, and TX). However, state-level and county-level percentages of believers and deniers were highly correlated between the two datasets (p < 0.001) (Fig. S5 B–E).

What type of people are climate change deniers?

We performed bivariate correlation analysis with data from multiple publicly available sources (see Online Methods S1 ) to characterize climate change deniers (Table 1 ). We evaluated the following characteristics of populations in those regions that were associated with the Twitter profiles for a positive or negative correlation with climate change denial: Political affiliation, race, income, education level, COVID-19 vaccination rates (proxy for belief in science in general), degree of carbon-intensity of the regional economy, degree of urbanization (county-level), and local weather patterns (Table 1 ). At both the county and state levels, populations with a high percentage of Republican voters had the strongest correlation with climate change deniers. Carbon dependency of the economy was also significantly high at the state level. The strongest negative correlations (at both state and county levels) were level of education and COVID-19 vaccination rates. Integrating these data into a weighted least squares regression model, we defined a profile of a "typical" climate change denier (Table 2 ). This was the typical profile : Republican, with no college degree and without COVID-19 vaccination living in a region with a high average annual temperature.

To gain additional insight into the geographical relationship between denialism and political affiliation at the county level, we used the bivariate LISA (Local Indicators of Spatial Association) model 20 to identify which counties with high rates of denialism or belief were spatially associated with high rates of Republican or Democratic voters. Clusters of deniers that coincided with high rates of Republican voters were spatially contiguous and covered large swaths of the interior West (Idaho, Montana, Wyoming), Central (Nebraska, Kansas, Oklahoma, Texas), and Appalachian regions (West Virginia, Tennessee) of the U.S. (Fig.  2 ). These findings are consistent with our regression modeling and bivariate correlations: these regions tend to have high rates of carbon dependency of the economy, low COVID-19 vaccination rates, and large rural populations. Conversely, clusters of believers and high rates of Democratic voters were most prevalent along the Pacific Coast (California, Washington), the New England Region, the Great Lakes, and the Southwest (Arizona), as well as in regions near metropolitan areas and technological hubs.

figure 2

Clusters of spatial association between climate change denial and belief in relation to political affiliation. Notes: Figure created using QGIS 3.30 ( https://www.qgis.org/ ). Spatial clustering analysis performed using Geoda 1.22 ( https://github.com/GeoDaCenter/geoda/ ).

Who are climate change influencers in the Twittersphere?

To delineate how polarized opinion forms in the Twittersphere, we constructed Twitter networks (based on the 1200 most retweeted users in the sample), analyzed how users interact, and identified key influencers (Fig.  3 ). To identify closely linked users assumed to share similar views, we evaluated co-retweeting, in which a single user retweets tweets from two or more other users 21 . Two distinct communities emerged, a denier community and a believer community (Fig.  3 A). The community of climate change believers (blue nodes) is larger, with 1029 users and ~ 224,000 co-retweets, giving it a broader reach and influence on Twitter than the denier community (red nodes), which has 171 users and ~ 15,000 co-retweets. The proportion of deniers among the top 1200 influential users (14.3%) aligned with the national percentage of climate change deniers identified in our model (14.8%).

figure 3

Influencers detected in climate change co-retweeted networks. ( A ) Co-retweeted networks formed by the 1200 most retweeted users in the U.S. The nodes represent unique accounts; the edges represent co-retweeted relationships. The size of nodes and the shade of the node color are proportional to their influence, as measured by eigenvector centrality scores. The high density of edges within the communities makes many individual edges not displayable. The top influencers in the community of climate change deniers ( B ) and believers ( C ) are labeled with the usernames. In panels ( B ) and ( C ), edges-to-users in the other community are not displayed.

Both believers and deniers mostly shared information and interacted within their own community. Users from the two communities were rarely co-retweeted, as illustrated by the distance between the cluster of nodes for each community and the low number of edges connecting the two communities. Among ~ 230,000 co-retweets, only 4083 (< 0.02%) were between users having opposite views on climate change. This low percentage of co-retweets of contrasting views highlights an echo-chamber effect . We found that a few nodes bridge the gap between the two communities, notably conservative news outlets such as Fox News and the Washington Examiner .

To identify the most influential users, we calculated the eigenvector centrality value per Twitter user. A high score means that a user was co-retweeted with many other users who also had high scores. Among climate change deniers, former U.S. President Donald Trump had the biggest influence (Fig.  3 B). Three groups of influential deniers were heavily co-retweeted with President Trump: (i) conservative media outlets that regularly broadcast contrarian views on climate change, including alt-right news and blogs such as The Daily Wire , Daily Caller , Breitbart and thebradfordfile ; (ii) mis/disinformation websites that publish misleading and false claims about climate change, including TownHall Media and the Climate Depot ; and (iii) right-wing producers, political commentators, and activists. Collectively, in concert with former President Trump and close colleagues, these three groups formed an organized and coordinated social media network, enabling climate change denialism to amplify and expand.

In contrast, the larger blue community is more diffuse. Politicians dominated the most influential users (Fig.  3 C). Of the top 30 influential believers, 15 accounts belong to members of the Democratic Party, such as Alexandria Ocasio-Cortez, Bernie Sanders, and Kamala Harris (Table S1 ). Eight of the top 30 nodes were popular media outlets, or websites, such as CNN, NBC, ABC, The Hill, The Washington Post, The New York Times . Other influential nodes included popular science communicators and entertainers advocating scientific consensus.

How does climate change-related tweeting and topic use vary over time?

To investigate the dynamics of tweeting activity for both communities and to understand how each perceived and responded to real-world events, we performed topic modeling and time series analysis of tweet volume. This analysis revealed how each group reacted selectively and opportunistically to the 17 events that occurred during the period of data collection (November 2017–May 2019).

Consistent with the larger size of the believer community, this community had a consistent pattern of climate change tweet activity throughout the sampling period (Fig.  4 A). In contrast, the denier community had lower activity overall. However, both communities had periods of high activity with spikes that exceeded the average pattern. The number of these high spikes was lower for the denier community. By manually identifying events that potentially triggered these large spikes, we found that deniers and believers do not always respond to the same events. Only six events triggered higher than average tweet volume by the denier community (Table S3 ): three were related to extreme cold weather events, two were related to United Nations activities about climate change —a report by the Intergovernmental Panel on Climate Change (IPCC) and an annual meeting of the United Nations Framework Convention on Climate Change (COP24), and the last was an attack on climate change deniers by Bill Nye in an HBO broadcast. Intriguingly, two of the highest spikes by the believer community occurred with events associated with President Trump that sparked high activity in the denier community, suggesting that these communities tried to influence or counter each other.

figure 4

Events that drive tweet volume among deniers and believers and topic prevalence for typical events. ( A ) Original tweet volume per day and locally weighted regression lines are plotted over time for both climate change deniers and believers. Events that sparked online discussions are labeled alongside tweet volume spikes numerically and detailed in the lower left. Red bubbles denote the events that a large group of deniers were actively involved with (> 1000 original tweets). The gap in November 2018 and between January and April 2019 was due to discontinued data collection. ( B ) Topic prevalence for typical major events 22 : Events 3 and 6 represent extreme cold weather events; Event 5 represents top denier influencer Donald Trump tweeting about cold weather and doubts about global warming; Event 12 represents the Trump Administration publicly refuting the validity of the IPCC 2018 report ( https://doi.org/10.1017/9781009157940 ); and Event 13, a UN climate change conference (COP24), represents an event that engaged both deniers and believers.

To gain further insight into whether the groups attempted to counter each other, we classified tweets of believers and deniers for these 17 events based on the five climate change narratives (see Fig.  4 B) proposed by Cook 22 . Overall, the major narrative in the believer community was “There is still time to adapt," representing 42% of the total tweets). In contrast, deniers focused tweeting activity on the message “Climate change is not real,” as indicated by 48% of the tweets falling into this category.

Although weather events were associated with spikes in tweets from both communities, events viewed as abnormal weather caused by climate change [the California Wildfires (Event 9) and Hurricane Florence (Event 10) triggered a high volume of tweets among believers and events viewed as colder-than-expected weather [a snowstorm in Texas (Event 3) and a blizzard in the Mid-Atlantic and New England regions (Event 6) triggered a surge in tweets amongst deniers. Both colder-than-expected weather events provided an opportunity for the deniers to espouse that climate change is not real (64% of the total tweets for both events), to delegitimize scientific consensus (12% of the total) and to reaffirm the claim that the changing climate is a normal geologic process and to foment doubt that human activities are a source of this change (13% of the total).

Consistent with an attempt to counter each other’s messages, the December 2017 tweet by Trump casting doubt on global warming due to a blizzard (Event 5) triggered the believer community to issue tweets emphasizing that climate change is unequivocal (32% of the total) and that there is clear scientific consensus (35% of the total). A common refrain among deniers was that climate change is a conspiracy theory or hoax (59% of the total) and a shadowy attempt to dupe the public into bearing the costs of decarbonization, while generating enormous wealth for Blue ‘elites’ (9% of the total). These tweets were heavily re-tweeted by conservative media (e.g., Daily Caller), right-wing activists (e.g., Chuck Woolery), and mis/disinformation sites (e.g., Watts Up With That?) (Table S1 ).

Conflicting messages were also common in response to COP24, consistent with an attempt to influence opinion. Believers overwhelmingly advocated for timely collective action or promoted campaigns showcasing impacts of and solutions to climate change (50% of the total). Deniers focused on conspiracy theories (climate change is not real, 46%) or a Democratic party agenda filled with impractical solutions (26%).

Study limitations

Our modeling approach has several limitations. First, X (formerly Twitter) has limited the number of posts a user can read in a day, ostensibly to prevent ‘data scraping’ practices from unauthorized users. If this policy persists, then it will hamper research that relies on social media data to assess online beliefs and sentiment. This has implications in terms of the replicability of our modeling to other countries, time frames, and topics 23 , 24 , 25 . A second limitation of our modeling is that uncertainty is higher in areas with low population densities (e.g., rural areas) due to small sample sizes 26 . This is a well-known limitation of using social media to assess consensus and is even more pronounced in countries where use of social media is limited. To minimize this effect in our study, we normalized input data by county population and employed a weighted approach using the total count of tweets as weights both for the calculation of bivariate relationships and for the regression models (see “Methods”). Third, our classification scheme labeled tweets as either believing or denying climate change. National surveys indicate a percentage of the population (5–15%) who remain neutral or may not have a particular opinion on the topic 27 . In our binary classification, we used those climate change-related keywords that indicated a clear position (for or against) on the issue. As a portion of our sample uses sarcastic or ambivalent language that is difficult for the model to distinguish, classifying these tweets can be challenging. To address this, we calculated confidence for each prediction (see “Methods”) to filter out those with low confidence (CI < 0.75).

Using data from Twitter (now X), we used AI techniques and network analysis to delineate a comprehensive social anatomy of climate change denialism in the U.S., at the state and county levels. We identified geographic clusters of climate change denial in Republican counties, especially rural ones, and among residents who do not have a college education. This provides critical knowledge for identifying segments of the population that would benefit from targeted efforts to expand awareness of the risks associated with climate change and strategies to increase local resilience.

The strong correlation between denialism and low COVID-19 vaccination rates indicated a broad skepticism of science generally amongst climate change deniers, which corresponds to resistance to science-based public policies such as shelter-in-place COVID-19 mandates 28 or mask usage 29 . This finding indicates that communities with a high prevalence of deniers are at risk for discounting other science-based health or safety recommendations. According to the theory of identity-protective cognition, people tend to selectively credit or discredit evidence in patterns that reflect beliefs that predominate in their group 30 , 31 . This theory helps explain why those who vote Republican are more likely to believe tweets from former President Trump regarding climate change than from other sources; it is identity affirming.

Classifying tweets based on the Cook’s five categories 22 enables identification of commonly deployed rhetorical strategies to promote climate misinformation, and in science denialism more broadly 32 . In our 7.3 million tweet sample, these techniques included fake experts, who have possessed little to no expertise about the underlying science but nonetheless conveyed messages that cast doubt. They serve as a credible messenger in which someone shares the same moral values and uses language consistent with existing beliefs 12 , 33 . One such example is the tweet by the Trump Administration casting doubts on the IPCC 2018 Climate Report, which was retweeted heavily by supporters. Then there are logical fallacies , such as a tweet by Trump questioning global warming because of an unusual cold weather event that went viral 34 . Other common strategies include impossible expectations as well as cherry picking to attack climate change science and scientists.

Combating misinformation requires effective refutation strategies 35 . Deploying such strategies on social media sites such as X, however, is challenging as denier and believer communities are isolated from each other, leading to echo chambers 19 . Only 0.02% of the co-retweets about climate change were between users having opposing views. Consequently, this leads one to conclude that believers have limited ability to reach deniers through the social media platform. One strategy is to label denialism tweets as misinformation. However, some evidence suggests that this can strengthen opposition rather than change attitudes 36 . Another option is to flag accounts that disseminate misinformation or dangerous information. For example, then-Twitter banned Donald Trump (his account has since been restored) from using the site because of tweets maintaining election fraud and supporting the January 6 capital riots 37 . Twitter also banned accounts for spreading COVID-19 misinformation and calling for violence against media 38 . As with COVID-19, climate change is a humanitarian crisis that will affect millions, albeit at a more elongated temporal scale. Based on current policy, X (formerly Twitter) does not appear to be imposing account bans or suspensions for promoting climate change denialism. This, and related policy changes associated with new ownership of the social media platform, may make it even more susceptible to the spread of misinformation.

Communities face increasing risks related to climate change, such as flooding, wildfire, heat stress, and sea-level rise. The scientific community has already identified especially vulnerable communities and regions 39 . Climate change denialism is also a risk, in the form of knowledge vulnerability. Those who discount climate change as a natural rather than human-induced process tend to underestimate their current (and future) risk to it. This renders them less likely to take necessary steps to mitigate and adapt to climate change.

Opinion data

As primary data, we used an open access dataset created by George Washington University that is available from the GWU Libraries Dataverse 40 . This dataset was created using the Twitter Stream API and contains ~ 40 million tweets related to climate change and global warming. It covers a two-year period from September 2017 to May 2019. We initially retrieved ~ 27.3 million raw tweets based on tweet IDs. The ~ 30% loss of tweets was due to deleted or inactive accounts since 2019.

To extract tweets located in the U.S., we developed a rule based on the geo-attributes in the raw data. We extracted the self-reported location information in an account profile. A large proportion of users (> 73%) provided the location information in our dataset. To standardize the addresses and improve the geocoding process, we first transformed all the user locations to lower case and removed the URL links, emojis, punctuation marks, and other non-ASCII characters. Next, we extracted all the unique user locations (~ 640,000 “clean” addresses) and standardized all the U.S. state and city abbreviations. As a final step, we manually inspected and removed national level and obviously fake user locations.

After the preprocessing, we used the Nominatim API server to geocode user locations based on the OpenStreetMap database 41 . We removed locations outside the U.S., and classified addresses within the U.S. into two levels: (1) county level with tweets from users reporting their local address, city, or county; (2) state level with tweets from users reporting only the state. In the state-level tweets, we also added the aggregated county-level tweets. We then rejoined these unique U.S. addresses and the corresponding geographical coordinates to the original datasets by spatial level. The geocoding yielded ~ 1.3 million unique users and ~ 5.2 million county-level tweets and ~ 7.4 million state-level tweets, from which ~ 2.2 million tweets had state-level only information. To reduce the incidence of non-human accounts in our sample (i.e. tweet bots), we removed users who tweeted more than 20 times a day. Figures S1 and S2 presents the data spatial distribution and representativeness analysis.

Tweet classification

To identify climate change opinions on Twitter, we built a tweet classifier based on the Transformer, a deep learning model in the field of natural language processing 42 . We parameterized the model to classify tweets as either believing in the existence of climate change (predicted as ‘for’) or denying that climate change is real (predicted as ‘against’). Instead of training a model de novo, the Transformer uses language models pre-trained on large text corpora in an unsupervised manner and then uses user-labeled training samples to fine-tune the model for specific natural language tasks. Our classifier was built upon OpenAI GPT-2, a large transformer-based language model pre-trained on a database of ~ 8 million web pages 43 . Previous studies found that the GPT-2 model performs well in classifying short text from social media 44 .

We built a training dataset of manually labeled tweets to fine tune the pre-trained GPT-2 model. Labeled samples were randomly extracted only from the 1.4 million original tweets, excluding retweets and quotes. Each tweet was reviewed independently by two members of the research team and labeled as either ‘against’ or ‘for’ climate change. In rare occasions where a tweet’s message was ambiguous and there was disagreement between the two members of the research team, the tweet was excluded from the model's training data due to its potential to introduce noise to the model.

We labeled training tweets as ‘for’ or ‘against’ climate change if they had one of the following viewpoints listed in Table 3 . This labeling resulted in a balanced sample of 6,500 tweets (3300 ‘for’ tweets and 3200 ‘against’ tweets) that we used as a training set for the model. Tweets with ambiguous messages, sarcastic language or tweets that were irrelevant to climate change were discarded from the training dataset.

Our model was built upon the Huggingface Transformers 45 Library and implemented in PyTorch 46 . To increase the model’s predictive accuracy, we fine-tuned the parameters that resulted in an optimum learning rate at 1e−5, with dropouts at 0.1. Tweets with sarcastic, ambiguous or irrelevant messages were evaluated with the model, but the predictions based on these tweets tended to be invalid or random. To overcome this limitation, we used the Softmax function embodied in PyTorch , which calculated the prediction confidence for every individual tweet. Based on this score, we removed predictions with low confidence (CI < 0.75). The final classification was performed on the complete set of 7.4 million tweets from the collection period. We then aggregated tweets at the county and state levels and calculated percentages of ‘against’ tweets and ‘for’ tweets as proxies of deniers and believers.

To evaluate the model’s performance, we performed a series of validation tests. We manually labeled an independent validation dataset to test model accuracy. To ensure the validation dataset was balanced across the two categories and was spatially representative, we randomly extracted 30 unique original tweets from each state. Our fine-tuned model achieved an overall accuracy of 0.91 and F1 score of 0.90 (Fig. S3 ). Our model predictions were compared with national estimates of climate change opinion based on representative surveys, showing that our model provided a percentage for U.S. climate change deniers within the range of those determined from the surveys (Fig. S4 ). To validate our results at the sub-national level, we used the Yale Climate Opinion Surveys. The Yale Surveys use a downscaling statistical model based on national survey data and are the only surveys that provide climate change opinion estimates at the state and county levels. We compared these data with our model results at both state and county levels by calculating the Pearson correlation coefficient. To normalize the data, we weighted the variables per population of each state and county (US Census 2018).

Correlation analysis

To examine what drives climate change opinion, we performed a series of correlation analyses. Studies have shown that climate change opinion is strongly correlated with political affiliation which is considered a major driver 47 , 48 . In addition to political ideology, studies have shown evidence that the socio-demographic profile also plays a role in climate change opinion as does the local microclimate 49 , 50 , and personal experience with extreme weather events, although the evidence is mixed and at times ambivalent 51 , 52 , 53 . Informed by this literature, therefore, we examined variables that have been reported to influence climate change opinion: political affiliation, COVID vaccination rate (proxy for belief in science in general), urbanization rate, education, income, race, carbon intensity of economy, natural hazard risk, and temperature anomalies.

We used the percentage of ‘against’ and ‘for’ tweets to reflect the prevalence of deniers and believers across the U.S. at the county and state levels. For political affiliation, we acquired 20 years (2000–2020) of county-level U.S. Presidential election returns from the MIT Election Data and Science Lab ( https://electionlab.mit.edu/data ). We calculated the average percentage of Democrats and Republicans per state and county, weighted by the county population. For science skepticism, we used the county-level COVID-19 vaccination rates as a proxy, using data from the CDC ( https://www.cdc.gov/coronavirus/2019-ncov/vaccines/distributing/reporting-counties.html ). For educational attainment, race, and income, we used data from the US Census Bureau's 2020 American Community Survey, which provides estimates of average characteristics from 2016 through 2020 at the state and county levels. Specifically, we used the number of people who have at least a Bachelor's college degree, number of people per race, and the median household income. For county-level natural hazard risk, we used the National Risk Index developed by FEMA ( https://www.fema.gov/flood-maps/products-tools/national-risk-index ). An overall risk score was calculated for each county, measuring the expected annual loss due to 18 types of natural hazards. For temperature anomalies, we acquired historic 30-year annual mean temperature (1981–2010) and the mean for recent years (2015–2019) from the PRISM climate group ( https://prism.oregonstate.edu/ ). County-level temperature anomalies were then obtained by calculating the standard deviation between annual mean temperature of recent years and the 30-year averages. To investigate the association between state-level carbon dependency of economy and climate change opinion, we used energy-related carbon emissions per gross domestic product (GDP) for each state from the Energy Information Administration ( https://www.eia.gov/environment/emissions/state/ ). The unit of carbon intensity is the metric tons of energy-related carbon dioxide per million dollars of GDP. A six-level urban–rural classification at the county level was from the National Center for Health Statistics data systems ( https://www.cdc.gov/nchs/data_access/urban_rural.htm ).

To account for variations in population across counties and states, we normalized all data expressed as counts. We adjusted the total county population as: Population Adj  = Total population/10,000. Then, we normalized each variable by population by dividing the counts of people for each variable by the adjusted population: Normalized Variable = Variable count/Population Adj . Based on the normalized data, we calculated bivariate weighted Pearson correlations between climate change opinion and each of these variables using the total count of tweets per county as the weight. The same data were used as predictors for the regression model. We used the weighted ordinary least squares for the total count of tweets per county as the universal weight.

To identify spatial clusters of climate change denialism or belief at the county level in relation to political affiliation (Republican or Democrat), we applied the bivariate Local Indicators of Spatial Association (LISA) 20 . We applied the second order Queen contiguity weights at the county level and ran the models with 999 permutations and significance at p < 0.05. This approach was executed in the open-source software Geoda 54 .

Co-retweeted network analysis

We constructed a co-retweeted network to delineate interactions and identify the most influential Twitter users from both sides. Co-retweeting is defined as the act of a single user retweeting two or more other users. We used these events to create undirected weighted edges between the co-retweeted accounts. The more users retweet two other users, the more weight the edge gains. Accordingly, we assumed that the more co-retweets two accounts receive, the more likely their views are related. The co-retweeted network represents engaged communities with similar opinions.

To construct the co-retweeted network, we first calculated the total sum of retweets as a measure of overall influence for each user account in our 7.2 million tweets dataset. We selected the 1200 most retweeted accounts for further processing, along with all the users who have retweeted them. We then constructed the retweet matrix A where the rows represent the 1200 top accounts, and the columns represent the rest of user accounts. Elements in matrix A are binary: A value of 1 means that the public account has retweeted the corresponding top influential account and 0 means the public account has not retweeted the top influential account. We then multiplied matrix A with its transposed matrix A T and transformed it into the co-retweeted square matrix B . Matrix B has 1200 rows and columns that represent the influential accounts. The upper and lower diagonal cells of matrix B contain the total number of times that two influential accounts are co-retweeted. We exported all the unique pairs of influential accounts and their co-retweets as the edge table for further network analysis.

Our co-retweeted network was visualized in Gephi , using the Force Atlas algorithm 55 , which clusters nodes based on their connections. The distance between two nodes was weighted by the number of co-retweets. We then applied the Louvain community-detection algorithm 56 and separated the nodes as two communities based on modularity scores. To detect opinion leaders in each community, we calculated the Eigen centrality values for each node based on the igraph package in R 57 . The number of co-retweets for each node was set as the weight. To facilitate visualization, we extracted the top 30 influencers from each community (Table S1 for deniers and Table S2 for believers). The eigenvalues are scaled to a maximum score of one.

Time-series analysis and topic modeling

To examine the dynamics of tweeting activity regarding climate change, we analyzed the tweet volume of both deniers and believers during September 2017 to May 2019 and identified 17 major climate change-related events that may spark controversy online. To identify real-world events, we first applied the z-score algorithm to detect spikes in tweets volume of deniers and believers separately, along with the daily time series. The algorithm detected peaks when the value of a new data-point is three standard deviations away from the moving mean of 5 days of observations. The algorithm identified nine peaks in deniers’ tweets and 16 peaks in believers’ tweets. We then applied the tidytext package in R to segment each day’s Twitter texts into separate words and calculate the term frequency of each word. Next, we reviewed these keywords and manually assigned each peak to one real-world event alone. Comparing independent coding results of the three coders, we got a Cohen's kappa score of 0.90, indicating good interrater reliability. Divergences were resolved following discussions among the coders. Table S3 presents the dates with spikes in tweet volume, the most prevalent keywords, and associated real-world events.

To delineate the major climate change-related topics discussed, to understand how the prevalence of each topic evolved over time, and to explore how each group perceived the event, we employed the Latent Dirichlet Allocation (LDA) algorithm 58 to automatically extract the main topics. We specified the number of topics before training the model. We devised a five-category classification scheme following Cook’s 22 categories of misinformation: (a) climate change is/is not real; (b) humans are/are not the main cause; (c) the impacts are/are not serious; (d) the experts agree/are unreliable; (e) there is still time to adapt/solutions offered are inefficient.

The model was implemented in Python’s Gensim package along with the Java-based package Mallet to accelerate data processing 59 . We ran topic modeling separately for tweets classified as from ‘believers’ or ‘deniers.’ We preprocessed the original ~ 7.2 million tweets, keeping original tweets and excluding retweets with the same text. We removed all the @mentions, hashtags, punctuation marks, and changed all characters to lower case. From keywords, we removed “climate change” and “global warming” because these words occurred too frequently and would dominate as distinct topics. After this pre-processing, we tokenized every tweet and created bigrams and trigrams because some words often occurred together as phrases. We reduced words to their common word stem and dropped duplicates to ensure the text corpora analyzed by the model was clean and distinct.

Data availability

Primary Twitter data and ancillary data were obtained from publicly available sources (see “ Methods ”). Secondary data that support the findings of this study are available from the corresponding author upon reasonable request.

Code availability

Code generated to perform this analysis is available at https://zenodo.org/record/8017765 .

McDonald, J., MacInnis, B. & Krosnick, J.A. Climate Insights 2020: Opinion in the States. Washington, DC: Resources for the Future. https://rb.gy/zbwo2k (2020).

Leiserowitz, A., Roser-Renouf, C., Marlon, J. & Maibach, E. Global Warming’s Six Americas: A review and recommendations for climate change communication. Curr. Opin. Behav. Sci. 42 , 97–103 (2021).

Article   Google Scholar  

Howe, P. D., Mildenberger, M., Marlon, J. R. & Leiserowitz, A. Geographic variation in opinions on climate change at state and local scales in the USA. Nat. Clim. Chang. 5 , 596–603 (2015).

Article   ADS   Google Scholar  

Hornsey, M. J., Harris, E. A., Bain, P. G. & Fielding, K. S. Meta-analyses of the determinants and outcomes of belief in climate change. Nat. Clim. Chang. 6 , 622–626 (2016).

Hornsey, M. J., Harris, E. A. & Fielding, K. S. Relationships among conspiratorial beliefs, conservatism and climate skepticism across nations. Nat. Clim. Chang. 8 , 614–620 (2018).

McCright, A. M. & Dunlap, R. E. Cool dudes: The denial of climate change among conservative white males in the United States. Glob. Environ.  Change  21 , 1163–1172 (2011).

Knight, K. W. Does fossil fuel dependence influence public awareness and perception of climate change? A cross-national investigation. Int. J. Sociol. 48 , 295–313 (2018).

Long, E. F., Chen, M. K. & Rohla, R. Political storms: Emergent partisan skepticism of hurricane risks. Sci. Adv. 6 , eabb7906 (2020).

Weckroth, M. & Ala-Mantila, S. Socioeconomic geography of climate change views in Europe. Glob. Environ. Change  72 , 102453 (2022).

Pennycook, G. et al. Shifting attention to accuracy can reduce misinformation online. Nature 592 , 590–595 (2021).

Article   ADS   CAS   PubMed   Google Scholar  

Johnson, N. F. et al. The online competition between pro- and anti-vaccination views. Nature 582 , 230–233 (2020).

Goldberg, M. H., Gustafson, A., Rosenthal, S. A. & Leiserowitz, A. Shifting republican views on climate change through targeted advertising. Nat. Clim. Change 11 , 573–577 (2021).

Zhang, B. et al. Experimental effects of climate messages vary geographically. Nat. Clim. Change 8 , 370–374 (2018).

Stieglitz, S. & Dang-Xuan, L. Social media and political communication: A social media analytics framework. Soc. Netw. Anal. Min. 3 , 1277–1291 (2013).

Kirilenko, A. P. & Stepchenkova, S. O. Public microblogging on climate change: One year of Twitter worldwide. Glob. Environ. Change   26 , 171–182 (2014).

Jaidka, K. et al. Estimating geographic subjective well-being from Twitter: A comparison of dictionary and data-driven language methods. Proc. Natl. Acad. Sci. USA 117 , 10165–10171 (2020).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Grossman, G., Kim, S., Rexer, J. M. & Thirumurthy, H. Political partisanship influences behavioral responses to governors’ recommendations for COVID-19 prevention in the United States. Proc. Natl. Acad. Sci. USA 117 , 24144–24153 (2020).

Bovet, A., Morone, F. & Makse, H. A. Validation of Twitter opinion trends with national polling aggregates: Hillary Clinton vs Donald Trump. Sci. Rep. 8 , 8673(2018).

Pennycook, G. & Rand, D. G. Fighting misinformation on social media using crowdsourced judgments of news source quality. Proc. Natl. Acad. Sci. USA 116 , 2521–2526 (2019).

Anselin, L. Local Indicators of Spatial Association—LISA. Geograph. Anal. 27 , 93–115 (1995).

Finn, S., Mustafaraj, E. & Metaxas, P. T. The co-retweeted network and its applications for measuring the perceived political polarization. In Proceedings of the 10th International Conference on Web Information Systems and Technologies 2 , 276–284 (2014).

Cook, J. Understanding and Countering Misinformation About Climate Change in Handbook of research on deception, fake news, and misinformation online (ed. Chiluwa, I.E. & Samoilenko, S.A.) 281–307 (IGI Global).

Davidson, B. I. et al. Platform-controlled social media APIs threaten open science. Nat. Hum. Behav. https://doi.org/10.1038/s41562-023-01750-2 (2023).

Article   PubMed   Google Scholar  

Walker, S., Mercea, D. & Bastos, M. The disinformation landscape and the lockdown of social platforms. Inf. Commun. Soc. 22 , 1531–1543 (2019).

Acker, A. & Kreisberg, A. Social media data archives in an API-driven world. Arch. Sci. 20 , 105–123 (2020).

Barberá, P. & Rivero, G. Understanding the political representativeness of twitter users. Soc. Sci. Comput. Rev. 33 (6), 712–729 (2015).

Leiserowitz, A. et al. Climate Change in the American Mind: Beliefs & Attitudes. Yale University and George Mason University. New Haven, CT: Yale Program on Climate Change Communication. (2023).

Brzezinski, A., Kecht, V., Van Dijcke, D. & Wright, A. L. Science skepticism reduced compliance with COVID-19 shelter-in-place policies in the United States. Nat. Hum. Behav. 5 , 1519–1527 (2021).

Merkley, E. & Loewen, P. J. Anti-intellectualism and the mass public’s response to the COVID-19 pandemic. Nat. Hum. Behav. 5 , 706–715 (2021).

Kahan, D. M. Misconceptions, misinformation, and the logic of identity-protective cognition. Cultural Cognition Project Working Paper Series No. 164, Yale Law School (2017).

Kahan, D. M. et al. Culture and identity-protective cognition: Explaining the white-male effect in risk perception. J. Emp. Legal Stud. 4 , 465–505 (2007).

Ceccarelli, L. Manufactured scientific controversy: Science, rhetoric, and public debate. Rhetoric Public Affairs 14 , 195–228 (2011).

Corner, A. et al. How do young people engage with climate change? The role of knowledge, values, message framing, and trusted communicators. WIREs Clim. Change 6 , 523–534 (2015).

Farmer, G. T. & Cook, J. Understanding Climate Change Denial in Climate Change Science: A Modern Synthesis: Volume 1 - The Physical Climate (eds Farmer, G. T. & Cook, J.) 445–466 (Springer Netherlands, 2013).

Schmid, P. & Betsch, C. Effective strategies for rebutting science denialism in public discussions. Nat. Hum. Behav. 3 , 931–939 (2019).

Christenson, D. P., Kreps, S. E. & Kriner, D. L. Contemporary presidency: Going public in an era of social media: Tweets, corrections, and public opinion. President. Stud. Q. 51 , 151–165 (2021).

Tollefson, J. Tracking QAnon: How Trump turned conspiracy-theory research upside down. Nature 590 , 192–194 (2021).

ADS   CAS   PubMed   Google Scholar  

Bak-Coleman, J. B. et al. Combining interventions to reduce the spread of viral misinformation. Nat Hum Behav 6 , 1372–1380 (2022).

Rentschler, J., Salhab, M. & Jafino, B. A. Flood exposure and poverty in 188 countries. Nat. Commun. 13 , 3527 (2022).

Littman, J. & Wrubel, L. Climate Change Tweets Ids. Harvard Dataverse (2019).

OpenStreetMap. https://www.openstreetmap.org/ (2020).

Devlin, J., Chang, M.-W., Lee, K. & Toutanova, K. BERT: Pre-training of deep bidirectional transformers for language understanding in Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies 1 , 4171–4186, Minneapolis, Minnesota. Association for Computational Linguistics. (2019).

Radford, A. et al. Language models are unsupervised multitask learners. OpenAI blog 1 , 9 (2019).

Google Scholar  

Fagni, T., Falchi, F., Gambini, M., Martella, A. & Tesconi, M. TweepFake: About detecting deepfake tweets. PLOS ONE 16 , e0251415 (2021).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Wolf, T. et al. Transformers: State-of-the-Art Natural Language Processing in P roceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations 38–45, Association for Computational Linguistics. (2020).

Paszke, A. et al. Pytorch: An imperative style, high-performance deep learning library. Adv. Neural Inf. Process. Syst. 32 (2019).

Poortinga, W., Whitmarsh, L., Steg, L., Böhm, G. & Fisher, S. Climate change perceptions and their individual-level determinants: A cross-European analysis. Glob. Environ. Change 55 , 25–35 (2019).

Roxburgh, N. et al. Characterising climate change discourse on social media during extreme weather events. Glob. Environ. Change 54 , 50–60 (2019).

McCright, A. M. & Dunlap, R. E. Cool dudes: The denial of climate change among conservative white males in the United States. Global Environ. Change 21 , 1163–1172 (2011).

Bedsworth, L. W. & Hanak, E. Climate policy at the local level: Insights from California. Glob. Environ. Change 23 , 664–677 (2013).

Howe, P. D., Marlon, J. R., Mildenberger, M. & Shield, B. S. How will climate change shape climate opinion?. Environ. Res. Lett. 14 , 113001 (2019).

Albright, E. A. & Crow, D. Beliefs about climate change in the aftermath of extreme flooding. Clim. Change 155 , 1–17 (2019).

Article   ADS   CAS   Google Scholar  

Howe, P. D., Markowitz, E. M., Lee, T. M., Ko, C.-Y. & Leiserowitz, A. Global perceptions of local temperature change. Nat. Clim. Change 3 , 352–356 (2013).

Anselin, L., Syabri, I. & Kho, Y. GeoDa: An Introduction to Spatial Data Analysis in Handbook of Applied Spatial Analysis: Software Tools, Methods and Applications (eds Fischer, M.M. & Getis, A.) 73–89 (Springer Berlin Heidelberg, 2010).

Jacomy, M., Venturini, T., Heymann, S. & Bastian, M. ForceAtlas2, a continuous graph layout algorithm for handy network visualization designed for the Gephi Software. PLOS ONE 9 , e98679 (2014).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Meo, P. D., Ferrara, E., Fiumara, G. & Provetti, A. Generalized Louvain method for community detection in large networks. In 11th International Conference on Intelligent Systems Design and Applications , 88–93 (2011).

Csardi, G. & Nepusz, T. The igraph software package for complex network research. Int. J. Complex Syst. 1695 , 1–9 (2006).

Blei, D. M., Ng, A. Y. & Jordan, M. I. Latent dirichlet allocation. J. Mach. Learn. Res. 3 , 993–1022 (2003).

McCallum, A. K. MALLET: A Machine Learning for Language Toolkit , http://mallet.cs.umass.edu (2002).

Download references

Acknowledgements

The authors are grateful to Jianxun Yang and Miaomiao Liu for their contribution to this manuscript. Yang performed data curation, classification modeling, network analysis and topic modeling. Liu supervised Yang and provided financial support. The authors also wish to thank the three anonymous reviewers for their constructive feedback on earlier versions of this manuscript and Nancy Gough for helping with the copy editing process.

This work was supported by a Propelling Original Data Science (PODS) Grant from Michigan Institute for Data Science (MIDAS) and from a Themes grants from the School for Environment and Sustainability (University of Michigan).

Author information

Authors and affiliations.

School for Environment and Sustainability, University of Michigan, Ann Arbor, MI, 48109, USA

Dimitrios Gounaridis & Joshua P. Newell

You can also search for this author in PubMed   Google Scholar

Contributions

Conceptualization: D.G.; J.N. Methodology: D.G.; J.N. Writing - Review & Editing: D.G.; J.N. Funding acquisition: D.G.; J.N.

Corresponding author

Correspondence to Joshua P. Newell .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Supplementary information., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Gounaridis, D., Newell, J.P. The social anatomy of climate change denial in the United States. Sci Rep 14 , 2097 (2024). https://doi.org/10.1038/s41598-023-50591-6

Download citation

Received : 27 June 2023

Accepted : 21 December 2023

Published : 14 February 2024

DOI : https://doi.org/10.1038/s41598-023-50591-6

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

  • Explore articles by subject
  • Guide to authors
  • Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

case control study is a type of

Log in using your username and password

  • Search More Search for this keyword Advanced search
  • Latest content
  • Current issue
  • BMJ Journals More You are viewing from: Google Indexer

You are here

  • Online First
  • Integrating large language models in systematic reviews: a framework and case study using ROBINS-I for risk of bias assessment
  • Article Text
  • Article info
  • Citation Tools
  • Rapid Responses
  • Article metrics

Download PDF

  • http://orcid.org/0000-0001-9531-4990 Bashar Hasan 1 , 2 ,
  • http://orcid.org/0000-0001-9225-1197 Samer Saadi 1 , 2 ,
  • Noora S Rajjoub 1 ,
  • Moustafa Hegazi 1 , 2 ,
  • Mohammad Al-Kordi 1 , 2 ,
  • Farah Fleti 1 , 2 ,
  • Magdoleen Farah 1 , 2 ,
  • Irbaz B Riaz 3 ,
  • Imon Banerjee 4 , 5 ,
  • http://orcid.org/0000-0002-9368-6149 Zhen Wang 1 , 6 ,
  • http://orcid.org/0000-0001-5502-5975 Mohammad Hassan Murad 1 , 2
  • 1 Kern Center for the Science of Healthcare Delivery , Mayo Clinic , Rochester , Minnesota , USA
  • 2 Public Health, Infectious Diseases and Occupational Medicine , Mayo Clinic , Rochester , Minnesota , USA
  • 3 Division of Hematology-Oncology Department of Medicine , Mayo Clinic , Rochester , Minnesota , USA
  • 4 Department of Radiology , Mayo Clinic Arizona , Scottsdale , Arizona , USA
  • 5 School of Computing and Augmented Intelligence , Arizona State University , Tempe , Arizona , USA
  • 6 Health Care Policy and Research , Mayo Clinic Minnesota , Rochester , Minnesota , USA
  • Correspondence to Dr Bashar Hasan, Mayo Clinic, Rochester, MN 55905, USA; Hasan.Bashar{at}mayo.edu

Large language models (LLMs) may facilitate and expedite systematic reviews, although the approach to integrate LLMs in the review process is unclear. This study evaluates GPT-4 agreement with human reviewers in assessing the risk of bias using the Risk Of Bias In Non-randomised Studies of Interventions (ROBINS-I) tool and proposes a framework for integrating LLMs into systematic reviews. The case study demonstrated that raw per cent agreement was the highest for the ROBINS-I domain of ‘Classification of Intervention’. Kendall agreement coefficient was highest for the domains of ‘Participant Selection’, ‘Missing Data’ and ‘Measurement of Outcomes’, suggesting moderate agreement in these domains. Raw agreement about the overall risk of bias across domains was 61% (Kendall coefficient=0.35). The proposed framework for integrating LLMs into systematic reviews consists of four domains: rationale for LLM use, protocol (task definition, model selection, prompt engineering, data entry methods, human role and success metrics), execution (iterative revisions to the protocol) and reporting. We identify five basic task types relevant to systematic reviews: selection, extraction, judgement, analysis and narration. Considering the agreement level with a human reviewer in the case study, pairing artificial intelligence with an independent human reviewer remains required.

  • Evidence-Based Practice
  • Systematic Reviews as Topic

Data availability statement

Data are available upon reasonable request. Search strategy, selection process flowchart, prompts and boxes containing included SRs and studies are available in the appendix. Analysed datasheet is available upon request.

https://doi.org/10.1136/bmjebm-2023-112597

Statistics from Altmetric.com

Request permissions.

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

Risk of bias assessment in systematic reviews is a time-consuming task associated with inconsistency. Large language models’ (LLMs) utilisation in systematic reviews may be helpful but largely unexplored.

WHAT THIS STUDY ADDS

This study introduces a structured framework for integrating LLMs into systematic reviews with four domains: rationale, protocol, execution and reporting.

The framework defines five possible task types for LLMs in systematic reviews: selection, data extraction, judgement, analysis and narration.

A case study about using LLMs for risk of bias assessments using Risk Of Bias In Non-randomised Studies of Interventions demonstrates fair agreement between LLM and human reviewers.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

The proposed framework can serve as a blueprint for future systematic reviewers planning to integrate LLMs into their workflow.

The case study suggests the need to pair LLMs assessing the risk of bias with a human reviewer.

Introduction

Systematic reviews are the key initial step in decision-making in healthcare. However, they are costly, require a long time to complete and become outdated, especially in areas of rapidly evolving evidence. Semi-automating systematic reviews and transitioning to living systematic reviews using the best contemporary available evidence are key priority areas of current evidence synthesis. 1–4 Recent advances in artificial intelligence (AI) have ushered in a new era of possibilities in healthcare practice and medical research, 5–7 including evidence synthesis and living systematic reviews. 8 9 By learning from human data analysis patterns (supervision), AI technologies offer the ability to automate, accelerate and enhance the accuracy of a wide array of research tasks, from data collection to analysis and even interpretation. 10

A recent AI advancement, large language models (LLMs) such as Meta AI LLaMA2 and OpenAI’s GPT-4, 11 are considered foundational models pre-trained in a self-supervised manner by leveraging a tremendous amount of free text data. The pre-training process allows them to acquire generic knowledge, and afterward, they can be fine-tuned on downstream tasks. With increasing model size, larger training data sets and longer training time, LLMs evolve emergent abilities such as zero-shot and few-shot in-context learning generalisation and have demonstrated significant capabilities in understanding and generating human-like text and processing data with minimal supervision, which may lead to meaningful participation in a systematic review. 12 13

Risk of bias (RoB) assessment is a significant step in systematic reviews that requires time, introduces inconsistencies and may be amenable to using AI and LLMs. 14 In this exposition, we propose a framework for incorporating LLMs into systematic reviews and employ GPT-4 for RoB assessment in a case study using the Cochrane Collaboration’s Risk Of Bias In Non-randomised Studies of Interventions (ROBINS-I) tool. 15 We chose the ROBINS-I tool for this case study because it is a modern tool that is quite detailed, relatively complicated, and requires a long time to apply, 16 which makes it an ideal candidate to explore whether models such as GPT-4 can improve its consistency and time requirements.

The reporting of this case study adheres to the guidelines of methodological research. 17

Search strategy and study identification

We searched Scopus to identify all systematic reviews (SRs) from the Cochrane Collaboration that cited the original publication of the ROBINS-I tool. 15 We limited our search to SRs conducted by Cochrane in the field of medicine that were fully published. All original non-randomised studies included in the identified SRs were included as long as the ROBINS-I tool was used for their RoB assessment in the SR.

Data entry into ChatGPT

We conducted several pilot tests to determine the most effective method of obtaining RoB assessments using ChatGPT (GPT-4). The initial approach involved directly uploading the study PDFs to GPT-4 via the Code Interpreter tool available to Plus users. However, the tool was unable to interpret the fragmented pieces of text from the PDFs. We then attempted to paste the full text of individual studies in the prompt, however, this was unsuccessful due to the current estimated 2500-word limit for GPT-4 prompts. Finally, we converted the PDF to a Word file and extracted only the Methods and Results sections from each study for RoB assessment because these are the sections on which human reviewers focus for RoB assessments. Prompts used to instruct ChatGPT are presented in the appendix. The processes of data entry and prompt development were done iteratively until data were appropriately uploaded and a sensical output was obtained (ie, these processes were not prespecified). Foreign-language studies were provided in their original language to GPT-4.

Statistical analysis

One reviewer extracted RoB judgements from each Cochrane SR and a second reviewer verified the extraction. We measured the agreement between Cochrane reviewers and GPT-4 comparing the ordinal judgements about RoB using raw per cent agreement, weighted Cohen’s kappa and Kendall’s τ for correlation. The magnitude of agreement based on values of a correlation or kappa coefficient was considered to be slight (0–0.20), fair (0.21–0.40), moderate (0.41–60), substantial (0.61–0.80) and almost perfect (0.81–1.0).

Analysis was conducted using R software package (R Core Team (2022). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria URL https://www.R-project.org ).

Initial screening and inclusion

The initial search yielded 98 SRs, from which 36 provided full ROBINS-I assessment. After deduplicating studies that appeared in multiple SRs, we finalised our sample with 307 unique individual studies ( online supplemental figure; box 1 and box 2 ).

Supplemental material

Agreement between cochrane reviewers and gpt-4.

Agreement measures are summarised in table 1 for each ROBINS-I domain and for overall judgements. Raw per cent agreement was the highest for the domain of ‘Classification of Intervention’. Kendall agreement coefficient was highest for the domains of ‘Participant Selection’, ‘Missing Data’ and ‘Measurement of Outcomes’, suggesting moderate agreement in these domains. Kappa coefficient was low across all domains. Agreement about the overall RoB across domains was fair (61% raw agreement, Kendall coefficient 0.35).

  • View inline

Performance metrics

Framework for incorporating LLM’s in a systematic review

Figure 1 outlines the proposed framework for integrating LLMs into a systematic review workflow. The framework has four domains that relate to establishing a rationale, incorporating LLM in the protocol of the systematic review, execution and reporting.

  • Download figure
  • Open in new tab
  • Download powerpoint

Framework for incorporating large language models in systematic reviews. LLM, large language model; RoB, risk of bias; SR, systematic review.

The first step is to establish the rationale (ie, why LLMs are needed, and whether they are capable of doing this specific task). In the protocol, the LLM model should be described with its version and whether it was off the shelf or used via other tools, applications or interfaces. For example, code interpreters or AI agents can be used. An LLM agent, such as a generative pre-trained transformer (GPT) agent, is a specialised system designed to execute complex, multistep tasks and can adapt to new tools not included in the general model’s training data or recently published tools.

The prompts for LLM need to be iteratively tested and refined and described in the protocol to the extent possible, realising that it will not be possible to prespecify or anticipate every step. The method of data entry (copy/paste vs uploading a file) also needs to be tested and described in the protocol. Metrics of success depend on the task type that is assigned to LLM. We identify five basic task types: selection (eg, of included studies), extraction (eg, of study characteristics and outcomes), judgement (eg, RoB assessment), analysis (quantitative and qualitative) and narration/editing (eg, writing a manuscript, abstract or a lay person or executive summary). The metrics of success and the extent of human interaction and supervision should also be specified in the protocol.

The execution of LLM engagement will likely lead to changes in some of the approaches specified in the protocol, which should be explicitly mentioned as revisions to the protocol. Reporting is the last part of the framework and is vital. The items mentioned above, which are beyond the usual reporting requirements from the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement and its extensions, should all be included in the manuscript. 18 19 Importantly, the AI model and interface used need to be explicitly reported along with a timestamp of when AI was used because the output may vary over time for the same input and prompts. The transparency in reporting and informing peer reviewers and journal editors about the details of using LLMs are critical for the credibility of the systematic review process and subsequent decisions made based on the evidence. The proposed framework is applied to the current case study in table 2 .

Applying the proposed framework to the case study

The current case study suggests an overall fair agreement between Cochrane reviewers and ChatGPT-4 in using ROBINS-I for assessing RoB in non-randomised studies of intervention. This work identifies several challenges for using general utility LLM models, such as handling file types, word token limits and the quality of prompt engineering. Nonetheless, our study provides an assessment of zero-shot performance and a rationale for training RoB-specific systematic review models. The proposed framework is just a starting point since this field is very dynamic.

The current study also provides insight into evaluating inter-rater agreement on ordinal variables. We found that the weighted kappa coefficient was low across all domains which likely reflects the skewed distribution of the ratings. Kappa accounts for agreement occurring by chance, while Kendall’s τ measures the strength and direction of the association between two ranked variables. A recent comparison of reliability coefficients for ordinal rating scales suggested that the differences between these measures can vary at different agreement levels. 20 Thus, using more than one measure is helpful to assess the robustness of results. While our findings suggest the potential of LLMs like GPT-4 to be used in systematic reviews, it is obvious that there is a certain rate of error and that duplication of RoB assessment is needed.

Some limitations of the case study should be mentioned. This study was feasible because of the availability of comprehensive systematic reviews from the Cochrane Collaboration that used the ROBINS-I tool and reported detailed judgements. While their RoB assessment is certainly not a reference standard and can be quite poor for some domains such as confounding, 21 the rigorous and multidomain evaluation conducted by pairs of independent reviewers in these reviews makes them a reasonable comparison for novel LLM application. It is possible also that some systematic reviews used ROBINS-I but did not cite its original paper and were not included in our sample. We also had to use ChatGPT to translate a few studies published in languages other than English, truncate text when it was too lengthy and convert files format, all may have affected RoB judgements.

Practical implications

Given its current capabilities, GPT-4 is arguably a very advanced text-analysing tool. A major advantage is its availability as a universal language model—one model that can perform any language-based extraction, retrieval or even reasoning-based tasks. However, this approach may not be suitable for application in every domain. Sensitive domains like medicine require precise use of language in a consistent manner. LLMs have displayed trends of inconsistency in performance—different output for the same input. LLMs have the propensity to generate favourable answers and to hallucinate. Hallucination is a major threat to the use of LLMs in research. In table 3 , we describe the phenomenon of artificial hallucinations in terms of definition, types and plausible causes. 22–24

The phenomenon of artificial hallucinations: definition, types and causes

Additional applications in systematic reviews can extend to other tasks such as aiding in screening studies, translating foreign-language studies in real-time, data extraction, meta-analysis and even generating decision aids or translational products. 25 However, a human reviewer remains needed as a duplicate independent reviewer.

This exploration of LLMs application in systematic reviews is a step toward integrating AI as a dynamic adjunct in research. The proposed framework, coupled with a case study on RoB assessment, underscores the potential of LLMs to facilitate research tasks. While GPT-4 is not without its limitations, its ability to assist in complex tasks under human supervision makes it a promising tool for assessing RoB in systematic reviews. Considering the agreement level with a human reviewer in the case study, pairing AI with an independent human reviewer remains required at present.

Ethics statements

Patient consent for publication.

Not applicable.

Ethics approval

  • Chu H , et al
  • Sipra QUAR ,
  • Naqvi SAA , et al
  • Ryu AJ , et al
  • Naqvi SAA ,
  • He H , et al
  • Kayaalp ME ,
  • Ollivier M , et al
  • Noorbakhsh-Sabet N ,
  • Zhang Y , et al
  • Ramkumar PN ,
  • Haeberle HS , et al
  • Kelly SE , et al
  • Feng Y , et al
  • van Dijk SHB ,
  • Brusse-Keizer MGJ ,
  • Bucsán CC , et al
  • Touvron H ,
  • Stone K , et al
  • Kolluri S ,
  • Liu R , et al
  • Jardim PSJ ,
  • Ames HM , et al
  • Sterne JA ,
  • Hernán MA ,
  • Reeves BC , et al
  • Jeyaraman MM ,
  • Rabbani R ,
  • Al-Yousif N , et al
  • Liberati A ,
  • Tetzlaff J , et al
  • de Raadt A ,
  • Warrens MJ ,
  • Bosker RJ , et al
  • Thirunavukarasu AJ ,
  • Elangovan K , et al
  • Alkaissi H ,
  • McFarlane SI
  • Blaizot A ,
  • Veettil SK ,
  • Saidoung P , et al

Supplementary materials

Supplementary data.

This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

  • Data supplement 1

Twitter @BasharHasanMD, @M_Hassan_Murad

Contributors MHM and BH conceived this study. BH, SS, MH, MA-K, FF, MF, ZW, IBR, IB and NSR participated in data identification, extraction and analysis. MHM, SS, IBR and IB wrote the first draft. All authors critically revised the manuscript and approved the final version. BH is the guarantor.

Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

Competing interests None declared.

Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Provenance and peer review Not commissioned; externally peer reviewed.

Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Read the full text or download the PDF:

U.S. flag

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

  • Publications
  • Account settings
  • Advanced Search
  • Journal List
  • Ann Indian Acad Neurol
  • v.16(4); Oct-Dec 2013

Design and data analysis case-controlled study in clinical research

Sanjeev v. thomas.

Department of Neurology, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Trivandrum, Kerala, India

Karthik Suresh

1 Department of Pulmonary and Critical Care Medicine, Johns Hopkins University School of Medicine, Louiseville, USA

Geetha Suresh

2 Department of Justice Administration, University of Louisville, Louiseville, USA

Clinicians during their training period and practice are often called upon to conduct studies to explore the association between certain exposures and disease states or interventions and outcomes. More often they need to interpret the results of research data published in the medical literature. Case-control studies are one of the most frequently used study designs for these purposes. This paper explains basic features of case control studies, rationality behind applying case control design with appropriate examples and limitations of this design. Analysis of sensitivity and specificity along with template to calculate various ratios are explained with user friendly tables and calculations in this article. The interpretation of some of the laboratory results requires sound knowledge of the various risk ratios and positive or negative predictive values for correct identification for unbiased analysis. A major advantage of case-control study is that they are small and retrospective and so they are economical than cohort studies and randomized controlled trials.

Introduction

Clinicians think of case-control study when they want to ascertain association between one clinical condition and an exposure or when a researcher wants to compare patients with disease exposed to the risk factors to non-exposed control group. In other words, case-control study compares subjects who have disease or outcome (cases) with subjects who do not have the disease or outcome (controls). Historically, case control studies came into fashion in the early 20 th century, when great interest arose in the role of environmental factors (such as pipe smoke) in the pathogenesis of disease. In the 1950s, case control studies were used to link cigarette smoke and lung cancer. Case-control studies look back in time to compare “what happened” in each group to determine the relationship between the risk factor and disease. The case-control study has important advantages, including cost and ease of deployment. However, it is important to note that a positive relationship between exposure and disease does not imply causality.

At the center of the case-control study is a collection of cases. [ Figure 1 ] This explains why this type of study is often used to study rare diseases, where the prevalence of the disease may not be high enough to permit for a cohort study. A cohort study identifies patients with and without an exposure and then “looks forward” to see whether or not greater numbers of patients with an exposure develop disease.

An external file that holds a picture, illustration, etc.
Object name is AIAN-16-483-g001.jpg

Comparison of cohort and case control studies

For instance, Yang et al . studied antiepileptic drug (AED) associated rashes in Asians in a case-control study.[ 1 ] They collected cases of confirmed anti-epileptic induced severe cutaneous reactions (such as Stevens Johnson syndrome) and then, using appropriate controls, analyzed various exposures (including type of [AED] used) to look for risk factors to developing AED induced skin disease.

Choosing controls is very important aspect of case-control study design. The investigator must weigh the need for the controls to be relevant against the tendency to over match controls such that potential differences may become muted. In general, one may consider three populations: Cases, the relevant control population and the population at large. For the study above, the cases include patients with AED skin disease. In this case, the relevant control population is a group of Asian patients without skin disease. It is important for controls to be relevant: In the anti-epileptic study, it would not be appropriate to choose a population across ethnicities since one of the premises of the paper revolves around particularly susceptibility to AED drug rashes in Asian populations.

One popular method of choosing controls is to choose patients from a geographic population at large. In studying the relationship between non-steroidal anti-inflammatory drugs and Parkinson's disease (PD), Wahner et al . chose a control population from several rural California counties.[ 2 ] There are other methods of choosing controls (using patients without disease admitted to the hospital during the time of study, neighbors of disease positive cases, using mail routes to identify disease negative cases). However, one must be careful not to introduce bias into control selection. For instance, a study that enrolls cases from a clinic population should not use a hospital population as control. Studies looking at geography specific population (e.g., Neurocysticercosis in India) cannot use controls from large studies done in other populations (registries of patients from countries where disease prevalence may be drastically different than in India). In general, geographic clustering is probably the easiest way to choose controls for case-control studies.

Two popular ways of choosing controls include hospitalized patients and patients from the general population. Choosing hospitalized, disease negative patients offers several advantages, including good rates of response (patients admitted to the hospital are generally already being examined and evaluated and often tend to be available to further questioning for a study, compared with the general population, where rates of response may be much lower) and possibly less amnestic bias (patients who are already in the hospital are, by default, being asked to remember details of their presenting illnesses and as such, may more reliably remember details of exposures). However, using hospitalized patients has one large disadvantage; these patients have higher severity of disease since they required hospitalization in the first place. In addition, patients may be hospitalized for disease processes that may share features with diseases under study, thus confounding results.

Using a general population offers the advantage of being a true control group, random in its choosing and without any common features that may confound associations. However, disadvantages include poor response rates and biasing based on geography. Administering long histories and questions regarding exposures are often hard to accomplish in the general population due to the number of people willing (or rather, not willing) to undergo testing. In addition, choosing cases from the general population from particular geographic areas may bias the population toward certain characteristics (such as a socio-economic status) of that geographic population. Consider a study that uses cases from a referral clinic population that draws patients from across socio-economic strata. Using a control group selected from a population from a very affluent or very impoverished area may be problematic unless the socio-economic status is included in the final analysis.

In case-controls studies, cases are usually available before controls. When studying specific diseases, cases are often collected from specialty clinics that see large numbers of patients with a specific disease. Consider for example, the study by Garwood et al .[ 3 ] which looked at patients with established PD and looked for associations between prior amphetamine use and subsequent development various neurologic disorders. Patients in this study were chosen from specialty clinics that see large numbers of patients with certain neurologic disorders. Case definitions are very important when planning to choose cases. For instance, in a hypothetical study aiming to study cases of peripheral neuropathy, will all patients who carry a diagnosis of peripheral neuropathy be included? Or, will only patients with definite electromyography evidence of neuropathy be included? If a disease process with known histopathology is being studied, will tissue diagnosis be required for all cases? More stringent case definitions that require multiple pieces of data to be present may limit the number of cases that can be used in the study. Less stringent criteria (for instance, counting all patients with the diagnosis of “peripheral neuropathy” listed in the chart) may inadvertently choose a group of cases that are too heterogeneous.

The disease history status of the chosen cases must also be decided. Will the cases being chosen have newly diagnosed disease, or will cases of ongoing/longstanding disease also be included? Will decedent cases be included? This is important when looking at exposures in the following fashion: Consider exposure X that is associated with disease Y. Suppose that exposure X negatively affects disease Y such that patients that are X + have more severe disease. Now, a case-control study that used only patients with long-standing or ongoing disease might miss a potential association between X and Y because X + patients, due to their more aggressive course of disease, are no longer alive and therefore were not included in the analysis. If this particular confounding effect is of concern, it can be circumvented by using incident cases only.

Selection bias occurs when the exposure of interest results in more careful screening of a population, thus mimicking an association. The classic example of this phenomenon was noted in the 70s, when certain studies noted a relationship between estrogen use and endometrial cancer. However, on close analysis, it was noted that patients who used estrogen were more likely to experience vaginal bleeding, which in turn is often a cause for close examination by physicians to rule out endometrial cancer. This is often seen with certain drug exposures as well. A drug may produce various symptoms, which lead to closer physician evaluation, thus leading to more disease positive cases. Thus, when analyzed in a retrospective fashion, more of the cases may have a particular exposure only insofar as that particular exposure led to evaluations that resulted in a diagnosis, but without any direct association or causality between the exposure and disease.

One advantage of case-control studies is the ability to study multiple exposures and other risk factors within one study. In addition, the “exposure” being studied can be biochemical in nature. Consider the study, which looked at a genetic variant of a kinase enzyme as a risk factor for development of Alzheimer's disease.[ 4 ] Compare this with the study mentioned earlier by Garwood et al .,[ 3 ] where exposure data was collected by surveys and questionnaires. In this study, the authors drew blood work on cases and controls in order to assess their polymorphism status. Indeed, more than one exposure can be assessed in the same study and with planning, a researcher may look at several variables, including biochemical ones, in single case-control study.

Matching is one of three ways (along with exclusion and statistical adjustment) to adjust for differences. Matching attempts to make sure that the control group is sufficiently similar to the cases group, with respects to variables such as age, sex, etc., Cases and controls should not be matched on variables that will be analyzed for possible associations to disease. Not only should exposure variables not be included, but neither should variables that are closely related to these variables. Lastly, overmatching should be avoided. If the control group is too similar to the cases group, the study may fail to detect the difference even if one exists. In addition, adding matching categories increases expense of the study.

One measure of association derived from case control studies are sensitivity and specificity ratios. These measures are important to a researcher, to understand the correct classification. A good understanding of sensitivity and specificity is essential to understand receiver operating characteristic curve and in distinguishing correct classification of positive exposure and disease with negative exposure and no disease. Table 1 explains a hypothetical example and method of calculation of specificity and sensitivity analysis.

Hypothetical example of sensitivity, specificity and predictive values

An external file that holds a picture, illustration, etc.
Object name is AIAN-16-483-g002.jpg

Interpretation of sensitivity, specificity and predictive values

Sensitivity and specificity are statistical measures of the performance of a two by two classification of cases and controls (sick or healthy) against positives and negatives (exposed or non-exposed).[ 5 ] Sensitivity measures or identifies the proportion of actual positives identified as the percentage of sick people who are correctly identified as sick. Specificity measures or identifies the proportion of negatives identified as the percentage of healthy people who are correctly identified as healthy. Theoretically, optimum prediction aims at 100% sensitivity and specificity with a minimum of margin of error. Table 1 also shows false positive rate, which is referred to as Type I error commonly stated as α “Alpha” is calculated using the following formula: 100 − specificity, which is equal to 100 − 90.80 = 9.20% for Table 1 example. Type 1 error is also known as false positive error is referred to as a false alarm, indicates that a condition is present when it is actually not present. In the above mentioned example, a false positive error indicates the percent falsely identified healthy as sick. The reason why we want Type 1 error to be as minimum as possible is because healthy should not get treatment.

The false negative rate, which is referred to as Type II error commonly stated as β “Beta” is calculated using the following formula: 100 − sensitivity which is equal to 100 − 73.30 = 26.70% for Table 1 example. Type II error is also known as false negative error indicates that a condition is not present when it should have been present. In the above mentioned example, a false negative error indicates percent falsely identified sick as healthy. A Type 1 error unnecessarily treats a healthy, which in turn increases the budget and Type II error would risk the sick, which would act against study objectives. A researcher wants to minimize both errors, which not a simple issue because an effort to decrease one type of error increases the other type of error. The only way to minimize both type of error statistically is by increasing sample size, which may be difficult sometimes not feasible or expensive. If the sample size is too low it lacks precision and it is too large, time and resources will be wasted. Hence, the question is what should be the sample size so that the study has the power to generalize the result obtained from the study. The researcher has to decide whether, the study has enough power to make a judgment of the population from their sample. The researcher has to decide this issue in the process of designing an experiment, how large a sample is needed to enable reliable judgment.

Statistical power is same as sensitivity (73.30%). In this example, large number of false positives and few false negatives indicate the test conducted alone is not the best test to confirm the disease. Higher statistical power increase statistical significance by reducing Type 1 error which increases confidence interval. In other words, larger the power more accurately the study can mirror the behavior of the study population.

The positive predictive values (PPV) or the precision rate is referred to as the proportion of positive test results, which means correct diagnoses. If the test correctly identifies all positive conditions then the PPV would be 100% and negative predictive value (NPV) would be 0. The calculative PPV in Table 1 is 11.8%, which is not large enough to predict cases with test conducted alone. However, the NPV 99.9% indicates the test correctly identifies negative conditions.

Clinical interpretation of a test

In a sample, there are two groups those who have the disease and those who do not have the disease. A test designed to detect that disease can have two results a positive result that states that the disease is present and a negative result that states that the disease is absent. In an ideal situation, we would want the test to be positive for all persons who have the disease and test to be negative for all persons who do not have the disease. Unfortunately, reality is often far from ideal. The clinician who had ordered the test has the result as positive or negative. What conclusion can he or she make about the disease status for his patient? The first step would be to examine the reliability of the test in statistical terms. (1) What is the sensitivity of the test? (2) What is the specificity of the test? The second step is to examine it applicability to his patient. (3) What is the PPV of the test? (4) What is the NPV of the test?

Suppose the test result had come as positive. In this example the test has a sensitivity of 73.3% and specificity of 90.8%. This test is capable of detecting the disease status in 73% of cases only. It has a false positivity of 9.2%. The PPV of the test is 11.8%. In other words, there is a good possibility that the test result is false positive and the person does not have the disease. We need to look at other test results and the clinical situation. Suppose the PPV of this test was close to 80 or 90%, one could conclude that most likely the person has the disease state if the test result is positive.

Suppose the test result had come as negative. The NPV of this test is 99.9%, which means this test gave a negative result in a patient with the disease only very rarely. Hence, there is only 0.1% possibility that the person who tested negative has in fact the disease. Probably no further tests are required unless the clinical suspicion is very high.

It is very important how the clinician interprets the result of a test. The usefulness of a positive result or negative result depends upon the PPV or NPV of the test respectively. A screening test should have high sensitivity and high PPV. A confirmatory test should have high specificity and high NPV.

Case control method is most efficient, for the study of rare diseases and most common diseases. Other measures of association from case control studies are calculation of odds ratio (OR) and risk ratio which is presented in Table 2 .

Different ratio calculation templates with sample calculation

An external file that holds a picture, illustration, etc.
Object name is AIAN-16-483-g003.jpg

Absolute risk means the probability of an event occurring and are not compared with any other type of risk. Absolute risk is expressed as a ratio or percent. In the example, absolute risk reduction indicates 27.37% decline in risk. Relative risk (RR) on the other hand compares the risk among exposed and non-exposed. In the example provided in Table 2 , the non-exposed control group is 69.93% less likely compared to exposed cases. Reader should keep in mind that RR does not mean increase in risk. This means that while a 100% likely risk among those exposed cases, unexposed control is less likely by 69.93%. RR does not explain actual risk but is expressed as relative increase or decrease in risk of exposed compared to non-exposed.

OR help the researcher to conclude whether the odds of a certain event or outcome are same for two groups. It calculates the odds of a health outcome when exposed compared to non-exposed. In our example an OR of. 207 can be interpreted as the non-exposed group is less likely to experience the event compared to the exposed group. If the OR is greater than 1 (example 1.11) means that the exposed are 1.11 times more likely to be riskier than the non-exposed.

Event rate for cases (E) and controls (C) in biostatistics explains how event ratio is a measure of how often a particular statistical exposure results in occurrence of disease within the experimental group (cases) of an experiment. This value in our example is 11.76%. This value or percent explains the extent of risk to patients exposed, compared with the non-exposed.

The statistical tests that can be used for ascertain an association depends upon the variable characteristics also. If the researcher wants to find the association between two categorical variables (e.g., a positive versus negative test result and disease state expressed as present or absent), Cochran-Armitage test, which is same as Pearson Chi-squared test can be used. When the objective is to find the association between two interval or ratio level (continuous) variables, correlation and regression analysis can be performed. In order to evaluate statistical significant difference between the means of cases and control, a test of group difference can be performed. If the researcher wants to find statically significant difference among means of more than two groups, analysis of variance can be performed. A detailed explanation and how to calculate various statistical tests will be published in later issues. The success of the research directly and indirectly depends on how the following biases or systematic errors, are controlled.

When selecting cases and controls, based on exposed or not-exposed factors, the ability of subjects to recall information on exposure is collected retrospectively and often forms the basis for recall bias. Recall bias is a methodological issue. Problems of recall method are: Limitations in human ability to recall and cases may remember their exposure with more accuracy than the controls. Other possible bias is the selection bias. In case-control studies, the cases and controls are selected from the same inherited characteristics. For instance, cases collected from referral clinics often exposed to selection bias cases. If selection bias is not controlled, the findings of association, most likely may be due to of chance resulting from the study design. Another possible bias is information bias, which arises because of misclassification of the level of exposure or misclassification of disease or other symptoms of outcome itself.

Case control studies are good for studying rare diseases, but they are not generally used to study rare exposures. As Kaelin and Bayona explains[ 6 ] if a researcher want to study the risk of asthma from working in a nuclear submarine shipyard, a case control study may not be a best option because a very small proportion of people with asthma might be exposed. Similarly, case-control studies cannot be the best option to study multiple diseases or conditions because the selection of the control group may not be comparable for multiple disease or conditions selected. The major advantage of case-control study is that they are small and retrospective and so they are economical than cohort studies and randomized controlled trials.

Source of Support: Nil

Conflict of Interest: Nil

Log in using your username and password

  • Search More Search for this keyword Advanced search
  • Latest content
  • Current issue
  • Hosted content
  • BMJ Journals More You are viewing from: Google Indexer

You are here

  • Online First
  • Proton pump inhibitors and the risk of inflammatory bowel disease: a Mendelian randomisation study
  • Article Text
  • Article info
  • Citation Tools
  • Rapid Responses
  • Article metrics

Download PDF

  • Hongjin An 1 ,
  • Min Zhong 1 ,
  • http://orcid.org/0000-0002-5736-1283 Huatian Gan 2 , 3
  • 1 Department of Gastroenterology and Hepatology, West China Hospital, Sichuan University , Chengdu , China
  • 2 Department of Geriatrics and National Clinical Research Center for Geriatrics, West China Hospital, Sichuan University , Chengdu , China
  • 3 Department of Gastroenterology and Laboratory of Inflammatory Bowel Disease, the Center for Inflammatory Bowel Disease, Clinical Institute of Inflammation and Immunology, Frontiers Science Center for Disease-related Molecular Network, West China Hospital, Sichuan University , Chengdu , China
  • Correspondence to Dr Huatian Gan, West China Hospital of Sichuan University, Chengdu, Sichuan, China; ganhuatian123{at}163.com

https://doi.org/10.1136/gutjnl-2024-331904

Statistics from Altmetric.com

Request permissions.

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

  • INFLAMMATORY BOWEL DISEASE

We read with great interest the population-based cohort study by Abrahami D et al , 1 in which they found that the use of proton pump inhibitors (PPIs) was not associated with an increased risk of inflammatory bowel disease (IBD). However, the assessment of causality in observational studies is often challenging due to the presence of multiple confounding factors. The existence of a causal relationship between PPIs and IBD remains unclear at present. Mendelian randomisation (MR) is a method of generating more reliable evidence using exposure-related genetic variants to assess causality, limiting the bias caused by confounders. 2 Therefore, we used a two-sample MR analysis to investigate the association between the use of PPIs and IBD including Crohn’s disease (CD) and ulcerative colitis (UC).

Supplemental material

Here, we mainly used the inverse-variance weighted 8 method for MR analysis with weighted median, 9 MR-Egger 10 and MR-PRESSO 5 as complementary approaches. Furthermore, we applied a series of sensitivity analyses to ensure the robustness of our results, with Cochran’s Q test to assess heterogeneity and the intercept of an MR-Egger regression to assess horizontal pleiotropy. The genetic prediction of omeprazole, esomeprazole, lansoprazole and rabeprazole use, as depicted in figure 1 , demonstrated no significant association with an increased risk of IBD after excluding pleiotropic SNPs (omeprazole, OR, 1.05; 95% CI, 0.88 to 1.25; p=0.587; esomeprazole, OR, 0.99; 95% CI, 0.92 to 1.07; p=0.865; lansoprazole, OR, 1.06; 95% CI, 0.89 to 1.26; p=0.537; and rabeprazole, OR, 1.00; 95% CI, 0.95 to 1.04; p=0.862). The IBD subtype analyses also did not reveal any evidence of an increased risk of CD or UC associated with the use of PPIs ( figure 1 ). These findings were robustly confirmed through complementary approaches employing rigorous methodologies that consistently yielded similar point estimates ( figure 1 ). Further sensitivity analyses showed the absence of heterogeneity (All P heterogeneity >0.05) and pleiotropy (All P pleiotropy >0.05), again demonstrating the robustness of the conclusions ( figure 1 ).

  • Download figure
  • Open in new tab
  • Download powerpoint

Mendelian randomisation estimates the associations between the use of different types of proton pump inhibitors and inflammatory bowel disease. IBD, inflammatory bowel disease; CD, Crohn’s disease; UC, ulcerative colitis; PPIs, proton pump inhibitors; IVW, inverse-variance weighted; MR, Mendelian randomisation.

In conclusion, the MR results corroborate Abrahami D et al ’s findings that PPIs were not associated with an increased risk of IBD. Nonetheless, further research is needed to elucidate the effects of more types, drug dosage, frequency and duration on IBD.

Ethics statements

Patient consent for publication.

Not applicable.

Ethics approval

  • Abrahami D ,
  • Pradhan R ,
  • Yin H , et al
  • Kathiresan S
  • Fang H , et al
  • van Sommeren S ,
  • Huang H , et al
  • Verbanck M ,
  • Neale B , et al
  • Tilling K ,
  • Davey Smith G
  • Brion M-JA ,
  • Shakhbazov K ,
  • Visscher PM
  • Burgess S ,
  • Timpson NJ , et al
  • Davey Smith G ,
  • Haycock PC , et al

Supplementary materials

Supplementary data.

This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

  • Data supplement 1

HA and MZ contributed equally.

Contributors All authors conceived and designed the study. HA and MZ did the statistical analyses and wrote the manuscript. HG revised the manuscript and is the guarantor. HA and MZ have contributed equally to this study.

Funding The present work was supported by the National Natural Science Foundation of China (No. 82070560) and 1.3.5 Project for Disciplines of Excellence, West China Hospital, Sichuan (No. ZYGD23013).

Competing interests None declared.

Provenance and peer review Not commissioned; externally peer reviewed.

Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Read the full text or download the PDF:

IMAGES

  1. What is a Case Control Study?

    case control study is a type of

  2. PPT

    case control study is a type of

  3. PPT

    case control study is a type of

  4. elements of case control study

    case control study is a type of

  5. how do case control studies work

    case control study is a type of

  6. PPT

    case control study is a type of

VIDEO

  1. Case Control Study (Lecture

  2. Case Study Part 3: Developing or Selecting the Case

  3. Case Studies

  4. Case Studies

  5. Case Control Study Part 1

  6. Case-control study design

COMMENTS

  1. What Is a Case-Control Study?

    Case-control studies are a type of observational study often used in fields like medical research, environmental health, or epidemiology. While most observational studies are qualitative in nature, case-control studies can also be quantitative, and they often are in healthcare settings.

  2. Case Control Studies

    A case-control study is a type of observational study commonly used to look at factors associated with diseases or outcomes. [1] The case-control study starts with a group of cases, which are the individuals who have the outcome of interest.

  3. Case Control Studies

    A case-control study is a type of observational study commonly used to look at factors associated with diseases or outcomes. The case-control study starts with a group of cases, which are the individuals who have the outcome of interest.

  4. Case-control study

    A case-control study (also known as case-referent study) is a type of observational study in which two existing groups differing in outcome are identified and compared on the basis of some supposed causal attribute.

  5. Case Control Study: Definition, Benefits & Examples

    A case control study is a retrospective, observational study that compares two existing groups. Researchers form these groups based on the existence of a condition in the case group and the lack of that condition in the control group.

  6. Case Control Study: Definition & Examples

    A case-control study is a research method where two groups of people are compared - those with the condition (cases) and those without (controls). By looking at their past, researchers try to identify what factors might have contributed to the condition in the 'case' group. Definition

  7. A Practical Overview of Case-Control Studies in Clinical Practice

    Case-control studies are one of the major observational study designs for performing clinical research. The advantages of these study designs over other study designs are that they are relatively quick to perform, economical, and easy to design and implement.

  8. Case-control study in medical research: Uses and limitations

    A case-control study is a type of medical research investigation often used to help determine the cause of a disease, particularly when investigating a disease outbreak or rare condition.

  9. Case-Control Studies

    Case-Control Studies Introduction Cohort studies have an intuitive logic to them, but they can be very problematic when one is investigating outcomes that only occur in a small fraction of exposed and unexposed individuals. They can also be problematic when it is expensive or very difficult to obtain exposure information from a cohort.

  10. Research Design: Case-Control Studies

    Case-control studies are observational studies in which cases are subjects who have a characteristic of interest, such as a clinical diagnosis, and controls are (usually) matched subjects who do not have that characteristic.

  11. Methodology Series Module 2: Case-control Studies

    Case-Control study design is a type of observational study. In this design, participants are selected for the study based on their outcome status. Thus, some participants have the outcome of interest (referred to as cases), whereas others do not have the outcome of interest (referred to as controls).

  12. A Practical Overview of Case-Control Studies in Clinical Practice

    Importance of Sample Size Selection in Case-Control Studies. In any type of observational study, using an appropriate sample size is a very crucial step because having the correct number of samples ensures the reliability of the study findings. Many studies have an insufficient number of samples, and in those studies it is difficult to ...

  13. Case-control and Cohort studies: A brief overview

    Introduction. Case-control and cohort studies are observational studies that lie near the middle of the hierarchy of evidence. These types of studies, along with randomised controlled trials, constitute analytical studies, whereas case reports and case series define descriptive studies (1). Although these studies are not ranked as highly as ...

  14. Research Guides: Study Design 101: Case Control Study

    Case control studies are also known as "retrospective studies" and "case-referent studies." Advantages Good for studying rare conditions or diseases Less time needed to conduct the study because the condition or disease has already occurred Lets you simultaneously look at multiple risk factors Useful as initial studies to establish an association

  15. Case-control study

    case-control study, in epidemiology, observational (nonexperimental) study design used to ascertain information on differences in suspected exposures and outcomes between individuals with a disease of interest (cases) and comparable individuals who do not have the disease (controls). Analysis yields an odds ratio (OR) that reflects the relative probabilities of exposure in the two populations.

  16. Case-Control Study- Definition, Steps, Advantages, Limitations

    A case-control study (also known as a case-referent study) is a type of observational study in which two existing groups differing in outcome are identified and compared on the basis of some supposed causal attribute. It is designed to help determine if an exposure is associated with an outcome (i.e., disease or condition of interest).

  17. LibGuides: Quantitative study designs: Case Control

    Case Control. In a Case-Control study there are two groups of people: one has a health issue (Case group), and this group is "matched" to a Control group without the health issue based on characteristics like age, gender, occupation. In this study type, we can look back in the patient's histories to look for exposure to risk factors that ...

  18. Case-control study: comparative studies

    A case-control study is a type of observational study. It looks at 2 sets of participants. One group has the condition you are interested in (the cases) and one group does not have it (the...

  19. Case-Control Study: Definition, Types and Examples

    One of the most widespread types of study nowadays is a case-control study. The purpose of the case-control study is to determine the causes of the onset and spread of the disease. In the case-control studies, the probability of the existence of a causal relationship is based not on the different incidence of morbidity, but on the different prevalence (occurrence) of the predicted risk factor ...

  20. Gout and incidence of 12 cardiovascular diseases: a case-control study

    In this matched case-control study, we used linked primary and secondary electronic health records from the UK Clinical Practice Research Datalink to assemble a cohort of individuals with a first-time diagnosis of gout between Jan 1, 2000 and Dec 31, 2017, who were aged 80 years or younger at diagnosis, and free of cardiovascular diseases up to 12 months after diagnosis.

  21. What types of studies are there?

    The main types of studies are randomized controlled trials (RCTs), cohort studies, case-control studies and qualitative studies. Go to: Randomized controlled trials If you want to know how effective a treatment or diagnostic test is, randomized trials provide the most reliable answers.

  22. Biomarker Changes during 20 Years Preceding Alzheimer's Disease

    Nested Case-Control Approach. Our study required that participants be observed for more than 15 years but not more than 20 years. ... the type of biomarker tests and their accuracy changed in ...

  23. Full article: Cancer, Diabetes, Survival and Glycemic Control: A Large

    The relationship between cancer and diabetes also has been examined in many case-control studies of individual solid and hematologic cancers: breast, lung, prostate, colorectal, pancreatic, gastroesophageal, uterine/ovarian, melanoma, lymphoma, leukemia, squamous cell and neuroendocrine cancers [Citation 6-17]. With the exception of ...

  24. JMIR Public Health and Surveillance

    Background: Maternal smoking during pregnancy (MSDP) is a known risk factor for offspring developing chronic obstructive pulmonary disease (COPD), but the underlying mechanism remains unclear. Objective: This study aimed to explore whether the increased COPD risk associated with MSDP could be attributed to tobacco dependence (TD). Methods: This case-control study used data from the nationwide ...

  25. Alteration Ocular Motility in Retinitis Pigmentosa: Case-Control Study

    In fact, in the control group, the ocular motility disturbances concerning accommodation-convergence were characterized by mild and unremarkable defects, whereas they were significant in the group of RP patients. Sensory strabismus is a type of strabismus that is established by organic or functional alteration of the visual system.

  26. Association of Obesity With Type 2 Diabetes Mellitus: A ...

    This study aimed to assess the association of obesity with T2DM among the patients visiting the Medicine Department of Ayub Teaching Hospital, Abbottabad, Pakistan. Method This hospital-based, unmatched case-control study was conducted from March 2022 to September 2022. A total of 200 patients (age ≥ 18) (100 cases and 100 controls) were ...

  27. The social anatomy of climate change denial in the United States

    Using data from Twitter (now X), this study deploys artificial intelligence (AI) and network analysis to map and profile climate change denialism across the United States. We estimate that 14.8% ...

  28. Integrating large language models in systematic reviews: a framework

    Large language models (LLMs) may facilitate and expedite systematic reviews, although the approach to integrate LLMs in the review process is unclear. This study evaluates GPT-4 agreement with human reviewers in assessing the risk of bias using the Risk Of Bias In Non-randomised Studies of Interventions (ROBINS-I) tool and proposes a framework for integrating LLMs into systematic reviews.

  29. Design and data analysis case-controlled study in clinical research

    At the center of the case-control study is a collection of cases. [ Figure 1] This explains why this type of study is often used to study rare diseases, where the prevalence of the disease may not be high enough to permit for a cohort study.

  30. Proton pump inhibitors and the risk of inflammatory bowel disease: a

    We read with great interest the population-based cohort study by Abrahami D et al ,1 in which they found that the use of proton pump inhibitors (PPIs) was not associated with an increased risk of inflammatory bowel disease (IBD). However, the assessment of causality in observational studies is often challenging due to the presence of multiple confounding factors. The existence of a causal ...