1
← Back to quizzesFree quiz
Epidemiologic Causality Principles
Understanding whether an observed association between an exposure and a health outcome truly reflects a causal relationship is a core skill for clinicians, public‑health professionals, and…
0 / 21
Score: 0%
2
In the context of epidemiologic studies, what does a 'dose‑effect gradient' indicate about a potential causal relationship?
3
A researcher finds that smokers have higher rates of both lung cancer and heart disease. Which causal criterion is most challenged by this observation?
4
Which type of bias arises when the probability of being included in a study differs between exposed and unexposed groups?
5
When an association is observed only at the ecological (population) level but not at the individual level, which problem does this illustrate?
6
A study shows a relative risk of 3.0 for disease X among exposed individuals. Which causal criterion does this primarily address?
7
Which of the following scenarios exemplifies a causal association conditional on two factors acting together?
8
Why are Koch's postulates not universally applicable to non‑infectious diseases?
9
In an epidemiologic investigation, which of the following would most directly demonstrate the plausibility criterion?
10
A researcher adjusts for age, sex, and smoking status in a multivariable model. Which type of bias is this attempt to control?
11
Which causal criterion is most directly assessed by replicating a study in different populations and settings?
12
In a case‑control study, which measure of association is most appropriate for evaluating a potential causal link?
13
Why might a long latency period between exposure and disease onset hinder causal inference?
14
Which of the following best illustrates an interaction (synergy) between two risk factors?
15
When an association is strong but biologically implausible, which causal criterion is violated?
16
A study reports that after removing exposure to a suspected toxin, disease incidence dropped sharply. Which causal evidence does this provide?
17
Which of the following best describes a 'direct causal association'?
18
Why is it said that no statistical method can definitively separate causal from non‑causal associations?
19
In evaluating causality, which criterion assesses whether the same association is observed across different study designs?
20
Which bias type is most likely when exposure status is misclassified due to inaccurate questionnaires?
21
A researcher observes that a factor is present in every case of a disease but also in many healthy individuals. Which causal criterion does this observation weaken?
Epidemiologic Causality Principles
Understanding whether an observed association between an exposure and a health outcome truly reflects a causal relationship is a core skill for clinicians, public‑health professionals, and researchers. This course unpacks the classic criteria used to evaluate causality, highlights common pitfalls such as bias and confounding, and illustrates each principle with clear examples drawn from the quiz questions.
Why Statistical Significance Is Not Sufficient for Causal Inference
Statistical significance tells us that an observed association is unlikely to be due to random chance alone (usually p < 0.05). However, significance does **not** eliminate other threats to validity:
- Confounding: An unmeasured third variable may be responsible for the observed link.
- Bias: Systematic errors in study design, data collection, or analysis can create spurious associations.
- Measurement error: Inaccurate exposure or outcome assessment can distort the true relationship.
For example, noticing that people who wear red hats are healthier might simply reflect that they spend more time outdoors (sunlight exposure), not the hat itself. Therefore, after establishing statistical significance, researchers must systematically assess these alternative explanations.
Key Causal Criteria (Bradford Hill Framework)
The Bradford Hill criteria provide a structured checklist for evaluating causality. While not all criteria need to be met, the more that are satisfied, the stronger the inference.
1. Strength of Association
A strong association (e.g., a relative risk of 3.0) suggests that the exposure may be a true cause, because large effects are less likely to be explained by confounding or bias.
2. Consistency
Repeated observations of the same association across different populations, study designs, and settings increase confidence in a causal link.
3. Specificity
Specificity refers to an exposure leading to a single disease outcome. When an exposure (like smoking) is linked to many diseases, the specificity criterion is weakened, challenging the causal interpretation.
4. Temporality
The exposure must precede the outcome. This is a non‑negotiable requirement; without it, reverse causation cannot be ruled out.
5. Dose‑Response Gradient
A dose‑effect gradient means that as exposure intensity increases, the risk of disease rises proportionally. This pattern, such as a linear increase in incidence with higher exposure levels, bolsters the argument for causality.
6. Biological Plausibility
There should be a credible biological mechanism linking exposure to disease. While not mandatory, plausibility strengthens the overall case.
7. Coherence, Analogy, and Experiment
These additional criteria consider whether the association fits with existing knowledge, resembles known causal relationships, or can be demonstrated experimentally.
Common Sources of Error in Causal Assessment
Selection Bias
Occurs when the probability of participating in a study differs between exposed and unexposed groups, potentially distorting the observed association.
Information (Measurement) Bias
Arises from systematic errors in how exposure or outcome data are collected, leading to misclassification.
Confounding
A confounder is a third variable associated with both the exposure and the outcome that can create a false association if not properly controlled.
Ecological Fallacy
When an association is observed at the group (population) level but not at the individual level, it may reflect an ecological fallacy—mistaking group‑level correlations for individual‑level causation.
Special Considerations for Non‑Infectious Diseases
Koch’s postulates, originally devised for infectious agents, are not universally applicable to chronic, non‑infectious conditions because such diseases often have multifactorial, non‑microbial etiologies. Modern epidemiology therefore relies on statistical and methodological criteria rather than strict laboratory proof.
Complex Interactions: Conditional Causality
Some diseases arise only when multiple factors act together. For instance, a disease may occur only when both factor A and factor B are present, illustrating a conditional (or synergistic) causal relationship. Recognizing these interactions is essential for accurate risk assessment and prevention strategies.
Applying the Concepts: Practice Questions
Review the quiz items below to see how each principle is tested.
- Question 1: Highlights why statistical significance alone is insufficient (confounding/bias).
- Question 2: Demonstrates the dose‑effect gradient concept.
- Question 3: Shows how lack of specificity challenges causal inference when an exposure is linked to multiple outcomes.
- Question 4: Identifies selection bias as a threat to validity.
- Question 5: Illustrates ecological fallacy.
- Question 6: Connects a relative risk of 3.0 to the strength of association criterion.
- Question 7: Explores conditional causality with two interacting factors.
- Question 8: Explains why Koch’s postulates are limited for non‑infectious diseases.
Key Take‑aways
- Statistical significance is a starting point, not proof of causality.
- Assess each Bradford Hill criterion systematically.
- Identify and control for selection bias, information bias, and confounding.
- Beware of ecological fallacy when interpreting population‑level data.
- Recognize that non‑infectious diseases often require multifactorial causal models.
By mastering these principles, you will be better equipped to design robust epidemiologic studies, critically appraise research findings, and make evidence‑based decisions in clinical and public‑health practice.
