What is the multifactorial nature of disease?
- Most diseases caused by multiple factors (not a single cause).
- Risk factors ↑ disease risk; protective factors ↓ disease risk.
- Factors are called disease determinants and vary by disease.
- Many chronic diseases share common determinants like poor diet, physical inactivity, smoking, and stress.
- Targeting these shared determinants can prevent multiple diseases.
What are disease determinants?
Factors that influence the occurrence and progression of diseases, including:
- Genetic factors
- Biological/physiological factors
- Lifestyle factors
- Psychosocial factors
- Sociodemographic factors
- Wider environmental factors
They can vary for different diseases.
What types of diseases are particularly affected by common determinants?
Chronic diseases such as:
- Cardiovascular diseases
- Cerebrovascular diseases
- Type 2 diabetes
- Cancer
- Neurodegenerative diseases
Define confounding.
A variable that distorts or biases the observed relationship between an exposure and an outcome.
What are the implications of confounding?
- It can create associations
- It can mask associations
- It can overestimate or underestimate associations
What is a spurious association?
False/misleading relationship between 2 variables that appears to be causal but is actually caused by a third, hidden factor.
What criteria must a potential confounder meet?
- It must be associated with the outcome of interest
- It must be associated with the exposure of interest
- It should not lie in the causal pathway between exposure and outcome.
True or False: A factor that is associated with both the outcome and the exposure but lies in the causal pathway is considered a confounder.
False
Such a factor is called a mediator.
What is the first step in dealing with confounding?
Identify potential confounders.
What does statistical adjustment for confounding aim to do?
Eliminate or reduce the confounding effect of potential confounders in any exposure-outcome association.
What is the standard method for adjusting for confounding?
Using regression analysis.
What is the purpose of stratification in confounding adjustment?
To control for confounding by examining exposure–outcome relationship within levels (strata) of a confounder.
Fill in the blank: Crude estimates are simply estimates for exposure-outcome associations (Odds ratio, mean difference etc.) while adjusted estimates are estimates for exposure-outcome associations after applying statistical adjustment for any _____ variable.
potential confounder.
What would an age-adjusted estimate indicate in a study?
The measure of association between smoking and dementia if everyone had the same age.
What is effect modification?
- Occurs when effect of an exposure on an outcome differs depending on level of another variable.
- A third variable changes the strength or direction of the effect
What is the Odds Ratio for lung cancer comparing high alcohol consumers to low consumers?
2.50 (95% CI: 2.30 - 2.90).
What would indicate that smoking is a potential confounder in the association between alcohol consumption and lung cancer?
If smoking is associated with both lung cancer and alcohol consumption, and does not lie in the causal pathway.
In the context of confounding, what does it mean to adjust for a confounder?
To recalculate estimates for the exposure-outcome association while controlling for the confounder.
What happens if the exposure-outcome association is different in the categories of the confounder?
The estimates cannot be combined and are presented separately.
What percentage of smokers were included in the study?
41%
What percentage of non-smokers were included in the study?
59%
What is the odds ratio for smokers in the study?
1.26
What is the odds ratio for non-smokers in the study?
1.09
What is the combined (adjusted) odds ratio estimate?
1.16
-
Section 1: Introduction to epidemiology & PH33
-
Section 2: Measures of disease frequency & mortality66
-
Section 3: Measure of association with binary outcomes I: Relative risk43
-
Section 4: measures of impact3
-
Section 5: observational study designs81
-
Section 6: Types of study design: Interventional studies42
-
Section 7: sampling, random error, statisical inference, selection bias & different sampling methods99
-
Section 8A: Systematic error in research II: information bias36
-
Section 8B: Sensitivity and specificity of screening and diagnostic tools32
-
Section 10: The Chernobyl and Fukushima Daiichi nuclear disasters17
-
Section 11B & 11C: Systematic Review and Meta-Analysis & literature search27
-
Section 12: Causality in research: association vs causation19
-
Section 13: wider/social determinants of health14
-
Section 14: Principles of disease prevention4
-
Section 15: Global Health Monitoring and Public Health Surveillance29
-
Section 9: confounding47
-
Section: internal & external study validity23
