Causation Concepts in Epidemiology

Causation

• Causation is defined as "anything producing an effect or a result"
• Direct causation: A causes B without intermediate effects (rare)
• Indirect causation: A causes B, but with intermediate effects
• The cause or causative factor must precede the outcome in time

Nature of Cause

• Rothman and Greenland (2005) developed the "multiple causation" concept
• They used the "causal pie" metaphor, where disease occurs if a circle (pie) is completed by one or more component parts ("pies").
• Each part is a cause for the disease.
• This suggests a disease may occur due to a combination of factors in various situations.

Types of Causes

• Necessary cause: Its presence is required for the disease to occur; it cannot occur in the absence of this cause.
• Sufficient cause: A complete causal mechanism; a set of minimal conditions and events that inevitably produce disease.
• Minimal conditions mean all conditions/events are necessary.
• Each cause that contributes to the pie is a component cause.

Different Component Causes

• Different component causes contribute to disease causation.
• For some diseases, there may be no necessary causes, even with multiple sufficient causes.
• Some causes are sufficient but not necessary, while others are necessary but not sufficient.

Causal Inference

• Most associations in scientific oral health literature are interpreted as causal.
• Potential causal associations should only be explored if the association is considered to be true.

Hill's Criteria

• Useful for establishing epidemiological evidence of a causal relationship between a presumed cause and an effect
• Strength: Stronger association, stronger argument for causation (large relative risk).
• Consistency: Repeated observation of association across different persons, places, circumstances, and times
• Specificity: Association limited to a particular exposure and outcome, with no association with other diseases (one cause, one outcome).
• Temporality: Cause must precede the outcome.
• Biological gradient: A dose-response curve showing an increase in exposure leading to increased disease risk. A stronger dose-response strengthens the causal argument.
• Plausibility: Association should be biologically plausible (understandable in biological terms).
• Coherence: Cause-and-effect interpretation shouldn't conflict with existing knowledge of the disease's natural history and biology.
• Experiment: Experimental or semi-experimental evidence can support a causal link.
• Analogy: Evidence from similar exposures or diseases can strengthen the analogy.

Epidemiologic Analysis

• The use of comparisons is quantified by measures of association, like risk ratios, rate ratios, and odds ratios.

Risk

• Risk is the probability of a disease or other health outcome occurring during a specified period.
• Measures of risk include: absolute measures (incidence rate, attack rate, prevalence rate), relative risk (RR), and attributable risk percentage (AR).

Interpretation of Risk Measures

• Relative Risk (RR): If RR = 1, risk of disease is the same in exposed vs unexposed groups. If RR > 1, exposure is a risk factor. If RR < 1, exposure is a protective factor.
• Odds Ratio (OR): Another association measure, especially useful in case-control studies. If OR > 1, indicates a stronger association. If OR = 1, exposure and disease are unrelated. If OR < 1, indicates no association.

Explore the fundamental concepts of causation in epidemiology, including direct and indirect causation. Learn about necessary and sufficient causes, as well as the 'multiple causation' theory by Rothman and Greenland. This quiz will enhance your understanding of how various factors contribute to disease.