Chapter 6: Association and Causality

Cause

variable that makes a health event occur (without it the disease would not have occurred) → modifies human health in some way

Examples:

• Microbe

• Behavior 

• Exposure to environmental factor

Association

linkage between or among variables

Exposure

contact with factors linked to adverse health outcomes

Deterministic

Cause/agent is always followed by some sort of health event

• Cause = independent variable

• Health outcome = dependant variable

• employs necessary and sufficient causes

Necessary

cause must be present in every case of disease

a factor [X] whose presence is required for the occurence of the effect [Y]

ex. you must be exposed to the flu virus to contract the flu

Sufficient

a factor [X] whose presence is sufficient by itself to produce the effect [Y]

ex. just because you are exposed to the flu virus doesn’t mean you are certain to contract the flu

Sufficient and necessary

Both X and Y are always present together

Everytime someone is exposed to X they contract Y

Nothing but X is needed to cause Y

(uncommon)

Sufficient but not necessary

X may or may not be preset when Y occurs

X is one of the causes of Y [however] there are other ways Y can come about

Necessary but not sufficient

X must be present for Y to occur, but X may be present without Y occuring

aka I can be exposed to X (infectious disease) but that doesn’t guarantee that I will contract the disease

Not sufficient nor necessary

X may or may not be present when Y occurs

X is a contributory cause of Y (often multiple contributory causes to disease of interest)

Type of cause: A sedentary lifestyle can lead to coronary heart disease

Neither necessary nor sufficient

Type of cause: The varicella-zoster virus [X] causes chicken pox [Y]

Necessary but not sufficient

Type of cause: A medication [X] resulted in atopic dermatitis [Y]

Sufficient but not necessary

Multifactorial cause aka sufficient component cause model aka causal pie model

Necessary + component causes = sufficient cause complex (multiple complexes can exist because variable component causes)

Necessary causes (causal pie model)

present in every example, common piece of the pie

A factor [X] whose presence is required for the occurrence of the effect [Y]

Component causes (causal pie model)

additional variables, unique to situation

Probabilistic causality

incorporates some element of randomness (stochastic causes)

cause is associated with the increased probability that an effect will happen

Association

 “is a particular exposure causally associated with a given outcome?”

Positive association

increasing one variable means other variable increases

negative associatoin

increasing one variable means other variable decreases

Spurious association

association that appears due to improper comparison

ex. polio and spongy tar

Bradford Hill criteria for causality

1. Strength

2. temportality

3. consistency

4. biological plausibility

5. biological gradient

6. coherence

7. experiment

8. specificity