Establishing Causality

Question 1

Could the association be observed by chance?

Answer 1

• if there is less than a 5% chance that the result obtained is due to chance, we accept it as a “real” association, and we can expect to be correct 95% of the time
• there is never a zero percent chance that a result happened by chance

Question 2

Could the association be due to bias?

Answer 2

• Bias refers to systematic errors made in data collection

Question 3

Is the relationship due to confounding? - def of Confounding

Answer 3

• is the distortion of the association between an exposure and health outcome by an extraneous, third variable

Question 4

To whom does the association apply? (3)

Answer 4

• associations are identified using data that is reflective of the population as a whole
• Non-random samples yield results that reflect the characteristics of the sample alone. Small samples cannot be generalized to the entire population with any confidence
• Samples that include a specific subset of the population (women aged 18-25, for example) cannot reliably tell us anything about other subsets of the population

Question 5

Direct Causation

Answer 5

• wherein a factor leads directly to the health event of interest
• is characteristic of many infectious diseases, but is atypical of non-infectious illness, except for things like traumatic injury

Question 6

Indirect Causation

Answer 6

• a precipitating factor requires the “assistance” of one or more intermediate steps to induce disease
• chronic diseases

Question 7

Necessary

Answer 7

• to produce the health outcome in question

- A necessary factor is one that must be present for the disease to occur – like varicella zoster for chicken pox

Question 8

Sufficient

Answer 8

• to do so on their own

- A factor is sufficient if its presence is enough to cause the disease

Question 9

Necessary and Sufficient

Answer 9

• in the presence of that factor, disease always develops
• without the factor, the disease never develops
• characteristic of some genetic disorders,

Question 10

Necessary but not Sufficient

Answer 10

• a factor must be present to cause the disease but is not capable of causing it on its own
• Infectious agents are necessary to induce the diseases attributable to them, but they often need some secondary factor, like a weakened immune system, to take hold and do their thing

Question 11

Sufficient but not Necessary

Answer 11

• the factor alone can cause the health outcome, but is not necessary for that outcome to occur
• but so can other factors that are acting alone
• Ex. decapitation

Question 12

Neither Necessary nor Sufficient

Answer 12

• a factor cannot, on its own, cause a disease, and it’s not the only thing capable of causing that disease
• one factor alone cannot do it alone
• most common in chronic diseases

Question 13

Temporal Relationship

Answer 13

• if a factor is believed to be the cause of a disease, exposure must occur BEFORE the disease develops
• important to consider the length of time that has passed after an exposure (it cannot be too short)

Question 14

Strength of Association

Answer 14

• the stronger the association between a health outcome and a potential risk factor, the more likely that it is a causal relationship.
  Ex. the relationship between smoking and lung cancer

Question 15

Dose-Response Relationship

Answer 15

• if the dose of an exposure increases, the health effects of that exposure will increase
  Ex. more cigarettes you smoke, the more likely you are to develop a variety of cancers
• there is a threshold dose beyond which the effects can get no worse
  Ex. to kill someone with arsenic, there is a threshold to its toxicity

Question 16

Evidence for a Causal Relationship (3)

Answer 16

1. organism is always found with the disease
2. organism is not found with any other disease
3. organism , who isolated from one who has the disease and cultured through several generations, produces the disease (in experimental animals)

Question 17

Replications of the Findings

Answer 17

• has to be found consistently in different studies and in different populations
• if findings can’t be replicated, then the association is likely due to chance

Question 18

Biological Plausibility

Answer 18

• must not violate what we understand about biology

Question 19

Consideration of Alternative Explanations

Answer 19

• always consider the role of confounders

Question 20

Cessation of Exposure

Answer 20

• If an exposure is causal, it’s quite reasonable to assume that elimination of the exposure will reduce the incidence of the disease

Question 21

Consistency with Other Knowledge

Answer 21

• to not being in opposition to what other sources of information are telling us

Question 22

Specificity of the Association

Answer 22

• one exposure should cause one disease

- but this is not common