Ch 6 - Association and Causality

what caused the disease?

• red door 1: witchcraft, gods and demons?

• yellow door 2: environmental influence

• purple door 3: miasma

enter door 4: the germ theory

• Pasteur and Koch developed and hypothesized

Pasteur’s test of spontaneous generation

• people believed it was life that spoiled things, so he takes 2 curved neck glasses and breaks the neck off of one to show the culture spread in the non-broken glass, while the broken one didn’t culture anything

  • one of the threads of the germ theory showed germs do not come out of nothing and are a manifestation of something in your body

• broth is boiled → broth remains free of microorganisms → curved neck is removed → microorganisms grow in broth

Koch’s postulates

1. the microorganism must be found in abundance in all organisms suffering from the disease, but should not be found in healthy organisms

2. the microorganism must be isolated from a diseases organisms and grown in pure culture

3. the cultured microorganism should cause disease when introduced into a healthy organism

4. the microorganism must be re-isolated from the inocluated diseased experimental host and identified as being identical to the original specific causative agent

epidemiology sets to discover whether

• exposure

  • contact with factors that usually may be linked to adverse outcomes

• → outcome

  • e.g. specific forms of morbidity and mortality

deterministic and probabilistic causality in epidemiology

• a concern of epidemiology is to assert that a casual association exists between an exposure factor and a disease (or other adverse health outcome)

• association: linkage between or among variables

• exposure: contact with factors linked to adverse health outcomes

deterministic causality

• a cause (exposure) is invariably followed by an effect (a health outcome) 

  • an exposure → an outcome

• necessary cause 

• sufficient cause

necessary cause

• factor whose presence is required for the occurrence of the effect 

• necessary in order to have occurrence

• ex: in order to have covid, the virus has to be in my body

sufficient cause

• cause that is sufficient by itself to produce the effect

• not necessary to cause a disease, but sufficient

necessary and sufficient

• a factor that is enough on its own to cause a disease and the disease will always happen once this factor is introduced 

• “both X and Y are always present together, and nothing but X is needed to cause Y…”

• X=factor; Y=outcome

sufficient but not necessary

• a factor that may be the cause of disease, but not other causes of the disease exist

• “X may or may not be present when Y is present, because Y has other causes and can occur without X.” In other words, X is one of the causes of the disease, but there are other causes

necessary but not sufficient

• a factor that is necessary for the disease to develop, but on its own does not cause the disease

• “X must be present when Y is present, but Y is not always present when X is.” This formulation means that X is necessary for causation of Y, but X by itself does not cause Y.

neither necessary nor sufficient

• a factor that along with other factors may cause the disease (in combination)

• “…X may or may not be present when Y is present. under these conditions, however, if X is present with Y, some additional factor must be present. here X is a contributory cause of Y.”

sufficient-component cause model (casual pie)

• a group of component causes, diagrammed as a pie

• one component cause is a necessary cause

• remaining component causes are not necessary

• together makes up a sufficient cause complex

• ex: pie charts on the slide

probability models and probabilistic (stochastic) causality

• that incorporates some element of randomness

• probabilistic causation describes the probability of an effect in mathematical terms, given a particular dose

• ex: radiation exposure and probability of carcinogenesis

• study the cycle of epidemiologic research slide

hypotheses come from two methods

• method of difference

• method of concomitant variation

method of difference

• all of the factors in two or more domains are the same except for a single factor

• the frequency of disease that varies across the two settings is hypothesized to result from variation in a single causative factor

the method of concomitant variation

• a type of association in which the frequency of an outcome increases with the frequency of exposure to a factor

• hypothesis = factor is associated with outcome

• EG dose-response relationship cigarettes smoked and mortality from lung cancer

• the more you do something, the more likely you’re going to get sick

operationalization

• process of defining measurement procedures for the variables used in a study

  • take 2 variables, make questionnaire to see possible association

• ex: in a study of the association between tobacco use and lung disease, the variables might be the number of cigarettes smoked and the occurrence of asthma 

  • variables operationalized by questionnaires and review of medical records

• study diagram

types of associations found among variables

• statistical association between X and Y?

  • no (X and Y are independent) 

  • yes → what kind of association?

    • noncausal (secondary)

    • causal → is it?

      • an indirect association

      • a direct association

criteria of causality

1. strength

2. consistency

3. specificity

4. biological gradient

5. plausibility

6. coherence

7. experimental evidence

8. analogy

strength

• how strong is the association?

consistency

• has the observed association been repeatedly observed by different persons, in different places, circumstances and times?

specificity

• is the outcome unique to the exposure?

  • is a biological relationship possible

temporality

• does exposure precede the outcome?

  • does x come before y, if it doesn’t, there’s no causality

biological gradient

• as the level of exposure increases, does the rate of disease also increase?

plausibility

• is a casual relationship biologically feasible?

coherence

• is a casual association compatible with the generally known facts of the natural history and biology of the disease

  • does it make sense?

experiment

• do interventions that modify exposure modify the outcome?

analogy

• has a similar casual relationship been observed with another exposure and/or disease?

multivariate causality

• a preponderance of the etiologies of diseases (particularly chronic diseases) involve more than one causal factor

  • personal, genetics, physical/mental health, diet, health behavior, household, close circle, family, friends, food at home, physical activity at home, school, work, access to food

• several things on several different levels work together to influence one health outcome

statistical significance

• asserting that an observed association valid (meaning is not likely to have occurred as a result of chance)

  • it is likely this relationship is not due to chance

• chance does not mean statistical significance

  • after all, an association can be coincidental

inferential statistics

• degree to which chance affects the conclusions that can be inferred from data

• inference: process of evolving from observations and axioms to generalizations

• reason: draw conclusions about a parent population from sample-based data

• sample = subset of the data that have been collected from a population

let’s say we want to find obesity prevalence (population) in society

• the more you sample, the closer your statistic is to the parameter

• we pick a sample of desired group and measure obesity prevalence → we find the percentage of the sample that are obese = STATISTIC → we use this to find the value for the population = parameter, which is a slight difference

point estimate

• a single value chosen to represent the population parameter

• does not exactly equal the population parameter because of sampling error

confidence interval

• a range of values that with a certain degree of probability (p-value) contain the population parameter

power

• statistical significance affected by the size of the sample

  • larger samples = more likely to produce meaningful significant results than smaller samples

• the ability of a study to demonstrate an association or effect if one exists

• how large an effect is observed

• good power: >80%; aim for 80%-90%

• when the effect is small and the sample size is large, the association may be statistically significant, but when the effect is large and the sample size is small, the association may not be significant merely because of the small sample size