3.4.1. Biostats Cohort Studies

Question 1

What study design is best for identifying people, measuring risk factors, then following forward to see if they develop disease?

Answer 1

Cohort study

Question 2

What particular study was to identify common factors or characteristics that contribute to CVD by following its development over a long period of time in a large group of participants who had not yet developed overt CVD symptoms or suffered a heart attack or stroke.

Answer 2

Framingham study

Question 3

The Framingham study measures what?

Answer 3

incidence of the CVD

Question 4

Factors that the Framingham study believed to contribute to CVD (exposures) are?

Answer 4

elevated cholesterol, smoking, habitual alcohol use, lower physical activity, increased thyroid function, high blood hemoglobin or hematocrit, high body weight, diabetes mellitis, gout, age (higher), sex (male), hypertension

Question 5

exposure -> outcome over time is what type of study

Answer 5

prospective cohort

Question 6

exposure and outcome have both occurred. What type of study can you do?

Answer 6

Historical or retrospective cohort. You can do this when incidence for a disease is low over a given time. For example, development of virus (outcome), take a serum, then look back at records to see if they were obese (exposure).

Question 7

State advantages/disadvantages for historical cohort.

Answer 7

1. Cheaper and faster
2. efficient with disease of long latency
3. Data may be inadequate

Question 8

Advantages/disadvantages to prospective cohort.

Answer 8

1. More expensive and time consuming.
2. Not efficient for diseases with long latent period
3. Better exposure and confounder data
4. Less vulnerable to bias.

Question 9

1. Rare exposures are good with what study? 2. Rare diseases are good with what study?

Answer 9

1. Historical cohort - used to identify special risk factors/exposures
2. case control

Question 10

Rare exposures are utilized with special cohorts. What groups are best utilized in these types of studies?

Answer 10

1. occupational groups. i.e. identifying a occupational hazard
2. groups with unusual dietary or lifestyle factors
3. groups undergoing particular medical treatments

Question 11

When are general cohorts best used?

Answer 11

With common exposures in groups such as occupation (nurse health study), geographic location (Framingham heart study) -> selected for ease of data collection.

Question 12

Restrictions in selecting unexposed for cohort

Answer 12

must be free of disease of interest at the time and must have had the opportunity for both exposure and disease

Question 13

Discuss morbidity ratio

Answer 13

Used in the early days of the Framingham heart study (1950’s) as a means to determine relative risk essentially. = observed case/ expected case x 100

Question 14

Define relative risk

Answer 14

= incidence in exposed/ incidence in unexposed. CI must not include 1.

Question 15

Define risk difference

Answer 15

= incidence in exposed- incidence in unexposed. CI must not include 0

Question 16

What is a risk factor?

Answer 16

It means there is a statistical association. It does not imply causal association. It only means there is a risk. It does not mean high blood pressure causes HBP, for example.

Question 17

Temporality

Answer 17

Cause must proceed the effect. If there is an expected delay between cause and effect, that delay must occur before the effect.

Question 18

consistency

Answer 18

association observed in different populations and circumstances

Question 19

strength of association

Answer 19

large relative risk or ARR

Question 20

dose response

Answer 20

more exposure leads to higher rate of disease

Question 21

plausibility

Answer 21

consistent with biological models

Question 22

reversibility

Answer 22

if exposure is reduced, disease rate falls

Question 23

specificity

Answer 23

one cause leads to one effect

Question 24

analogy

Answer 24

cause-effect relationship established for similar exposure or disease

Question 25

What test can measure temporality

Answer 25

Cohort and randomized controlled trials