Cohort Studies Notes

Cohort studies compare outcomes between exposed and unexposed groups.
If there's a positive association between exposure and disease, the exposed group is expected to have a higher incidence of the disease than the unexposed group.
Two ways to generate study groups:
1. Select groups based on exposure status.
2. Select a defined population before exposures are identified, then categorize into exposed and unexposed.
Essential characteristic: comparison of outcomes in exposed vs. unexposed groups.

Prospective Cohort Study:
- The investigator identifies the original population at the beginning of the study, follows subjects concurrently through calendar time, and monitors who develops the disease.
- Exposure and non-exposure are determined as they occur during the study.
Retrospective (Historical) Cohort Study:
- Uses historical data to shorten the study timeframe.
- Exposure is ascertained from past records, and the outcome is determined when the study begins.
- Still compares exposed and unexposed groups.
- Exposure is ascertained from past records, the outcome (development or no development of disease) is determined when the study is begun.
Combination Approach:
- Exposure ascertained from past records (historical cohort study).
- Follow-up and outcome measurement continue into the future.

The Framingham Study:
- Began in 1948 to study cardiovascular disease.
- Followed a population of residents in Framingham, Massachusetts, aged 30-62.
- Examined multiple exposures and their interactions using multivariable techniques.
Incidence of Breast Cancer and Progesterone Deficiency:
- A retrospective cohort study examined women with hormonal abnormalities (progesterone deficiency) and compared their breast cancer incidence to women without such abnormalities.
- The study found a higher risk of premenopausal breast cancer in women with hormonal abnormalities.

Experiences and exposures during fetal life may have long-lasting effects.
Studies can begin:
- At birth and rely on retrospective data for prenatal exposures.
- In the prenatal period.
- Prior to conception to gather concurrent data about exposures.

Selection Biases
- Nonparticipation and nonresponse can introduce biases.
- Loss to follow-up can be a serious problem.
Information Biases
1. Differences in the quality of information obtained from exposed vs. unexposed persons can introduce bias.
2. The person deciding whether the disease has developed should be masked to the subject's exposure status.
3. Preconceptions of epidemiologists and statisticians may unintentionally introduce biases into data analyses and interpretations.

Cohort study is indicated when prior good evidence suggests an association of a disease with a certain exposure or exposures (evidence obtained from either clinical observations or case-control or other types of studies).

Hybrid design combining elements of case-control and cohort studies.
Nested Case-Control Studies:
- Controls are a sample of individuals who are at risk for the disease at the time each case of the disease develops.
- Cases and controls are matched on calendar time and length of follow-up.
Case-Cohort Studies:
- Controls are randomly chosen from the defined cohort.
- Allows studying different diseases using the same subcohort for controls.

Data are obtained before any disease has developed, eliminating recall bias.
Laboratory values are obtained before the development of clinical disease.
More economical to conduct.
Cases and controls are derived from the same original cohort, so there is likely to be greater comparability between the cases and the controls.