Cohort Studies
Researchers may use cohort, or follow-up, studies to investigate the association between exposures or potential risk factors and patient outcomes. These studies identify subjects who are free of the disease of interest and classify them by the presence or absence of potential risk factors. Then the study follows these subjects for subsequent development of the disease of interest (see Fig 1-3).
The Los Angeles Latino Eye Study (LALES) is an example of a population-based cohort study with prospective data collection. This study examined and interviewed approximately 6000 residents of Los Angeles, California, and followed them longitudinally for the incidence of ocular disease. Researchers explored potential risk factors for diseases such as AMD, diabetic retinopathy, glaucoma, and cataract using the residents’ exposures at the beginning of the study and the incidence of the diseases years later. For example, the investigators discovered new cases of macular degeneration over a 4-year period in Latino individuals. The study found that older age and pulse pressure (difference between systolic and diastolic pressure) were independently associated with new onset early AMD, soft indistinct drusen, and retinal pigmentary abnormalities. This is an example of prospectively assessing the risk factor (blood pressure and age) for an outcome of interest (incidence of AMD).
Cohort studies can provide associations between risk factors and disease. The primary weakness of this study design is that participants with the risk factor of interest may differ in many ways from those without the risk factor, and those other characteristics may affect the incidence of the disease. One example relates to the higher incidence of graft failure among patients with interrupted sutures. It would be inaccurate to conclude that use of interrupted sutures increases the risk of graft failure, because ophthalmologists use interrupted sutures in patients with a preexisting risk of graft failure, such as stromal vascularization. In this example, stromal vascularization is a confounding factor. Statistical analysis techniques, such as stratified analysis and regression analysis, can adjust for the effect of known confounding factors. However, quite often investigators do not understand all the factors that affect the incidence of a disease. For this reason, cohort studies may identify associations and disease incidence, but these associations are not considered causal.
-
Choudhury F, Varma R, McKean-Cowdin R, Klein R, Azen SP; Los Angeles Latino Eye Study Group. Risk factors for four-year incidence and progression of age-related macular degeneration: the Los Angeles Latino Eye Study. Am J Ophthalmol. 2011;152(3):385–395.
Excerpted from BCSC 2020-2021 series: Section 1 - Update on General Medicine. For more information and to purchase the entire series, please visit https://www.aao.org/bcsc.