作者: W. C. Willett , G. A. Colditz
DOI: 10.1093/OXFORDJOURNALS.EPIREV.A017975
关键词:
摘要: Numerous cohort studies have been initiated over the past 50 years to address lifestyle and biochemical determinants of specific diseases. A key feature these is that information on primary exposures potential confounders required from individual subjects themselves, rather than routinely collected data. In this brief overview we will discuss some theoretical practical issues be considered in development conduct a large modern study. Earlier cohorts, such as those occupational medical including health effects ionizing radiation (1, 2), frequently did not require contact with participants, and, thus, could only very were limited ability account for confounding. These detail here. Some first established specifically evaluate one or few factors, smoking (3-5) physical activity (6, 7). Still, other simple variables provided valuable opportunities examine obesity (8), alcohol consumption (9, 10), oral contraceptives (11, 12). Early population-based cohorts included work Archie Cochrane who, 1950, began study Rhondda Fach which encompassed eight mining towns. Of adult population, 89 percent x-rayed, 2,764 miners 3,055 ex-miners (13, 14). Though initial focus was causes pneumoconiosis progressive massive fibrosis among living communities (14), rheumatoid arthritis examined (15), by follow-up 2'/2 after baseline data collection, psychiatric conditions added disability (16). Cardiovascular risk factors endpoints studied early life