The value of comparing health outcomes in cohort studies: An example of self-rated health in seven studies including 79 653 participants

Objective: To demonstrate the value of comparing data from multiple cohort studies using the example of self-rated health (SRH). Methods: Seven Australian cohort studies including comparable data on SRH were identified. Comparisons of the distributions of SRH were conducted, and logistic regression was used to evaluate age, sex and education effects within studies. A nationally representative survey was used as a statistical reference to determine how studies differed from the Australian population in frequencies of responses. Results: Ratings of SRH declined with increasing age. Low education was associated with higher frequencies of fair/poor SRH even in young adulthood but there were no sex differences. Results for smaller studies did not necessarily differ from nationally representative studies. Conclusion: Collaborative reanalysis of Australian cohort permits analysis of health outcomes from a large numbers of participants. Health outcomes and their sociodemographic determinants may be more comprehensively evaluated through such collaborative projects.