Abstract
A large body of evidence suggests that exposure to childhood adversities increases risk for poor quality physical health in adulthood. Much of this evidence is based on retrospective measures which are believed to be contaminated by the limitations and biases of autobiographical memory. Using longitudinal data on 454 African Americans (61 percent female) this study examines the corroboration between prospective and retrospective measures of childhood adversities gathered approximately two decades apart, and the relative ability of the measures to predict self-reported illnesses and a biomarker of 30-year cardiovascular disease risk. Comparisons indicated that the retrospective and prospective measures demonstrated weak convergence and did not provide completely equivalent information about self-reported adverse childhood experiences. A series of regression models indicated that the two measures of adversities exhibited similar associations with the cardiovascular disease biomarker but divergent associations with self-reported illnesses. Furthermore, both the prospective and retrospective measures simultaneously predicted cardiovascular disease risk in adulthood. That the prospective measure did not significantly predict perceived illnesses after adjusting for the retrospective measure is evidence that childhood adversities predict self-reported health burden insofar as respondents remember those adversities as adults. The findings provide evidence that retrospective self-report measures of childhood adversities do not closely converge with prospective measures, and that retrospective measures may not provide valid estimates of the association between childhood adversities and perceived illnesses in adulthood.