The first term on the right hand side is the incidence of events. The second term is the survival rate, and the third term is the health effect among survivors. The change in disability rates is arithmetically related to the change in one or more of these factors.4 The marginal survivors theory focuses on the second term: the change in the probability of survival after an acute event, i.e. as more patients survive acute events, the pool of people at risk of disability expands. The other theories focus on the first and third terms, i.e. either reduced incidence of disease or better health among survivors. other
Table 2 shows cardiovascular disease event probabilities, survival rates, and conditional disability rates for each of our three time periods. To measure the cardiovascular disease event rate, we consider the population surveyed at the beginning of the 5-year interval, and look at events in those five years. For example, the cardiovascular disease event rate for the 1989 cohort is the share of the population aged 65 and older in 1984 that had a cardiovascular disease hospital admission in the subsequent five years.
The first row shows that the share of people who had a hospitalization for cardiovascular disease was relatively constant over the time period, at about 26 percent. This is somewhat surprising given the reduction in event rates noted in other surveys such as the Framingham Heart Study (Sytkowski, D’Agostino et al. 1996) and the Minnesota Heart Survey (McGovern, Pankow et al. 1996; McGovern, Jacobs et al. 2001). It may be that some of the admissions among the later cohorts in our study were done explicitly to perform surgical operations such as angioplasty or bypass surgery, and thus contribute to an increased reporting of cardiovascular disease. However, another recent U.S. study of subjects aged 35 to 74 reported little change in the incidence of first myocardial infarction between 1987 and 1994 (Rosamond, Chambless et al. 1998). Alternatively, it may be that less severe cases of these conditions are being diagnosed over time.
In addition, the Framingham and Minnesota studies included patients younger than 65, and results from these studies may have been driven by a decline in heart disease among the younger population.