Unobserved individual-level variation in cardiovascular mortality in birth cohorts born before and after 1945

Cardiovascular disease (CVD) mortality has recently declined in many populations. It is not known if inequalities between individuals after accounting for risk factors have become more important. We investigated if the variation in risk due to unobserved factors – termed frailty - is larger in cohorts born after compared to before 1945. 446,053 individuals born 1930–1945 and 1946–1960, participating in Norwegian health surveys (1974–1997) were followed for premature CVD mortality. We used proportional hazards models with Weibull baseline distributions to evaluate mortality after adjusting for CVD risk factors. Models included gamma frailty to quantify variation in risk due to unobserved factors between individuals within each birth cohort. Additionally, we calculated Gini coefficients and plotted Lorenz curves to compare mortality inequalities across cohorts. Hazard ratio estimates for each CVD risk factor were similar in the two cohorts except for smoking. Frailty estimates after adjustment for all measured risk factors were 4.08 in the older cohort and 4.69 in the younger cohort, corresponding to nearly identical Gini coefficients of 0.79 and 0.77, respectively. Contrary to our hypothesis, we did not observe a substantial difference in unobserved individual-level variation in CVD mortality among those born after 1945. The nearly identical Gini coefficients indicate stable inequality over time. After accounting for established risk factors, CVD mortality remained highly concentrated, with roughly 80% of cumulative risk observed among approximately of 20% of individuals in the population.