the leading causes, including diabetes (Hummer et al., 2004; Williams, 2001b). However, this is not the case for Pacific Islanders specifically.
Among the leading causes not shown in the figures, there is one further interesting contrast: whites have a substantially higher death rate from Alzheimer’s disease than any other group. For older white women, for instance, the rate is 75 percent higher than the rate for blacks and more than double the rate for any other group (Hummer et al., 2004).
One way to summarize these contrasts, at least between whites and black, is to estimate how much each cause of death contributes to the racial and ethnic differences. We cannot look specifically at older people, but analysis has been done for all ages combined. Across all ages, deaths from hypertension contribute the most to black-white differences (15 percent), followed by HIV (11 percent), diabetes (8 percent), and homicide (8 percent) (Wong et al., 2002). At older ages, two of these causes—HIV and homicide—should be less significant, which may have some influence on the convergence of black and white mortality rates.
In some respects, comparisons of health and disability across racial and ethnic groups are consistent with the mortality comparisons, but in others they are not. Surveys often ask people to rate their health as excellent, very good, good, fair, or poor, and these ratings generally correlate with subsequent mortality. Table 1-2 shows the percentages of people who report fair or poor health in pooled National Health Interview Surveys from 1989 to 1994 (Hummer et al., 2004). Though there is some variation at the oldest ages, self-rated health for blacks is generally the worst among all racial and ethnic groups (Hayward and Heron, 1999), and self-rated health for Asians is generally the best. Between these extremes, rankings are different from those for mortality. Whites have better self-rated health than Hispanics and much better self-rated health than American Indians and Alaska Natives.
Comparisons of groups in regard to activity limitations (Table 1-2) parallel those for self-rated health. The limitations indicator includes individuals who report being unable to perform their major activity, being limited in it, or being limited in other activities. The rankings by racial and ethnic groups on this indicator are essentially the same as for self-rated health, except that American Indians and Alaska Natives are worse than blacks for some age and gender categories.
In contrast to mortality rates, black and white morbidity indicators do not cross over at the oldest ages. Even at the oldest ages, blacks report more severe problems with disability and functioning than whites.
Health and disability comparisons could differ from mortality comparisons for substantive reasons, or the differences could be artifactual. The