Data and Methods Summary and Discussion

4. Results

4.1 Change in Disease Prevalence

There are statistically significant increases between 1984 and 1994 in the reported prevalence of most diseases among the American population 70 years of age and over. Odds ratios indicating the relative likelihood of having a condition in 1994 and 1984 after age is controlled are shown in Figures 1 (mortal conditions) and 2 (non-mortal conditions). The odds ratios are shown only where there is significant or measurable change over time; ratios for males are on the left and females are on the right. For almost all diseases the ratios are above 1 indicating a significantly higher level in 1994. There is only one significant decline - the prevalence of hypertension for females. Among females, there is no change in the prevalence of stroke survivors. For the other mortal diseases and for hypertension among males, increases in prevalence are observed. Odds ratios range from 1.12 for hypertension in males to 2.0 for cancer meaning that the relative likelihood of having a disease increased between 1984 and 1994 from 12 percent to 100 percent. For both men and women the increase in the relative likelihood of having had cancer was the greatest. The increase in the relative likelihood of reporting heart disease was the next largest.

The equations which generated the odds ratios can also be used to predict the absolute probability of having each disease at the two dates. The predicted probability of having each disease at age 70 is shown in Table 2. The two major causes of death in this age range are heart disease and cancer. The estimated reported prevalence of heart disease at age 70 has increased by 6 percentage points for men and 4 percentage points for women. The increase in reported cancer survivors is 9 percentage points for men and 5 percentage points for women.

Figure 2 shows the increase for the non-fatal conditions and impairments. For both sexes there are statistically significant increases in the prevalence of arthritis, osteoporosis, cataracts and glaucoma. There are significant increases in blindness among women. Some of the non-fatal conditions do not appear on the figure because there is no significant change in prevalence. There is no change in the prevalence of having a broken hip, hearing trouble or dizziness among either sex.

The highest odds ratios are for osteoporosis indicating that this condition has the greatest relative increase over time for both men and women. It is possible that the increase in osteoporosis prevalence represents, at least in part, the dissemination of the use of this diagnosis to describe the condition. For most of the diseases the odds ratios for men are slightly higher than those for women, indicating more relative increase over time among men. The exceptions are for the eye conditions of glaucoma and blindness where women have more relative increase than men.

The estimated probability of having arthritis at age 70 increases from .44 to .49 for men and from .59 to .62 for women at age 70. The probability of having cataracts at age 70 has increased by .03 for both sexes. Even though the odds ratio indicated that the size of the increase in the relative likelihood of having osteoporosis was similar for men and women, the absolute increase (7 % at age 70 for women) is much greater among women.

4.2 Change in Functioning and Disability among those with Diseases

In order to indicate change in the severity of diseases and conditions among those who have each disease, we examine change between 1984 and 1994 in the levels of disability and functioning loss among those who have each specific disease or condition in Table 3. The average change in the number of Nagi functions (out of 9), in the number of ADLs (out of 7), and IADLs (out of 5) one is unable to perform is presented after controlling for the age composition of the samples of people having each of the 13 diseases. There are striking sex differences in the change over ten years in functioning and disability. For men with disease there is almost no change in functioning and disability and the only significant changes indicate increased functioning problems in 1994. For women, the change is quite different. Ability to perform Nagi functions and IADLs clearly improved over the decade. Women with almost all of the conditions experienced significantly less functioning loss and IADL impairment. The decreases in Nagi functioning deficits were largest among women who had heart diseases, strokes, arthritis, and osteoporosis; the decreases in IADL functioning were largest for those with heart disease, stroke, and osteoporosis. In general the greatest improvement was in the Nagi indicators, or the indicators of the most moderate functioning loss. Only for women with diabetes, broken hips, and blindness was there no improvement in these types of functioning.

Any significant change in ADL functioning for both men and women, represented deterioration in functioning over time. For men almost all of the significant functioning changes represented increases in ADL disability in 1994. For women most of the increases were small. Examination of functioning change for those with no disease indicates no change over time in any type of functioning ability among either men or women. Change over the ten-year period in the total sample, including people with any disease and without disease, indicates improvement in Nagi and IADL functioning among women and slight deterioration of ADL functioning among both genders.

4.3 Co-morbidity

To this point we have treated diseases individually, not acknowledging the potential for any individual to have multiple diseases or conditions. It is possible that there has been a change over time in the number of diseases people have, in the likelihood of having multiple diseases as well as in the likelihood of having any one disease. We examine the change in the number of reported diseases per person: mortal conditions (range 0 to 5), morbid conditions (range 0 to 4), and impairments (range 0 to 4). The regression coefficients representing the increase from 1984 to 1994 from sex-specific equations with age controlled are shown in Table 4. In 1994 both men and women report an average of .3 diseases more than in 1984. For men most of the increase is due to increase in the number of mortal conditions. For women the increase in the number of morbid conditions is greater.

We also use logistic regression to estimate the effect of the time change on the likelihood of having no disease and the likelihood of having extensive comorbidity (3+ diseases). Where there was significant change between the two surveys, we use the results to estimate the probability that a 70 year old man or woman would have no disease or would have three or more diseases in 1984 and 1994 (Table 5). It is clear that in 1994, people are less likely to be without disease. About one third (32%) of men had no mortal disease in 1994; in 1984, the figure was 41 percent. For women, the percentage having no morbid disease fell from 38% to 34%. The number of people reporting 3 or more diseases at age 70 increased for both men and women and for both genders the increase was concentrated in mortal diseases.

Some combinations of disease are particularly interesting. The likelihood of persons with heart disease also being cancer survivors was twice as great for men in 1994 as in 1984. For women the increase was about 50 percent.

Among those who have a mortal disease, there is an increase in the likelihood of also having a morbid condition. Among those with one of the mortal conditions, the likelihood of having at least three comorbid conditions increased by 25% over the ten years for both men and women.


Data and Methods Summary and Discussion

Change in the Prevalence of Diseases among Older Americans: 1984-1994
Eileen M. Crimmins, Yasuhiko Saito
© 2000 Max-Planck-Gesellschaft ISSN 1435-9871