The Contribution of Medication Use to Recent Trends in Old-Age Functioning


U.S. Department of Health and Human Services

The Contribution of Medication Use to Recent Trends in Old-Age Functioning

Executive Summary

Vicki A. Freedman, Ph.D., and Hakan Aykan, Ph.D.

Polisher Research Institute, Madlyn and Leonard Abramson Center for Jewish Life
(formerly the Philadelphia Geriatric Center)

August 2002

This report was prepared under contract between the U.S. Department of Health and Human Services (HHS), Office of Disability, Aging and Long-Term Care Policy (DALTCP) and the Philadelphia Geriatric Center. For additional information about this subject, you can visit the DALTCP home page at or contact the office at HHS/ASPE/DALTCP, Room 424E, H.H. Humphrey Building, 200 Independence Avenue, S.W., Washington, D.C. 20201. The e-mail address is: The Project Officer was William Marton.

This research was funded by a supplement from the Office of the Assistant Secretary for Planning and Evaluation of the U.S. Department of Health and Human Services to the National Institute on Aging (Grant No. R01-14346). The views expressed are those of the authors alone and do not reflect the opinions of the funding agencies.


After nearly a decade of debate among academics and policy makers, a consensus has emerged that disability rates among older Americans have declined over the last fifteen years. However, the implications of such trends for publicly funded programs such as Medicare, Medicaid, and Social Security remain far from clear. In order to understand the consequences of disability declines for such programs, better insight into the causes driving the trend is needed.

Several existing studies raise the possibility that changes in the management of chronic disease--and in changes in medication use in particular--could be a potentially important explanation for observed improvements in functioning. Indeed, drug treatment has become an increasingly important aspect of medical care for older Americans; currently nine out of ten older Americans take one or more prescription drugs daily and many consume multiple medications. During the 1980s and early 1990s there also have been major shifts in the classes of drugs prescribed for some of the more debilitating chronic conditions.

The purpose of this study is to explore the extent to which changes in medication use account for improvements in functioning among older Americans. Using several waves of the Health and Retirement Study (HRS), a nationally representative survey of non-institutionalized Americans ages 51-61, we examine changes during the 1990s in the prevalence of functional limitations and medication use associated with five highly prevalent and often debilitating chronic conditions: hypertension, diabetes, lung disease, stroke, and arthritis.

For Americans of pre-retirement age (51-61) and for subgroups of this age group with specific chronic conditions we explore the following questions:

  • Has functioning improved over time?

  • Has medication use increased over time?

  • Do changes in medication use account for improvements in functioning?

  • Have improvements over time been greater for those groups reporting medication use?


Drawing upon multiple waves of the HRS, we compare estimates of functional limitations and medication use for the population ages 51-61 in 1992 (N=9,573) and in 1998 (N=7,099). We also model the number of functional limitations as a function of year, chronic conditions, and various demographic and socioeconomic characteristics. Because medication use may be a reflection of the severity of an underlying functional problem, we also present evidence about changes in functioning between 1994 (N=8,618) and 2000 (N=6,486), using medication use and other predictors from 1992 and 1998, respectively.

In all four waves, respondents were asked to report about difficulty with 12 functional limitation items. However, in 1992 respondents were asked "How much difficulty do you have _____" whereas in subsequent waves respondents were asked "Do you have any difficulty…" We handle this change in question wording in two ways. First, we limit our analysis to 7 of the 12 functional limitation items (sitting for about 2 hours; getting up from a chair after sitting for long periods; lifting or carrying weights over 10 pounds; stooping, kneeling, or crouching; picking up a dime from a table; reaching or extending arms above shoulder level; and pulling or pushing large objects). Our analysis of experimental module data included in the 1994 wave (presented in detail in Appendix I) suggests that the 7-item scale is less sensitive to question wording changes without sacrificing internal consistency or predictive validity. Second, we 'correct' comparisons between 1992 and 1998 for changes in question wording based on a correction factor developed from our analysis of the experimental module (see Appendix II for details).

In 1992 and 1998, the HRS obtained information in an identical fashion about five chronic conditions--hypertension, diabetes, chronic lung disease, stroke, and arthritis. For arthritis, survey respondents were asked if they ever had or a doctor ever told them they have arthritis. For all other conditions, they were asked to report whether a doctor ever told them they have the given condition. In both years, respondents reporting a given condition were asked follow-up questions about medication use. Medication use questions were tailored to each condition but were essentially identical in 1992 and 1998.


After correcting for changes in wording, we find the average number of functional limitations declined from 1.57 in 1992 to 1.34 in 1998 (p<0.01). Statistically significant improvements were also found among those reporting hypertension, diabetes, and arthritis. Improvements were experienced widely with the important exception of non-Whites whose functioning remained constant over this time period.

When we limit the analysis to 53-63 year olds in 1994 and 2000, we find no significant improvement in functioning (1.39 in 1994 vs. 1.40 in 2000), except among those reporting no arthritis, those who were not married, and those with no liquid assets (savings, checking, CD, or stocks). Functioning was significantly worse among those with arthritis and among those with more than a high school education.

We also find large increases in the reports of medication use among older Americans reporting hypertension (from 64% in 1992 to 75% in 1998) and reporting diabetes (from 62% in 1992 to 77% in 1998). However, for the five conditions considered here, changes in medication use do not appear to account for any of the improvements in functioning between 1992 and 1998 or the stability between 1994 and 2000. Instead, increases in educational attainment appear to be offsetting increases in the prevalence of chronic conditions, most notably obesity (which increased by over 25% in just six years) and arthritis.

Finally, we find no evidence that improvements in functioning were larger for Americans taking medications than for those with the same condition but not taking medication. Thus, it does not appear that medications have become more effective during the 1990s in averting functional limitations among Americans of pre-retirement age.


We conclude that changes in medication use for five highly prevalent conditions do not appear to explain improvements in functioning among Americans nearing old age. Our analysis also provides a cautionary tale for policy makers. The year-to-year variation we observe in whether there have been improvements in functioning for this cohort underscores the need for analysis of multiple data sets and time points before drawing conclusions about trends. More importantly, given that educational increases will be smaller in the future than they have been in the past, and that increases in the prevalence of obesity are likely to continue as younger cohorts enter old-age, caution is warranted in projecting forward improvements in old-age functioning.

The Full Report is also available from the DALTCP website ( or directly at