Health and Retirement Study (HRS)

University of Michigan
Years Available
1992 to present
Mode of Collection
Multi-stage area probability sample survey, conducted using in-person and telephone interviews.
The University of Michigan Health and Retirement Study (HRS) is a longitudinal panel study that surveys a representative probability sample of approximately 20,000 Americans over the age of 50 every two years. Supported by the National Institute on Aging (NIA) and the Social Security Administration (SSA), the HRS originally explored the changes in labor force participation and the health transitions that individuals undergo toward the end of their work lives and in the years that follow. Since its launch in 1992, the study has substantially broadened its scope and now collects information about income, work, assets, pension plans, health insurance, disability, physical health and functioning, cognitive functioning, and health care expenditures, as well as genomic, blood-based biomarker, and psychosocial data. Through its unique and in-depth interviews, the HRS provides an invaluable and growing body of multidisciplinary data that researchers can use to address important questions about the challenges and opportunities of aging. Survey responses for Medicare recipients are linked to Medicare Claims Data, providing a more extensive analytic data set.
Selected Content
The study collects information about income, work, assets, pension plans, health insurance, disability, physical health and functioning, cognitive functioning, and health care expenditures as well as genomic, blood-based biomarker, and psychosocial data.
Population Covered
U.S. population aged 50 years and older
The HRS is a longitudinal survey with a complex panel structure and sample design. Participants are ages 51 to 61 years when they are invited to join the study and are generally followed until death. Both members of a couple are invited to join and are followed together. Survey participants are community- dwelling persons at baseline but are retained in the study if they move to long term care facilities. New cohorts are added every 2 years; the sample is refreshed with younger cohorts every 6 years. Sample size has ranged from 18,000 to 23,000 in any given wave. In the baseline year, half of the core sample is assigned to a face-to-face interview, including physical and biological measures and a mail-back psychosocial questionnaire. The other half of the sample receives a telephone interview in the baseline year and an enhanced face-to-face interview in the follow-up year. Respondents generally answer for themselves, although proxy respondents (usually spouse or other family member) may be used if the respondent is unable to participate. The core survey, which includes the main part of the data, takes place every 2 years and lasts 90 minutes to 3 hours. Studies on a range of topics take place in the "off year" and are linked to the core. HRS core data are linked to various sources of administrative data, such as Medicare, the Veteran's Administration, and the National Death Index. A randomly selected half of participants over age 65 are invited to participate in the Healthy Cognitive Aging Project which collects more intensive cognitive data valuable for detecting dementia or cognitive impairment in sample participants, which enhances our understanding of risk and protective factors for Alzheimer's disease and related dementias.
Response Rates and Sample Size
The response rates for the initial interview are about 73%; follow-up response rates range from 80% to 90%+.
Interpretation Issues
All HRS respondents who are Medicare beneficiaries are asked to provide their identification number; over 80 percent consent to do so. Of those who provided an identification number, 98 percent have been successfully matched to Medicare files. During face-to-face interviews conducted every four years, participants may consent to the collection of data on height, weight, blood pressure, grip strength and other physical performance measures as well as DNA for genotyping and (currently) venous blood for biomarker analyses. Other health information obtained through HRS are self-report data.
Although the HRS oversamples African Americans and Census-defined Hispanics, it does not oversample other racial or ethnic minorities.
Health and Retirement Study, An Introduction (slides and video).