The Health and Retirement Study (HRS) is an unparalleled resource for studying the aging population in the United States. Its success was the impetus for the creation of comparable, longitudinal surveys in over 40 countries around the world.

With over 30 years of data collection and one of the most thorough, multidisciplinary surveys in the world, there is a vast amount of data available for researchers. This amount of data can be overwhelming for many users, especially for those new to using HRS. The RAND Center for the Study of Aging led initial efforts to make this data more approachable and easier to use by creating the RAND HRS Longitudinal File in 2003, known as the RAND HRS. The RAND HRS established data structure and variable naming conventions that have been adopted by several other longitudinal datasets.

The RAND HRS has become an important asset for HRS users.  The primary goal of the RAND HRS is to facilitate longitudinal research using HRS. However, it was not designed for cross-country research, which has rapidly expanded along with the evolution of the Health & Retirement Study - International Network of Studies (HRS-INS). Moreover, HRS introduced additional questions on many emerging topics over time, and such expansion has not been fully reflected in the RAND HRS. To facilitate comparative studies with other HRS-INS and to increase topical coverage, the Gateway team built a supplement to the RAND HRS, which we call the Harmonized HRS. Please see the figure below summarizing the topics added by the Harmonized HRS at its inception.

The Harmonized HRS provides a host of research-ready variables for users looking to conduct longitudinal and cross-study research. The Harmonized HRS maintains the HRS person- and household-level identifiers for easy merging with the RAND HRS and the original HRS survey data. The Harmonized HRS offers important demographic and geographic variables, including internationally comparable education variables, whether the respondent lives in an urban or rural area, their house type, and citizenship status. While the RAND HRS offers a host of health variables, the Harmonized HRS expands those by adding variables capturing vision and hearing, falls, sleep, pain, heart events, medication usage, age at disease diagnosis, CAGE alcohol problem screening, smoking history, and vaccinations. The Harmonized HRS also includes several daily activities and functional limitations summary variables designed to be comparable with other HRS-INS studies.

For users interested in intergenerational research and family dynamics, the Harmonized HRS offers new indicators of parental socio-economic status, number of grandchildren, who the respondent lives with, and frequency of contact with children and parents. The Harmonized HRS also provides important measures of religious and social participation to help capture social engagement. For research into working conditions, the Harmonized HRS provides variables capturing job characteristics, job satisfaction, and job search activities for both employed and unemployed respondents.

The Harmonized HRS also includes variables capturing all regularly conducted biomarker tests conducted as part of the HRS Health Assessments, which were not initially included in the RAND HRS.  Similarly, the Harmonized HRS greatly expanded the caregiving and assistance variables available to researchers who want to understand care needs, paid and unpaid caregiving received by respondents, as well as the care the respondents provide to others. Working with the NIA Stress Network and the NIA Emotional Wellbeing Network, the Gateway team has developed comparable stress and psychosocial variables in the Harmonized HRS. Finally, the Harmonized HRS includes several variables capturing whether and how respondents plan for their end of life, including wills, trusts, life insurance, and power of attorney designations. For additional details, please see our HRS webinar from March, 2022.

Along with the expansion of the core Harmonized datasets, the Gateway to Global Aging Data also developed additional Harmonized data products focused on the other questionnaires the HRS fields, including their exit interviews, Harmonized Cognitive Assessment Protocol, COVID-19 questionnaires, and life history questionnaires. Each of these are (or will be) released as separate data products that are easily merged with other Harmonized HRS data products, the RAND HRS, and original HRS data, making an even greater portion of HRS data accessible to researchers around the world.

The Harmonized HRS End of Life summarizes information collected in the exit and post exit interviews conducted with an informed proxy after the respondent has passed away. Like with the core data, it incorporates variables from all available waves. It currently includes variables from the questions most commonly asked across the HRS-INS end of life questionnaires and will expand to cover even more topics, including those specific to the HRS exit interview, in the coming years. Comparable Harmonized End of Life datasets are available for Mexico, England, Europe, South Korea, and China, with Japan coming this year, and they include a codebook section describing the differences between this data and the Harmonized HRS End of Life.

The Harmonized Cognitive Assessment Protocol (HCAP), an in-depth cognitive assessment and informant interview, was created by the HRS and adapted to a growing number of HRS-INS countries.  The Harmonized HRS-HCAP, available on the Gateway to Global Aging Data Enclave, includes variables that have been made comparably with those from Mexico, England, and India, with China and Chile coming soon. Codebooks describe differences in test administration or variable creation across all available datasets, not just with the HRS-HCAP.  These Harmonized HCAP datasets also include statistically harmonized factor score estimates of underlying general and domain-specific cognitive functioning that have been created for comparing cognition scores across studies.

When the COVID-19 pandemic began, the HRS and other HRS-INS studies quickly developed questionnaires to assess its impact on the lives of the older population. In addition to almost immediately adding questions to the ongoing core interview in 2020, the HRS also developed an additional self-completed questionnaire that was mailed out the following year. The Harmonized HRS COVID incorporates responses from both questionnaires and includes comparable variables to those in England and Europe, with Ireland and Northern Ireland coming in 2025. Additionally, the Harmonized COVID data includes (or will include) linkages to the Oxford Covid-19 Government Response Tracker policy variables that have been matched to the year and month of interview when available.

The HRS sent out life history mail surveys in 2015, 2017, and 2019, which collect detailed information on various aspects of the respondent’s life prior to joining the HRS panel. The Harmonized HRS Life History will be released soon. Like existing Harmonized data for England and Europe, it is built in a user-friendly sequence format, allowing users to know the state of the respondent’s relationships, work, etc. at each age from 18 to present.

A major goal of the Gateway to Global Aging Data is to enable cross-country and longitudinal research on aging using the network of Health and Retirement Studies. To do so, we produce various Harmonized data products using consistent variable naming and data structure, as well as detailed documentation on cross-wave and cross-study differences to reduce barriers to international research. As the HRS-INS expands to more countries over time and the questionnaires expand to cover more and more topics, the Gateway team will continue to produce resources that facilitate comparisons to the HRS, paying homage to the study that started this impressive collection of data.