Cognition Data

Measurement of cognition has been part of the core HRS survey since the beginning in 1992. Two separate supplemental studies provide in depth cognition measurement. Four researcher contributed datasets using the core data are also available.

These sources are listed below with relevant documentation. The cognition data in the core are available as a part of the public data. The HCAP and ADAMS data are available as sensitive data products, which requires an extra step to access.

Core interview data (1992-2018)

User Guides

Harmonized Cognitive Assessment Protocol (HCAP)

HRS-HCAP is part of an ongoing international research collaboration (HCAP Network) funded by the National Institute on Aging to measure and understand dementia risk within ongoing longitudinal studies of aging around the world that have aimed to harmonize methods and content to facilitate cross-national comparisons.

Data Product

User Guide

Aging, Demographics, and Memory Study (ADAMS)

ADAMS is a supplemental study in the HRS that conducted in-person clinical assessments to gather information on their cognitive status. A diagnosis of dementia, cognitive impairment but not demented (CIND), or non-case was assigned on the basis of this assessment.

Data Product

User Guide

Researcher Contributions

  • Langa-Weir Classification of Cognitive Function. A researcher contributed dataset that provides a total summary score for cognition using measures from the core HRS interview as well as three derived categories: Normal, Cognitively Impaired but not Demented (CIND), and Demented, described in Crimmins et al. (2011).

  • Dementia Predicted Probabilities Files. A researcher contributed dataset that provides predicted probabilities of dementia in respondents aged 70 and older in the HRS based on a methodology developed for the study of the monetary cost of dementia in Hurd et al. (2013).

  • Gianattasio-Power Predicted Dementia Probability Scores and Dementia Classifications. A researcher contributed data file containing predicted dementia probabilities and classifications for 2000-2014 HRS respondents aged 70+ with self-reported race/ethnicity non-Hispanic white, non-Hispanic black, or Hispanic, using three newly developed algorithms.

  • Predicted Cognition and Dementia Measures. This is a researcher contributed dataset that provides predicted cognition and dementia measures developed in Hudomiet, Hurd, and Rohwedder (2022). This dataset contains five measures: 1) estimated probability of dementia of the person-year observation, 2) estimated probability of “cognitive impairment, not dementia” (CIND) 3) estimated probability of normal cognitive status (i.e., neither dementia nor CIND), 3) expected value of latent cognitive ability, and 4) the standard deviation of latent cognitive ability (i.e., the precision of the Cog estimate).

Disparities in dementia rates are an area of growing research and policy interest.

From the beginning, a key element of the HRS sample design has been oversampling of African American and Hispanic populations in each new cohort to serve the important goal of supporting racial and ethnic disparities research (Jackson, Lockery, & Juster, 1996). Ofstedal and Weir (2011) demonstrated that HRS has been successful in both recruiting and retaining African American and Hispanic populations. Building on this success, HRS implemented additional screening in the 2010 and 2016 cohort enrollments to expand the number of minority participants, with similar plans for 2022 underway. The current active sample of approximately 22,500 includes approximately 4,700 African-American and 2,600 Hispanic participants. With the addition of data from the Harmonized Cognitive Assessment Protocol (HCAP), HRS is now also a major resource for studying racial and ethnic disparities in Alzheimer’s Disease and Alzheimer’s Related Dementias (AD/ADRD). HCAP was specifically designed to enable epidemiological research on variation in the incidence, prevalence, and impact of AD and related dementias (Langa et al., 2019).


Jackson JS, Lockery SA, F. Juster T. Minority perspectives from the Health and Retirement Study. Introduction: health and retirement among ethnic and racial minority groups. Gerontologist. 1996;36(3):282-4.

Langa KM, Ryan LH, McCammon RJ, et al. The Health and Retirement Study Harmonized Cognitive Assessment Protocol Project: Study Design and Methods. Neuroepidemiology. 2019.

Ofstedal MB, Weir DR. Recruitment and retention of minority participants in the health and retirement study. Gerontologist. 2011;51 Suppl 1:S8-20.

Video Tutorial



Health and Retirement Study: Data on Cognition

A webinar took place in March 2018 that provides an overview of these data resources with information on how to get started using them.

The Harmonized Cognitive Assessment Protocol: A New HRS Data Resource

A webinar held in March 2020 that provides an overview of the design and content of the HCAP study followed by a question and answer portion.