Associations of Summary Measures of Longitudinal Frailty Assessments with Health Outcomes

Benjamin Seligman; Dae Hyun Kim; Ariela Orkaby

doi:10.1093/gerona/glaf161

Associations of Summary Measures of Longitudinal Frailty Assessments with Health Outcomes

J Gerontol A Biol Sci Med Sci. 2025 Sep 26:glaf161. doi: 10.1093/gerona/glaf161. Online ahead of print.

Authors

Benjamin Seligman^{1

2}, Dae Hyun Kim^{3

4}, Ariela Orkaby^{5

6}

Affiliations

¹ Geriatric Research, Education, and Clinical Center, VA Greater Los Angeles Health Care System, Los Angeles, CA, USA.
² Division of Geriatric Medicine, Department of Medicine, UCLA David Geffen School of Medicine, Los Angeles, CA, USA.
³ Hinda and Arthur Marcus Institute for Aging Research, Hebrew SeniorLife, Boston, MA.
⁴ Harvard Medical School, Boston, MA, USA.
⁵ New England Geriatric Research, Education, and Clinical Center, VA Boston Health Care System, Boston, MA, USA.
⁶ Division of Aging, Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA.

PMID: 41003684
DOI: 10.1093/gerona/glaf161

Abstract

Background: Assessment of frailty has become common in clinical settings to risk-stratify older adults. Understanding how to use repeated measurements answers important questions both for the clinical use of serial assessments and understanding frailty trajectories.

Methods: Using 2012-16 Health and Retirement Study data, we calculated six summary measures of assessments of frailty index (FI, 3 assessments) and Fried Frailty Phenotype (FFP, 2 assessments): most recent value, maximum, minimum, mean, standard deviation (SD), and delta. We assessed the association of scaled values with mortality and institutionalization between 2016 and 2018, and three measures of epigenetic aging collected in 2016 using Cox, logistic, and linear regression respectively with adjustment for age, sex, and smoking. We then used LASSO regression to determine which summary measures were most often retained.

Results: 14,451 and 2,196 individuals had complete data for FI and FFP respectively. The maximum and most recent frailty values had the strongest associations with the outcomes considered, while those of SD and delta were weakest. In LASSO regressions, the maximum and most recent values were most commonly retained (12-13 of 20 regressions), followed by SD (7), mean (6), and minimum and delta (1 each).

Conclusions: These findings show that maximum and most recent values of frailty tend to be most strongly associated with mortality. For clinicians, this means that the most recent assessment may be sufficient for many purposes and if historical data are unavailable.

Funding: National Institute on Aging; Office of Research and Development Department of Veterans Affairs).

Keywords: Frailty; Longitudinal; Machine Learning.

© The Author(s) 2025. Published by Oxford University Press on behalf of the Gerontological Society of America. All rights reserved. For commercial re-use, please contact reprints@oup.com for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site—for further information please contact journals.permissions@oup.com.