Frailty and healthy working life expectancy: a multicohort modelling study.
Abstract
Background
Raising the retirement age and extending working years have become key strategies for many countries to mitigate the challenges of global population ageing. However, the association of frailty, a dynamic and reversible clinical condition, with healthy working life expectancy (HWLE) remains unclear.Methods
We used data from five harmonised health and retirement cohorts from across 33 countries, from which we tracked participants aged ≥50 years from baseline until their last available data point or death. Using this data, we evaluated frailty through the Frailty Index based on the cumulative deficit model, which classifies frailty status as robust, pre-frail, or frail. Then, we used discrete-time multistate Markov models to explore the association between frailty and HWLE, and further stratified our analyses by individual and country-level factors.Results
Among 180,314 participants, HWLE at the age of 50 years was 11.87 years (95% confidence interval (CI) = 11.74-12.00) for participants in the robust group, 7.34 years (95% CI = 7.19-7.50) for participants in the pre-frail group, and 3.77 years (95% CI = 3.57-3.97) for participants in the frail group. Compared with participants in the robust group, HWLE was 4.53 years shorter for those in the pre-frail group (Bonferroni-adjusted 95% confidence interval (95% CIBA) = -4.76, -4.29) and 8.11 years shorter for those in the frail group (95% CIBA = -8.38, -7.84). Men had longer HWLE than women in the robust and pre-fail groups. Higher educational attainment and current non-smoking, as well as favourable national economic and political factors, were associated with longer HWLE across frailty strata.Conclusions
While frailty was generally associated with shorter HWLE at the age of 50 years, individual and country-level inequalities were also evident across frailty statuses. These results highlight a need to integrate targeted frailty prevention and intervention programmes with broader initiatives aimed at fostering national economic and political development, thereby promoting healthier, voluntary, and supported work participation in later life.
Related
- Country- and individual-level factors associated with healthy working life expectancy: A multinational longitudinal study
- Evaluation of the FI-Lab, a laboratory-based, automated frailty index for acute care: A multicohort study
- Cognitively healthy life expectancy of older women and men in India.
- Healthy life expectancy and its trends among Brazilian older adults from 2003 to 2019: an observational study using PNAD and PNS data
- Dietary diversity and healthy life expectancy among Chinese older adults: a multistate analysis of CLHLS, 2008-2018.
BibTeX
@article{shui2026frailty,
title = {{Frailty and healthy working life expectancy: a multicohort modelling study.}},
author = {Yuhang Shui and Jingjie Zhu and Zeyu Huang and Huazhen You and Junling Gao},
journal = {Journal of global health},
year = {2026},
month = oct,
doi = {10.7189/jogh.16.04338},
url = {https://europepmc.org/article/MED/42895270},
note = {CC BY},
}
RIS
TY - JOUR
TI - Frailty and healthy working life expectancy: a multicohort modelling study.
AU - Yuhang Shui
AU - Jingjie Zhu
AU - Zeyu Huang
AU - Huazhen You
AU - Junling Gao
PY - 2026
DA - 2026/10/19
JO - Journal of global health
DO - 10.7189/jogh.16.04338
UR - https://europepmc.org/article/MED/42895270
AB - Background Raising the retirement age and extending working years have become key strategies for many countries to mitigate the challenges of global population ageing. However, the association of frailty, a dynamic and reversible clinical condition, with healthy working life expectancy (HWLE) remains unclear.Methods We used data from five harmonised health and retirement cohorts from across 33 countries, from which we tracked participants aged ≥50 years from baseline until their last available data point or death. Using this data, we evaluated frailty through the Frailty Index based on the cumulative deficit model, which classifies frailty status as robust, pre-frail, or frail. Then, we used discrete-time multistate Markov models to explore the association between frailty and HWLE, and further stratified our analyses by individual and country-level factors.Results Among 180,314 participants, HWLE at the age of 50 years was 11.87 years (95% confidence interval (CI) = 11.74-12.00) for participants in the robust group, 7.34 years (95% CI = 7.19-7.50) for participants in the pre-frail group, and 3.77 years (95% CI = 3.57-3.97) for participants in the frail group. Compared with participants in the robust group, HWLE was 4.53 years shorter for those in the pre-frail group (Bonferroni-adjusted 95% confidence interval (95% CIBA) = -4.76, -4.29) and 8.11 years shorter for those in the frail group (95% CIBA = -8.38, -7.84). Men had longer HWLE than women in the robust and pre-fail groups. Higher educational attainment and current non-smoking, as well as favourable national economic and political factors, were associated with longer HWLE across frailty strata.Conclusions While frailty was generally associated with shorter HWLE at the age of 50 years, individual and country-level inequalities were also evident across frailty statuses. These results highlight a need to integrate targeted frailty prevention and intervention programmes with broader initiatives aimed at fostering national economic and political development, thereby promoting healthier, voluntary, and supported work participation in later life.
SN - 2047-2978
SN - 2047-2986
ER -