Longitudinal changes in frailty and incident diabetes in middle-aged and older adults: evidence from CHARLS

Abstract

Objectives

To investigate how changes in frailty affect diabetes risk and whether sex and age influence this relationship.

Study design

This prospective cohort study, based on the China Health and Retirement Longitudinal Study (CHARLS), included 11,221 adults aged 45 and over without diabetes before the second survey. The frailty index (FI), with scores ranging from 0 to 1, was administered at baseline (2011) and during the second survey (2013), categorizing participants as robust (FI ≤ 0.10), pre-frail (FI 0.10–0.25), or frail (FI ≥ 0.25). Frailty changes were determined from the two assessments.

Main outcome measures

The primary outcome was incident diabetes. Cox regression was used to assess the association of frailty changes with diabetes and whether this association differed by sex and age (≤65 years and >65 years).

Results

During a median 7-year follow-up, 1197 developed diabetes. Those remaining pre-frail/frail had nearly double the risk of diabetes (HR 1.99, 95 % CI 1.70-2.33) compared with those remaining robust. Progressing from robust to pre-frail/frail increased risk (HR 1.80, 95 % CI 1.48-2.19), and improving from pre-frail/frail to robust still had an elevated risk (HR 1.32, 95 % CI 1.06-1.66). Higher FI tertile changes were linked to increased diabetes risk (HR 1.44, 95 % CI 1.24-1.67), with a significant trend (P < 0.001). Women were at higher risk (P = 0.004), but age did not significantly modify the association (P = 0.972).

Conclusions

Frailty progression is strongly linked to increased risk of diabetes, especially in women, highlighting the need for targeted frailty management in diabetes prevention.