Prediabetes and glycemic transitions as determinants of frailty and functional decline in adults aged 50 years and older: A longitudinal analysis from five multinational aging cohorts
Malagon-Liceaga, A.; Basile-Alvarez, M. R.; Fermin-Martinez, C. A.; Ramirez-Rivera, D. L.; Perezalonso Espinosa, J.; Diaz-Sanchez, J. P.; Garcia-Gonzalez, S. B.; Carrillo-Herrera, K. B.; Cabrera-Quintana, L. A.; Antonio-Villa, N. E.; Gomes-Goncalves, N.; Garcia-Pena, C.; Bello-Chavolla, O. Y.
Show abstract
BackgroundPrediabetes is highly prevalent in older adults and is characterized by heterogeneous clinical trajectories, including regression to normoglycemia and progression to diabetes. While prediabetes has been associated with impaired physical function and frailty, the longitudinal impact of both a single diagnosis and dynamic glycemic transitions on functional outcomes remains unclear. We aimed to evaluate associations between baseline prediabetes and glycemic transitions over time with trajectories of functional capacity and frailty in older adults. MethodsWe conducted a pooled analysis of harmonized data from five nationally representative longitudinal aging cohorts (MHAS, HRS, CHARLS, ELSA, CRELES) within the Gateway to Global Aging Data, including adults aged [≥]50 years with [≥]1 HbA1c measurements. Prediabetes was defined per ADA criteria (HbA1c 5.7-6.4%). Functional outcomes included activities of daily living (ADL), instrumental ADL (IADL), and frailty assessed using Fried phenotype, FRAIL scale, and a deficit-accumulation Frailty Index (FI). Mixed-effects Poisson models estimated incidence rate ratios (IRRs) for baseline prediabetes, while generalized estimating equations assessed time-varying glycemic status and transition trajectories. Models were adjusted for age, sex, cohort, and time-varying covariates, with sensitivity analyses including BMI, smoking, and alcohol intake. FindingsAmong 18,571 participants (median follow-up 13.6 years), baseline prediabetes was associated with increased progression of functional deficits and frailty compared with normoglycemia, including higher FI values and accelerated FI progression. Prediabetes was associated with higher incidence of ADL, IADL, and multimorbidity deficits from early follow-up, although time-dependent changes in incidence rates were not significant. In time-varying analyses (n=7,840), both prediabetes and diabetes were associated with higher incidence of functional deficits compared with normoglycemia, with diabetes showing the strongest effects across all outcomes. Diabetes was associated with greater FI burden and accelerated progression, whereas prediabetes showed a smaller increase, with attenuation over time. Among individuals with baseline prediabetes, regression to normoglycemia occurred in 20.8% and was associated with increased incidence of ADL and frailty deficits. In contrast, progression to diabetes occurred in 24.3%, and was associated with lower risk of incident ADL and Fried frailty deficits compared to stable prediabetes. InterpretationPrediabetes is associated with increased risk of functional decline, frailty, and deficit accumulation in older adults, independent of progression to diabetes. Regression to normoglycemia was associated with higher risk of functional deterioration. These findings suggest that prediabetes reflects a state of metabolic vulnerability linked to biological aging rather than solely a precursor to diabetes and highlights a need to reframe its clinical significance in older populations. FundingThis research was supported by Instituto Nacional de Geriatria in Mexico. RESEARCH IN CONTEXTO_ST_ABSEvidence before this studyC_ST_ABSWe searched PubMed up to April 1, 2026, using the terms ("prediabetes") AND ("frailty" OR "functional decline" OR "activities of daily living" OR "mortality") AND ("older adults" OR "aging" OR "longitudinal" OR "cohort"). Existing evidence from predominantly single-country and cross-sectional studies suggests that prediabetes is associated with increased risk of frailty, disability, and adverse health outcomes in older adults. However, findings have been inconsistent, particularly regarding incident versus prevalent outcomes and the role of competing risks in aging populations. Importantly, few studies have used harmonized multinational longitudinal data, incorporated repeated measures of glycemic status, or evaluated multiple complementary frailty constructs. The impact of dynamic glycemic transitions, including regression to normoglycemia and progression to diabetes, on functional trajectories remains poorly understood. Added value of this studyThis study leverages harmonized longitudinal data from five nationally representative aging cohorts from Mexico, England, China, Costa Rica and the United States to examine both baseline prediabetes and time-varying glycemic status in relation to trajectories of functional decline and frailty. By integrating multiple validated measures of functional capacity and frailty, and applying longitudinal modeling strategies, we provide a comprehensive assessment of the relationship between prediabetes and age-related outcomes. We further characterize glycemic transition trajectories and show that regression to normoglycemia among older adults with prediabetes is associated with increased risk of frailty, while progression to diabetes does not uniformly confer additional risk beyond stable prediabetes. Implications of all the available evidenceOur results support a reconceptualization of prediabetes in older adults as a marker of systemic vulnerability rather than solely a precursor to diabetes. Glycemic status in later life appears to also reflect underlying physiological reserve, with dynamic changes potentially signaling deterioration rather than improvement. Our findings suggest a need to incorporate glycemic markers into geriatric risk stratification and to interpret their changes in the context of aging biology. Future research should focus on elucidating the mechanisms linking dysglycemia with frailty and functional decline, including the roles of sarcopenia, inflammation, and multimorbidity, and on evaluating interventions tailored to older populations across diverse global contexts.
Matching journals
The top 2 journals account for 50% of the predicted probability mass.