Type 2 diabetes (T2D) is associated with frailty in older people. We aim to explore changes in frailty status after ≥18 months of intervention.

298 community-dwelling older adults (>70 years) participating in MIDFRAIL followed-up for 18–24 months were randomly allocated by trial site (cluster) to intervention IG (16-weeks resistance exercise program, nutritional-educational sessions, optimization of diabetes care), or usual care group (UCG). Frailty status was assessed by the Fried Frailty Phenotype criteria at baseline and final visit. Functional status was assessed by the SPPB in every visit. We used multivariate linear and logistic regression for continuous and dichotomous outcomes. This study was registered at Clinicaltrials.gov (NCT01654341).

Mean age was 77.7 (SD 5.54), 47% were male, 32.9% frail and 67.1% prefrail. The probability of improving the frailty status and decreasing the number of Fried’s frailty criteria was higher in the IG than in the UCG (OR 2.6, 95%CI 1.3–5.4; p = 0.009 and OR 1.9; 95%CI 1.1−3.1; p = 0.02, respectively). IG participants more frequently improved ≥1 point in SPPB score (OR 1.85; 95%CI 1.09–3.12; p = 0.022). These benefits were mainly accounted for the prefrail participants.

The MIDFRAIL intervention improved frailty status and physical function at long-term follow-up in older people with T2D.