Response rates and associated factors after a multicomponent intervention in frail older adults with diabetes

Background: Type 2 diabetes mellitus (T2DM) and frailty are associated with functional decline in older population. Objective: To explore the individual response to a multimodal intervention on functional performance. Design: A cluster-randomised multicentre clinical trial. Setting: Outpatients in hospital or primary care. Subjects: 843 (77.83 years, 50.65% men) prefrail and frail individuals ≥70 years with T2DM. Methods: Participants were allocated to usual care group (UCG) or a multicomponent intervention group (IG): 16-week progressive resistance training, seven nutritional and diabetological educational sessions and achievement of glycated haemoglobin (7-8%) and blood pressure (<150 mmHg) targets. Functional performance was assessed with the Short Physical Performance Battery (SPPB) at 1 year. We used multivariate binomial and multinomial logistic regression models to explore the effect of the IG, and adherence on the outcomes studied, in several adjusted models. Results: 53.7% in the IG versus 38.0% in the UCG improved by at least 1 point in their SPPB score [OR (95% CI): 2.07 (1.43, 2.98), P value <0.001]. Age, SPPB score and number of frailty criteria met decreased the probability of improving the SPPB score. Factors associated with worsening were pertaining to IG (decreased), age, SPPB score and the number of frailty criteria (increased). An adherence ≥84% was needed to achieve benefits, reaching the peak in the probability of improving SPPB when this was ≥85% [OR(95%CI): 2.38 (1.29, 4.79), P value 0.014]. Conclusions: Factors predicting the likelihood of improvement in a multimodal programme in pre-frail and frail older adults with diabetes are age, basal SPPB score, the number of frailty criteria and adherence.