Abstract
Introduction: The representation of frailty in type 2 diabetes trials is unclear. This study used individual patient data (IPD) from trials of newer glucose-lowering therapies to quantify frailty and assess the association between frailty and treatment efficacy and adverse events.
Methods: We analysed IPD from 34 trials of sodium glucose cotransporter 2 inhibitors, glucagon-like peptide-1 receptor analogues and dipeptidyl peptidase-4 inhibitors. Frailty was quantified using a cumulative deficit frailty index (FI). For each trial we assessed the distribution of the FI; interactions between frailty and treatment efficacy (HbA1c and major adverse cardiovascular events [MACE]); and associations between FI and non-completion, adverse events, and hypoglycaemic episodes before pooling results using random-effects network meta-analysis.
Findings: Trial participants numbered 25,208. Mean age 53.8-74.2 years. Median frailty prevalence (FI>0.24) was 1.9% (IQR 0.8% to 6.1%). There was no heterogeneity in treatment efficacy by FI for MACE or HbA1c in the primary analysis (high uncertainty for MACE). A 0.1-point increase in the FI was associated with adverse events (incidence rate ratio, IRR 1.43, 95% confidence interval 1.34-1.53), treatment-related adverse events (1.35, 1.22-1.50), serious adverse events (2.04, 1.80-2.30), hypoglycaemia (1.18, 1.04-1.34), MACE (hazard ratio 3.02, 2.49-3.68) and early withdrawal (odds ratio 1.45, 1.30-1.62).
Conclusions: Frailty is associated with similar efficacy of treatment but with greater incidence of both adverse events and MACE. Frailty was rare in most trials. While these findings support calls to relax HbA1c-based targets in people living with frailty, they also highlight the need for inclusion of people living with frailty in trials.