Dual effects of insulin resistance on mortality and function in non-diabetic older adults: findings from the Toledo Study of Healthy Aging

Abstract

Insulin signalling declines with increasing age and impacts skeletal muscle function and longevity in animal models. Our aim was to assess the relationships between insulin resistance (IR) and frailty and mortality in a unique community-dwelling cohort of older people. 991 non-diabetic subjects from the Toledo Study of Healthy Ageing (TSHA) cohort were included. IR was estimated by the homeostasis model assessment index (HOMA-IR) at baseline while frailty was determined by frailty phenotype (FP) and Frailty Trait Scale (FTS) at baseline and after 5-year follow-up. Deaths were also determined. Multivariate regression models were used to analyze the effects of HOMA-IR on outcomes. Age, gender, BMI, education level, cardio- and cerebro-vascular disease, glomerular filtration rate, and disability were included as potential confounding variables in progressive adjustment models. IR determined as increasing log HOMA-IR was inversely associated with risk of mortality. The association remained significant for all adjustment models (HR: 0.64-0.69). When we analyzed survival curves, the higher the HOMA-IR tertile, the lower the mortality rate (highest vs lowest tertile, p = 0.0082). In contrast, IR increased the risk of incident frailty determined by FP (OR 1.81 [1.14-2.87]) as well as deterioration of frailty status determined by worsening in FTS score (OR 1.28 [1.01-1.63]) at 5-year follow-up. In non-diabetic older subjects, IR significantly increases the risk for frailty and functional decline but decreased the risk of death at 5-year follow-up. This finding raises the need of assessing the effect of biomarkers on different outcomes before establishing their role as biomarkers of aging.

Keywords: Aging; Frailty; Functional performance; Insulin resistance; Mortality.

Fig. 1

Baseline insulin resistance in highest tertile increases probability of survival at 5-year follow-up. Kaplan–Meier analysis of survival probability along 5-year follow-up in non-diabetic older adults (n = 991) segregated into tertiles of homeostasis model assessment of insulin resistance (HOMA-IR) values at baseline. T1, T2, and T3 indicate tertiles with the lowest, the intermediate and the highest HOMA-IR values. Ranges of HOMA-IR values for each tertile is in parentheses. For comparison, survival probability for the same follow-up in a sub-population of 502 diabetic subjects (75.9 ± 6.1 years-old, 274 (54.6% women) from Toledo Study for Healthy Aging cohort. A log-rank pairwise comparison was performed: * indicates p < 0.01 for T3 vs T1. No significant differences were obtained in comparisons of T1 vs. T2 or T2 vs. T3. † indicates p < 0.01 vs. non-diabetic subjects, either for T1, T2 or T3