Feasibility of home-based tracking of insulin resistance from vascular stiffness estimated from the photoplethysmographic finger pulse waveform

Abstract

Objective. In this study we explored the utility of post-prandial vascular stiffness as a surrogate measure for estimating insulin resistance, which is a pre-diabetic condition. Approach. A cohort of 51 healthy young adults with varying body mass index (BMI) values was studied using fasting plasma values of insulin and glucose, fasting and post-meal finger photoplethysmography (PPG) and electrocardiogram (ECG). Insulin resistance was estimated by homeostatic model assessment for insulin resistance 2 (HOMA-IR2) using fasting plasma insulin and glucose. Vascular stiffness was estimated by reciprocal of pulse arrival time (rPAT) from ECG and finger PPG at five time points from fasting to 2 h post-oral glucose ingestion. We examined if insulin resistance correlates with meal-induced vascular stiffness changes, supporting the feasibility of using finger PPG to estimate insulin resistance. Main results. HOMA-IR2 was positively correlated with an early rise (0 to 30 min post-meal) and delayed fall (30 to 120 min post-meal) of rPAT. Correlation persisted even after the effect of BMI has been partialled out in subgroup analysis. We conclude that finger PPG-based pulse waveform and single-lead ECG has the potential to be used as a non-invasive method for the assessment of insulin resistance. Significance. As both signals, namely ECG and PPG, can be easily acquired using wearable and other low-cost sensing systems, the present study can serve as a pointer to develop accessible strategies for monitoring and longitudinal tracking of insulin resistance in health and pathophysiological states.