High intensity interval training (HIIT) produces small improvements in fasting glucose, insulin, and insulin resistance in sedentary older men but not masters athletes

The aim of this study was to investigate whether 6 weeks' high intensity interval training (HIIT; 6 × 30 s sprints at 40% peak power, once every five days) following 6 weeks' of aerobic preconditioning could favourably affect fasting insulin, glucose, and the homeostatic model assessment of insulin resistance (HOMA1-IR) in sedentary older men compared with masters athletes. A secondary aim was to establish whether lifelong exercisers (LEX) exhibited improved fasting insulin, glucose, and HOMA1-IR, compared to sedentary older males (SED). Twenty-two males (62 ± 2 years) comprised the SED group and 17 males (60 ± 5 years) were enrolled as LEX. Participants were tested at phase A (baseline), B (after preconditioning), and C (post-HIIT). There was no effect of time (P = 0.116) or interaction (P = 0.727) on insulin. However, there was an effect of group (P < 0.001). In terms of magnitude, HIIT induced a small decrease in SED insulin compared to baseline (15.8 ± 8.1 uIU·ml⁻¹ at baseline and 14.0 ± 7.8 uIU·ml⁻¹ post-HIIT; Cohen's d = 0.23) and compared to post-preconditioning (17.5 ± 9.7 uIU·ml⁻¹; Cohen's d = 0.40). LEX insulin was unchanged throughout (all differences were trivial). Insulin was lower in LEX than SED at phase A (P < 0.001, Cohen's d = 1.31), B (P = 0.023, Cohen's d = 0.78), and C (P = 0.004, Cohen's d = 1.01). There was no effect of time (P = 0.290), group (P = 0.166), or interaction (P = 0.153) for glucose. In terms of magnitude, HIIT produced a small reduction in SED glucose compared to baseline (5.7 ± 1.3 mmol·l⁻¹ at baseline and 5.3 ± 0.9 mmol·l⁻¹ post-HIIT; Cohen's d = 0.36), and compared to phase B (5.6 ± 0.8 mmol·l⁻¹, Cohen's d = 0.35). LEX glucose was unchanged throughout (all changes were trivial). SED had moderately higher blood glucose than LEX at phase A (Cohen's d = 0.49), and B (Cohen's d = 0.63), but only a trivial difference existed at phase C (Cohen's d = 0.15). There was no effect of time (P = 0.110), or interaction (P = 0.569) on HOMA1-IR. However, there was an effect of group (P = 0.002). In terms of magnitude, SED HOMA1-IR was unchanged from phase A to B (4.2 ± 3.0 and 4.5 ± 2.9 arbitrary units respectively [Cohen's d = 0.10]). However, at C (3.5 ± 2.6) there was a small decrease compared to B (Cohen's d = 0.36), and A (Cohen's d = 0.25). LEX experienced a small increase in HOMA1-IR from phase A to B (1.6 ± 1.3 and 2.3 ± 2.8 respectively [Cohen's d = 0.32]), followed by a small decrease from B to C (1.7 ± 1.1 at phase C [Cohen's d = 0.28]), and a trivial change from A to C (Cohen's d = 0.08). HOMA1-IR was lower in LEX than SED at baseline (P = 0.002, Cohen's d = 1.12), after preconditioning (P = 0.024, Cohen's d = 0.77), and post-HIIT (P = 0.014, Cohen's d = 0.90). Results of this study provide preliminary evidence that HIIT preceded by preconditioning can induce small improvements in fasting insulin, glucose, and HOMA1-IR in sedentary older men compared with masters athletes.