Unconditioned and conditioned effects of intravenous insulin and glucose on heart rate variability in healthy men

We examined whether an injection of intravenous insulin and intravenous glucose would affect frequency-domain measures of heart rate variability (HRV), i.e., the high-frequency (HF-) band and the ratio of the low frequency (LF-) to the HF-band in healthy humans. Using a classical conditioning protocol, we also assessed whether the measures of HRV are subject to classical conditioning. Thirty healthy men were divided into three groups, given a conditioned stimulus (CS) and an intravenous injection of either insulin (0.05 IU/kg) in Group 1, glucose (15%, 0.5 g/kg) in Group 2, or placebo (physiological saline [0.9%]) in Group 3 during the 4-day acquisition phase. All subjects were given an olfactory CS (rosewood-peppermint smell) and placebo injection on day 5 (test). Due to their high inter-individual variability, HF and LF/HF-ratio were analysed by intragroup comparisons, using a pre-injection baseline interval (min -15 to -5), and three functional post-injection intervals: a) the interval to the maximum insulin level, i. e. insulin peak (min 0-5) in Groups 1 and 2, b) the interval to the maximum of insulin-induced hypoglycaemia (min 20-25) in Group 1, and c) the end of the session (min 70-75). On days 1 to 4, we found significant increases of the HF-band from baseline to interval min 0-5 in Group 1, and an even more pronounced increase in the glucose-treated Group 2. At the test (Day 5), both experimental groups responded with an HF-increase in the interval of the former insulin peak, and also at the other measurement intervals, reflecting some general increase of vagal activity remaining as a conditioned response. On days 1 to 4, the HF-band was positively correlated with the change of peripheral insulin levels in Group 1, reaching statistical significance on days 3 and 4. This pattern only emerged in tendency on Day 4 in Group 2. In conclusion, insulin triggers an increase in parasympathetic tone at maximum hyperinsulinaemia, and our data support the notion that this response pattern can become classically conditioned.