The effects of insulin infusion on cardiac autonomic nervous system (ANS) a
ctivity were investigated in healthy subjects (n = 15) and in patients with
various types of insulin-resistance, such as obese subjects (n = 20) and t
hose with hypertension (n = 15) or type II (non-insulin-dependent) diabetes
(n = 22). Healthy subjects and patients underwent euglycaemic hyperinsulin
aemic glucose clamp, and cardiac ANS and haemodynamic changes were investig
ated by continuous recording of heart rate variability by the Holter techni
que and by venous occlusion plethysmography respectively. At baseline, heal
thy subjects had the highest values for total spectral power and the low-fr
equency (LF) component, and the lowest value for the high-frequency (HF) co
mponent. In the pooled data (n = 72), the fasting plasma insulin concentrat
ion was correlated with baseline total spectral power (r = -0.37; P < 0.001
) and the LF/HF ratio (r = -0.35; P < 0.003). Such correlations were still
significant (P < 0.01 for both) after adjustment for body fat and mean arte
rial blood pressure. In a multivariate linear stepwise analysis (n = 72), a
model including body fat, waist/hip ratio, fasting plasma glucose concentr
ation and insulin-mediated glucose uptake explained 47% of the variability
of the change in the LF/HF ratio, with body fat (t = -3.11; P < 0.01) and i
nsulin-mediated glucose uptake (t = -3.48; P < 0.008) being significantly a
nd independently associated with insulin-mediated changes in the LF/HF rati
o. Insulin infusion reduced the total spectral power and increased the LF/H
F ratio in healthy subjects, but not in insulin-resistant patients. In conc
lusion, our study demonstrates that insulin fails to stimulate cardiac ANS
activity in insulin-resistant patients, independently of the causes of insu
lin resistance.