POWER SPECTRAL-ANALYSIS OF HEART-RATE IN HYPERTHYROIDISM

The aim of the present study was to evaluate the impact of hyperthyroi dism on the cardiovascular system by separately analyzing the sympathe tic and parasympathetic influences on heart rate. Heart rate variabili ty was evaluated by autoregressive power spectral analysis. This metho d allows a reliable quantification of the low frequency (LF) and high frequency (HF) components of the heart rate power spectral density; th ese are considered to be under mainly sympathetic and pure parasympath etic control, respectively. In 10 newly diagnosed untreated hyperthyro id patients with Graves' disease, we analyzed power spectral density o f heart rate cyclic variations at rest, while lying, and while standin g. In addition, heart rate variations during deep breathing, lying to standing, and Valsalva's maneuver were analyzed. The results were comp ared to those obtained from 10 age-, sex-, and body mass index-matched control subjects. In 8 hyperthyroid patients, the same evaluation was repeated alter the induction of stable euthyroidism by methimazole. H eart rate power spectral analysis showed a sharp reduction of HF compo nents in hyperthyroid subjects compared to controls [lying, 13.3 +/- 4 .1 us. 32.0 +/- 5.6 normalized units (NU; P < 0.01); standing, 6.0 +/- 2.7 vs. 15.0 +/- 4.0 NU (P < 0.01); mean +/- SEM]. On the other hand, LF components were comparable in the 2 groups (lying, 64.0 +/- 6.9 vs . 62.0 +/- 6.5 NU; standing, 77.0 +/- 6.5 vs. 78.0 +/- 5.4 NU). Hence, the LF/HF ratio, which is considered an index of sympathovagal balanc e, was increased in hyperthyroid subjects while both lying (11.3 +/- 4 .5 us. 3.5 +/- 1.1; P < 0.05) and standing (54.0 +/- 12.6 vs. 9.8 +/- 2.6; P < 0.02). This parameter was positively correlated with both T-3 (r = 0.61; P < 0.05) and free T-4 (r = 0.63; P < 0.05) serum levels. Among traditional cardiovascular autonomic tests, the reflex response of heart rate during lying to standing was significantly lower in hype rthyroid patients than in controls (1.12 +/- 0.03 us, 1.31 +/- 0.04; P < 0.002). No statistically significant difference in reflex responses between the two groups was found in deep breathing or Valsalva's mane uver. In the 8 patients reexamined after methimazole treatment, we obs erved complete normalization of altered cardiovascular parameters, wit h slight predominance of the vagal component compared with controls. T hese results suggest that thyroid hormone excess may determine reduced parasympathetic activity and, thus, a relative hypersympathetic tone.