Limited utility of the phenylephrine-nitroprusside sigmoid curve method of measuring baroreflex function after myocardial infarction

Abstract

Background: Conventional testing of baroreflex function with phenylephrine bolus injection measures reflex vagal tone.

Objective: To evaluate an alternative pharmacological method of baroreflex testing, which might provide more comprehensive evaluation of the baroreflex function in patients with uncomplicated myocardial infarction.

Methods: Forty-eight patients, aged 59 +/- 7 years, were evaluated 5-7 days after myocardial infarction and after 6 weeks of rehabilitation. Baroreflex testing was performed with 10-14 incremental injections of nitroprusside and phenylephrine. The peak heart rate and blood pressure point from each injection were fitted to a four parameter (upper plateau, lower plateau, gain and median blood pressure) sigmoid logistic function. The baroreflex sensitivity (BRS) was also measured by conventional linear analysis of the response of the R-R interval to the initial rise in systolic blood pressure induced by phenylephrine bolus injection.

Results: Non-linear curve fitting of four logistic curve parameters was possible for results from 91% of tests; in the remaining tests the upper plateau was held constant at the maximum heart-rate response to allow estimation of the other three parameters. When all four parameters had been estimated, the gain parameter could not be precisely determined (the coefficient of variation of the gain parameter was 85 +/- 10%). The upper (R = 0.72, P < 0.001) and lower plateaux (R = 0.76, P < 0.001) were strongly related to the resting heart rate and weakly related to measures of heart-rate variability. There was also a significant inverse correlation between the lower plateau and the BRS (R = -0.57, P < 0.001). The gain parameter was not related to the BRS or any measure of the heart-rate variability. After 6 weeks' rehabilitation there was a significant decrease in the lower plateau (from 54 +/- 2 to 48 +/- 1 beats/min, P < 0.001), an increase in BRS (from 9.4 +/- 0.8 to 12.1 +/- 0.8 ms/mmHg, P < 0.001) and no change in the other three sigmoid curve parameters.

Conclusions: Non-linear curve fitting of the heart-rate-blood-pressure relationship for patients after myocardial infarction is feasible but technical limitations and the lack of correlation between the gain parameter and other measures of autonomic function suggest that this method has limited usefulness. The lower plateau is related strongly to the BRS; both are vagal measures of cardiac autonomic function.