exercise levels PREDICT cardiac vagal tone as measured by the neuroscope

P.S. Mckechnie, R.M. Hamilton, P.W. Macfarlane
University Department of Medical Cardiology, Queen Elizabeth Building, Glasgow Royal Infirmary,
10 Alexandra Parade, Glasgow G31 2ER, UK

Abstract: Four factors which may predict cardiac vagal tone (CVT) as measured by the NeuroScope were examined in 36 subjects: age, sex, mean resting heart rate (HR) and exercise levels. Subjects described their typical weekly exercise habits and had a 5-minute NeuroScope recording done from which average HR and CVT could be calculated. CVT values ranged from 0.6 to 43.8 units of the linear vagal scale. Stepwise multiple regression showed significant effects of exercise level and resting HR on CVT; the effects of age and sex were not significant.

INTRODUCTION

The NeuroScope is a new instrument which selectively measures cardiac vagal tone (CVT) [1]. Little has been published on the factors which determine the absolute values of resting CVT as measured by the NeuroScope. In one paper [2], the normal range was given as 5-10 units of the linear vagal scale (LVS), although it has been suggested that higher values may be obtained in athletes [3]. It was therefore decided to investigate some of the factors likely to influence CVT: age, sex, resting heart rate (HR) and exercise level.

METHODS

36 subjects were recruited, comprising Glasgow Royal Infirmary cardiac patients and students and staff of the University of Glasgow. All subjects were in sinus rhythm. The procedure was explained and once consent was obtained the subject’s age, sex and an estimate of the typical weekly amount of exercise was recorded. A fitness scoring system was developed, based on the length of time spent exercising in minutes per week and the intensity of exercise. A score of 1 was given to subjects who were unable to walk any great distance and unable to take part in any leisure exercise. Subjects who had no exercise limitation but chose not to do any leisure exercise were given a score of 2. A score of 3 was given to those who exercised at a light intensity level, or at moderate intensity for less than 100 min per week. A score of 4 was given to subjects who exercised at a moderate intensity for over 100 min per week while those who exercised at an intense level were given a score of 5.

Subjects lay supine while three electrodes (3M Red Dot) were placed on the chest (upper right, and upper and lower left chest) and attached to the NeuroScope. The NeuroScope was connected via a COM port to a laptop PC running a specialized software program, Vagusoft 3.12, which displays CVT and HR simultaneously. The NeuroScope calculates CVT from the RR intervals: the greater the rate of change of RR interval from one beat to the next, the higher the CVT.

There was a 2-min run-in period to allow adjustment to the supine posture then a 5-min recording was made using the NeuroScope. Subjects were required to remain still and in silence throughout.

Data were exported from the NeuroScope laptop into Microsoft Excel for analysis. 5-min averages were taken of CVT and HR and tabulated along with subjects' age, sex and exercise score to enable statistical analysis.

RESULTS

In the study group of 36 subjects, ten were female and 26 male and the ages ranged from 20 to 69 years. Five of the subjects were cardiac patients; the remaining 31 were healthy volunteers. Five subjects (four patients and one volunteer) had a CVT < 5, 13 subjects (one patient) had a CVT in the range 5.0 - 10.0, and 18 had a CVT > 10.0 (no patients).

The possible effect of exercise on CVT was examined by dividing subjects into two groups - those who participated in leisure exercise (exercise scores 3 - 5) and those who did not (scores 1 and 2). The means for CVT, age and HR in these two groups are shown in Table 1, together with the standard error (SEM) and the results of two-tailed t-tests for the comparison of means with unequal variances.

TABLE I
Comparison between 'leisure exercise' and
'no leisure exercise' groups

	CVT (LVS)	Age (yr)	HR (bpm)
	x          SEM	x          SEM	x         SEM
Leisure exercise	14.0      1.67	25.8       2.23	65.4     2.07
No leisure exercise	4.5        0.75	35.5       5.52	69.1     2.16
t	5.19	-1.63	-1.22
p	0.000	0.147	0.240

Table 1 shows that there was a significant difference in the mean CVT between the two different groups, those taking part in weekly leisure exercise having a mean CVT of 14.0, and those who did not having a mean CVT of 4.5. The mean ages and heart rates in the two groups were not significantly different. Nine out of 30 in the 'leisure exercise' group were female; one out of six in the 'no leisure exercise' group was female.

In order to examine the contribution of exercise to CVT more closely, CVT was plotted against exercise score as shown in Figure 1.

Figure 1. The relationship between CVT and exercise score

Figure 1 shows that as exercise level increased, CVT increased (r = 0.62). The relationship between the other possible predictors and CVT was weaker: for age and CVT, r = -0.41; and for HR and CVT, r = -0.50. The mean CVT for females (n = 10) was 13.8 ± 2.17 (SEM) and for males 11.8 ± 3.13.

Stepwise linear regression of CVT on exercise score yielded the following equation (shown in Figure 1):

                CVT = (4.90 * exercise score) - 4.51.                  (1)

Equation (1) explained 39.0% of the variance in CVT; addition of the HR term gave (2) which explained 46.5% of the variance in CVT (see Table 2):

       CVT = (3.97 * exercise score) - (0.26 * HR) + 15.61.  (2)     

Addition of further terms did not significantly improve the regression equation.

TABLE 2
Regression parameters for (1) and (2)

Equation	r2 (%)	Residual SD	p (slope)	SE (slope)
(1)	39.0	7.21	0.000	1.05
(2)   exercise HR	46.5	6.85	0.001 0.038	1.09 0.12

DISCUSSION

The previously published normal limits for CVT as measured by the NeuroScope are 5 - 10 units of the LVS in a population aged 13 - 79 years [2]. Our study provides some support for these in that 13 of our subjects had a CVT within these limits. However, 18 subjects had a CVT > 10 (none of them patients) which implies that the upper limit of normal needs to be extended. It could be argued that those with exercise score 5 (all of whom had a CVT > 10) are actually 'supra-normal' in that several of them were athletes involved in sports at a competitive level. Some of the subjects with exercise scores 3 and 4, however, also had a CVT > 10. Of those subjects with CVT < 5, four were cardiology patients and the fifth was a normal volunteer, which provides indirect support for the lower limit of normal being reasonable.

Dividing the subjects into two groups according to whether they did any leisure exercise outwith normal daily life produced interesting results. Resting HR, which is linked to cardiovascular fitness, was lower in the 'exercise' group but this was not statistically significant (Table 1). CVT, on the other hand, was significantly higher in the 'exercise' group. CVT may therefore be a more specific indicator of some aspect of cardiovascular fitness than resting HR. In a future study, 'exercise' could be differentiated into aerobic or anaerobic to investigate the specificity of CVT for 'fitness'.

Support for a causal relationship between exercise and CVT was provided by the regression analysis (Figure 1). Both CVT and HR were significant predictors of CVT as shown in Table 2 (p < 0.05 for both slope terms in (2)). Age and sex did not predict CVT and were not included in the regression equation. The error in the model could be reduced in a future study by having subjects keep an activity diary of actual exercise in a given period rather than retrospectively being asked about typical weekly exercise. 

Overall these results suggest that of the factors examined in this study, exercise score was the most predictive of CVT, followed by resting HR; age and sex were not predictive for CVT. The study also supports the contention by previous authors that CVT may be higher in athletes than in the rest of the population.

Acknowledgments:  We are grateful to the staff and students of Glasgow University and the patients of Glasgow Royal Infirmary for their help with this study.

REFERENCES

[1]    C.J.L. Little, P.O.O. Julu, S. Hansen, et al.. “Real-time measurement of cardiac vagal tone in conscious dogs,” American Journal of Physiology, vol. 276, pp. H758-H765, 1999.

[2]    P.O.O. Julu, M.O. McCarron, S. Hansen, et al.. “Selective defect of baroreflex blood pressure buffering with intact cardioinhibition in a woman with familial aniridia,” Neurology, vol. 49, pp. 1705-1707, 1997.

[3]    P.O.O. Julu. “A linear scale for measuring vagal tone in man,” Journal of Autonomic Pharmacology, vol. 12, pp. 109-115, 1992.