Transcript
Clinical Science (1996) 91, 59-64 (Printed i n Great Britain)
59
Abnormal baroreceptor-cardiac reflex sensitivity is not detected by conventional tests of autonomic function in patients with insulidependent diabetes me1Iitus Philip J. WESTON, Martin A. JAMES*, Ronnie PANERAIT, Paul G. McNALLY, John F. POllER*, Herbert THURSTON and John D. SWALES Department of Medicine and Therapeutics, Clinical Sciences Building, Leicester Royal Infirmary, Leicester, *Division of Medicine for the Elderly, Glenfield General Hospital, Leicester, and ?Department of Medical Physics, Leicester Royal Infirmary, U.K. (Received 6 November 1995/1 I March 1996; accepted 15 March 1996)
~
1. Autonomic neuropathy is a common complication of diabetes mellitus and is a major risk factor for sudden death. 2. A group of 30 patients with insulin-dependent diabetes mellitus and 30 age-, sex- and blood pressure-matched control subjects underwent traditional tests of autonomic function. Resting supine R-R interval and systolic blood pressure variability were assessed using spectral analysis methods. In addition, we assessed the baroreceptor-cardiac reflex from the linear relation of the change in R-R interval to the increasing systolic blood pressure measured noninvasively with the Finapres monitor during phase 4 of the Valsalva manoeuvre and from resting heart rate and systolic blood pressure power spectra. 3. Whereas standard tests of autonomic function revealed no differences between patients with insulindependent diabetes mellitus and control subjects, there was a significant reduction in power spectral density of heart rate variability around the highfrequency region (125.2 & 112.9 versus 459.3 & 189.8 ms2, mean & SD). Furthermore, reductions in baroreflex sensitivity calculated from the Valsalva manoeuvre were detected in diabetics compared with versus 9.5 & 2.5 ms/mmHg, controls (3.3f 1.6 mean f SD, P< 0.00001). There were significant relations between impairment of the baroreflex and duration of diabetes ( PC 0.001) and poor diabetic control ( P < 0.05). 4. In summary, autonomic dysfunction occurs much more frequently in diabetic patients than conventional tests would suggest. Abnormal baroreceptorcardiac reflex sensitivity in patients with insulindependent diabetes mellitus may in part be explained by abnormal parasympathetic function. This unrecognized abnormality may have a role in the increased incidence of sudden death seen in young diabetic subjects.
INTRODUCTION
Autonomic dysfunction has been reported in 2040% of unselected diabetic patients when assessed using standard bedside tests [11. Severely impaired cardiac autonomic function in diabetics is associated with a poor prognosis and an increased risk of sudden death [2]. For the past 25 years evidence for autonomic dysfunction has been sought by using a battery of non-invasive cardiovascular reflex tests [l]. These include the ratio of maximal and minimal heart rate variation during deep breathing (I/E ratio), the ratio of maximum and minimum R-R interval during the Valsalva manoeuvre, the heart rate response to standing (30:15 ratio) and the blood pressure responses to standing and sustained hand grip. Using spectral analysis methods to study heart rate variability allows quantification of individual vagal and sympathetic influences on the heart [3, 41. The power spectrum of heart rate variability has been shown to consist of three major peaks: low, mid and high frequency. Heart rate fluctuations in the mid-frequency band (0.05-0.15 Hz) are said to represent predominantly sympathetic function with a contribution from the parasympathetic system, whereas the high-frequency peak (0.2-0.35 Hz) is said to reflect vagal, or parasympathetic, activity [S]. Furthermore, baroreceptor-cardiac reflex sensitivity can be assessed from the simultaneous spectral analysis of R-R interval and systolic blood pressure variations [6, 71. In this study, the baroreceptorcardiac reflex from the spectral data is represented by the square root of the ratio of the powers of RR interval and systolic blood pressure [6, 81. This method of evaluating baroreceptor-cardiac reflex sensitivity (BRS) has proved similar to more established methods of assessing BRS, e.g. measuring the increase in R-R interval reflexly produced in res-
Key words: autonomic function, baroreflex sensitivity, insulindependent diabetes mellitus, sudden death. Abbreviations: BP, blood pressure; BRS, baroreceptor-cardiac reflex sensitivity; IDDM, insulindependent diabetes mellitus. Correspondence: D r P. Weston, Department of Medicine and Therapeutics, Clinical Sciences Building, Leicester Royal Infirmary, Leicester LE2 7LX, U.K.
60
P. J. Weston
ponse to acute pharmacologically induced changes in blood pressure [9]. Previously, this technique has required arterial cannulation for the measurement of beat-to-beat changes in blood pressure (BP). However, this limitation has recently been overcome with the development of non-invasive beat-to-beat BP measurement systems such as the Finapres device [lo]. This apparatus uses an inflatable finger cuff equipped with a infrared photoplethysmograph to measure the finger artery blood volume under the cuff. The observed blood pressure is similar to that obtained by brachial intra-arterial measurement with an offset that remains fairly constant during physiological fluctuations such as those seen during the Valsalva manoeuvre [ 111. BRS can also be measured from the relation of BP change to R-R interval lengthening during phase 4 of the Valsalva manoeuvre (from the lowest BP after release of the strain to the peak BP observed a few seconds later) as well as from resting heart rate and blood pressure data. The Valsalva method has also proved to be an accurate measure of BRS by comparison with traditional pharmacological methods [12, 131. Accordingly, we examined a group of patients with insulin-dependent diabetes mellitus (IDDM) using standard tests of autonomic function and looked at individual components of the autonomic nervous system using spectral analysis of heart rate variability. Finally, BRS was assessed using these non-invasive methods. The aim of the study was to assess if BRS testing revealed abnormalities in autonomic function before traditional tests of autonomic function became abnormal. Furthermore, we aimed to identify the possible site of abnormalities in the baroreflex. PATIENTS A N D METHODS Subjects Thirty patients with IDDM aged between 20 and 54 (mean+SD 34.8k5.4 years) and 30 age-, sexand BP-matched non-diabetic control subjects were studied. Patients were recruited from the diabetic clinic at Leicester Royal Infirmary and all gave informed consent. The study was approved by the local ethics committee. The diabetic patients all had a history of ketonuria at the time of diagnosis. Patients with evidence of microvascular disease (i.e. those with microalbuminuria, retinopathy or clinical evidence of neuropathy) were excluded, as were those with clinical evidence of peripheral vascular disease or a history of cerebrovascular disease. Patients were all normotensive with clinic BP of 140/90 on at least three occasions and were not taking medication other than insulin. Control subjects were recruited from volunteers among the hospital staff. All had a random blood glucose of 6mmol/l and were excluded if they had any underlying medical conditions or if clinic BP as 140/90.
et
al.
Protocol
Subjects attended the test laboratory at least 2 h after a meal, and the tests were performed in a quiet room with the temperature controlled between 20 and 24°C. All subjects had height and weight measured, from which body mass index was calculated, and waist/hip ratio was also measured. After resting for 10min the subjects performed the five standard non-invasive tests of autonomic function [l]. All subjects performed the Valsalva manoeuvre by performing a forced expiration at 40mmHg for 15s: the ratio of the maximum to minimum R-R interval is the Valsalva ratio. The subjects also breathed deeply at six breaths per minute: the mean difference between maximum and minimum heart rates for six breaths was taken as the inspiration/expiration (I/E) ratio. The heart rate response to standing was assessed by asking the subjects to rise from supine, and the ratio of the longest R-R interval (usually around 30th beat from standing) to the shortest (around the 15th beat) is the 30:15 ratio. Postural changes in systolic blood pressure were assessed by measuring the difference between supine and standing systolic blood pressure. Finally, the blood pressure response to hand grip was assessed by asking the subjects to maintain 30% of maximum hand grip for up to 5min using a hand grip dynamometer. The blood pressure was measured each minute and the difference between diastolic BP just before release of the hand grip and before starting was recorded. The Valsalva manoeuvre was then repeated with the Finapres cuff applied to the middle finger of the left hand and a surface ECG to record R-R interval. Subjects were sitting with the arm at the level of the right atrium and performed a forced expiration at 40mmHg for 15 s. Mouth pressure was measured with a pressure transducer, and after release of the strain data were collected for a further 60s, during which time the subjects were asked not to talk. Finally, after resting for a further lOmin, resting blood pressure and heart rate data were recorded for 15min with the patients supine. Sphygmomanometer blood pressures were measured before and after the Finapres recording to compare with the Finapres readings. Data were converted from analogue to digital form using an on-line personal computer sampling at 200 Hz. A QRS detection algorithm automatically marked the R-R intervals (from the surface ECG), and phase 4 of the Valsalva manoeuvre was identified from the point of release of the strain. BRS was determined from the regression of R-R interval on systolic BP including all beats in phase 4 with no lag. Only regression lines with a correlation coeffcient of 0.8 or significant at P
