Author + information
- Received December 1, 2001
- Revision received March 21, 2001
- Accepted April 5, 2001
- Published online July 1, 2001.
- Naohiko Takahashi, MD∗,* (, )
- Mikiko Nakagawa, MD†,
- Tetsunori Saikawa, MD†,
- Tatsuhiko Ooie, MD∗,
- Kunio Yufu, MD∗,
- Sakuji Shigematsu, MD∗,
- Masahide Hara, MD∗,
- Hiroshi Sakino, MD∗,
- Isao Katsuragi, MD∗,
- Toshimitsu Okeda, MD∗,
- Hironobu Yoshimatsu, MD∗ and
- Toshiie Sakata, MD∗
- ↵*Reprint requests and correspondence:
Dr. Naohiko Takahashi, Department of Internal Medicine I, School of Medicine, Oita Medical University, 1-1 Idaigaoka, Hasama, Oita 879-5593, Japan
The aim of this study was to examine the effects of essential hypertension on cardiac autonomic function in type 2 diabetic patients.
Hypertension is common in type 2 diabetic patients and is associated with a high mortality. However, the combined effects of type 2 diabetes and essential hypertension on cardiac autonomic function have not been fully elucidated.
Thirty-three patients with type 2 diabetes were assigned to a hypertensive diabetic group (n = 15; age: 56 ± 8 years, mean ± SD) or an age-matched normotensive diabetic group (n = 18, 56 ± 6 years). Cardiac autonomic function was assessed by baroreflex sensitivity (BRS), heart rate variability (HRV), plasma norepinephrine concentration and cardiac 123I-metaiodobenzylguanidine (MIBG) scintigraphic findings.
Baroreflex sensitivity was lower in the hypertensive diabetic group than it was in the normotensive diabetic group (p < 0.05). The early and delayed myocardial uptake of 123I-MIBG was lower (p < 0.01 and p < 0.05, respectively), and the percent washout rate of 123I-MIBG was higher (p < 0.05) in the hypertensive diabetic group. However, the high frequency (HF) power and the ratio of low frequency (LF) power to HF power (LF/HF) of HRV and plasma norepinephrine concentration were not significantly different. The homeostasis model assessment index was higher in the hypertensive diabetic group than it was in the normotensive diabetic group (p < 0.01).
Our results indicate that essential hypertension acts synergistically with type 2 diabetes to depress cardiac reflex vagal and sympathetic function, and the results also suggest that insulin resistance may play a pathogenic role in these processes.
Diabetes mellitus and hypertension frequently occur concurrently and probably have synergistic detrimental effects on the cardiovascular system (1–3). The mortality in diabetic patients with hypertension is several fold higher than it is in normotensive diabetic patients (4). Progression of atherosclerosis to occlusive coronary artery disease can explain the high mortality in hypertensive diabetic patients. However, sudden and unexpected deaths, probably caused by ventricular tachyarrhythmias, have often been attributed to diabetic cardiac neuropathy (5–7). Previous studies have demonstrated that both diabetes and hypertension impair cardiac autonomic function (8–12), suggesting that the association of these two diseases promotes cardiac autonomic dysfunction. However, the combined effects of diabetes and hypertension on cardiac autonomic control have not been fully elucidated.
Technical advances, including the measurement of baroreflex sensitivity (BRS), heart rate variability (HRV) and cardiac 123I-metaiodobenzylguanidine (MIBG) scintigraphy, have allowed the precise assessment of cardiac autonomic function (8,13). In this study, we determined BRS, HRV and cardiac 123I-MIBG scintigraphic findings in diabetic patients in order to determine the impact of essential hypertension on cardiac vagal and sympathetic function in type 2 diabetic patients.
Thirty-three Japanese patients with type 2 diabetes mellitus (age: 56 ± 7 years, mean ± SD; 15 women and 18 men) who were admitted to our department were enrolled in this study. The clinical characteristics of these patients are summarized in Table 1. None of the patients had organic heart disease as determined by physical examination, chest X-ray, 12-lead electrocardiogram (ECG), echocardiography, treadmill exercise ECG testing and 201thallium cardiac scintigraphy. Essential hypertension was defined as diastolic blood pressure (BP) ≥90 mm Hg, systolic BP ≥140 mm Hg or self-reported use of antihypertensive medication (12). Fifteen of the 33 patients met these criteria and were assigned to the hypertensive diabetic group, while the remaining 18 patients were assigned to the normotensive diabetic group. In the hypertensive diabetic group, eight patients were treated with calcium channel antagonists, four patients with angiotensin-converting enzyme inhibitors and one patient with both. These medications were discontinued at least two days before cardiac autonomic function testing. No patients were treated with diuretics, beta-adrenergic blocking agents or alpha blockers. Patients who were treated with insulin or had albuminuria (≥500 mg/day) or abnormal plasma creatine concentrations (≥1.2 mg/dl) were excluded from the study.
The study was approved by the ethics review board of our institution, and written informed consent was obtained from all patients.
M-mode two-dimensional echocardiography and cardiac Doppler recordings were obtained by means of a phase-array echo-Doppler system. Echocardiograms were obtained in a standard manner using standard parasternal, short-axis and apical views. Left ventricular (LV) mass was calculated according to a previous study (14): where LVIDd = LV internal dimension at end-diastole; IVSTd = intraventricular septal thickness at end-diastole and PWTd = posterior wall thickness at end-diastole. Left ventricular mass was divided by body surface area to calculate the LV mass index. Pulsed Doppler recordings were made from the standard apical four-chamber view. Mitral inflow velocity was recorded with the sample volume at the mitral annulus level; the average of ≥3 cardiac cycles was taken. The following measurements were made: peak velocity of early ventricular filling (E), peak velocity of late ventricular filling (A), their ratio (E/A) and deceleration time.
Measurement of BRS and plasma norepinephrine (NE) concentration
All subjects were studied while in supine position in a quiet room with dimmed lights between 9 and 11 am(15). A catheter was inserted in the right cubital vein, and arterial BP was recorded noninvasively using tonometry (Jentow-7700, Nihon Colin, Komaki, Japan). The tonometric sensor was attached over the left radial artery. The accuracy of continuous BP monitoring using this system has been demonstrated previously (16). Arterial BP and the standard 12-lead ECG were monitored simultaneously, and data were stored in a PCM data recorder (RD-200T; TEAC, Tokyo, Japan). Three-lead precordial Holter ECG recordings (model-459, Del Mar Avionics, Irvine, California) were also obtained throughout the procedure for analysis of HRV.
After a waiting period of 30 min to allow cardiovascular baroreflex mechanisms to stabilize, patients were asked to breathe at a rate of 15 breaths/min using a metronome to maximize regularity between respiration and cardiovascular function. A blood sample was obtained from the venous catheter for measurement of plasma NE concentration. Baroreflex sensitivity was assessed by the phenylephrine method, as previously described (15). Phenylephrine (2 to 3 μg/kg) was injected over 15 s to obtain a 15 to 40 mm Hg systolic BP increase. Baroreflex sensitivity was calculated as the slope of the linear regression line relating systolic BP changes to RR interval changes. Regression lines with more than 20 data points and a correlation coefficient (r) >0.8 were accepted for analysis. The mean of the two slope values was taken as the BRS value.
Heart rate variability was analyzed using a 300-s interval of Holter ECG recording immediately before phenylephrine injection. The power spectrum of the RR interval was computed by a fast Fourier transform and expressed as the area under the power spectrum (17). We calculated the power of two spectral bands, the low-frequency component at 0.04 to 0.15 Hz (LF) and the high-frequency component at 0.15 to 0.40 Hz (HF). Measured values of HRV were transformed using a natural logarithm because their distributions were skewed. The ratio of LF to HF (LF/HF) was also computed.
Cardiac 123i-MIBG scintigraphy
Planar and single photon emission computed tomography studies were performed both 15 min (early) and 4 h (delayed) after the injection of 111 MBq of 123I-MIBG using a rotating gamma camera (ZLC 7500; Siemens, Munich, Germany). Data were analyzed with computer-based analysis software (SCINTIPAC; Shimadzu, Kyoto, Japan). The anterior planar images of the early and delayed 123I-MIBG studies were analyzed visually. For semiquantitative analysis, regions of interest were drawn over the whole heart and a 10-mm × 10-mm area over the upper mediastinum on the early and delayed planar images to calculate the mean heart-to-mediastinum (H/M) ratio. After correcting for the physical decay of 123I, the percent washout rate (WR) of the tracer from the myocardium was determined for the 4-h period.
Insulin resistance was evaluated by the homeostasis model assessment (HOMA) index = (fasting plasma insulin [pmol/l] × fasting plasma glucose [mmol/l])/22.5 (18).
Data are presented as mean ± SD. Differences between the two groups were analyzed by unpaired Student ttest, chi-square test or Fisher exact probability test. A value of p <0.05 was considered statistically significant.
Clinical characteristics of the patients
As demonstrated in Table 1, the mean age was similar between the groups. No significant differences were observed between the two groups with respect to gender, duration of diabetes or number of patients treated with sulfonylurea, alpha glucosidase inhibitors and troglitazone. Calcium channel antagonists and angiotensin-converting enzyme inhibitors were used in the hypertensive diabetic groups but not in the normotensive diabetic group (p < 0.0005 and p < 0.05, respectively). Body mass index was greater in the hypertensive diabetic group than it was in the normotensive diabetic group (p < 0.05). There were no significant differences in the fasting plasma glucose and hemoglobin A1c between the groups. The hemodynamic data in Table 1were obtained immediately before the BRS assessment. Although the resting heart rate was not significantly different, systolic and diastolic BP were higher in the hypertensive diabetic group than in the normotensive diabetic group (p < 0.001 and p < 0.05, respectively). Both the fasting plasma insulin concentration and HOMA index were higher in the hypertensive diabetic group than in the normotensive diabetic group (p < 0.05 and p < 0.01, respectively). Plasma total cholesterol, triglyceride and HDL-cholesterol were not significantly different between the two groups. However, the level of uric acid was higher in the hypertensive diabetic group than it was in the normotensive diabetic group (p < 0.001).
Table 2summarizes the echocardiographic findings. There were no significant differences in the LV dimensions at end-diastole and end-systole, ejection fraction or posterior wall thickness at end-diastole between the two groups. In contrast, left atrial dimension and intraventricular septal thickness at end-diastole were greater in the hypertensive diabetic group than in the normotensive diabetic group (p < 0.001 and p < 0.005, respectively). There was no significant difference in LV mass index between the groups. With respect to the LV diastolic function, the peak velocity of A was increased, and the E/A ratio was decreased more in the hypertensive diabetic group than it was in the normotensive diabetic group (p < 0.005 and p < 0.05, respectively). However, there was no significant difference in the deceleration time between the groups.
Cardiac autonomic function tests
Figure 1summarizes the findings of the cardiac autonomic function tests. Baroreflex sensitivity was lower in the hypertensive diabetic group than it was in the normotensive diabetic group (5.4 ms/mm Hg ± 4.2 ms/mm Hg vs. 9.5 ms/mm Hg ± 4.7 ms/mm Hg, p < 0.05, Fig. 1A). The plasma NE concentration in the hypertensive diabetic group was similar to that in the normotensive diabetic group (227 pg/ml ± 89 pg/ml vs. 225 pg/ml ± 84 pg/ml, p = NS, Fig. 1B). The analysis of HRV revealed that the HF power and the LF/HF ratio in the hypertensive diabetic group were not significantly different from those in the normotensive diabetic group (4.1 ln-ms2± 1.0 ln-ms2vs. 4.1 ln-ms2± 1.0 ln-ms2, p = NS, 1.1 ln-ms2± 1.1 ln-ms2vs. 1.3 ln-ms2± 1.1 ln-ms2, p = NS, respectively, Fig. 1C). Cardiac 123I-MIBG scintigraphy revealed that the H/M ratios at early and delayed phases were lower in the hypertensive diabetic group than they were in the normotensive diabetic group (1.97 ± 0.29 vs. 2.29 ± 0.32, p < 0.01 and 1.90 ± 0.29 vs. 2.19 ± 0.26, p < 0.05, respectively, Fig. 1D). The percent WR of 123I-MIBG was higher in the hypertensive diabetic group compared with the normotensive diabetic group (40.7% ± 9.1% vs. 33.2% ± 10.0%, p < 0.05, Fig. 1D).
To our knowledge, this is the first study that compares cardiac autonomic function between diabetic patients with and without essential hypertension by simultaneous assessment of BRS, HRV and 123I-MIBG scintigraphy. The main findings were that type 2 diabetic patients with essential hypertension had a lower BRS and showed lower myocardial uptake and enhanced clearance of 123I-MIBG compared with those patients without essential hypertension. These observations suggest that both cardiac reflex vagal and sympathetic function in type 2 diabetic patients may be impaired when associated with essential hypertension.
Assessment of cardiac vagal function
Analysis of HRV has been used as a measure of cardiac autonomic control. Vagal activity is certainly the contributor to the HF component of HRV. Baroreflex sensitivity with the phenylephrine method also estimates cardiovascular vagal activity. These two tests have been widely applied to identify patients at high risk for sudden cardiac death after myocardial infarction (MI) (19,20). Although both HF power and BRS are indexes of vagal activity, they reflect different physiological aspects (i.e., HF power is a marker of tonic vagal activity whereas BRS is a marker of reflex vagal activity) (15,19–21). For instance, De Ferrari et al. (21)assessed HRV and BRS in patients with and without episodes of sustained ventricular tachycardia or ventricular fibrillation after MI and demonstrated that patients with tachyarrhythmias have significantly lower BRS compared with those without tachyarrhythmias, while no significant difference was found in measures of HRV. Furthermore, the ATRAMI (Autonomic Tone and Reflex After MI) investigators demonstrated that BRS has strong independent prognostic value after MI (20). These findings suggest that analysis of BRS rather than HRV may be more helpful in risk stratification after MI. In this study, BRS was significantly lower in the hypertensive diabetic group than it was in the normotensive diabetic group in spite of a similar HF power. This finding indicates that BRS may also be more useful in detecting cardiac vagal dysfunction in hypertensive or diabetic patients.
Although our study showed that HF power is not affected by the presence of essential hypertension, the Atherosclerosis Risk In Communities (ARIC) study showed that hypertension clustered with diabetes has an inverse additive effect on HRV (12). It is difficult to explain the difference between the two studies. In the ARIC study, most enrolled patients were American. Therefore, there may be racial differences between the two studies. The ARIC study also included patients with a history of coronary heart disease and those treated with beta-adrenergic blocking agents. Severe hypertensive patients may also have been included in the former study, while our study included only patients with mild essential hypertension. These factors may explain the differences observed between the two studies.
Assessment of cardiac sympathetic function
In this study, cardiac sympathetic function was estimated by three different methods (plasma NE concentration, HRV and 123I-MIBG scintigraphy). Plasma NE concentration was similar between the groups. The LF/HF ratio did not differ between the groups. In contrast, the myocardial uptake of 123I-MIBG was reduced, and its clearance was enhanced in the hypertensive diabetic group. Although no technique is currently considered to be the “gold standard” for assessing human sympathetic function with which other techniques might be compared (22), these findings suggest that 123I-MIBG scintigraphy may be fairly sensitive in detecting cardiac sympathetic dysfunction, at least in the population we studied (23).
Detrimental effects of essential hypertension on diabetic neuropathy—possible role of insulin resistance
The two groups examined in our study were age-matched and had a similar duration of diabetes. The plasma glucose and hemoglobin A1c were similar between the groups. Furthermore, the LV mass index showed no significant difference. The main difference between the two groups was the arterial BP. Therefore, the depressed cardiac autonomic function in the hypertensive diabetic group observed in this study can be explained by the additive adverse effects of hypertension. Liao et al. (12)reported that hypertension clustered with diabetes synergistically influenced cardiac autonomic control when assessed by HRV. Tamura et al. (24)demonstrated that the percent WR of 123I-MIBG was increased in hypertensive diabetic patients compared with healthy subjects, whereas the difference was not statistically different between normotensive diabetic patients and healthy subjects. Our study did not show a significant difference in HRV between the two groups. However, BRS and 123I-MIBG scintigraphy detected depressed cardiac autonomic function in the hypertensive diabetic group.
With respect to the pathogenesis of essential hypertension clustered with type 2 diabetes, it is noteworthy that the body mass index and the HOMA index were higher in the hypertensive diabetic group than in the normotensive diabetic group. These findings suggest the involvement of insulin resistance in influencing cardiac autonomic control. In fact, experimental studies have revealed that insulin-resistant rats with hypertension showed both impaired reflex vagal activity and increased the WR of 123I-MIBG (25,26). The report by Liao et al. (12)demonstrated that depressed HRV in hypertensive diabetic patients is associated with an increased fasting plasma insulin concentration. Pikkujamsa et al. (11)also demonstrated that HRV in hypertensive individuals with insulin resistance syndrome is lower than it is in those without insulin resistance syndrome. Taken together, our observations suggest that insulin resistance may serve as the mechanism for the clustering of essential hypertension and type 2 diabetes.
In this study, patients with essential hypertension plus type 2 diabetes had depressed cardiac diastolic function compared with those without essential hypertension. These findings are consistent with the recent studies demonstrating that diastolic dysfunction is related to depressed BRS (27)and to insulin resistance (28)in patients with hypertension.
There are several limitations in this study. First, our study did not include patients with only essential hypertension or individuals without diabetes and hypertension. Second, 13 patients in the hypertensive diabetic group were treated with calcium channel antagonists or angiotensin-converting enzyme inhibitors. For ethical reasons, we did not discontinue the antihypertensive drugs for an extended period. Antihypertensive therapy has been reported to improve cardiac autonomic function in hypertensive patients (29,30). However, the discontinued medication could cause a rebound sympathetic effect, which could, in turn, mask HRV. Thus, the influences of antihypertensive therapy and its discontinuation should be carefully considered. Third, no patients enrolled in this study underwent coronary angiography. Although ischemic heart disease could not be completely excluded, severe coronary artery disease was unlikely to be present in view of normal treadmill exercise ECG testing and 201thallium cardiac scintigraphy. Finally, we attributed the depressed cardiac autonomic function in hypertensive diabetic patients to the effects of hypertension. However, the converse interpretation may also be valid (i.e., cardiac autonomic dysfunction could contribute to the development of essential hypertension).
Our findings suggest that essential hypertension acts synergistically with type 2 diabetes to depress cardiac reflex vagal and sympathetic function, and they suggest that insulin resistance may play a pathogenic role in these processes. Baroreflex sensitivity and 123I-MIBG scintigraphy provide sensitive assessments of cardiac autonomic control in this population. Future studies are required to assess the prognostic value of BRS and 123I-MIBG scintigraphy in diabetic and hypertensive patients.
The authors thank Miss Emi Noguchi, MT, for her excellent technical assistance.
- baroreflex sensitivity
- high frequency component (0.15 to 0.45 Hz)
- homeostasis model assessment
- heart rate variability
- low frequency component (0.04 to 0.15 Hz)
- left ventricular
- washout rate
- Received December 1, 2001.
- Revision received March 21, 2001.
- Accepted April 5, 2001.
- American College of Cardiology
- Dupree E.A.,
- Meyer M.B.
- Ewing D.J.
- Liao D.,
- Cai J.,
- Barnes R.W.,
- et al.
- Pikkujamsa S.M.,
- Huikuri H.V.,
- Airaksinen K.E.,
- et al.
- Liao D.,
- Sloan R.P.,
- Cascio W.E.,
- et al.
- Spallone V.,
- Menzinger G.
- Devereux R.B.,
- Alonso D.R.,
- Lutas E.M.,
- et al.
- Takahashi N.,
- Nakagawa M.,
- Saikawa T.,
- et al.
- Nakagawa M.,
- Iwao T.,
- Ishida S.,
- et al.
- La Rovere M.T.,
- Specchia G.,
- Mortara A.,
- Schwartz P.J.
- Schnell O.,
- Muhr D.,
- Weiss M.,
- et al.
- Pitzalis M.V.,
- Passantino A.,
- Massari F.,
- et al.
- Sakata K.,
- Shirotani M.,
- Yoshida H.,
- Kurata C.