Clin Auton Res (2014) 24:275–283 DOI 10.1007/s10286-014-0253-y

RESEARCH ARTICLE

Autonomic control during acute hypoglycemia in type 1 diabetes mellitus Jacqueline K. Limberg • Kathryn E. Farni • Jennifer L. Taylor • Simmi Dube • Ananda Basu • Rita Basu • Erica A. Wehrwein • Michael J. Joyner

Received: 10 April 2014 / Accepted: 25 June 2014 / Published online: 27 September 2014 Ó Springer-Verlag Berlin Heidelberg 2014

Abstract Purpose We hypothesized that adults with type 1 diabetes mellitus (T1DM) would exhibit impaired heart rate variability (HRV), QT interval, T-wave amplitude, and baroreflex sensitivity (BRS) when compared with healthy controls. In addition, we hypothesized that acute hypoglycemia would result in further adverse changes in measures of autonomic and cardiovascular function. Methods A single 180-min hyperinsulinemic (2 mU/kg TBW/min), hypoglycemic (*3.3 umol/mL) clamp was completed in 10 healthy adults and 13 adults with T1DM. Counterregulatory hormones were assessed and measures of heart rate (electrocardiogram) and blood pressure (intraarterial catheter or finger photoplethysmography) were analyzed at baseline and during the hypoglycemic clamp for measures of HRV, QT interval, T-wave amplitude, and spontaneous cardiac BRS (sCBRS). Results Baseline measures of HRV, sCBRS, and T-wave amplitude were blunted in adults with T1DM when compared with healthy controls. Hypoglycemia resulted in significant reductions in HRV, sCBRS, and T-wave amplitude and prolonged QT intervals; these changes were

J. K. Limberg (&)  K. E. Farni  J. L. Taylor  M. J. Joyner Department of Anesthesiology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905, USA e-mail: [email protected] S. Dube  A. Basu  R. Basu Department of Endocrinology, Mayo Clinic, Rochester, MN, USA E. A. Wehrwein Department of Physiology, Michigan State University, East Lansing, MI, USA

not different between adults with T1DM and healthy controls. Conclusions Results from the current study show that adults with T1DM exhibit impaired autonomic and cardiovascular function. Additionally, novel findings highlight an effect of acute hypoglycemia to further reduce measures of autonomic and cardiovascular function similarly between adults with T1DM and healthy controls. These results suggest that acute hypoglycemia may worsen impairments in autonomic and cardiovascular control in patients with T1DM, thus increasing the risk of ventricular arrhythmias and cardiovascular mortality. Keywords Diabetes

Baroreflex sensitivity  Heart rate variability 

Introduction Glycemic control is crucial in the management of diabetes. However, as a consequence of rigorous control, recent reports suggest 82 % of patients with type 1 diabetes mellitus (T1DM) experience one or more hypoglycemic events each month [6]. Antecedent hypoglycemia has been shown to contribute to impaired autonomic function in healthy adults [2] and blunted autonomic responses to subsequent hypoglycemia in adults with T1DM [5]. Impairments in autonomic control strongly predict cardiovascular mortality [29]. Thus, it is reasonable to propose that hypoglycemia-mediated impairments in autonomic and cardiovascular control may lead to sudden, fatal arrhythmias and contribute to the underlying pathophysiology in ‘‘dead in bed’’ syndrome [25]. Despite evidence that prior exposure to hypoglycemia may contribute to autonomic and cardiovascular

123

276

dysfunction, the effects of a single bout of hypoglycemia in adults with T1DM have not been well studied. Such measurements may provide novel information about variations in autonomic function during acute hypoglycemia. Taken together, we sought to examine whether adults with T1DM exhibit impaired heart rate variability (HRV), QT interval, T-wave amplitude, and baroreflex sensitivity (BRS) when compared with healthy controls. In addition, we hypothesized that acute hypoglycemia would result in further adverse changes in measures of autonomic and cardiovascular function.

Materials and methods Ethical approval The institutional review board at the Mayo Clinic approved all experiments and procedures, and studies conformed to the 1964 Declaration of Helsinki and its later amendments. Written informed consent was obtained from all subjects prior to study enrollment. Data regarding HRV and BRS during hypoglycemia in control subjects were published previously [14]. Subjects All subjects were between the ages of 21–60 years, nonobese (BMI \40 kg/m2), non-smokers, normotensive, and non-pregnant/non-breastfeeding. Control subjects were healthy. Inclusion criteria for subjects with T1DM included HbA1c B10 %, creatinine B1.5 mg/dL, insulin pump (n = 9) or multiple daily insulin injection (n = 4) therapies. Investigators were not blinded to subject condition (control, patient). Exclusion criteria included unstable diabetic retinopathy/nephropathy, a history of unstable macrovascular disease, seizure disorder (or on anti-seizure medications), diagnosed autonomic disorder, and/or other chronic diseases. Subjects could not be taking medications that altered glucose metabolism (i.e., corticosteroids, opiates, barbiturates, anticoagulants). Four subjects with T1DM were taking mediations for hypertension (angiotensin-converting enzyme inhibitor, lisinopril). Subjects who engaged in regular physical exercise programs or were actively losing weight were excluded. Body composition was measured using dual energy X-ray absorptiometry (DEXA, Lunar iDXA software version 6.10, GE Healthcare Technologies, Madison, WI). Research dieticians provided advice to subjects to ensure that body weight was maintained for 2 weeks prior to the study visit. Subjects refrained from exercise, alcohol, and caffeine for at least 24 h prior to the study visit. Medications were not withheld.

123

Clin Auton Res (2014) 24:275–283

Hypoglycemic clamps Subjects were admitted to the Clinical Research Unit at 17 h on the evening prior to the study. A standard 10 cal/kg meal (55 % carbohydrate, 30 % fat, and 15 % protein) was eaten between 18 and 1830 hours and subjects fasted thereafter until the end of the study. Adults with T1DM administered their customary dose of insulin according to their schedule for the evening meal. At 2030 hours, adults with T1DM had two intravenous catheters placed in the forearm for insulin infusion and plasma glucose monitoring. Intravenous insulin was started at 21 h and hospital protocol was followed overnight to maintain euglycemia. For those using an insulin pump (n = 9), the pump was discontinued prior to starting the intravenous insulin. Those subjects on multiple daily injection regimen (n = 4) were provided 50 % of their nighttime basal insulin dose. In control subjects, two intravenous catheters were placed at 0600 the following morning. At 7 h (-120 min), a primedcontinuous infusion of [3-3H] glucose (12 lCi and 0.12 lCi/min, Perkin Elmer) was started in all subjects and continued until the end of the study (1200 noon; T180). At 9 h, the overnight variable insulin infusion was discontinued in T1DM subjects and a constant infusion (NovolinÒ, Novo Nordisk Inc., Princeton NJ) was started in all subjects at a rate of 2.0 mU/kg TBW/min from protocol time T0 to T180 min. Exogenous glucose (50 % dextrose containing [3-3H] glucose) was infused in amounts sufficient to maintain glucose concentrations at hypoglycemic levels (*3.3 umol/mL). Monitoring In control subjects, a 20-gauge, 5-cm brachial artery catheter was placed under ultrasound guidance and after local anesthesia for blood sampling and blood pressure monitoring. Due to increased risk of wound complications in subjects with T1DM, monitoring by a brachial artery catheter was forgone and a venous catheter was introduced in a retrograde fashion into a hand vein. The hand was placed in a heated plexiglass box (55 °C) for sampling of arterialized venous blood. This method is commonly used during metabolic clamps and has been shown to provide comparable results [15]. In subjects with T1DM, blood pressure was monitored using finger plethysmography (Nexfin, BMEYE, Amsterdam, The Netherlands). In all subjects, heart rate was monitored with a 5-lead electrocardiogram (ECG; Cardiocap/5, Datex-Ohmeda). Spontaneous cardiac baroreflex sensitivity (SCBRS) Spontaneous cardiac baroreflex sensitivity was assessed from 30 to 60 min sections of data during euglycemia

Clin Auton Res (2014) 24:275–283

[T60–T0 (baseline)] and steady-state hypoglycemia [T120–T180 (clamp)]. The distances between all R-wave peaks of the ECG recording were calculated and paired with the systolic pressure wave amplitude of the preceding beat. A computer software program (LabChart7; ADinstruments, Colorado Springs, CO) selected all sequences of three or more successive heart beats in which there were concordant increases or decreases in systolic blood pressure and R–R interval. The recordings were reviewed and non-sinus beats and segments with artifacts were removed. A linear regression was applied to each of the sequences and only relationships with an R2 [ 0.80 were accepted. An average regression slope was then calculated for the acceptable sequences. The slope represented the sCBRS (ms/mmHg). Responses were also evaluated by plotting the changes in systolic pressures with heart rate (beat/min/mmHg) to take into consideration the mathematical constraint of the hyperbolic relationship between R–R interval and heart rate. Given blood pressure was monitored using different methods between groups, it is important to note: (1) beatto-beat changes in non-invasive measurements closely follow changes in intra-arterial pressure [19, 20] and (2) invasive and non-invasive baroreflex sensitivity measurements are highly correlated (r = 0.91–0.98; [10, 19, 21]) and have shown to provide equivalent information [21]. Heart rate variability (HRV) Short-term (*5-min) data selections from a 5-lead ECG recording were analyzed during baseline [T30–T0] and steady-state hypoglycemia [T150–T180] following standard procedures recommended by the Task Force of the European Society of Cardiology and North American Society of Pacing and Electrophysiology [1]. Physiologically stable conditions were confirmed by visual checks, ensuring that only stationary segments were selected for analysis (avoiding ectopic beats, arrhythmias, missing data, and noise). A computer program (HRV Module, LabChart7, ADInstruments Pty Ltd, Australia) was used to assess both time and frequency (fast Fourier transformation, Welsh windowing function) domains. When alterations in total variability (total power) occur (e.g., during hypoglycemia), frequency power spectral densities should be calculated in units normalized to total spectral power (normalized units, nu). Thus, data reported include: mean NN interval (time between normal cardiac cycles, reported in ms), low-frequency normalized [LF nu; range: 0.04–0.15 Hz; reported in normalized units (nu)], highfrequency normalized [HF nu; range: 0.15–0.4 Hz; reported in normalized units (nu)], and low-frequency/ high-frequency ratio.

277

T-wave morphology QT intervals and T-wave amplitudes were assessed from the same short-term ECG data selections used for HRV analysis. Physiologically stable conditions were confirmed by visual checks during baseline [T30–T0] and steady-state hypoglycemia [T150–T180], ensuring that ectopic beats, arrhythmias, and noise were excluded. QT intervals were electronically selected from ECG tracings, beginning at the first deflection of the QRS complex and ending with the return of the T-wave to baseline (LabChart7, ADInstruments Pty Ltd, Australia) and were confirmed by visual checks. Heart rate-adjusted QT intervals (QTc) were calculated according to Bazett’s formula [3]. Maximum T-wave amplitude was defined as the maximum difference between the T-wave and baseline, which was automatically identified by a computer program and confirmed by visual checks (LabChart7, ADInstruments Pty Ltd, Australia). Analytical methods Plasma glucose was measured every 5–10 min at the bedside using a glucose oxidase method (Analox Instruments USA Inc., Lunenberg, Massachusetts). Additional arterial (or arterialized venous) blood was drawn for measures of glucose, insulin, cortisol, glucagon, growth hormone, epinephrine, and norepinephrine during baseline euglycemia and throughout the hyperinsulinemic hypoglycemic clamp. Baseline values are reported as an average of T30, -20, -10, and 0 min samples and clamp values as an average of T150, 160, 170, and 180 min [26]. All blood samples were immediately placed on ice and centrifuged at 4 °C after which time the plasma was removed and stored at -80 °C until analysis. Plasma insulin and growth hormone were assessed using a two-site immunoenzymatic assay performed on the D 9 I automated immunoassay system (Beckman Instruments, Chaska, MN). Plasma catecholamines (epinephrine, norepinephrine) were measured with reverse phase high-performance liquid chromatography with electrochemical detection after extraction with activated alumina. Glucagon was measured by radioimmunoassay and cortisol by a competitive binding immunoenzymatic assay (Dxl 800 automated immunoassay system). Data analysis and statistics All statistical analyses were completed by a biostatistician (Darrell Schroeder, M.S.) using SAS (SAS Institute Inc.; Cary, NC, USA). The number of participants (n = 13) was determined a priori by a power test equation with a = 0.05 and power = 0.80, using group differences (10.5 ± 9.3 ms/mmHg; Mean ± SD) in BRS from previously

123

278

published research in patients with T1DM [28]. Two-way repeated measures analysis of variance was performed to compare the effect of time (baseline, clamp) and group (control, T1DM) on main (BRS) and secondary (HRV, QT interval, T-wave amplitude, hemodynamics, counterregulatory hormones) outcome variables. Changes (D) in main outcome variables were calculated as (clamp - baseline). Distributional assumptions were assessed and non-parametric methods were used as appropriate; however, given conclusions were not impacted significantly by a lack of normality, results presented are from parametric analyses only. When appropriate, Pearson’s product–moment correlations were used to determine the association between the main outcome variables (e.g., HRV) and descriptive measurements (e.g., plasma catecholamines). Statistical significance was determined a priori at the a = 0.05 level. When a direction of a difference was hypothesized a priori (e.g., HRV is blunted in patients with T1DM vs. control), a one-tailed test was used. Data are reported as mean ± SE of the mean.

Clin Auton Res (2014) 24:275–283 Table 1 Subject demographics Characteristics

Control

Sex (male/female)

7/3

7/6

Age (years)

25 ± 1

43 ± 12*

Height (cm)

177 ± 2

172 ± 3 76 ± 5

Weight (kg)

75 ± 3

BMI (kg/m2)

24 ± 1

26 ± 2

Body fat (%)

25 ± 2

27 ± 3

Fat-free mass (kg)

56 ± 3

54 ± 3

Glucose (mg/dL)

83 ± 2

129 ± 11*

HbA1c (%)



7±1

Duration of diabetes (year)



17 ± 5

Data are presented as mean ± SEM. T1DM (n = 13, unless noted): height (n = 11), body fat (n = 12), duration of diabetes (n = 11). BMI body mass index, HbA1c = glycosylated hemoglobin. Effect of group: *p \ 0.05 vs. control

Table 2 Changes hypoglycemia

Results Subjects Ten healthy control subjects (7 M/3F) and 13 adults with T1DM (6 M/7F) participated in the current study (Table 1). Groups were matched for weight, BMI, and body composition. Adults with T1DM were significantly older than the control group (25 ± 1 vs. 43 ± 3 years); however, age did not have a significant effect on main outcome variables, and thus non-adjusted results are presented.

in

counterregulatory

hormones

during

Baseline

Clamp

D

Control

4±1

132 ± 8 

128 ± 8

T1DM

12 ± 2

145 ± 18 

132 ± 17

Insulin (uU/mL)

Glucose (umol/mL) Control

5.4 ± 0.1

3.4 ± 0.1 

-2.0 ± 0.1

T1DM

8.5 ± 1.2*

3.4 ± 0.1*, 

-5.1 ± 1.2à

Glucose infusion rate (umol/kg/min) Control – 28 ± 4 T1DM





29 ± 5



9±1

19 ± 2 

10 ± 2

 

16 ± 3

5 ± 2à

101 ± 11 

38 ± 9

Cortisol (ug/dL) Control

Counterregulatory hormone response to hypoglycemia

T1DM

T1DM

11 ± 1

Glucagon (pg/mL)

The results are presented in Table 2. As expected, baseline glucose concentrations were higher in T1DM (main effect of group, p \ 0.01) as compared to controls. As designed, the hyperinsulinemic hypoglycemic clamp resulted in higher insulin concentrations and lower glucose concentrations in both groups (main effect of time, p \ 0.01) as compared to baseline. The glucose infusion rate required to maintain hypoglycemia was not different between groups (p = 0.42). Hypoglycemia resulted in a significant increase in plasma catecholamines (norepinephrine and epinephrine) and counterregulatory hormones cortisol, glucagon, and growth hormone (main effect of time, p B 0.05); however, any rise in epinephrine, cortisol, and glucagon was blunted in adults with T1DM when compared with controls (interaction between group and time, p B 0.05).

123

Control

63 ± 3

T1DM

53 ± 5*

49 ± 7*, 

-4 ± 5à

Growth hormone (ng/mL) Control

1±1

11 ± 2à

10 ± 2

T1DM

1±1

9 ± 3à

8±3

Norepinephrine (pg/mL) Control

206 ± 20

355 ± 40 

149 ± 29

T1DM

184 ± 14

325 ± 28 

140 ± 25

748 ± 101 

720 ± 99

342 ± 102*, 

315 ± 103à

Epinephrine (pg/mL) Control 28 ± 4 T1DM

27 ± 3*

Data are presented as mean ± SEM. T1DM (n = 13, unless noted): glucagon (n = 10). D = clamp - baseline Effect of time:  p \ 0.05 vs. baseline. Effect of group: *p \ 0.05 vs. control. Interaction of group and time: àp \ 0.05

Clin Auton Res (2014) 24:275–283

279

Hemodynamic response to hypoglycemia Blood pressure was not different between groups (main effect of group, p [ 0.05). Systolic and mean blood pressure did not change with hypoglycemia (main effect of condition, p [ 0.05). In healthy control subjects, there was a reduction in diastolic blood pressure and a rise in heart rate with hypoglycemia; however, the effect of hypoglycemia on changes in diastolic blood pressure and heart rate were blunted in patients with T1DM (interaction between group and hypoglycemia, p B 0.05; see Table 3). Spontaneous cardiac baroreflex sensitivity (sCBRS) Spontaneous cardiac baroreflex sensitivity was blunted in adults with T1DM when compared with healthy controls (main effect of group, p \ 0.05). There was a significant reduction in sCBRS from baseline levels during steadystate hypoglycemia (main effect of time, p \ 0.05); although, the effect of hypoglycemia on baroreflex sensitivity was not specific to group (interaction between group and time, p [ 0.05; see Fig. 1). Heart rate variability (HRV) Adults with T1DM exhibited greater low-frequency HRV (main effect of group, p \ 0.01), lower high-frequency Table 3 Changes in hemodynamic variables during hypoglycemia Baseline

D

Clamp

Systolic blood pressure (mmHg) Control

128 ± 4

131 ± 9

3±6

T1DM

123 ± 4

127 ± 4

4±5

Diastolic blood pressure (mmHg) Control

65 ± 2

57 ± 3 

-8 ± 2

T1DM

67 ± 2

65 ± 2 

-2 ± 2à

HRV (main effect of group, p \ 0.01), and greater lowfrequency to high-frequency ratio (main effect of group, p \ 0.01) when compared with healthy controls. Mean NN interval was significantly reduced during hypoglycemia (main effect of condition, p \ 0.01), with a greater reduction observed in healthy control subjects (interaction between group and time, p = 0.04). In addition, the effect of hypoglycemia on the low-frequency to high-frequency ratio was different between groups (interaction between group and time, p = 0.04), with a greater increase observed in controls; see Fig. 2. Importantly, any effect of hypoglycemia on respiratory rate was not different between groups (baseline: control 16 ± 1, T1DM 16 ± 1; clamp: control 18 ± 1, T1DM 17 ± 1; interaction between group and time, p = 0.38). QT interval and T-wave amplitude QT intervals were not different between groups (main effect of group, p = 0.49). QT intervals were prolonged with hypoglycemia (main effect of time, p \ 0.01), and the effect of hypoglycemia to prolong QT intervals was more pronounced in healthy controls when compared with adults with T1DM (interaction between group and time, p = 0.03). The amplitude of the T-wave was lower in adults with T1DM when compared with controls (main effect of group, p = 0.05) and T-wave amplitudes significantly decreased during hypoglycemia in both groups (main effect of time, p \ 0.01; see Table 3). Interestingly, QT interval and T-wave amplitude were both linearly related to measures of HRV (mean NN interval: r = 0.64, p \ 0.01 and r = 0.63, p \ 0.01, respectively) and changes in QT interval and T-wave amplitude with hypoglycemia (D) were linearly related to changes in epinephrine (D; r = 0.60, p \ 0.01 and r = -0.54, p \ 0.01, respectively).

Mean blood pressure (mmHg) Control

86 ± 3

82 ± 4

-4 ± 2

T1DM

87 ± 3

86 ± 2

-1 ± 3

59 ± 2 63 ± 2

75 ± 4  73 ± 3 

16 ± 3 10 ± 2à

Control

0.37 ± 0.01

0.44 ± 0.02 

0.07 ± 0.02

T1DM

0.39 ± 0.01

0.43 ± 0.01 

0.04 ± 0.01à

Heart rate (beat/min) Control T1DM QTc (s)

T-wave amplitude (V) Control T1DM

0.34 ± 0.03 0.26 ± 0.03*

0.20 ± 0.02 

-0.14 ± 0.03 , 

0.16 ± 0.03*

Discussion Results from the current study show that adults with T1DM exhibit impaired autonomic and cardiovascular function. Additionally, novel findings highlight an effect of acute hypoglycemia to further reduce measures of autonomic and cardiovascular function similarly between adults with T1DM and their healthy counterparts. These results suggest that acute hypoglycemia may worsen impairments in autonomic and cardiovascular control in patients with T1DM.

-0.10 ± 0.01

Data are presented as mean ± SEM. T1DM (n = 13, unless noted): QTc (n = 12), T-wave amplitude (n = 12). D = clamp - baseline Effect of time:  p \ 0.05 vs. baseline. Effect of group: *p \ 0.05 vs. control. Interaction of group and time: àp \ 0.05

T1DM and hypoglycemia Normal defenses against hypoglycemia include a reduction in insulin, with concomitant increases in glucagon and

123

280

Clin Auton Res (2014) 24:275–283

Fig. 1 Effect of acute hypoglycemia on spontaneous cardiac baroreflex sensitivity in control and patients with T1DM. Mean ± SE. Control n = 10, T1DM n = 12 (one subject with T1DM was excluded due to inability to achieve sufficient number of sequences). a Cardiac baroreflex sensitivity was measured as the slope of the relationship between the distance between R-wave peaks and paired systolic blood pressure measurements (ms/mmHg). b Cardiac baroreflex sensitivity was also evaluated by plotting changes in systolic

pressures with heart rate (beat/min/mmHg) to take into consideration the mathematical constraint of the hyperbolic relationship between R– R interval and heart rate. Spontaneous cardiac baroreflex sensitivity was blunted in patients with T1DM when compared with healthy controls (*main effect of group, p \ 0.05). Hypoglycemia resulted in a significant reduction in baroreflex sensitivity in both groups ( main effect of hypoglycemia, p \ 0.05)

epinephrine. As evidenced in the present study, the rise in both glucagon and epinephrine in response to hypoglycemia is markedly impaired in individuals with T1DM (Table 2). Interestingly, despite blunted counterregulatory responses to hypoglycemia in patients with T1DM, the glucose infusion rate during the hyperinsulinemic, hypoglycemic clamp was not significantly different between groups (Table 2). This observation may be the result of subject variability, the mild hypoglycemic exposure (60 mg/dL), higher baseline glucose levels in adults with T1DM (Table 2), and/or the presence of reduced insulin sensitivity commonly observed in adults with T1DM [30].

glycemic control have been observed [28]. However, the effect of acute hypoglycemia on sCBRS was previously unexamined. Novel findings from the current study support impaired sCBRS in patients with T1DM at baseline and uncovered significant reductions in sCBRS with acute hypoglycemia (Fig. 1).

Cardiovascular responses to hypoglycemia The majority of cardiovascular changes during hypoglycemia are often attributed to activation of the autonomic nervous system. For example, in response to a reduction in blood pressure, the baroreflex initiates reflex increases in heart rate, contractility, vascular resistance, and venous return to maintain blood pressure at optimal levels. Consistent with this finding, healthy adults exhibited expected reductions in diastolic blood pressure during hypoglycemia, which were accompanied by reflex increases in heart rate. In contrast, the fall in diastolic blood pressure was blunted in T1DM subjects and any change in heart rate was also blunted from control levels (Table 3). Impairments in sCBRS at baseline have been shown repeatedly in patients with T1DM [8, 23, 27, 28] and significant relationships between BRS impairment and duration of T1DM and poor

123

Heart rate variability and hypoglycemia Although BRS may be considered a more sensitive measure of autonomic function when compared with conventional tests [8], analysis of HRV is commonly used to assess autonomic regulation of cardiovascular function. Similar to impairments in sCBRS, baseline measures of HRV are known to be reduced in patients with T1DM [4, 7, 9, 12, 17, 23, 28, 31] and results from the present study confirm this notion (Fig. 2). Specifically, cardiac vagal activity [as measured by the high-frequency (HF) component of HRV] was reduced in adults with T1DM and impairments in the sympathovagal balance (LF/HF ratio) were observed. Although baseline HRV is known to be altered in patients with T1DM, controversy exists regarding the effect of acute hypoglycemia on measures of HRV. Specifically, exposure to a hyperinsulinemic, hypoglycemic clamp has been shown to increase [24], decrease [12], or have no effect [13] on cardiac vagal activity in healthy adults. In the present study, we observed a reduction in mean NN interval with hypoglycemia. However, the decrease in mean NN interval (D) was not related to

Clin Auton Res (2014) 24:275–283

281

Fig. 2 Effect of acute hypoglycemia on heart rate variability in control and patients with T1DM. Mean ± SE. Control n = 10, T1DM n = 13. Heart rate variability was altered in patients with T1DM when compared with healthy controls. Specifically, adults with T1DM exhibited greater low-frequency HRV (a *main effect of group, p \ 0.01), lower high-frequency HRV (b *main effect of group, p \ 0.01), and greater low-frequency to high-frequency ratio

(main effect of group, p \ 0.01). Hypoglycemia resulted in significant changes in heart rate variability that were different between groups (àinteraction between group and hypoglycemia, p \ 0.05). Specifically, healthy adults exhibited a greater reduction in mean NN interval (c) and a significant increase in the low-frequency to highfrequency ratio (d)

changes in plasma catecholamines (D; norepinephrine: r = 0.25, p = 0.26; epinephrine: r = -0.16, p = 0.45), supporting the idea that adrenomedullary sympathoexcitation is not the primary cause of changes in HRV and rather responses are likely a result of reduced cardiac vagal outflow [12]. An attenuation of vagal protection is associated with sudden arrhythmic death and increased risk of mortality [11]; thus the present results suggest that patients with T1DM exhibit large impairments in vagal autonomic function at baseline that have the potential to worsen during hypoglycemia—putting them at increased risk of sudden arrhythmic death.

potassium. Consistent with this concept, QT interval and T-wave amplitude were both linearly related to measures of HRV (mean NN interval: r = -0.64, p \ 0.01 and r = 0.63, p \ 0.01, respectively) and changes in QT interval and T-wave amplitude with hypoglycemia (D) were linearly related to changes in epinephrine (D; r = 0.60, p \ 0.01 and r = -0.54, p \ 0.01, respectively). To our knowledge, this is the first study to report such relationships. Importantly, standard precautions were taken to control potassium levels (i.e., co-infusion of 0.9 % NaCl ? KCl and insulin). The observed alterations in cardiac electrical properties during hypoglycemia in healthy adults and adults with T1DM again point to increased risk of cardiac arrhythmias and sudden death with hypoglycemic exposure.

T-wave morphology and hypoglycemia Hypoglycemia is known to affect cardiac electrical properties. QT interval is a commonly used measure of cardiac repolarization and predictor of arrhythmia risk and sudden death [16]. Confirming previous findings [18, 22], QT interval was prolonged and T-wave amplitude was reduced during hypoglycemia in healthy adults and individuals with T1DM (Table 3). QT prolongation may occur as a result of: (1) impaired cardiac autonomic regulation, (2) altered sympathoadrenal activation, and (3) lower serum

Experimental considerations Given both sexes are known to be affected by T1DM, the present investigation examined differences between groups of both men and women, between the ages of 21 and 60 years. Given the complexity of the study design, we were unable to control for menstrual cycle phase. Further, although the majority of data from both cohorts (control vs.

123

282

patient) were collected concurrently (Control 2010–2012; T1DM 2012), for patient safety the first few experiments were conducted in healthy adults only and investigators were not blinded to the group. Importantly, groups were matched for weight, BMI, and body composition (Table 1); however, adults with T1DM were significantly older than the control group. Evidence supports an age-related decline in measures of HRV and BRS; therefore it is reasonable to question whether age may impact present findings rather than T1DM per se. Ideally, we would have studied agematched control subjects; but importantly, when our analyses were statistically adjusted for age, there was no significant effect on main outcome variables and derived conclusions. It is also important to note that insulin itself can activate the sympathetic nervous system and decrease vagal influence of the heart. Although any independent effect of insulin is unlikely to have an effect on differences in measures of autonomic and cardiovascular control at baseline, future studies will be necessary to examine the differential effect of hyperinsulinemia in healthy adults and adults with T1DM.

Conclusion This is the first study to systematically examine the acute effects of hypoglycemia on autonomic and cardiovascular responses in adults with T1DM. We have shown that adults with T1DM exhibit impairments in autonomic and cardiovascular control at baseline. We further observed reductions in measures of autonomic control with acute hypoglycemia in both healthy adults and adults with T1DM. These results suggest that acute hypoglycemia may worsen impairments in autonomic and cardiovascular control in patients with T1DM, thus increasing the risk of ventricular arrhythmias, cardiovascular mortality, and sudden death in this population. Acknowledgments Our deepest appreciation and thanks to Dr. Robert Rizza for his valuable and constructive suggestions with study design. The authors also wish to thank Drs. Timothy Curry and John Eisenach (Department of Anesthesiology) for placement of brachial artery catheters. The authors further wish to acknowledge the contributions of the nursing and technical staff: Cheryl Shonkwiler, Barbara Norby, Shelly Roberts, Karen Krucker, Sarah Wolhart, Jean Knutson, Brent McConahey, Pamela Reich, Nancy Meyer, Pam Engrav, and Christopher Johnson of the Mayo Clinic. In addition, we thank the Clinical Research Unit staff at Mayo Clinic, the Immunochemical Core Laboratory at Mayo Clinic, in particular Hilary Blair. We are deeply indebted to our research participants. We thank Brandon Bucher and Brenton Nelson at ADinstruments for the development of the Spontaneous Cardiac Baroreflex Analysis Program. Funding sources: NIH DK090541 (MJJ, RB), NIH NS32352 (MJJ), NIH T32 DK07352 (EAW, JKL), NIH F32 DK84624 (EAW), NIH 1 UL1 RR024150 (Mayo Clinic CTSA, MJJ), and NIH DK29953 (RB).

123

Clin Auton Res (2014) 24:275–283 Conflict of interest The authors have no conflicts of interest and there are no concurrent submissions.

References 1. (1996) Heart rate variability: standards of measurement, physiological interpretation and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Circulation 93:1043–1065 2. Adler GK, Bonyhay I, Failing H, Waring E, Dotson S, Freeman R (2009) Antecedent hypoglycemia impairs autonomic cardiovascular function: implications for rigorous glycemic control. Diabetes 58:360–366 3. Bazett HC (1920) An analysis of time-relations of the electrocardiogram. Heart 7:353–370 4. Bellavere F, Balzani I, De Masi G, Carraro M, Carenza P, Cobelli C, Thomaseth K (1992) Power spectral analysis of heart-rate variations improves assessment of diabetic cardiac autonomic neuropathy. Diabetes 41:633–640 5. Davis SN, Mann S, Briscoe VJ, Ertl AC, Tate DB (2009) Effects of intensive therapy and antecedent hypoglycemia on counterregulatory responses to hypoglycemia in type 2 diabetes. Diabetes 58:701–709 6. Donnelly LA, Morris AD, Frier BM, Ellis JD, Donnan PT, Durrant R, Band MM, Reekie G, Leese GP (2005) Frequency and predictors of hypoglycaemia in Type 1 and insulin-treated Type 2 diabetes: a population-based study. Diabet Med 22:749–755 7. Ewing DJ, Neilson JM, Shapiro CM, Stewart JA, Reid W (1991) Twenty four hour heart rate variability: effects of posture, sleep, and time of day in healthy controls and comparison with bedside tests of autonomic function in diabetic patients. Br Heart J 65:239–244 8. Frattola A, Parati G, Gamba P, Paleari F, Mauri G, Di Rienzo M, Castiglioni P, Mancia G (1997) Time and frequency domain estimates of spontaneous baroreflex sensitivity provide early detection of autonomic dysfunction in diabetes mellitus. Diabetologia 40:1470–1475 9. Freeman R, Saul JP, Roberts MS, Berger RD, Broadbridge C, Cohen RJ (1991) Spectral analysis of heart rate in diabetic autonomic neuropathy. A comparison with standard tests of autonomic function. Arch Neurol 48:185–190 10. Huang CL, Liu CC, Tseng HC, Wang YP, Tsai SK (2000) Comparison of invasive and non-invasive measurement of spontaneous baroreflex during anesthesia. Acta Anaesthesiol Sin 38:149–153 11. Kleiger RE, Miller JP, Bigger JT Jr, Moss AJ (1987) Decreased heart rate variability and its association with increased mortality after acute myocardial infarction. Am J Cardiol 59:256–262 12. Koivikko ML, Salmela PI, Airaksinen KE, Tapanainen JS, Ruokonen A, Makikallio TH, Huikuri HV (2005) Effects of sustained insulin-induced hypoglycemia on cardiovascular autonomic regulation in type 1 diabetes. Diabetes 54:744–750 13. Laitinen T, Huopio H, Vauhkonen I, Camaro C, Hartikainen J, Laakso M, Niskanen L (2003) Effects of euglycaemic and hypoglycaemic hyperinsulinaemia on sympathetic and parasympathetic regulation of haemodynamics in healthy subjects. Clin Sci Lond 105:315–322 14. Limberg JK, Taylor JL, Dube S, Basu R, Basu A, Joyner MJ, Wehrwein EA (2014) Role of the carotid body chemoreceptors in baroreflex control of blood pressure during hypoglycaemia in humans. Exp Physiol 99(4):640–650 15. Liu D, Moberg E, Kollind M, Lins PE, Adamson U, Macdonald IA (1992) Arterial, arterialized venous, venous and capillary blood glucose measurements in normal man during

Clin Auton Res (2014) 24:275–283

16.

17. 18.

19.

20.

21.

22.

23.

hyperinsulinaemic euglycaemia and hypoglycaemia. Diabetologia 35:287–290 Lykke JA, Tarnow L, Parving HH, Hilsted J (2008) A combined abnormality in heart rate variation and QT corrected interval is a strong predictor of cardiovascular death in type 1 diabetes. Scand J Clin Lab Invest 68:654–659 Malpas SC, Maling TJ (1990) Heart-rate variability and cardiac autonomic function in diabetes. Diabetes 39:1177–1181 Marques JL, George E, Peacey SR, Harris ND, Macdonald IA, Cochrane T, Heller SR (1997) Altered ventricular repolarization during hypoglycaemia in patients with diabetes. Diabet Med 14:648–654 Parati G, Casadei R, Groppelli A, Di Rienzo M, Mancia G (1989) Comparison of finger and intra-arterial blood pressure monitoring at rest and during laboratory testing. Hypertension 13:647–655 Petersen ME, Williams TR, Sutton R (1995) A comparison of non-invasive continuous finger blood pressure measurement (Finapres) with intra-arterial pressure during prolonged head-up tilt. Eur Heart J 16:1641–1654 Pinna GD, La Rovere MT, Maestri R, Mortara A, Bigger JT, Schwartz PJ (2000) Comparison between invasive and noninvasive measurements of baroreflex sensitivity; implications for studies on risk stratification after a myocardial infarction. Eur Heart J 21:1522–1529 Robinson RT, Harris ND, Ireland RH, Macdonald IA, Heller SR (2004) Changes in cardiac repolarization during clinical episodes of nocturnal hypoglycaemia in adults with Type 1 diabetes. Diabetologia 47:312–315 Rosengard-Barlund M, Bernardi L, Fagerudd J, Mantysaari M, Af Bjorkesten CG, Lindholm H, Forsblom C, Waden J, Groop PH

283

24.

25.

26.

27.

28.

29.

30. 31.

(2009) Early autonomic dysfunction in Type 1 diabetes: a reversible disorder? Diabetologia 52:1164–1172 Schachinger H, Port J, Brody S, Linder L, Wilhelm FH, Huber PR, Cox D, Keller U (2004) Increased high-frequency heart rate variability during insulin-induced hypoglycaemia in healthy humans. Clin Sci Lond 106:583–588 Tu E, Twigg SM, Semsarian C (2010) Sudden death in type 1 diabetes: the mystery of the ‘dead in bed’ syndrome. Int J Cardiol 138:91–93 Wehrwein EA, Curry TB, Basu A, Rizza RA, Basu R, Joyner MJ (2012) Do the carotid bodies modulate hypoglycemic counter regulation and baroreflex control of blood pressure in humans? Adv Exp Med Biol 758:129–135 Weston PJ, James MA, Panerai RB, McNally PG, Potter JF, Thurston H (1998) Evidence of defective cardiovascular regulation in insulin-dependent diabetic patients without clinical autonomic dysfunction. Diabetes Res Clin Pract 42:141–148 Weston PJ, Panerai RB, McCullough A, McNally PG, James MA, Potter JF, Thurston H, Swales JD (1996) Assessment of baroreceptor-cardiac reflex sensitivity using time domain analysis in patients with IDDM and the relation to left ventricular mass index. Diabetologia 39:1385–1391 Xhyheri B, Manfrini O, Mazzolini M, Pizzi C, Bugiardini R (2012) Heart rate variability today. Prog Cardiovasc Dis 55:321–331 Yki-Jarvinen H, Koivisto VA (1986) Natural course of insulin resistance in type I diabetes. N Engl J Med 315:224–230 Ziegler D, Laude D, Akila F, Elghozi JL (2001) Time- and frequency-domain estimation of early diabetic cardiovascular autonomic neuropathy. Clin Auton Res 11:369–376

123

Autonomic control during acute hypoglycemia in type 1 diabetes mellitus.

We hypothesized that adults with type 1 diabetes mellitus (T1DM) would exhibit impaired heart rate variability (HRV), QT interval, T-wave amplitude, a...
330KB Sizes 0 Downloads 9 Views