Hindawi

Cardiovascular erapeutics
Volume 2020, Article ID 8563135, 9 pages
https://doi.org/10.1155/2020/8563135

Research Article
Heart Rate Variability in Patients with Hypertension: the Effect of
Metabolic Syndrome and Antihypertensive Treatment

Małgorzata Maciorowska , Paweł Krzesiński, Robert Wierzbowski, and Grzegorz Gielerak

Department of Cardiology and Internal Diseases, Military Institute of Medicine, Szaserow Street 128, 04-141 Warsaw 44, Poland

Correspondence should be addressed to Małgorzata Maciorowska; [email protected]

Received 16 June 2020; Revised 10 September 2020; Accepted 25 September 2020; Published 14 October 2020

Academic Editor: Nicholas B. Norgard

Copyright © 2020 Małgorzata Maciorowska et al. This is an open access article distributed under the Creative Commons
Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work
is properly cited.

Metabolic syndrome (MetS) is a combination of factors which, collectively, increase cardiovascular risk to a greater extent than each
of them separately. Previous studies showed high cardiovascular risk to be associated with autonomic nervous system dysfunction.
The purpose of this study was to assess the effects of antihypertensive treatment on heart rate variability (HRV) in patients with
hypertension (HTN), depending on cooccurrence of MetS. 118 patients with uncontrolled HTN were enrolled to the study.
HRV was compared among patients with and without MetS (MetS [+], n = 70) at baseline and following 12 months
antihypertensive treatment. The HRV indices measured from RR intervals recorded form using 24-hour ambulatory
electrocardiography. The measured HRV domains were the standard deviation of the average of NN intervals [SDNN], square
root of the mean of the sum of the squares of differences between adjacent NN intervals [rMSSD], percentage of NN50
[pNN50], low frequency [LF], high frequency [HF], total power of variance of all NN intervals [TP], and LF/HF ratio. Baseline
parameters: SDNN, rMSSD, pNN50, and HF were significantly lower in the MetS[+] compared to the MetS[-] subgroup
(p < 0:05). After a 12-month antihypertensive treatment, MetS[+] patients achieved a significant improvement in parameters:
SDNN, rMSSD, pNN50, and TP (p < 0:05), while the changes in HRV observed in the MetS[-] subgroup were not statistically
significant. The cooccurrence of HTN and other components of MetS is associated with disturbances of the autonomic balance.
HTN control has a beneficial effect on HRV, with the effect being more evident in patients with MetS.

1. Introduction prevalence of MetS from 50.4% in 2001 to 58.1% in 2010

[7]. European MetS prevalence, using the International Dia-
Metabolic syndrome (MetS) is a combination of cardiovascu- betes Federation diagnostic criteria, has been estimated as
lar risk factors, such as abdominal obesity, hypertension 41% in men and 38% in women [8]. Previous studies show
(HTN), abnormal glucose metabolism, and atherogenic dys- that MetS is a significant predictor of cardiovascular morbid-
lipidemia (hypertriglyceridemia and low HDL levels) [1]. ity and mortality [3, 4, 8]. Mediterranean Hypertensive Pop-
Combined, these factors were shown not only to adversely ulation patients with three or more components of MetS had
affect cardiovascular hemodynamics [2] but, above all, to threefold higher risk for cardiac events, 2.59 for cerebrovas-
increase cardiovascular risk to a greater extent than each of cular, and 2.26 for total cardiovascular events compared with
them individually [3–5]. those with no other component [9]. Consequently, the total
MetS is very common in general population, but the cost of the complications of the syndrome including the cost
prevalence is highly influenced by different diagnostic criteria of health care and loss of potential economic activity is huge.
used. According to the 2003-2012 data from NHANES prev- Earlier studies demonstrated the complex pathophysiol-
alence of the MetS in the United States was 33%, with preva- ogy of MetS by identifying the role of genetic and environ-
lence growing with age, 18.3% among those 20-39 years to mental factors, insulin resistance, inflammation, and
46.7% in those aged 60 years or older [6], the survey con- oxidative stress [10, 11]. Therefore, MetS is unequivocally a
ducted in the Chinese elderly population showed increase systemic condition. There are a number of papers showing
2 Cardiovascular Therapeutics

MetS-associated autonomic nervous system (ANS) imbal- significant arrhythmias, significant valvular heart disease,
ance, manifesting as elevated sympathetic and diminished chronic obstructive pulmonary disease, previously diagnosed
parasympathetic activity. This phenomenon has been ob- diabetes mellitus, polyneuropathy, and peripheral vascular
served for all MetS components, including HTN [12–14]. disease; (4) age < 18 years and > 75 years; (5) body mass
The sympathetic hyperexcitability appears to have primarily index ðBMIÞ > 40 kg/m2 ; (6) psychiatric disorders precluding
consequences for the development of obesity and insulin the patient’s cooperation; (7) any nonsinus hearth rhythm
resistance as well as hypertension, what is connected with (including permanent cardiac pacing); and (8) ECG tracings
elevated urinary and plasma noradrenaline levels, TNFα con- containing > 300 premature complexes and artifacts. The
tribution, elevation of adipokine levels, renal upregulation of study protocol had been approved by the Institutional
glucose transporters, β-adrenoceptor sensitization, and Review Board at Military Institute of Medicine (Approval
angiotensin II release [15]. No. 21/WIM/2011), and each patient had provided his or
There are many methods that allow both direct and indi- her written consent.
rect assessment of the ANS function. An indirect method,
relatively easily accessible, is evaluation of heart rate variabil- 2.2. History and Physical Examination. History-taking and
ity (HRV) in ambulatory electrocardiography. It has been physical examination focused particularly on cardiovascular
postulated that a noninvasive assessment of ANS activity, risk factors: age, sex, office systolic blood pressure (SBP),
e.g., via analyzing heart rate variability (HRV), may be useful office diastolic blood pressure (DBP), smoking, family his-
in identifying patients at risk of developing MetS in the future tory of heart disease, and BMI. The following parameters were
[16]. Particularly, 24-hour recordings seem to be more reli- measured in each patient: fasting blood glucose (mg/dL), cre-
able to clarify to what extent HRV is altered in MetS. HRV atinine (mg/dL), high-density lipoprotein (HDL) cholesterol
is frequently abnormal in patients with clinically overt car- (mg/dL), low-density lipoprotein (LDL) cholesterol (mg/dL),
diovascular conditions, such as coronary artery disease and triglyceride levels (mg/dL), and estimated GFR (MDRD
heart failure [17, 18], strongly related to poorly controlled eGFR) (mL/min/1.73m2). Metabolic syndrome was diagnosed
cardiometabolic risk factors. based on the International Diabetes Federation (IDF) criteria
There are studies which have examined the association [20]: central obesity—waist circumference > 94 cm for Euro-
between metabolic syndrome and heart variability, but fewer pean men and >80 cm for European women plus any two of
take the challenge to evaluate the effectiveness of applied the following four factors: triglyceride levels ≥ 150 mg/dL
treatment in primarily not treated hypertensives. Therefore, (≥1.7 mmol/L), or treatment for hypertriglyceridemia; HDL
the purpose of this study was the assumption that in the case − cholesterol levels < 40 mg/dL (<1.03 mmol/L) in men or
of patients with HTN and MetS, it seems of clinical impor- <50 mg/dL (<1.29 mmol/L) in women, or treatment for low
tance to determine how much the concomitant metabolic HDL; SBP ≥ 130 mmHg or DBP ≥ 85 mmHg, or treatment
disturbances affect HRV and whether or not antihyperten- for previously diagnosed HTN; fasting plasma glucose ≥ 100
sive treatment modifies HRV to the same extent as in hyper- mg/dL (≥5.6 mmol/l), or previously diagnosed diabetes
tensive patients without MetS. mellitus.

2. Methods 2.3. 24-Hour Ambulatory Electrocardiography (Holter

Monitoring). All patients underwent 24-hour ambulatory
2.1. Study Population. This study analyzed the data collected electrocardiography with 3-channel digital LifeCard CF
from patients recruited for the FINEPATH research study recorders (Del Mar Reynolds Medical - Spacelabs Healthcare;
(ClinicalTrials.gov Identifier NCT01996085), which had US) to assess any arrhythmias, minimum, mean, maximum
been conducted at the Department of Cardiology and Inter- heart rate (HR), and HRV.
nal Diseases of Military Institute of Medicine, in the period The time-domain and spectral (frequency-domain) HRV
2011–2014. The FINEPATH study enrolled 144 patients with parameters were analyzed with the use of the Impresario
uncontrolled HTN, defined as elevated blood pressure Symphony Holter Analyzer system (Del Mar Reynolds Med-
(≧140/90 mmHg) for at least 3 months prior to study enroll- ical, Spacelabs Healthcare Ltd/UK). The preliminary process-
ment, without pharmacotherapy at baseline. The key patient ing of the obtained ECG tracings included a review and
characteristics were presented in one of the earlier publica- correction of wrongly classified beats, artifact elimination,
tions by our team [19]. The FINEPATH study was a and evaluation of any arrhythmias and ST-segment changes.
prospective, randomized, controlled study to assess a novel Only the R-R intervals between normal QRS complexes were
HTN treatment. The following drug classes were used after analyzed, with the R-R intervals preceding and following
baseline assessment: beta-blockers, angiotensin converting ventricular premature complexes excluded from analysis.
enzyme (ACE) inhibitors, angiotensin receptor blockers The author who analyzed the HRV data was blinded to which
(ARB), calcium channel blockers (CCB), and diuretics, either patients had metabolic syndrome and which patients did not.
alone or in combination. The final follow-up visit was The patients were asked to avoid intense physical activity,
conducted 12 months after treatment initiation. The exclu- smoking, and drinking alcohol. They were recommended to
sion criteria were (1) secondary HTN; (2) chronic kidney dis- stop their activity at 10 p.m. and sleep till 6 a.m. The exami-
ease (glomerular filtration rate ðGFRÞ < 60 mL/min/1:73m2 nations were performed in hospital settings that limited the
calculated by the MDRD formula); (3) other severe comor- influences of other confounding factors, such as diet and
bidities, including systolic heart failure, cardiomyopathy, work stress.
Cardiovascular Therapeutics 3

Recruited patients (FINEPATH)

n = 144
n=5
HRV calculation impossible (lack of
Holter or presence of arrhythmias)

HRV analysis in 139 patients

n = 18
Lost to follow–up
121 patients returned
for follow–up visit
n=3
HRV calculation impossible (lack of
Holter or presence of arrhythmias)

HRV analysis in 118 patients

Figure 1: Study patient flow chart.

2.4. Analysis of HRV Time-Domain Parameters. The auto- comparison analysis was conducted for two subgroups:
matically detected time-domain parameters included in our MetS[+] (patients with other MetS factors apart from
analysis were daytime (parameter_day), nighttime (parame- HTN) and MetS[-] (patients not diagnosed with MetS). Stu-
ter_night), and 24-hour (parameter_24h) HRV parameters. dent’s t-test was used for normally distributed data, whereas
The time-domain analysis of HRV provides mainly quantita- the Mann-Whiney U-test was used for the data with nonnor-
tive data, illustrating the extent of variability. The following mal distribution. Spearman’s rank correlation coefficient was
parameters were used in our comprehensive HRV assessment: performed to investigate the relations between changes in BP
the standard deviation of the average of NN intervals in milli- and HRV parameters. The assessment of treatment effects
seconds (SDNN)and - for assessing the parasympathetic com- for subgroups separately involved the use of the Wilcoxon
ponent in the area under the curve - the squares for assessing signed-rank test. And the nonparametric Friedman test as an
the parasympathetic component in the area under the alternative to the two-way repeated measures ANOVA was
curve—, the square root of the mean of the sum of the squares performed in order to determine whether there is a significant
of differences between adjacent NN intervals in milliseconds interaction between MetS and effect of time (treatment). The p
(rMSSD), and the percentage of NN50 (pNN50) [21]. value of <0.05 was considered statistically significant.

2.5. Analysis of Frequency-Domain HRV Parameters. The 3. Results

analysis of frequency-domain parameters was conducted with
fast Fourier transform (FFT). Out of the total recorded spec- Out of the 144 patients included in the FINEPATH study,
trum, our analysis included the normalized low-frequency 139 underwent Holter monitoring with HRV analysis, with
(LF) (0.05–0.15 Hz) and high-frequency (HF) (0.15–0.4 Hz) data from 118 patients (who returned for the final follow-
values, LF/HF ratio, and total power of variance of all NN up visit) included in the final analysis. Eighteen patients
intervals [TP]). A spectral analysis was conducted for each had been lost to follow-up (they failed to return for the visit
hour out of the 24-hour period. Subsequently, the mean day- after 12 months), and the HRV of 3 patients could not be cal-
time and nighttime values were calculated and the day/night culated due to a lack of Holter monitoring or the presence or
ratio. The HF parameter was considered to be an indicator arrhythmias) (Figure 1).
of the parasympathetic activity. LF values depend on the effect
of both the vagus nerve and sympathetic tone. At rest, LF 3.1. Baseline Characteristics. The study group comprised
shows a combined effect of the sympathetic and parasympa- mostly males (69%). The mean age was 46 years, mean HR
thetic nervous systems, whereas following sympathetic stimu- 74 bpm, and mean blood pressure 141/90 mmHg. All patients
lation (e.g., standing up, exercise, and psycho-emotional enrolled to the study were Caucasian. More than half (59%)
stress), LF reflects mainly the activity of the sympathetic ner- of the patients met the MetS criteria (Table 1). Those subjects
vous system. The relationship between LF and HF (LF/HF were slightly (borderline p) and more frequently males.
ratio) reflects the sympathetic-parasympathetic balance [21].
3.2. Comparison of Baseline HRV Values. Table 2 presents
2.6. Statistical Analysis. Statistical analyses were conducted HRV parameters in patients stratified by the presence or
with Statistica 12.0 (StatSoft Inc.). The distribution and normal- absence of MetS prior to antihypertensive treatment initia-
ity of data were assessed visually and with the Kolmogorov- tion. In comparison with the MetS[-] subgroup, the patients
Smirnov test. Continuous variables were presented as the from the MetS[+] subgroup of comparable age, HR, and
mean ± standard deviation (SD), whereas categorical variables blood pressure values showed significantly lower values of
were presented as absolute and relative values (percentages). A the following time-domain HRV parameters: SDNN_24h
4 Cardiovascular Therapeutics

Table 1: Baseline patient characteristics (the entire study population).

All [n = 118] MetS [+], n = 70 MetS [-], n = 40 p

Age [years] 46 ± 10 48 ± 10 44 ± 11 0.054
Males 80 (68) 55 (79) 25 (52) 0.002
HR [bpm] 74 ± 11 74 ± 10 73 ± 12 0.382
OSBP [mmHg] 141 ± 13 142 ± 11 140 ± 15 0.284
ODBP [mmHg] 90 ± 9 90 ± 9 89 ± 10 0.500
MetS (IDF) 70 (59) 70 (100) 0 (0) —
Creatinine [mg/dL] 0:83 ± 0:16 0:85 ± 0:14 0:82 ± 0:19 0.342
eGFR [mL/min/1.73m2] 99 ± 17 99 ± 16 99 ± 19 0.683
Glucose [mg/dL] 99 ± 12 103 ± 12 93 ± 7:2 <0.001
Total cholesterol [mg/dL] 225 ± 39 226 ± 40 222 ± 37 0.828
HDL [mg/dL] 59 ± 18 52 ± 15 68 ± 19 <0.001
LDL [mg/dL] 144 ± 34 148 ± 31 139 ± 39 0.294
TG [mg/dL] 152 ± 76 186 ± 73 103 ± 48 <0.001
BMI [kg/m2] 29 ± 4 30 ± 4 27 ± 4 <0.001
Smokers 23 (19) 10 (21) 13 (19) 0.761
Statin 3 (2.5) 3 (4) 0 (0) 0.146
Data presented as mean ± SD/n (%). BM: body mass index; eGFR: estimated glomerular filtration rate; HR: heart rate; IDF: International Diabetes Foundation
criteria; MetS: metabolic syndrome; ODBP: office diastolic blood pressure; OSBP: office systolic blood pressure; TG: triglycerides.

Table 2: Comparison of HRV parameters in subgroups stratified by

(p = 0:048), SDNN_day (p = 0:015), rMSSD_24h (p = 0:002), concomitant MetS (before treatment initiation).
rMSSD_day (p = 0:002), rMSSD_night (p = 0:020), pNN50_
24h (p = 0:0008), pNN50_day (p = 0:0004), pNN50_night MetS [+], n = 70 MetS [-], n = 48
p value
(p = 0:018), and the spectral parameter HF_night (p = 0:041). Mean ± SD Mean ± SD
SDNN_24h [ms] 135.5 149.1 0.048
3.3. Assessment of 12-Month Treatment Effects. After 12-
SDNN_day [ms] 109.8 121.5 0.015
months of antihypertensive treatment, there was significant
reduction in blood pressure both in the MetS[+] and SDNN_night [ms] 90.5 94.6 0.473
MetS[-] subgroups, down to 120.1/77.2 mmHg (p < 0:001) rMSSD_24h [ms] 31.5 38.2 0.002
and 121.2/77.6 mmHg (p < 0:001), respectively. A similar rMSSD_day [ms] 27.3 32.7 0.002
effect was also observed in terms of HR, with HR of rMSSD_night [ms] 40.0 49.0 0.020
69.1 bpm (p < 0:001) and 66.5 bpm (p < 0:001), respec- pNN50_24h [%] 6.80 10.79 0.0008
tively. The BP control (<140/90 mmHg) was achieved in pNN50_day [%] 4.66 7.63 0.0004
68 (97%) pts with MetS and 42 (88%) pts without MetS
pNN50_night [%] 12.65 19.28 0.018
(p = 0:041). The 12-month follow-up showed no changes
in body weight either in the MetS[+] (91:1 ± 13:5 kg before LF/HF_day [-] 4.50 3.88 0.625
vs. 91:3 ± 13:4 kg after treatment; p = 0:73) or in the LF/HF_night [-] 3.04 2.01 0.048
MetS[-] subgroup (87:0 ± 15:9 kg before vs. 86:6 ± 15:3 kg LF_day [n.u.] 70.5 70.0 0.939
after treatment; p = 0:84). LF_night [n.u.] 62.3 55.8 0.061
Antihypertensive monotherapy was used in 47.1% of HF_day [n.u.] 23.0 24.1 0.524
patients (a CCB in 0.8%, beta-blocker in 5.0%, ARB in HF_night [n.u.] 32.3 38.3 0.041
4.2%, diuretic in 3.4%, ACE inhibitor in 33.6%). Combina-
TP_day [ms2] 2.747 3.248 0.152
tion antihypertensive therapy was used in 50.4% of
TP_night [ms2] 2.923 3.174 0.616
patients, with two-drug combination regimens of ACE
inhibitor plus diuretic in 16.8% of patients, ACE inhibitor DBP: diastolic blood pressure; HF: power in the high frequency range; HR:
plus beta blocker in 10.9%, ACE inhibitor plus CCB in heart rate; LF: power in the low frequency range; n.u. : normalized units;
pNN50: percentage of NN50; rMSSD: square root of the mean of the sum
7.6%, ARB plus diuretic in 3.4%, ARB plus beta-blocker of the squares of differences between adjacent NN intervals; SBP: systolic
in 0.8%, and ARB plus CCB in 0.8%. Three-drug combi- blood pressure; SDNN: standard deviation of the average of NN intervals;
nation regimens, used in 7.5% of patients, included ACE TP: total power of variance of all NN intervals.
inhibitor plus beta-blocker plus CCB in 2.5% of patients,
ARB plus CCB plus diuretic in 0.8%, ACE inhibitor plus tions were used in 2.3% of patients, who only received
beta-blocker plus diuretic in 2.5%, and CCB plus beta- nonpharmacological recommendations. Statins were intro-
blocker plus diuretic in 1.7%. No antihypertensive medica- duced in 16 patients with MetS (22.9%).
Cardiovascular Therapeutics 5

Table 3: Comparison of HRV parameters before and after 12-month treatment, in patients with HTN stratified by concomitant MetS.

MetS[+], n = 70 MetS[+], n = 70 MetS[-], n = 48 MetS[-], n = 48

Difference Difference
Mean ± SD Mean ± SD p value Mean ± SD Mean ± SD p value
(MetS[+]) (MetS[-])
Before After Before After
HR [L/min] 74.1 69.1 -5.0 0.023 72.3 66.5 -5.8 0.007
OSBP [mmHg] 142.1 120.1 -22.0 <0.0001 139.5 121.2 -18.3 <0.0001
ODBP [mmHg] 90.5 77.2 -13.3 <0.0001 89.3 77.6 -11.7 <0.0001
SDNN_24h [ms] 135.5 140.2 4.7 0.012 149.1 146.3 -2.8 0.665
SDNN_day [ms] 109.8 114.6 5.1 0.042 121.5 123.2 1.7 0.312
SDNN_night [ms] 90.5 95.2 4.7 0.189 94.6 96.7 2.1 0.885
rMSSD_24h [ms] 31.5 34.7 3.2 0.003 38.2 39.4 1.2 0.470
rMSSD_day [ms] 27.3 29.7 2.4 0.001 32.7 35.0 2.3 0.855
rMSSD_night [ms] 40.0 43.8 3.8 0.042 49.0 48.6 -0.4 0.112
pNN50_24h [%] 6.80 8.03 1.23 0.0002 10.79 11.42 0.69 0.665
pNN50_day [%] 4.66 5.56 0.90 0.001 7.63 8.91 1.28 0.885
pNN50_night [%] 12.65 14.29 1.64 0.051 19.28 18.13 -1.15 0.105
LF/HF_day [-] 4.50 4.02 -0.48 0.082 3.88 3.64 -0.24 0.136
LF/HF_night [-] 3.04 2.41 -0.63 0.550 2.01 1.88 -0.13 0.470
LF_day [n.u.] 70.5 67.9 -2.6 0.457 70.0 66.1 -3.9 0.136
LF_night [n.u.] 62.3 59.9 -2.4 0.403 55.8 53.7 -2.1 0.470
HF_day [n.u.] 23.0 26.1 3.1 0.082 24.1 28.4 4.3 0.074
HF_night [n.u.] 32.3 34.6 2.3 0.189 38.3 40.8 2.5 0.470
TP_day [ms2] 2.747 3.803 1.056 0.026 3.248 4.054 806 0.307
TP_night [ms2] 2.923 3.483 560 0.402 3.174 3.377 203 0.665
ODBP: office diastolic blood pressure; HF: power in the high frequency range; HR: heart rate; LF: power in the low frequency range; n.u.: normalized units;
pNN50: percentage of NN50; rMSSD: square root of the mean of the sum of the squares of differences between adjacent NN intervals; OSBP: office systolic
blood pressure; SDNN: standard deviation of the average of NN intervals; TP: total power of variance of all NN intervals.

Table 3 presents a comparison of HRV parameters in the MetS patients were those who seem to benefit more from
patients with HTN stratified by the presence or absence the treatment with respect to sympatovagal balance.
of concomitant MetS at 12 months of antihypertensive The baseline values of HRV parameters obtained in our
treatment. MetS[+] patients achieved a significant study and the impact of MetS are consistent with the data
improvement in their HRV as shown by time-domain reported in the available literature [12–14, 22–28]. A 2013
parameters: SDNN_24h (p = 0:012), SDNN_day (p = 0:042), study by Li et al., which aimed to assess the relationship
rMSSD_24h (p = 0:003), rMSSD_day (p = 0:001), rMSSD_ between MetS severity and ANS function, demonstrated
night (p = 0:042), pNN50_24h (p = 0:0002), pNN50_day independent negative correlations of two MetS components
(p = 0:001), and the frequency-domain parameter of TP_day (fasting plasma glucose and HTN) with ANS function [23].
(p = 0:026). The results achieved in the MetS[-] subgroup also Moreover, an earlier study (Twins Heart Study) conducted
suggest a favorable effect of treatment; however, the observed in 288 pairs of twins showed a relationship between MetS
differences did not reach the adopted level of significance and decreased HRV parameters, both in individual analyses
and were lower than in the MetS[+] subgroup also in terms and in the analyses of the twin pairs. Additionally, HRV
of absolute values (Table 3 and Figure 2). Friedman’s test parameters were found to be decreased in individuals with
revealed the significant interaction between MetS and effect more MetS components [24]. American researchers [25]
of treatment for SDNN_day, rMSSD_day, pNN50_24h, reached a similar conclusion while assessing the HR and
pNN50_day, and TP_day. The significant correlations were HRV parameters in patients stratified by their fasting glucose
observed for 12-month changes in diastolic blood pressure (FG) levels and other concomitant MetS components. This
and some HRV parameters (SDNN, rMSDD, and pNN50) American study demonstrated lowering of most of the evalu-
in both MetS[+] and MetS[-]. The effect on systolic blood ated HRV parameters (particularly the SDNN, standard
pressure was less related to HRV [Supplementary Table 1]. deviation of the 5-minute average NN intervals [SDANN],
TP, ultra-low frequency [ULF], and very low frequency
4. Discussion [VLF] power) in patients with markedly elevated FG (6.1–
6.9 mmol/L) and type 2 diabetes (with FG > 6:9 mmol/L or
Our findings indicate a considerable effect of metabolic dis- on antidiabetic medication or insulin) in comparison with
orders on the HRV in patients with HTN. Implemented the patients with normal (4.5–5.5 mmol/L) and slightly ele-
hypertensive therapy was effective in both subgroups, but vated (5.6–6.0 mmol/L) FG. The patients with normal to
6 Cardiovascular Therapeutics

155 ns
ns 125
150
120
145 p = 0.042
p = 0.012 115
140
110
135
130 105
125 100
MS [+] MS [–] MS[+] MS[–]

SDNN–24 h [ms] before treatment SDNN–day [ms] before treatment

SDNN–24 h [ms] after treatment SDNN–day [ms] after treatment
45 40

ns ns
40 35
p = 0.003 p = 0.001
35 30

30 25

25 20
MS [+] MS [–] MS [+] MS [–]

rMSSD–24 h [ms] before treatment rMSSD–24 h [ms] before treatment

rMSSD–24 h [ms] after treatment rMSSD–24 h [ms] after treatment
12 ns
10 ns

10 8
p = 0.0002 p = 0.001
8 6

6 4

4 2
MS [+] MS [–] MS [+] MS [–]

pNN50–24 h [ms] before treatment pNN50–24 h [ms] before treatment

pNN50–24 h [ms] after treatment pNN50–24 h [ms] after treatment

Figure 2: Comparison of selected HRV parameters before and after 12-month treatment, in patients with HTN stratified by concomitant
MetS.

slightly elevated FG who met more than 2 MetS criteria HRV parameters correlated only with fasting glucose levels,
showed decreased HRV (SDNN, SDANN, TP, and ULF) in with no differences between the groups in terms of the
comparison with the patients meeting at most one MetS cri- remaining 4 diagnostic criteria of MetS (notably, both study
terion. In patients with diabetes or markedly elevated FG, groups included patients with HTN).
MetS was associated with decreased HRV compared with Our study demonstrated the effects of antihypertensive
the HRV in patients without MetS. These American findings treatment on HRV parameters to be beneficial, particularly
were consistent with those of the Finnish authors whose 1998 in the group of patients with MetS. The more altered HRV
study demonstrated significantly decreased HRV parameters at baseline may partly explain greater reduction after 12
(SDANN, TP, VLF, LF) in hypertensive patients with insulin months of treatment in MetS. No other mechanism can be
resistance in comparison with both hypertensive patients identified basing on our data. Our findings are consistent
without insulin resistance and normotensive patients. The with earlier reports indicating beneficial effects of antihyper-
HF parameter (p < 0:001) and baroreflex sensitivity tensive treatment on HRV. However, there is no clear con-
(p < 0:05) were diminished in both hypertensive groups sensus which hypotensive drugs are the most beneficial in
[28]. A prospective study by Balcioğlu et al. [29] (n = 240) terms of the sympatovagal balance. Some earlier studies dem-
showed significantly decreased HRV parameters (SDNN, onstrated beta-blockers and ARB to be particularly beneficial
SDNN index, SDANN, rMSSD, pNN50) in 24-hour Holter in that respect [30–32]. Moreover, some other reports indi-
recordings in comparison with those in the control group. cated that ARB treatment yielded better effects than treat-
Unlike in the studies mentioned above, the lowering of ment with ACE inhibitors and beta-blockers [33–35]. One
Cardiovascular Therapeutics 7

prospective, randomized Japanese study compared the pressure control seems to complement the intervention
effects of ARB treatment in MetS patients randomized into based on diet and physical activity to reverse MetS and pre-
three therapeutic groups (telmisartan, candesartan, diet vent cardiac autonomic neuropathy [43].
therapy) [35]. At 6 months, the study showed a comparable
lowering of blood pressure in both drug-treated groups.
However, ARB treatment yielded increased baroreflex sensi- 4.1. Strengths and Limitations. The strength of our study is
tivity, increased high-molecular-weight adiponectin levels, the enrollment of hypertensive subjects, some of them with
and improved endothelial dysfunction (in this last respect, MetS, but no significant comorbidities. Moreover, there were
a more pronounced effect was achieved with telmisartan). no bias of previous hypotensive treatment at baseline assess-
Moreover, the telmisartan group showed significantly ment. Some limitations should be also considered. One is the
decreased norepinephrine levels, blood pressure variability, small size of the study population and thus a small size of
and the spectral HRV parameter of the LF/HF ratio individual subgroups. Therefore, our analyses may be under-
(p < 0:05). A study by Menzes et al. showed improvement powered, and the findings need to be confirmed in a larger
in all HRV parameters (SDNN, pNN50, LF; p < 0:001) fol- study group. Another limitation is the difference in sex pro-
lowing a 3-month treatment with an ACE inhibitor (enala- portions between MetS subgroups that could bias baseline
pril or ramipril) in contrast with the control group [31]. comparison. Moreover, we would like to point out that the
Petretta et al. assessed the effect of a 12-month lisinopril absence of any monitoring in terms of pharmacotherapy,
treatment on HRV and, for the entire study population, other than the patients’ antihypertensive treatment report,
observed an increase only in the nighttime HF parameter may have affected our results. Our study assessed neither
in comparison with baseline values [33]. However, the sub- patients’ physical activity nor the effect of treatment on
group of patients with left ventricular mass normalization the metabolic dysfunction. In terms of Discussion, we
showed increased both daytime and nighttime HF, as well would like to emphasize the issues with comparing indi-
as increased nighttime TP and VLF. A 2010 study by Pavi- vidual studies due to the differences in study protocols,
thran et al. examined 150 patients newly diagnosed with the adopted diagnostic criteria of MetS, and the length of
HTN, divided into five 30 patient groups, each receiving analyzed electrocardiographic recordings. Moreover, only
one of the following: amlodipine, atenolol, enalapril, hydro- a handful of studies included separate groups of patients
chlorothiazide, or an amlodipine+atenolol combination. with uncomplicated HTN and those with HTN and con-
Only the amlodipine+atenolol group showed a significant comitant metabolic disorders.
change in HRV (increased total variability of RR intervals
and HF spectral power) [36]. There were also studies
attempting to evaluate the effect of individual CCB medi- 5. Conclusions
cines on HRV parameters [37–40]. The available data on
Our findings confirm that the cooccurrence of HTN and
the effect of amlodipine on the HRV are contradictory.
other components of MetS is associated with differences in
Individual authors report either an insignificant-to-absent
HRV and its modulation by hypotensive medicines. Blood
effect of this drug on the ANS activity, or enhanced sympa-
pressure control has a beneficial effect on HRV, with the
thetic activity, or—conversely—vagus nerve stimulation
effect being more evident in patients with MetS.
[37–39]. A prospective, randomized study by Karas et al.
(n = 57) evaluated the effects of treatment with amlodypine,
ramipril, and telmistartan on HRV spectral analysis and
plasma norepinephrine and epinephrine level measure-
Data Availability
ments [41]. Following amlodipine treatment, an increased The datasets used and analysed during the current study are
daytime sympathetic activity and decreased nighttime para- available from the corresponding author on reasonable
sympathetic activity, together with increased plasma norepi- request.
nephrine levels, were observed. Telmisartan treatment
yielded considerably increased parasympathetic activity
without changes in plasma norepinephrine levels, whereas Conflicts of Interest
ramipril increased the parasympathetic activity only during
the day. The authors declare that there is no conflict of interest
For both baseline comparison and treatment effects, regarding the publication of this paper.
time-domain HRV parameters revealed to better diversify
the presence of metabolic burden than frequency-domain
HRV parameters. The frequency-domain HF power is Acknowledgments
assumed to correspond to the frequency of breathing, reflect-
ing respiratory sinus arrhythmia. More controversies con- The authors would like to thank the team from the Depart-
cern LF power. Some authors undermine that LF power is ment of Cardiology and Internal Diseases who took part in
and index of cardiac sympathetic tone and are even more recruiting patients and collecting data for analysis. The
willing to claim that it reflects baroreflexes [42]. FINEPATH project was financed from the Military Institute
Considering high prevalence of MetS around the world, of Medicine/Ministry of Science and Higher Education
our finding may concern a wide range of patients. Blood research grant (No. 148/WIM).
8 Cardiovascular Therapeutics

Supplementary Materials and HPA dysfunction in the metabolic syndrome,” Frontiers

in Neuroscience, vol. 10, p. 142, 2016.
Supplementary Table 1. The correlations between changes [16] L. R. Wulsin, P. S. Horn, J. L. Perry, J. M. Massaro, and R. B.
(delta) in BP and HRV parameters after 12-month treatment. DʼAgostino Sr., “The contribution of autonomic imbalance
(Supplementary Materials) to the development of metabolic syndrome,” Psychosomatic
Medicine, vol. 78, no. 4, pp. 474–480, 2016.
References [17] K. C. Bilchick, B. Fetics, R. Djoukeng et al., “Prognostic value
of heart rate variability in chronic congestive heart failure
[1] K. G. M. M. Alberti, P. Zimmet, and J. Shaw, “Metabolic syn- (veterans affairs’ survival trial of antiarrhythmic therapy in
drome–a new world-wide definition. A consensus statement congestive heart failure),” The American Journal of Cardiology,
from the international diabetes federation,” Diabetic Medicine, vol. 90, no. 1, pp. 24–28, 2002.
vol. 23, no. 5, pp. 469–480, 2006. [18] R. E. Kleiger, J. P. Miller, J. T. Bigger, and A. J. Moss,
[2] P. Krzesiński, A. Stańczyk, K. Piotrowicz, G. Gielerak, “Decreased heart rate variability and its association with
B. Uziębło-Zyczkowska, and A. Skrobowski, “Abdominal obe- increased mortality after acute myocardial infarction,” The
sity and hypertension: a double burden to the heart,” Hyper- American Journal of Cardiology, vol. 59, no. 4, pp. 256–262,
tension Research, vol. 39, no. 5, pp. 349–355, 2016. 1987.
[3] A. Scuteri, S. S. Najjar, C. H. Morrell, and E. G. Lakatta, “The [19] P. Krzesiński, G. Gielerak, A. Stańczyk et al., “The effect of
metabolic syndrome in older individuals: prevalence and pre- hemodynamically-guided hypotensive therapy in one-year
diction of cardiovascular events: the cardiovascular health observation: Randomized, prospective and controlled trial
study,” Diabetes Care, vol. 28, no. 4, pp. 882–887, 2005. (FINEPATH study),” Cardiology Journal, vol. 23, no. 2,
[4] P. W. F. Wilson and J. B. Meigs, “Cardiometabolic risk: a pp. 132–140, 2016.
Framingham perspective,” International Journal of Obesity, [20] IDF Clinical Guidelines Task Force, “Global guideline for type
vol. 32, no. S2, pp. S17–S20, 2008. 2 diabetes: recommendations for standard, comprehensive,
[5] J. D. Tune, A. G. Goodwill, D. J. Sassoon, and K. J. Mather, and minimal care,” Diabetic Medicine, vol. 23, no. 6,
“Cardiovascular consequences of metabolic syndrome,” pp. 579–593, 2006.
Translational Research, vol. 183, pp. 57–70, 2017. [21] M. Malik, J. T. Bigger, G. Breithardt et al., “Task force of the
[6] M. Aguilar, T. Bhuket, S. Torres, B. Liu, and R. J. Wong, European Society of Cardiology and the North American Soci-
“Prevalence of the metabolic syndrome in the United States, ety of Pacing and Electrophysiology Heart Rate Variability.
2003-2012,” JAMA, vol. 313, no. 19, pp. 1973-1974, 2015. Standards of Measurement, Physiological Interpretation, and
Clinical Use,” European Heart Journal, vol. 17, pp. 354–381,
[7] M. Liu, J. Wang, B. Jiang et al., “Increasing prevalence of met- 1996.
abolic syndrome in a Chinese elderly population: 2001–2010,”
PLoS One, vol. 8, no. 6, article e66233, 2013. [22] D. Liao, R. P. Sloan, W. E. Cascio et al., “Multiple metabolic
syndrome is associated with lower heart rate variability. The
[8] The DECODE Study Group, “Does the constellation of risk atherosclerosis risk in communities study,” Diabetes Care,
factors with and without abdominal adiposity associate with vol. 21, no. 12, pp. 2116–2122, 1998.
different cardiovascular mortality risk?,” International Journal
[23] Z. Li, Z. H. Tang, F. Zeng, and L. Zhou, “Associations between
of Obesity, vol. 32, no. 5, pp. 757–762, 2008.
the severity of metabolic syndrome and cardiovascular auto-
[9] E. Andreadis, G. Tsourous, C. Tzavara et al., “Metabolic syn- nomic function in a Chinese population,” Journal of Endocri-
drome and incident cardiovascular morbidity and mortality nological Investigation, vol. 36, no. 11, pp. 993–999, 2013.
in a Mediterranean hypertensive population,” American Jour-
[24] A. K. Gehi, R. Lampert, E. Veledar et al., “A twin study of
nal of Hypertension, vol. 20, no. 5, pp. 558–564, 2007.
metabolic syndrome and autonomic tone,” Journal of Car-
[10] S. L. Samson and A. J. Garber, “Metabolic syndrome,” Endocri- diovascular Electrophysiology, vol. 20, no. 4, pp. 422–428,
nology and Metabolism Clinics of North America, vol. 43, no. 1, 2009.
pp. 1–23, 2014.
[25] P. K. Stein, J. I. Barzilay, P. P. Domitrovich et al., “The relation-
[11] M. Tanaka, “Improving obesity and blood pressure,” Hyper- ship of heart rate and heart rate variability to non-diabetic fast-
tension Research, vol. 43, no. 2, pp. 79–89, 2020. ing glucose levels and the metabolic syndrome: the
[12] M. I. Stuckey, M. P. Tulppo, A. M. Kiviniemi, and R. J. Petrella, cardiovascular health study,” Diabetic Medicine, vol. 24,
“Heart rate variability and the metabolic syndrome: a system- no. 8, pp. 855–863, 2007.
atic review of the literature,” Diabetes/Metabolism Research [26] E. J. Brunner, H. Hemingway, B. R. Walker et al., “Adrenocor-
and Reviews, vol. 30, no. 8, pp. 784–793, 2014. tical, autonomic, and inflammatory causes of the metabolic
[13] L. R. Wulsin, P. S. Horn, J. L. Perry, J. M. Massaro, and R. B. syndrome: nested case–control study,” Circulation, vol. 106,
D'Agostino, “Autonomic imbalance as a predictor of meta- no. 21, pp. 2659–2665, 2002.
bolic risks, cardiovascular disease, diabetes, and mortality,” [27] C. J. Chang, Y. C. Yang, F. H. Lu et al., “Altered cardiac auto-
The Journal of Clinical Endocrinology and Metabolism, nomic function may precede insulin resistance in metabolic
vol. 100, no. 6, pp. 2443–2448, 2015. syndrome,” The American Journal of Medicine, vol. 123,
[14] J. F. Thayer, S. S. Yamamoto, and J. F. Brosschot, “The rela- no. 5, pp. 432–438, 2010.
tionship of autonomic imbalance, heart rate variability and [28] S. M. Pikkujämsä, H. V. Huikuri, K. E. Airaksinen et al., “Heart
cardiovascular disease risk factors,” International Journal of rate variability and baroreflex sensitivity in hypertensive sub-
Cardiology, vol. 141, no. 2, pp. 122–131, 2010. jects with and without metabolic features of insulin resistance
[15] E. Lemche, O. S. Chaban, and A. V. Lemche, “Neuroendorine syndrome,” American Journal of Hypertension, vol. 11, no. 5,
and epigentic mechanisms subserving autonomic imbalance pp. 523–531, 1998.
Cardiovascular Therapeutics 9

[29] A. S. Balcioğlu, S. Akinci, D. Çiçek et al., “Which is responsible [42] D. S. Goldstein, O. Bentho, M. Y. Park, and Y. Sharabi, “Low-
for cardiac autonomic dysfunction in non-diabetic patients frequency power of heart rate variability is not a measure of
with metabolic syndrome: prediabetes or the syndrome cardiac sympathetic tone but may be a measure of modulation
itself?,” Diabetes & Metabolic Syndrome: Clinical Research & of cardiac autonomic outflows by baroreflexes,” Experimental
Reviews, vol. 10, no. 1, pp. S13–S20, 2016. Physiology, vol. 96, no. 12, pp. 1255–1261, 2011.
[30] S. Guzzetti, E. Piccaluga, R. Casati et al., “Sympathetic pre- [43] S. M. Williams, A. Eleftheriadou, U. Alam, D. J. Cuthbertson,
dominance in essential hypertension: a study employing spec- and J. P. H. Wilding, “Cardiac autonomic neuropathy in obe-
tral analysis of heart rate variability,” Journal of Hypertension, sity, the metabolic syndrome and prediabetes: a narrative
vol. 6, no. 9, pp. 711–717, 1988. review,” Diabetes Therapy, vol. 10, no. 6, pp. 1995–2021, 2019.
[31] A. S. Menzes Jr., H. G. Moreira, and M. T. Daher, “Analysis of
heart rate variability in hypertensive patients before and after
treatment with angiotensin II-converting enzyme inhibitors,”
Arquivos Brasileiros de Cardiologia, vol. 83, no. 2, pp. 169–
172, 2004.
[32] R. Vesalainen, I. Kantola, K. Airaksinen, K. Tahvanainen, and
T. Kaila, “Vagal cardiac activity in essential hypertension: the
effects of metoprolol and ramipril,” American Journal of
Hypertension, vol. 11, no. 6, Part 1, pp. 649–658, 1998.
[33] M. Petretta, D. Bonaduce, F. Marciano et al., “Effect of 1 year of
lisinopril treatment on cardiac autonomic control in hyperten-
sive patients with left ventricular hypertrophy,” Hypertension,
vol. 27, no. 3, pp. 330–338, 1996.
[34] C. M. Chern, H. Y. Hsu, H. H. Hu, Y. Y. Chen, L. C. Hsu, and
A. C. Chao, “Effects of atenolol and losartan on baroreflex sen-
sitivity and heart rate variability in uncomplicated Essential
hypertension,” Journal of Cardiovascular Pharmacology,
vol. 47, no. 2, pp. 169–174, 2006.
[35] T. Kishi, Y. Hirooka, S. Konno, and K. Sunagawa, “Angioten-
sin II receptor blockers improve endothelial dysfunction asso-
ciated with sympathetic hyperactivity in metabolic syndrome,”
Journal of Hypertension, vol. 30, no. 8, pp. 1646–1655, 2012.
[36] P. Pavithran, E. S. Prakash, T. K. Dutta, and T. Madanmohan,
“Effect of antihypertensive drug therapy on short-term heart
rate variability in newly diagnosed essential hypertension,”
Clinical and Experimental Pharmacology & Physiology,
vol. 37, no. 2, pp. e107–e113, 2010.
[37] T. Hamada, M. Watanabe, T. Kaneda et al., “Evaluation of
changes in sympathetic nerve activity and heart rate in essen-
tial hypertensive patients induced by amlodipine and nifedi-
pine,” Journal of Hypertension, vol. 16, no. 1, pp. 111–118,
1998.
[38] J. de Champlain, M. Karas, P. Nguyen et al., “Different effects
of nifedipine and amlodipine on circulating catecholamine
levels in essential hypertensive patients,” Journal of Hyperten-
sion, vol. 16, no. 9, pp. 1357–1369, 1998.
[39] R. Žaliūnas, J. Braždžionytė, V. Zabiela, and R. Jurkevičius,
“Effects of amlodipine and lacidipine on heart rate variability
in hypertensive patients with stable angina pectoris and iso-
lated left ventricular diastolic dysfunction,” International Jour-
nal of Cardiology, vol. 101, no. 3, pp. 347–353, 2005.
[40] B. M. Psaty, S. R. Heckbert, T. D. Koepsell et al., “The risk of
myocardial infarction associated with antihypertensive drug
therapies,” JAMA, vol. 274, no. 8, pp. 620–625, 1995.
[41] M. Karas, Y. Lacourcière, A. R. LeBlanc et al., “Effect of the
renin-angiotensin system or calcium channel blockade on the
circadian variation of heart rate variability, blood pressure
and circulating catecholamines in hypertensive patients,” Jour-
nal of Hypertension, vol. 23, no. 6, pp. 1251–1260, 2005.