Abnormalities of Anthropometric, Hemodynamic, and Autonomic Variables in Offspring of Hypertensive Parents

Abstract

Young adult offspring of hypertensive parents (pHTN⊕) are a good model for assessing abnormalities of anthropometric, cardiometabolic, and autonomic variables prior to clinical hypertension. The objectives of this study were to determine whether these variables and autonomic responses to oral carbohydrates were altered in offspring of pHTN⊕. Two hundred consecutive patients, including 100 pHTN⊕, were evaluated, with 29 patients, including 14 pHTN⊕, given a 70‐gram carbohydrate load. The pHTN⊕ group had higher blood pressure, pulse pressure, abdominal circumference (AC), weight, body mass index, and basal metabolic rate than offspring of normotensive parents (pHTN∅). At baseline, the low‐frequency (LF, sympathetic) to high‐frequency (HF, parasympathetic) ratio, assessed by spectral analysis of heart rate variability, was similar in both groups. After the carbohydrate load, the LF/HF ratio was greater in offspring of pHTN⊕. pHTN⊕ individuals have abnormalities of anthropometric and hemodynamic variables at baseline and autonomic responses to oral carbohydrates before developing hypertension.

Hypertension is associated with a large and growing health and economic burden of cardiovascular and renal diseases.1, 2 High arterial pressure causes target organ changes in the brain, heart, vessels, and kidneys. The etiology of arterial hypertension is unknown in most cases, although genetic factors play an important role.3 Environmental and behavioral factors also participate and include excessive salt and alcohol intake, stress, smoking, sedentary lifestyle, and obesity. Obesity is commonly associated with hypertension and appears to begin early in offspring of hypertensive parents (pHTN⊕).4, 5

The offspring of pHTN⊕ are more likely to develop hypertension than those of normotensive parents (pHTN∅).4 Normotensive offspring of pHTN⊕ have higher casual arterial pressure than normotensive offspring of pHTN∅.4 The difference in arterial pressure between children of pHTN∅ and children of pHTN⊕ has been reported in infancy, childhood, adolescence, and adulthood.6 Anthropometric data such as body mass index (BMI), waist circumference, and neck circumference might be related to the pathogenesis of hypertension in the offspring of pHTN⊕. However, there is no consensus regarding this issue.7, 8, 9 The interrelation between offspring of pHTN⊕ and both anthropometric and environmental variables can result in increased blood pressure (BP) in patients with familial predisposition, ie, gene‐environment interactions.10 Although associations between hypertension and anthropometric, metabolic, and autonomic nervous system variables were reported,11, 12, 13 few studies have addressed this association in offspring of pHTN⊕ parents.

The sympathetic nervous system plays an important role in the pathogenesis of arterial hypertension. Greater activity of the sympathetic nervous system leads to lower heart rate variability (HRV) and higher risk for cardiovascular events, such as myocardial infarction and sudden death.14, 15 Conversely, higher parasympathetic nervous system activity increases HRV and is associated with lower heart‐related mortality. Autonomic nervous system regulation in hypertensive patients appears to be altered, and the changes can be detected through spectral analysis.15

Autonomic nervous system changes can be affected by environmental factors including food intake, and may be modulated by family history. Inappropriate increase in sympathetic activity after any kind of challenge in normotensive patients could be indicative of autonomic imbalance preceding the onset of hypertension. In one report, normotensive offspring of pHTN⊕ show increased sympathetic activity at baseline and after isometric exercise.16 Carbohydrate consumption can have a negative effect on sympathetic activity in hypertensive patients.17 However, the impact of carbohydrate ingestion on autonomic balance has not been reported in normotensive offspring of pHTN⊕.

The aim of this study was to assess anthropometric, hemodynamic, and metabolic variables, including basal metabolic rate, in pHTN⊕ and pHTN∅ offspring prior to the onset of hypertension. In a subset of pHTN⊕ and pHTN∅ offspring, autonomic function was assessed under fasting conditions and following an oral carbohydrate load.

Patients and Methods

Study design

The study protocol was approved by the research ethics committee of the Hospital Campo Limpo under the number CAAE: 05505012.7.1001.5511. All participants signed the informed consent document, according to resolution 466/12.

We recruited 255 participants (hospital workers) with a pHTN⊕ and pHTN∅ background. Patients were considered offspring of pHTN⊕ if they were aware that one or both parents were taking antihypertensive medicine. The medical records of parents were also used to better characterize history of hypertension. Participants did not have major medical conditions and were not receiving any kind of medication therapy. The screening phase included a history, physical examination, anthropometric measures, and biochemical tests.

Inclusion criteria included Caucasian and non‐Caucasian men and women aged 18 to 35 years with BP <140/<90 mm Hg, BMI<40 kg/m2, low‐density lipoprotein cholesterol <190 mg/dL, triglycerides <400 mg/dL, normal renal function, and no systemic diseases were included. Exclusion criteria were hypertension, valvular heart disease, cardiac pacemaker, severe arrhythmia, hypothyroidism or hyperthyroidism, collagen disease, Crohn's disease and celiac disease, any type of cancer or disabling chronic disease, smoking, and diabetes. Participants who completed the screening phase and met inclusion/exclusion criteria were included in the study.

Two hundred individuals including 100 pHTN⊕ and 100 pHTN∅ were selected and matched for age, sex, and race (Figure 1).

Figure 1.

Flowchart of study participants.

Biochemical testing

All volunteers underwent biochemical testing including uric acid, high‐sensitivity C‐reactive protein, brain natriuretic peptide, glycated hemoglobin, insulin, tetraiodothyronine, triiodothyronine, sodium, potassium, urea, and creatinine using commercial assays. Total cholesterol, high‐density lipoprotein (HDL) cholesterol, triglycerides, blood glucose, sodium, and potassium levels were tested with the enzymatic colorimetric method (ABI PRISM 310, Applied Biosystems, Carlsbad, CA).

Anthropometric and metabolic measures such as weight, height, BMI, abdominal circumference, hip circumference, and neck circumference were obtained.18 Lean mass, percent body fat, and basal metabolic rate (BMR) were evaluated by bioimpedance (BIA 450 Bioimpedance Analyzer, Bio Dyncorp, Shoreline, WA).

Blood pressure

Casual BP was measured with patients in the sitting position after 5 minutes of rest using a calibrated mercury sphygmomanometer. Heart rate, measured by palpating the radial pulse, was measured in patients after at least 5 minutes of rest in the seated position.

Noninvasive hemodynamics, including BP, stroke volume, cardiac output and index, total vascular resistance, and distensibility of large (vascular compliance) and small (endothelial function) arteries were evaluated using the HDI/PulseWave CR‐2000 (Hypertension Diagnostics, Eagan, MN).19

All patients were instructed to fast for 4 hours before testing and to abstain from alcohol and caffeinated beverages as well as vigorous physical activity for 24 hours before testing. They emptied their bladder 30 minutes before the assessment. They were also instructed to rest for 5 minutes and remove all metal objects.

Power spectral analysis before and after carbohydrate loading

Forty participants, 21 with pHTN∅ and 19 with pHTN⊕, were selected for assessing sympathovagal balance before and after an oral carbohydrate load. They were matched for age, sex, race and BP. From these 40 participants, 29 (15 pHTN∅ and 14 were pHTN⊕) had acceptable electrocardiographic (heart rate) signal quality for inclusion in the carbohydrate substudy.

Oral carbohydrate load

Hemodynamic and autonomic variables were assessed before and after ingesting approximately 70 g of high glycemic index carbohydrates (a can of cola or guarana soft drink and two slices of white bread with jam). This test was performed in the morning following an overnight fast.

Hemodynamic and autonomic balance evaluation

Beat‐to‐beat arterial BP was measured with a Finometer (FMS, Finapres Medical System BV, Holland, Netherlands) and electrocardiographic findings were also recorded for 20 minutes before and again between 40 and 60 minutes after the mixed carbohydrate meal.

Spectral analysis

Heart rate and BP recorded at baseline and after the oral carbohydrate load were used in the spectral analysis. First, manual editing of the signals was performed via the detection of the systolic events (peak) of the systolic BP beat‐to‐beat signal. Pulse interval was estimated based on the interval between consecutive systoles. Next, each heartbeat was identified using an algorithm with the Matlab program (Mathwors Inc, Natick, MA; Baum‐Welch algorithm) generating the final result of the spectral analysis with the relevant bands. The frequency band relevant for the spectral analysis in humans lies between 0.0 Hz and 0.4 Hz approximately.20

• High frequencies (HFs) between 0.4 Hz and 0.15 Hz (parasympathetic)

• Low frequencies (LFs) between 0.15 Hz and 0.04 Hz (sympathetic)

• Autonomic balance: LF/HF ratio

The components of HRV in the frequency domain were analyzed and presented in their normalized form (normalized units [nu]):

• LF nu = power of LF/(total power/ms²−VLF) ×100

• HF nu = power of HF/(total power/ms²−VLF) ×100

• LF/HF = LF ms²/HF ms² ratio

The variability of systolic and diastolic BP and heart rate were calculated in the domain time frequency using the normalized coefficient of variation for HF and the normalized coefficient of variation for LF and the LF/HF ratio for BP and heart rate. Baroreflex sensitivity was estimated using pulse interval power and systolic BP power in the LF band of the spectral analysis.

Statistical Analysis

Sample size estimates

To estimate the sample size for the study we used BMI and heart rate. A sample of 100 patients for each group permits to detect a significance of 5%, a difference of 0.5 kg/m2 between the BMI means in the two groups with a statistical power of 94% using Student t test for independent samples. A standard deviation of 1 kg/m2 for both groups was adopted.4

For the carbohydrate loading and spectral analysis of HRV, resting supine values for low and high frequency in healthy adult volunteers are in the range of 40 nu to 50 nu.21 Data on changes in spectral analysis with carbohydrate loading in humans are unavailable. A modest gravitational challenge, eg, head‐up tilt at 15 degrees induces changes in the 8 nu to 10 nu range.21 We decided to calculate power based on a 10 nu difference between groups with a pooled standard deviation of 10 nu. With these assumptions and 14 patients per group, then β=0.26 (74% power) to detect this level of difference with α≤0.05.21

Data analysis

The chi‐square test was used to compare the nonnumerical data. The values obtained in the anthropometric assessment from direct measurements and bioimpedance, biochemical variables, and hemodynamic data were compared using the paired and unpaired Student t tests for statistical evaluation of intragroup and intergroup differences, respectively. Pearson correlations and linear regression analysis were performed. For stepwise linear regression, BMR was the dependent variable and systolic BP, diastolic BP, triglycerides, abdominal circumference, glucose, HDL cholesterol, and uric acid were independent variables. For comparisons of pHTN∅ and pHTN⊕ subgroups regarding components of spectral analysis, multivariate analyses of variance with repeated measures were conducted. All tests were done with IBM‐SPSS Statistic version 20 (IBM Corporation, Armonk, NY). Data are expressed as mean±standard deviation. Statistical significance was set at P<.05.

Results

The pHTN⊕ and pHTN∅ groups did not differ with regard to height and neck circumference. However, weight, abdominal circumference, BMI, and hip circumference were significantly higher in the pHTN⊕ than in the pHTN∅ group (Table 1). Lean and fat mass evaluated by bioimpedance did not differ between groups. BMR was significantly higher in the pHTN⊕ than in the pHTN∅ group (1750±384 cal/d vs 1613±334 cal/d, respectively; P=.02). Casual systolic and diastolic BP were significantly higher in the pHTN⊕ than in the pHTN∅ group. Heart rate did not differ between the two groups (Figure 2).

Table 1.

Demographic and Anthropometric Data of pHTN∅ and pHTN⊕ Offspring

Variables	pHTN∅ Offspring (n=100)	pHTN⊕ Offspring (n=100)	P Value
Age, y	27.7±5.1	28.3±4.4	.27
Male/female	30/70	40/60	.59
Caucasian/non‐Caucasian	19/81	28/72	.55
Weight, kg	69.8±13.6	75.8±16.3	<.01
Height, m	165.6±9.5	167.7±9.9	.14
Body mass index, kg/m	25.5±4.3	26.8±4.6	.06
Abdominal circumference, cm	85.9±12.0	89.8±13.6	.04
Hip circumference, cm	100.8±10.1	105.5±10.2	<.01
Cervical circumference, cm	34.5±3.5	35.4±4.0	.14

Abbreviations: pHTN∅, offspring of normotensive parents; pHTN⊕, offspring of hypertensive parents. Data are expressed as mean±standard deviation.

Figure 2.

Arterial blood pressure and heart rate in the offspring of normotensive parents (pHTN∅) and the offspring of hypertensive parents (pHTN⊕) (mean±standard deviation). SBP indicates systolic blood pressure; DBP, diastolic blood pressure; PP, pulse pressure; HR, heart rate; bpm, beats per minute.

Systolic BP (122±14 mm Hg vs 117±12 mm Hg; P=.019) and pulse pressure were higher in the pHTN⊕ than in the pHTN∅ group (Figure 2). The remaining variables including diastolic BP, cardiac index, systemic vascular resistance, and large and small artery distensibility did not differ between the pHTN⊕ (68±10 mm Hg, 3.3±0.3 L/min/m², 1151±218 dynes. sec/cm⁵, 16.5± 4.6 mL/mm Hg ×100, 7.3±2.4 mL/mm Hg ×100, respectively) and pHTN∅ (66±9 mm Hg, 3.3±0.5 L/min/m², 1189±232 dynes. sec/cm⁵, 16.1±4.9 mL/mm Hg ×100, 7.0±2.6 mL/mm Hg ×100, respectively) groups.

Total, low‐density lipoprotein, and HDL cholesterol; triglycerides; blood glucose; sodium; potassium; uric acid; urea; creatinine; glycated hemoglobin; brain natriuretic peptide; tetraiodothyronine; thyroid‐stimulating hormone; and C‐reactive protein did not differ between the pHTN⊕ and pHTN∅ groups. In all patients, systolic BP correlated positively with abdominal circumference (r=0.292, P<.001), weight (r=0.417, P<.001), and BMI (r=0.243, P=.001). HDL cholesterol correlated negatively with both abdominal (r=–0.264, P<.001) and neck (r=−0.310, P<.001) circumferences. In stepwise linear regression analysis, significant independent correlations were identified between abdominal circumference, HDL cholesterol, and uric acid with BMR (Table 2).

Table 2.

Linear Regression Analysis Showing an Independent Correlation of Metabolic Syndrome Components and Uric Acid With BMR

Variables	β (95% CI)	P Value
Systolic BP, mm Hg	1.6 (−2.9–5.8)	.476
Diastolic BP, mm Hg	3.8 (−3.3–11.0)	.287
Abdominal circumference, cm	10.2 (6.1–14.4)	<.001
Glucose, mg/dL	1.6 (−1.9–6.1)	.429
Triglycerides, mg/dL	−0.7 (−1.7–0.17)	.177
HDL cholesterol, mg/dL	−6.8 (−11.3–−2.9)	.002
Uric acid, mg/dL	94.6 (61.4–129.6)	<.001

Abbreviations: BP, blood pressure; BMR, basal metabolic rate; CI, confidence interval; HDL, high‐density lipoprotein.

The pHTN⊕ and pHTN∅ groups selected for carbohydrate loading did not differ in age, sex, race, and BP. Heart rate did not differ between groups at baseline (Table 3). After carbohydrate load, heart rate increased four beats in the pHTN⊕ group (66±9 beats per minute [bpm] to 70±9 bpm, P<.001) and four beats in the pHTN∅ (64±9 bpm to 68±9 bpm, P<.001).

Table 3.

Demographic Data, Arterial Pressure, and Heart Rate in pHTN∅ and pHTN⊕ Offspring Selected for Carbohydrate Overload

Variables	pHTN∅ Offspring (n=15)	pHTN⊕ Offspring (n=14)	P Value
Age, y	33±1	32±4	.16
Male/female	5/10	7/7	.72
Caucasian/non‐Caucasian	4/11	2/12	.07
Systolic arterial pressure, mm Hg	116±15	112±11	.37
Diastolic arterial pressure, mm Hg	68±8	65±7	.27
Heart rate, beats per min	66±9	64±9	.57

Abbreviations: pHTN∅, offspring of normotensive parents; pHTN⊕, offspring of hypertensive parents.

The baseline LF component was similar in the pHTN⊕ and pHTN∅ groups (41±14 nu vs 36±16 nu, respectively; P=.439). Respective values for the HF (59±14 nu vs 64±16 nu, P=.440) and the LF/HF ratio (1.1±0.7 nu vs 0.9±0.7 nu, P=.539) also did not differ between the pHTN⊕ and pHTN∅ groups at baseline. After the carbohydrate load, the LF component was higher (45±15 nu vs 34±12 nu, P=.046) and the HF component was lower (55±15 nu vs 67±15 nu, P=.014) in the pHTN⊕ than in the pHTN∅ group, respectively. The LF/HF ratio in the pHTN∅ and pHTN⊕ groups, which did not differ at baseline (Figure 3), increased significantly in the pHTN⊕ group (P=.046) after the oral carbohydrate load (Figure 3). Baroreflex sensitivity did not differ between the pHTN⊕ and pHTN∅ groups (12±3 ms/mm Hg vs 11±3 ms/mm Hg, P=.257) at baseline or after the carbohydrate load (11±3 ms/mm Hg vs 11±3 ms/mm Hg, P=.982).

Figure 3.

Low‐frequency (LF)/high‐frequency (HF) component ratio in the offspring of normotensive parents (pHTN∅) and the offspring of hypertensive parents (pHTN⊕) before (a) and after (b) carbohydrate load (CHL) (mean±standard deviation).

Discussion

The primary purpose of our study was to assess the anthropometric, hemodynamic, and autonomic variables and their interrelationships in young adults with positive (pHTN⊕) and negative (pHTN∅) family histories for hypertension prior to the onset of clinical hypertension. Another key objective was to assess autonomic balance in a subset of positive (pHTN⊕) and negative (pHTN∅) participants at baseline and after a carbohydrate load. The two groups were well matched in the substudy as the only significant measured difference between groups was family history of hypertension.

No differences in autonomic control of HRV were detected at fasting baseline. However, after the carbohydrate load, the pHTN⊕ group had greater values for the LF component, smaller values for the HF component, and a higher LF/HF ratio than the pHTN∅ group. These data suggest cardiac sympathetic activation and parasympathetic inhibition in the pHTN⊕ group 40 minutes after carbohydrate loading.

Sympathovagal imbalance in patients with arterial hypertension and diabetes has been reported.22, 23 However, the sympathovagal imbalance resulting from carbohydrate loading in normotensive adult offspring of pHTN⊕ is a novel finding.

Nutritional regulation of sympathetic function has been studied. A diet rich in carbohydrates increased sympathetic activity in lean women, whereas a diet rich in fat did not change sympathetic activity in lean and obese women.17 Vollenweider and colleagues24 showed an increase in sympathetic activity following the infusion of insulin and glucose in normal volunteers. In a recent review, Klein and Kial25 discussed a possible association between oral fructose consumption and sympathetic activation, based mainly on animal models of fructose‐induced hypertension. Although many studies evaluated the effect of carbohydrates on the autonomic nervous system, we are not aware of any study in which autonomic responses to a mixed glucose load were evaluated in normotensive offspring of hypertensive parents.

The preponderance of the sympathetic component at the onset of hypertension is well established14 and is likely a pathogenetic factor in the initiation and maintenance of hypertension. In a study by Thiyagarjan and colleagues,26 young prehypertensive individuals had a greater LF component and lesser HF component than young normotensive individuals at baseline.

Our objective to assess anthropometric, hemodynamic, and autonomic variables and their interrelationships in young pHTN⊕ and pHTN∅ adults prior to the onset of clinical hypertension also generated important findings. The 100 pHTN⊕ patients had higher weight, BMI, abdominal circumference, BMR, and casual systolic and diastolic BP than the 100 pHTN∅ patients. Moreover, the pHTN⊕ group exhibited a larger abdominal circumference and higher BMI than the pHTN∅ group. We previously reported a higher BMI in normotensive offspring of malignant hypertensive parents than offspring of normotensive parents.4

The correlation of abdominal circumference and BMI with systolic BP did not differ between the pHTN⊕ and pHTN∅ groups. However, abdominal circumference, weight, and BMI were higher in the pHTN⊕ group. This finding suggests that greater values for key anthropometric measures in the pHTN⊕ group may contribute to systolic BP differences between groups. Of note, individuals in the pHTN⊕ and pHTN∅ groups were matched for age, sex, and race to minimize confounding for key demographic differences.

Pulse pressure was higher in the pHTN⊕ compared with the pHTN∅ group. The association of pulse pressure with BMI and central obesity has been previously reported.27, 28 However, differences with regard to family history of hypertension have not been described in the literature. The greater pulse pressure observed in the offspring of hypertensive parents may be a sign of arterial stiffness in this population, although this was not confirmed with measurements of arterial distensibility.

A positive correlation was found between neck circumference and systolic and diastolic arterial pressure, pulse pressure, urea, creatinine, uric acid, and triglycerides, as well as a negative correlation between neck circumference and HDL cholesterol. The correlation of neck circumference with metabolic variables was slightly stronger than the correlations with abdominal circumference. Moreover, there may be fewer barriers to measuring neck than abdominal circumference in clinical medicine. Neck circumference is less dependent on the examiner's technique and does not require undressing the patient.

The relationship of neck circumference to other anthropometric as well as metabolic variables has been reported. Neck circumference was a reliable marker of high BMI in children.29 An association was also reported between neck circumference and BMI with hypertension in children.8 However, in a large population study, neck circumference, adjusted for BMI and waist circumference, was not a predictor of prehypertensionen.9

Ben‐Noun and Laor reported a positive correlation between neck circumference and metabolic syndrome variables.30 In a recent study with repeated longitudinal measurements, neck circumference was associated with HDL cholesterol, glucose, triglycerides, and uric acid after stratification for BMI and abdominal circumference. These observations suggest that neck circumference is better correlated to metabolic syndrome components than BMI and abdominal circumference.31 However, it is important to note that neck circumference was not different between the pHTN⊕ and pHTN∅ groups in our study.

Uric acid was higher in the pHTN⊕ group than in the pHTN ∅ group in our study. Uric acid is often associated with metabolic changes and has been identified as a predictor of cardiovascular risk in young individuals.32 Uric acid has also been implicated in the pathogenesis of hypertension.33 In a meta‐analysis of 55,607 participants, uric acid was associated with an increased risk of incident hypertension.34 In a randomized, double‐blind, placebo‐controlled, crossover trial of 30 adolescents with stage 1 hypertension, allopurinol lowered BP in hyperuricemic patients.35

BMR correlated negatively with HDL cholesterol and positively with uric acid. In addition to the correlation between BMR and biochemical parameters, we observed higher BMR in the offspring of hypertensive parents. BMR is influenced by variables including body composition and sympathetic activity. The variables used to calculate BMR included sex, age, height, and weight. The equation used is accurate, according to previous findings.36 The slightly greater body weight in the pHTN⊕ group may have contributed to their higher BMR relative to the pHTN ∅ group.

Conclusions

A novel finding of our study was that offspring of hypertensive parents, relative to offspring of normotensive parents, exhibited evidence of sympathetic activation and vagal withdrawal after a carbohydrate load. Sympathovagal imbalance has been described in prehypertension, is implicated in the pathogenesis of hypertension, and is also associated with increased risk for cardiovascular events including sudden death. In addition, normotensive offspring of hypertensive parents exhibit higher casual BP, pulse pressure, abdominal circumference, weight, BMI, and higher BMR than matched offspring of normotensive parents. While both neck and abdominal circumferences were associated with several cardiometabolic syndrome—related abnormalities, associations were stronger for neck circumference. Neck circumference is easier and more reliably obtained in the clinical setting, which has implications for clinical practice given the worldwide epidemic of obesity and metabolic syndrome–related disease.

Financial disclosure

The authors report no specific funding in relation to this research and no conflicts of interest to disclose.

J Clin Hypertens (Greenwich). 2016;18:942–948. DOI: 10.1111/jch.12800. © 2016 Wiley Periodicals, Inc.