Abstract
The aims of this study were to assess associations of body fat levels and distribution with metabolic profiles and 24‐hour blood pressure in young adults with primary hypertension. Visceral fat (VF) was estimated using dual‐energy X‐ray absorptiometry. VF was highly significantly associated with a high frequency of overweight/obesity, impaired fasting glucose, increased levels of triglycerides and LDL‐cholesterol, and lowered level of HDL‐cholesterol. The value of systolic blood pressure (SBP) nocturnal fall was similar between patients receiving RAAS inhibitors, beta‐blockers, and calcium channel blockers. In multiple regression, the VF/weight ratio after adjusting for age, gender, total fat, and chronotherapeutic drug delivery was associated with the percentage SBP nocturnal fall (β = −.3108; 95% CI: −0.5923; −0.0980; P = .013). In males, excess VF increased the odds by 2.3 times for non‐dipping blood pressure. Our results suggest that in young adult hypertensives, the VF/weight ratio might be associated with non‐dipping blood pressure.
Keywords: body composition, cardiovascular risk factors, hypertension, non‐dipping blood pressure, Visceral fat
1. INTRODUCTION
Hypertension is associated with target organ damage, which can lead to premature loss of health, absence from work, and an increased risk for cardio‐ and cerebrovascular deaths. According to the Global Burden of Disease Study estimates, elevated blood pressure is the most robust single predictor for premature death, giving stronger prediction than being overweight, obesity, smoking, alcohol use, or having unhealthy diet.1, 2 The prevalence of hypertension increases with age, reflecting the age‐dependent changes in systolic blood pressure (SBP) and pulse pressure that lead to arterial stiffness.3 However, the global rates of hypertension among young adults also have been dramatically increasing but its causes have not yet been fully elucidated. Aside from distinct features in the pathophysiology of hypertension in youth, there is growing evidence that adiposity, along with environmental factors related to present‐day life‐styles, can play a role in the development of elevated blood pressure,4 just as in older populations. If this is the case, it means that not only excess total fat, but also its unfavorable visceral distribution could provide key predictors for hypertension in young individuals. It has been shown that individuals with excess visceral fat (VF) are prone to develop abnormal lipid profiles, abnormal glucose tolerance, elevated blood pressure, and other abnormalities frequently associated with metabolic syndrome.5, 6 Some studies have suggested that high body mass index could be associated with non‐dipping hypertension in children7 and in adults.8 We hypothesized that non‐dipping blood pressure could be associated not only with adiposity but also with excess VF. There have been no previous reports on the potential impact of VF on a fall in nocturnal blood pressure.
The aims of this study were therefore to assess associations of body fat levels and distribution with metabolic profiles and 24‐hour blood pressure in young adults with primary hypertension. We were particularly interested in possible deterioration of diurnal blood profiles, including physiological nocturnal blood pressure fall, associated with excess VF.
2. MATERIAL AND METHODS
2.1. Study participants
The study group comprised 206 young adult patients (132 males; 74 females) with hypertension who were consecutively hospitalized between 1st of November 2016 to the end of October 2018 in our tertiary care unit (Department of Hypertension & Internal Diseases, Pomeranian Medical University in Szczecin) to rule out secondary causes of hypertension as well as for the assessment of target organ damage. We included patients aged 18‐35 years with body mass index (BMI) 18.5‐35.0 kg/m2, in whom the diagnosis of primary hypertension was finally established. The exclusion criteria were as follows: (a) thyrotoxicosis, type 1 diabetes, and other endocrine disorders; (b) history of malignancy within proceeding 5 years; (c) pregnancy or lactation; (d) polycystic ovary syndrome (PCOS) or symptoms suggesting PCOS (oligomenorrhea/amenorrhea and clinical or biochemical hyperandrogenism; (e) use of medications or dietary supplements known to affect body composition (glucocorticoids, insulin, anabolic steroids, protein supplements, etc); (f) nutrition disorders, including anorexia and bulimia; (g) elevated blood pressure reading without diagnosis of hypertension (R03.0 code in the ICD‐10 classification); and (h) rapid weight changes (above 5 kg) within the last 12 months. None of the patients was diagnosed with spurious isolated systolic hypertension.
All subjects received standard antihypertensive treatments8, 9 with renin‐angiotensin‐aldosterone system (RAAS) inhibitors: angiotensin‐converting enzyme or AT1 receptor inhibitors (in males) and β‐blockers, calcium channel blockers (CCBs), diuretics, and α1‐blockers (in both females and males). No patients received treatments from other drug classes. Among males, 80 received RAAS inhibitors as a monotherapy. RAAS inhibitors were administered also in combination with CCBs (n = 12), β‐blockers (n = 7), diuretics (n = 2), and α1‐blockers (n = 3). β‐blockers were administered in 21 of males: in monotherapy (n = 16), in combination with CCBs (n = 2), and in combination with CCBs and α1‐blockers (n = 3). CCBs were given in monotherapy (n = 19) and in combination with diuretics or α1‐blockers (n = 12). Among females, 60 patients were treated with CCBs (in 42 of patients given as a monotherapy, eight in combination with diuretics, and 10 in a triple combination with β‐blockers, α1‐blockers, or diuretics). Fourteen females received β‐blockers (10 in monotherapy and 4 in combination with α1‐blockers or diuretics). Overall, 49 patients were treated with RAAS inhibitors, 26 with β‐blockers, and 61 with CCBs given in monotherapy. Seventy (34%) of the patients were treated with medications given twice daily (in the morning and at bedtime), including RAAS inhibitors (in males) and α1‐blockers or CCBs (in the both genders) given at bedtime.
The study complied with all applicable institutional and governmental regulations regarding to the ethical use in human volunteers and in the terms of Declaration of Helsinki. The Pomeranian Medical University Ethics Committee approved the study protocol, and all the recruited volunteers gave their written consent.
2.2. Procedures
We measured height, weight, and waist and hip circumferences. BMI was calculated as weight (kg) divided by height (m) squared. Twenty‐four‐hour blood pressure monitoring (ABPM) was recorded using the Spacelabs device (model 90207; Spacelabs Healthcare). Automated blood pressure measurements were performed every 20 minutes during the day and every 30 minutes during nighttime. The night period in ABPM was set by default between 22.00 and 06.00. Dipping (nocturnal SBP fall by 10%‐20%) and non‐dipping including extreme dipping (a fall of >20%) and reverse dipping (<10%) diurnal profiles were defined as a percent difference between daytime and nocturnal mean values of systolic blood pressure (SBP).9, 10
In all study participants, lipid profiles including serum levels of triglycerides (TG) and total, low‐density lipoprotein (LDL)‐, and high‐density lipoprotein (HDL)‐cholesterol, glucose, insulin (taken after an overnight fast), and 120 minutes glucose and insulin levels during a 75 g oral glucose tolerance test (OGTT) were measured. From fasting insulin and glucose measurements, we calculated a homeostatic model assessment‐insulin resistance index (HOMA‐IR). From,11, 12 we used the value of HOMA‐IR of ≥2.5 as a marker for insulin resistance.
To evaluate the presence of metabolic risk factors, we used the International Diabetes Federation (IDF) gender‐specific diagnostic criteria for metabolic syndrome in populations of European descent: (a) waist circumference ≥94 cm in men and ≥80 cm in women; (b) high TG concentration ≥150 mg/dL (≥1.7 mmol/L); (c) low HDL‐cholesterol level <50 mg/dL (<1.29 mmol/L) in women and <40 mg/dL (<1.03 mmol/L) in men; and (d) raised fasting glucose ≥100 mg/dL (≥5.6 mmol/L).13 For the definition of raised total and LDL‐cholesterol levels, we used the following cutoffs: >200 mg/dL (>5.17 mmol/L) and ≥115 mg/dL (≥2.97 mmol/L), respectively. From waist circumference and TG level, the lipid accumulation product (LAP) was calculated using the following, gender‐specific equations: LAP (cm × mmol/L) = (waist circumference−65) × TG level in males; and LAP (cm × mmol/L) = (waist circumference−58) × TG level in females.14
Body composition was assessed using dual‐energy X‐ray absorptiometry (DXA) (GE Healthcare Lunar Prodigy Advance; enCORE software version 14.10) using the automatic whole‐body scan mode. All scans were performed and analyzed by a single operator using the Advanced Body Composition Assessment tool of the software system according to the standard protocol given by the manufacturer. In the whole‐body scan, the following parameters were analyzed: (a) total body fat (TBF) expressed in mass units (kg) or as percentage mass (TBF%); (b) abdominal fat (Android) measured in the android region of interest (ROI), which comprises VF and android subcutaneous fat; (c) gynoid fat (Gynoid); and (d) lean mass (LM), which in DXA is a surrogate measure of muscle mass. From height and LM, we calculated a lean mass index (LMI) using the following formula: LMI (kg/m2) = LM (kg)/height (m2).
Visceral fat mass and volume were computed in the android ROI by the instrument‐specific application (CoreScan) dedicated to GE Healthcare DXA devices. CoreScan computes VF by subtracting android subcutaneous fat from android total fat. VF computed by this method strongly correlates with VF in computed tomography (R = 0.96) and magnetic resonance imaging (R = 0.82‐0.86).15, 16, 17 From VF measurements, we calculated the ratios VF/TBF, VF/weight, VF/LM, and VF/Android fat. Body composition parameters were analyzed using age‐, gender‐, race‐, and instrument‐specific reference values for VF,5, 18 TBF,19 and LMI.20
2.3. Statistical analyses
Data were checked for normality using Shapiro‐Wilk's test. Descriptive statistics included means ± standard deviation (SD) for continuous variables and frequency distributions for categorical variables. Variables with normal distribution were compared using parametric Student's t tests; otherwise, non‐parametric Mann‐Whitney U‐tests and Kruskal‐Wallis tests were used. Differences in means of nocturnal blood pressure fall between the patients receiving in monotherapy antihypertensive drugs from main three drug classes were compared using ANOVA. To determine the relationship between qualitative variables, Fisher's exact tests were used. Correlations between pairs of quantitative variables were assessed using Pearson's linear correlations or Spearman's rho correlations for normally and non‐normally distributed variables, respectively. A receiver operating characteristic (ROC) curve analysis was used to evaluate the diagnostic ability and discrimination thresholds of VF and TBF in predicting metabolic abnormalities. The Youden index, which is a commonly used measure of overall diagnostic effectiveness, was calculated as sensitivity (c) + specificity (c) − 1. Logistic regression was used to calculate odds ratios (ORs) for abnormal fasting glucose, HOMA‐IR, blood lipids, or an absence of a fall in nocturnal SBP associated with excess VF. Predictors of nocturnal blood pressure fall were assessed using stepwise regression models. Using post hoc analysis, the statistical power of the study with 206 subjects was sufficient to detect with 80% probability the true effect size of association between quantitative variables corresponding to correlation coefficient value of 0.20. Statistical analyses were performed using Statistica (StatSoft, Poland; version 13) and the R Statistical Platform (version 3.5.1; https://cran.r-project.org).
3. RESULTS
Baseline characteristics of the study population are shown in Table 1. Overall, nearly 55% of subjects had BMI >25.0 kg/m2 but the frequencies of overweight patients and obesity were significantly higher in men. In the whole group, BMI was positively correlated with age (r = .223; P = 00009). In comparison with females, males had significantly higher total and LDL‐cholesterol, TG, and LAP but lower HDL‐cholesterol. They also had a higher frequency of abnormal lipid profiles (elevated concentrations of TG, and total and LDL‐cholesterol). Interestingly, males had a lower mean diastolic blood pressure (DBP) but a more pronounced nocturnal fall in SBP. Using fasting glucose and OGTT measurements, we detected one new case with type 2 diabetes, two with impaired glucose tolerance (IGT), and 22 (11%) with impaired fasting glucose (IFG). As summarized in Table 2, body composition assessments revealed gender differences in LM as well as in TBF and its distribution. Generally, males had higher LM, Android fat, and VF (not only as mass or volume estimates but also as expressed in proportions to TBF, mass, or LM). The most common abnormalities found from body composition analyses were VF, TBF%, and a VF/LM ratio above the reference value (Table 3). These affected approximately 50% of the study population with a comparable frequency in both genders. About 15% of patients had a deficit in muscle mass defined as LMI <13.6 kg/m2 in females and <16.8 kg/m2 in males.
Table 1.
Descriptive statistics of the study population
| All (n = 206) | Women (n = 74) | Men (n = 132) | P | |
|---|---|---|---|---|
| Age (y) | 27.40 ± 5.1 | 27.55 ± 4.9 | 26.75 ± 5.1 | .0509 |
| Weight (kg) | 83.24 ± 17.9 | 71.27 ± 17.1 | 89.95 ± 14.6 | <.0001 |
| Height (m) | 1.744 ± 0.1 | 1.651 ± 0.1 | 1.797 ± 0.1 | <.0001 |
| BMI (kg/m2) | 27.20 ± 4.9 | 26.05 ± 5.7 | 27.84 ± 4.3 | .0022 |
| <24.9 kg/m2 (n) | 93 (45.14%) | 38 (51.35%) | 39 (29.54%) | .0019 |
| 25.0‐29.9 kg/m2 (n) | 77 (37.37%) | 17 (22.97%) | 60 (45.45%) | .0022 |
| >30.0 kg/m2 (n) | 36 (17.47%) | 19 (25.67%) | 33 (25.0%) | .9522 |
| Waist circumference (cm) | 93.99 ± 14.4 | 87.54 ± 14.4 | 97.61 ± 13.1 | <.0001 |
| ≥80 cm (F); ≥94 cm (M) (n) | 118 (57.3%) | 45 (60.8%) | 73 (55.3) | .5351 |
| Hip circumference (cm) | 100.0 ± 11.1 | 101.9 ± 9.0 | 96.62 ± 13.5 | .0051 |
| Waist‐to‐hip ratio | 0.938 ± 0.1 | 0.906 ± 0.1 | 0.955 ± 0.1 | <.0001 |
| Family history of hypertension (n) | 97 (47.08%) | 37 (50.0%) | 60 (45.45%) | .6301 |
| 24‐h Systolic blood pressure (mm Hg) | 131.2 ± 14.7 | 128.72 ± 14.5 | 132.59 ± 14.7 | .0519 |
| Daytime | 135.82 ± 14.9 | 133.91 ± 14.7 | 136.83 ± 14.9 | .0601 |
| Nighttime | 122.71 ± 13.7 | 121.73 ± 13.6 | 123.42 ± 13.7 | .0945 |
| 24‐h Diastolic blood pressure (mm Hg) | 77.31 ± 11.4 | 81.43 ± 12.6 | 75.00 ± 10.0 | .0014 |
| Daytime | 83.24 ± 12.1 | 85.90 ± 12.9 | 80.13 ± 11.9 | .0142 |
| Nighttime | 74.01 ± 11.3 | 77.71 ± 11.4 | 72.06 ± 11.2 | .0356 |
| SBP nocturnal fall (%) | −10.28 ± 4.8 | −9.400 ± 4.9 | −10.80 ± 4.7 | .0182 |
| Absence of SBP nocturnal fall (n) | 83 (40.29%) | 37 (50.0%) | 46 (34.84%) | .0478 |
| Fasting insulin (μIU/mL) | 12.59 ± 12.7 | 11.89 ± 6.6 | 12.96 ± 15.0 | .5121 |
| Insulin 120’ OGTT (μIU/mL) | 46.58 ± 55.3 | 45.40 ± 38.8 | 47.62 ± 67.0 | .1210 |
| Fasting glucose (mg/dL) | 88.96 ± 9.7 | 87.32 ± 9.3 | 89.85 ± 9.9 | .1180 |
| Glucose 120’ OGTT (mg/dL) | 94.86 ± 25.5 | 98.79 ± 29.7 | 98.11 ± 20.8 | .3491 |
| HOMA‐IR | 2.853 ± 3.7 | 2.598 ± 1.6 | 2.991 ± 4.5 | .7751 |
| >2.5 (n) | 78 (37.9%) | 30 (40.5%) | 48 (36.4%) | .5537 |
| Total cholesterol (mg/dL) | 181.71 ± 36.7 | 172.69 ± 31.1 | 186.55 ± 38.6 | .0283 |
| >200 mg/dL (n) | 51 (24.8%) | 8 (10.8%) | 43 (32.6%) | .0004 |
| HDL‐cholesterol (mg/dL) | 51.86 ± 14.5 | 59.38 ± 14.7 | 47.81 ± 12.8 | <.0001 |
| <50 mg/dL (F); <40 mg/dL (M) (n) | 58 (28.2%) | 19 (25.7%) | 39 (29.5%) | .6291 |
| LDL‐cholesterol (mg/dL) | 116.39 ± 33.9 | 105.67 ± 31.2 | 122.16 ± 33.9 | .0024 |
| ≥115 mg/dL (n) | 90 (43.7%) | 16 (21.6%) | 74 (56.1%) | .0001 |
| Triglycerides (mg/dL) | 136.51 ± 71.3 | 115.23 ± 58.1 | 147.96 ± 75.2 | .0090 |
| >150 mg/dL (n) | 63 (30.6%) | 12 (16.2%) | 51 (38.6%) | .0009 |
| LAP (cm × mmol/L) | 54.59 ± 45.6 | 43.35 ± 36.0 | 60.46 ± 49.0 | .0488 |
| Type 2 diabetes (n) | 1 (0.48%) | 1 (1.4%) | 0 (0%) | .3590 |
| Impaired glucose tolerance (n) | 2 (0.97%) | 1 (1.4%) | 1 (0.75%) | 1.0 |
| Impaired fasting glucose (n) | 22 (10.7%) | 4 (5.4%) | 18 (13.6%) | .0980 |
P‐value refers to comparison between women and men.
Table 2.
Body composition analysis
| All (n = 206) | Women (n = 74) | Men (n = 132) | P | |
|---|---|---|---|---|
| Total fat (kg) | 26.539 ± 10.6 | 26.729 ± 11.5 | 26.433 ± 10.0 | .7750 |
| Total fat (%) | 32.422 ± 8.2 | 37.140 ± 7.6 | 29.777 ± 7.3 | <.0001 |
| Android fat (kg) | 2.446 ± 1.4 | 2.222 ± 1.4 | 2.571 ± 1.3 | .0412 |
| Gynoid fat (kg) | 4.406 ± 2.2 | 4.948 ± 2.2 | 4.101 ± 1.4 | .0449 |
| VF (kg) | 0.920 ± 0.7 | 0.580 ± 0.5 | 1.110 ± 0.7 | <.0001 |
| Visceral fat (cm3) | 980.65 ± 753.8 | 627.10 ± 594.0 | 1178.85 ± 763.2 | <.0001 |
| Visceral fat/Android fat ratio (%) | 33.817 ± 16.0 | 21.363 ± 11.7 | 40.799 ± 13.7 | <.0001 |
| Visceral fat/weight ratio (%) | 1.507 ± 0.6 | 1.712 ± 0.5 | 1.172 ± 0.6 | <.0001 |
| Visceral fat/total fat ratio (%) | 3.123 ± 1.8 | 1.793 ± 1.1 | 3.869 ± 1.7 | <.0001 |
| Lean mass (kg) | 53.733 ± 10.8 | 42.631 ± 6.4 | 59.957 ± 7.2 | <.0001 |
| Lean mass index (kg/m2) | 17.488 ± 2.3 | 15.603 ± 1.9 | 18.544 ± 1.8 | <.0001 |
Table 3.
Frequencies of abnormal values in body composition parameters
| All (n = 206) | Women (n = 74) | Men (n = 132) | P | |
|---|---|---|---|---|
| Low lean mass indexa | 31 (15.0%) | 11 (14.9%) | 20 (15.2%) | 1.0 |
| High visceral fat/lean mass ratiob | 102 (49.5%) | 35 (47.3%) | 67 (50.8%) | .6649 |
| High visceral/total fat ratioc | 94 (45.6%) | 35 (47.3%) | 59 (45.4%) | .7714 |
| High visceral fat/weight ratiod | 100 (48.5%) | 35 (47.3%) | 65 (49.2%) | .8846 |
| High visceral fate | 103 (50.0%) | 35 (47.3%) | 68 (51.5%) | .6612 |
| High total fatf | 98 (47.6%) | 36 (48.6%) | 62 (47.0%) | .8847 |
< 13.6 kg/m2 (women), <16.8 kg/m2 (men).
> 1.2% (women), >1.67% (men).
> 1.96% (women), >4.03% (men).
> 0.71% (women), >1.12% (men).
> 484 g (women), >993 g (men).
> 37.7% (women), >30.6% (men).
As expected, both TBF and the fat depots significantly correlated with insulin, HOMA‐IR, blood lipids, and, except for Gynoid fat, also with age (Table 4). However, VF, both expressed in mass units and as mass percentages, showed the strongest correlations with TG, HDL, LDL, and HOMA‐IR. In contrast to other fat compartments, VF showed a weak but significant correlation with fasting glucose. In addition, when we compared subjects with normal and high VF (Table 5), those with high VF had significantly higher waist and hip circumferences, LM, insulin (both fasting and from OGTT), fasting glucose, HOMA‐IR, total and LDL‐cholesterol, and TG as well as lower HDL‐cholesterol. Rather unexpectedly, there were also significant correlations between LM and blood lipids with HOMA‐IR, although these associations were weaker.
Table 4.
Correlations between body composition parameters, metabolic profiles, and blood pressure
| TBF (kg) | TBF % | VF (kg) | VF/weight (%) | Android fat (kg) | Gynoid fat (kg) | Lean mass (kg) | |
|---|---|---|---|---|---|---|---|
| Insulin (μIU/mL) | 0.4426 d | 0.3482 d | 0.4613 d | 0.3233 d | 0.4613 d | 0.3628 d | 0.4066 d |
| Glucose (mg/dL) | 0.2070 | 0.1112 | 0.2284 a | 0.2304 b | 0.2274 | 0.0908 | 0.1735 |
| HOMA‐IR | 0.3872 d | 0.3009 d | 0.4148 d | 0.5257 d | 0.4148 d | 0.3088 d | 0.3540 d |
| HDL (mg/dL) | −0.3663 d | −0.2102 | −0.4234 d | −0.4736 c | −0.4234 d | −0.2141 c | −0.3009 d |
| LDL (mg/dL) | 0.3147 d | 0.1828 | 0.3622 d | 0.3475 b | 0.3622 d | 0.1317 c | 0.1374 a |
| TG (mg/dL) | 0.4243 d | 0.2692 c | 0.4915 d | 0.4545 d | 0.4915 d | 0.2068 c | 0.3093 d |
| LAP (cm × mmol/L) | 0.7046 d | 0.4897 d | 0.7588 d | 0.7625 d | 0.7588 d | 0.4137 d | 0.5211 d |
| 24‐h SBP (mm Hg) | −0.1063 | −0.1632 | −0.0784 | −0.1090 | −0.0784 | −0.1411 | 0.0705 |
| 24‐h DBP (mm Hg) | −0.0646 | 0.0746 | −0.0771 | −0.1450 | −0.0771 | −0.0566 | −0.0523 |
| Nocturnal SBP fall (%) | 0.0001 | −0.0249 | 0.0124 | −0.0973 | 0.0124 | −0.0725 | −0.0518 |
| Age (y) | 0.1853 a | 0.2490 b | 0.2114 b | 0.1838 a | 0.2114 b | 0.0111 | 0.1669 a |
P < .05.
P < .01.
P < .001.
P < .0001.
Table 5.
Comparisons between subjects with normal and high visceral fat mass
| Women | P | Men | P | |||
|---|---|---|---|---|---|---|
| VF < 0.484 kg | VF ≥ 0.484 kg | VF < 0.993 | VF ≥ 0.993 kg | |||
| Weight (kg) | 59.05 ± 7.6 | 84.89 ± 14.2 | <.0001 | 80.27 ± 9.1 | 99.34 ± 12.7 | <.0001 |
| Waist circumference (cm) | 77.05 ± 6.1 | 99.23 ± 11.7 | <.0001 | 88.22 ± 7.5 | 106.7 ± 10.8 | <.0001 |
| Hip circumference (cm) | 87.12 ± 7.4 | 107.2 ± 10.7 | <.0001 | 96.14 ± 6.2 | 107.56 ± 7.6 | <.0001 |
| Lean mass (kg) | 39.32 ± 22.6 | 46.31 ± 26.5 | <.0001 | 57.85 ± 23.2 | 61.93 ± 25.2 | <.0001 |
| Fasting insulin (μIU/ml) | 8.901 ± 4.9 | 15.26 ± 6.6 | <.0001 | 8.421 ± 4.0 | 17.45 ± 19.8 | <.0001 |
| Insulin 120’ OGTT (μIU/mL) | 28.78 ± 18.1 | 57.71 ± 45.3 | .0048 | 25.13 ± 17.4 | 67.71 ± 26.5 | .0425 |
| Fasting glucose (mg/dL) | 86.13 ± 8.4 | 88.66 ± 10.2 | .2602 | 88.54 ± 9.0 | 91.14 ± 10.6 | .1307 |
| Glucose 120’ OGTT (mg/dL) | 98.79 ± 12.4 | 101.0 ± 13.4 | .1980 | 86.12 ± 9.5 | 96.97 ± 14.1 | .0520 |
| HOMA‐IR | 1.940 ± 1.2 | 3.340 ± 1.7 | .0003 | 1.796 ± 0.2 | 4.131 ± 1.9 | <.0001 |
| Total cholesterol (mg/dL) | 166.4 ± 27.7 | 179.1 ± 33.5 | .0857 | 175.4 ± 31.2 | 197.3 ± 42.1 | .0009 |
| HDL‐cholesterol (mg/dL) | 65.67 ± 12.4 | 52.92 ± 14.2 | .0002 | 51.89 ± 12.4 | 43.86 ± 11.9 | .0002 |
| LDL‐cholesterol (mg/dL) | 93.57 ± 26.4 | 118.1 ± 31.3 | .0007 | 111.6 ± 27.9 | 132.3 ± 36.3 | .0004 |
| Triglycerides (mg/dL) | 90.98 ± 32.9 | 140.1 ± 67.7 | .0003 | 181.1 ± 79.0 | 113.8 ± 53.1 | <.0001 |
| LAP (cm × mmol/L) | 20.86 ± 13.3 | 67.90 ± 37.0 | <.0001 | 30.80 ± 19.1 | 89.25 ± 52.0 | <.0001 |
| SBP (mm Hg) | 130.6 ± 17.4 | 126.5 ± 10.2 | .2170 | 134.8 ± 12.1 | 130.5 ± 16.5 | .0521 |
| DBP (mm Hg) | 81.26 ± 14.5 | 78.21 ± 9.3 | .0557 | 75.05 ± 9.9 | 74.96 ± 10.2 | .9584 |
| Nocturnal SBP fall (%) | −9.888 ± 1.0 | −8.869 ± 0.9 | .3864 | −10.14 ± 1.2 | −11.34 ± 1.2 | .246 |
As shown in Table 6, in both females and males excess VF substantially increased the odds ratios (ORs) for fasting glucose ≥100 mg/dL, HOMA‐IR ≥2.5 and abnormal lipid profiles, and especially: TG ≥ 150 mg/dL (OR = 6.7 in females and OR = 6.0 in males). Interestingly, excess VF was associated with a lack of fall in nocturnal SBP, but only in males (OR = 2.26; P = .033). Similarly, in ROC analysis (Table 7), VF was a robust predictor for high TG and LDL‐cholesterol levels, high HOMA‐IR values, and low HDL‐cholesterol levels. The VF cutoff values above which the risk of these abnormalities increased were 0.79‐0.88 kg in females and 1.3‐1.9 kg in males.
Table 6.
Crude odds ratios for metabolic abnormalities and non‐dipping blood pressure profiles associated with high visceral fat mass
| VF ≥ 0.484 kg (women) | VF ≥ 0.993 kg (men) | |||||
|---|---|---|---|---|---|---|
| OR | 95% CI | P | OR | 95% CI | P | |
| Fasting glucosea | 1.734 | 0.27; 11.03 | .5598 | 0.814 | 0.29; 2.261 | .797 |
| HOMA‐IRb | 5.812 | 2.07; 16.28 | .0008 | 5.208 | 2.33; 11.63 | <.0001 |
| Total cholesterolc | 4.153 | 1.02; 16.85 | .0463 | 3.286 | 1.51; 7.122 | .0029 |
| HDL‐cholesterold, e | 4.631 | 1.54; 13.84 | .0061 | 2.682 | 1.21; 5.942 | .0203 |
| LDL‐cholesterolf | 5.666 | 1.64; 19.52 | .0060 | 3.620 | 1.73; 7.561 | .0007 |
| Triglyceridesg | 6.652 | 1.97; 47.08 | .0050 | 5.967 | 2.67; 13.33 | <.0001 |
| SBP nocturnal fallh | 1.065 | 0.41; 2.707 | .8934 | 2.266 | 1.11; 4.606 | .0331 |
Table 7.
Visceral fat predictive values for metabolic abnormalities (ROC analysis)
| Visceral fat (kg) | AUC | 95% CI | P | Cutoff | Sensitivity | Specificity | Youden index |
|---|---|---|---|---|---|---|---|
| LDL‐Cholesterol ≥ 115 mg/dL | |||||||
| Women | 0.692 | 0.612; 0.872 | .0001 | 0.776 | 0.859 | 0.485 | 0.344 |
| Men | 0.630 | 0.387; 0.825 | .0001 | 1.320 | 0.735 | 0.364 | 0.312 |
| HDL‐Cholesterol (mg/dL) | |||||||
| Women < 50 | 0.715 | 0.621; 0.910 | .0001 | 0.859 | 0.897 | 0.462 | 0.359 |
| Men < 40 | 0.837 | 0.627; 0.947 | .0001 | 1.560 | 0.578 | 0.980 | 0.559 |
| Fasting glucose ≥ 100 mg/dL | |||||||
| Women | 0.608 | 0.408; 0.915 | .1352 | 1.125 | 0.333 | 0.888 | 0.222 |
| Men | 0.567 | 0.262; 0.908 | .0510 | 1.430 | 0.795 | 0.379 | 0.236 |
| Triglycerides > 150 mg/dL | |||||||
| Women | 0.801 | 0.746; 0.908 | .0001 | 0.882 | 0.615 | 0.869 | 0.485 |
| Men | 0.824 | 0.752; 0.902 | .0001 | 1.352 | 0.773 | 0.757 | 0.575 |
| HOMA‐IR ≥ 2.5 | |||||||
| Women | 0.708 | 0.632; 0.991 | .0001 | 0.890 | 0.649 | 0.732 | 0.382 |
| Men | 0.731 | 0.640; 0.912 | .0001 | 1.612 | 0.592 | 0.790 | 0.320 |
The magnitude of nocturnal blood pressure fall was comparable between patients receiving medications from three main drug classes in monotherapy (−10.24 ± 4.11%, −10.32 ± 4.47%, and −10.21 ± 4.36% for RAAS inhibitors, CCBs, and β‐blockers, respectively). As expected, patients treated with two doses of antihypertensive drugs (ie, given in the morning and at bedtime) had a higher fall in nocturnal SBP in comparison with those treated with a single morning dose (−11.55 ± 4.01% vs −10.18 ± 3.88%, respectively; P = .039). In multiple regression, the VF/weight and Android fat/Gynoid fat ratios adjusted for age, gender, and TBF were associated with the percentage SBP nocturnal fall (β = −0.3747; 95% CI: −0.7169 to −0.0422; P = .0152 and β = −0.2840; 95% CI: −0.5041 to −0.0781; P = .0449, respectively). However, after the further adjustment for antihypertensive treatments given at bedtime, the association of a fall in nocturnal SBP with the Android fat/Gynoid fat ratio disappeared, but still remained significant with the VF/weight ratio (β = −0.3108; 95% CI: −0.5923 to −0.0980; P = .013).
4. DISCUSSION
This study was carried out on a carefully selected group of young adult patients with primary hypertension who, per study protocol, were consecutively referred to our tertiary care unit. The study group comprised many more males than females (132 to 74) suggesting gender disparities in the prevalence of hypertension among subjects aged 18‐35 years. Indeed, the Polish national epidemiological survey WOBASZ II 21 and WHO estimates for Poland 22 have shown similar differences in the rates of hypertension in young Polish adults.
Body composition was assessed using DXA along with the new CoreScan application, which provides a quantification of VF. In previous studies, this method has mainly been used to determine the precision of the instrument and for validation against CT, MRI, and bioelectrical impedance devices.15, 16, 23 It has also been used in a few clinical studies to evaluate VF‐dependent cardiometabolic risks in different populations.19, 24 Here, we used this method for the first time in a population of young adults with hypertension. Using our pre‐defined age‐ and gender‐specific VF reference values derived from the healthy population,5, 18 we found excess VF mass in 52% of males and 47% of females. DXA‐VF was a relatively small fat depot, which constituted only 1.5% of patient weight and 3.1% of TBF. However, it was highly significantly correlated with fasting insulin, glucose, total cholesterol, HDL, LDL, and TG levels. In both genders, VF was a strong predictor of insulin resistance and unfavorable lipid profiles, especially high TG concentrations. Based on ROC analysis, these abnormalities were predicted by a VF mass above approximately 0.776‐0.890 kg in females and 1.320‐1.612 kg in males.
In our sample, the prevalence of non‐dipping hypertension was 50%. Similar or even higher rates were found among young white participants in the CARDIA Study (32%), Jackson Heart Study (73%), and African‐PREDICT Study (34%).25, 26 A novel finding of our study is that VF, known to be strongly associated with a range of metabolic abnormalities even in healthy individuals,5, 18, 24 is also suggested here to have had an impact on the magnitude of fall in nocturnal blood pressure. VF (expressed as mass percentage), after adjusting for confounders, was found associated with an absence of physiological nocturnal fall in SBP. In addition, in young males (but not in females) excess VF increased the odds by 2.3 times for non‐dipping blood pressure. These findings may have potential clinical implications. Our results suggest that young adult hypertensives with abdominal obesity, even without knowing the exact VF value, should be screened early with ABPM for non‐dipping blood pressure. Early identification of such cases could be helpful in the selection of more tailored treatments, for instance based on chronotherapy. In addition, it has previously been shown that VF was strongly associated with the severity of obstructive sleep apnea (OSA),27 which, in turn, can predispose to non‐dipping hypertension.28
Our study has some limitations. Firstly, ABPM was performed in a hospital setting. Because diurnal blood pressure is regulated mainly by the autonomic nervous system, hospitalization‐related factors such as stress (associated with hospitalization per se, or the companionship of seriously sick patients), diet, different times of meals, lack of privacy, or lower physical activity could influence SBP, DBP, and nocturnal blood pressure fall. In addition, the night period in ABPM was set by default, which does not always correspond to the real period of sleep. Secondly, we did not analyze the influence of antihypertensive treatments on metabolic parameters. Although routine antihypertensive medications seem not to influence DXA‐derived body composition, some of them may alter insulin sensitivity, glucose tolerance, and lipid profiles (eg, with thiazide diuretics) or sympathetic activity (with β‐blockers and α1‐blockers). Thirdly, we defined a non‐dipping blood pressure profile based solely on SBP as in some previous reports.29, 30, 31 However, other studies have used both the SBP and DBP day‐to‐night differences for defining non‐dippers.32 On the other hand, Birkenhäger et al33 have suggested that the mean arterial pressure (MAP) would be the most accurate parameter because MAP is a primary measure in the oscillometric method used in ABPM, while SBP and DBP are computed by the software. Using MAP, the systematic error as well as the impact of individual variability in pulse pressure and heart rate (in atrial fibrillation and other arrhythmias) could be minimalized. Fourthly, adipose tissue can be accumulated in depots other than VF ectopic fat, for example, as perirenal, perivascular, pericardial and intramuscular fat, and/or fat infiltrating the liver or pancreas. These conceivably might produce both local and systemic metabolic effects similar to VF.34, 35, 36 DXA does not differentiate these other regions of ectopic fat and includes them with TBF, subcutaneous fat, or VF. This issue may be relevant in the quantification of VF in patients for example with non‐alcoholic (or alcoholic) fatty liver disease, pancreas steatosis or with an excess of periaortal fat (which in the DXA‐CoreScan is automatically included with VF). Finally, some studies demonstrated modest reproducibility of nocturnal blood pressure fall in ABPM.37, 38 Therefore, conclusions drawn from the analysis of single ABPM measurement may not correspond precisely with true dipping/non‐dipping patterns.
In conclusion, our results have confirmed the close association of VF with the occurrence of cardiometabolic risk factors, as reported in previous studies. In addition, our results suggest that in young adult hypertensives, VF might be associated with non‐dipping blood pressure.
CONFLICT OF INTEREST
None.
AUTHOR CONTRIBUTIONS
TM and AT involved in study research and design; AT and BM involved in collection and assembling of data; TM, AT, and BM analyzed and interpreted the data; TM drafted the manuscript; and AT and BM involved in critical revision of the article.
Miazgowski T, Taszarek A, Miazgowski B. Visceral fat, cardiometabolic risk factors, and nocturnal blood pressure fall in young adults with primary hypertension. J Clin Hypertens. 2019;21:1406–1414. 10.1111/jch.13639
Funding information
This work was supported by scientific grants obtained from Pomeranian Medical University in Szczecin (Grant for Young Researchers No. MB‐315‐227‐17) and Polish Osteoporosis Foundation BO in Szczecin. The funding organizations had no role in the design, analysis, and decision to submit the manuscript for publication.
REFERENCES
- 1. GBD 2015 Risk Factors Collaborators . Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet. 2016;388(10053):1659‐1724. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Forouzanfar MH, Liu P, Roth GA, et al. Global burden of hypertension and systolic blood pressure of at least 110 to 115 mm Hg, 1990–2015. JAMA. 2017;317(2):165‐182. [DOI] [PubMed] [Google Scholar]
- 3. Pinto E. Blood pressure and ageing. Postgrad Med J. 2007;83:109‐114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Juhola J, Oikonen M, Magnussen CG, et al. Childhood physical, environmental, and genetic predictors of adult hypertension: the cardiovascular risk in young Finns study. Circulation. 2012;126(4):402‐409. [DOI] [PubMed] [Google Scholar]
- 5. Miazgowski T, Krzyżanowska‐Świniarska B, Dziwura‐Ogonowska J, Widecka K. The associations between cardiometabolic risk factors and visceral fat measured by a new dual‐energy X‐ray absorptiometry‐derived method in lean healthy Caucasian women. Endocrine. 2014;47:500‐505. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Kurniawan L, Bahrun U, Hatta M, Arif M. Body mass, total body fat percentage, and visceral fat level predict insulin resistance better than waist circumference and body mass index in healthy young male adults in Indonesia. J Clin Med. 2018;7(5):E96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Westerståhl M, Hedvall Kallerman P, Hagman E, Ek AE, Rössner SM, Marcus C. Nocturnal blood pressure non‐dipping is prevalent in severely obese, prepubertal and early pubertal children. Acta Paediatr. 2014;103(2):225‐230. [DOI] [PubMed] [Google Scholar]
- 8. Araújo S, Rouxinol‐Dias A, Mesquita‐Bastos J, Silva J, Barbosa L, Polónia J. Ambulatory blood pressure monitoring profiles in a cross‐sectional analysis of a large database of normotensive and true or suspected hypertensive patients. Rev Port Cardiol. 2018;37(4):319‐327. [DOI] [PubMed] [Google Scholar]
- 9. Polish Society of Hypertension . 2015 guidelines for the management of hypertension. Part 1–7. Arterial Hypertens. 2015;19(2):53‐83. [Google Scholar]
- 10. Williams B, Mancia G, Spiering W, et al. 2018 ESC/ESH guidelines for the management of arterial hypertension. Eur Heart J. 2018;39(33):3021‐3104. [DOI] [PubMed] [Google Scholar]
- 11. Gutch M, Kumar S, Razi SM, et al. Assessment of insulin sensitivity / resistance. Indian J Endocrinol Metab. 2015;19(1):160‐164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Tang Q, Li X, Song P, Xu L. Optimal cutoff values for the homeostasis model assessment of insulin resistance (HOMA‐IR) and pre‐diabetes screening: developments in research and prospects for the future. Drug Discov Ther. 2015;9(6):380‐385. [DOI] [PubMed] [Google Scholar]
- 13. Alberti KG, Eckel RH, Grundy SM, et al. Harmonizing the metabolic syndrome: a joint interim statement of the international diabetes federation task force on epidemiology and prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation. 2009;120(16):1640‐1645. [DOI] [PubMed] [Google Scholar]
- 14. Kahn HS. The "lipid accumulation product" performs better than the body mass index for recognizing cardiovascular risk: a population‐based comparison. BMC Cardiovasc Disord. 2005;5:26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Cheung AS, de Rooy C, Hoermann R, et al. Correlation of visceral adipose tissue measured by Lunar Prodigy dual X‐ray absorptiometry with MRI and CT in older men. Int J Obesity. 2014;40:1325‐1328. [DOI] [PubMed] [Google Scholar]
- 16. Rothney MP, Xia Y, Wacker WK, et al. Precision of a new tool to measure visceral adipose tissue (VAT) using dual‐energy X‐Ray absorptiometry (DXA). Obesity. 2013;21:E134‐E136. [DOI] [PubMed] [Google Scholar]
- 17. Neeland IJ, Grundy SM, Li X, et al. Comparison of visceral fat mass measurement by dual X‐ray absorptiometry and magnetic resonance imaging in multiethnic cohort: the Dallas heart study. Nutr Diabetes. 2016;6(7):e221. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Miazgowski T, Kucharski R, Sołtysiak M, et al. Visceral fat reference values derived from healthy European men and women aged 20–30 years using GE Healthcare dual‐energy x‐ray absorptiometry. PLoS ONE. 2017;12(7):e0180614. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Imboden MT, Welch WA, Swartz AM, et al. Reference standards for body fat measures using GE dual energy x‐ray absorptiometry in Caucasian adults. PLoS ONE. 2017;12(4):e0175110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Imboden MT, Swartz AM, Finch HW, et al. Reference standards for lean mass measures using GE dual energy x‐ray absorptiometry in Caucasian adults. PLoS ONE. 2017;12(4):e0176161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Niklas A, Flotyńska A, Puch‐Walczak A, et al. Prevalence, awareness, treatment and control of hypertension in the adult Polish population – Multi‐center National population health examination surveys – WOBASZ studies. Arch Med Sci. 2018;14(5):951‐961. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. World Health Statistics . WHO Library Cataloguing‐in‐Publication data; 2015. Accessed on 2019 Feb 20 from: http://apps.who.int/iris/bitstream/handle/10665/170250/9789240694439_eng.pdf;jsessjonxml:id=05660e16a1143ca55f094ecea21ab5d7?sequence=1
- 23. Day K, Kwok A, Evans A, et al. Comparison of a bioelectrical impedance device against the reference method dual energy x‐ray absorptiometry and anthropometry for the evaluation of body composition in adults. Nutrients. 2018;10(10):E1469. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Ding WQ, Liu JT, Shang YX, et al. DXA‐measured visceral fat mass and lean body mass reflect abnormal metabolic phenotypes among some obese and nonobese Chinese children and adolescents. Nutr Metab Cardiovasc Dis. 2018;28(6):618‐628. [DOI] [PubMed] [Google Scholar]
- 25. Sakhuja S, Booth Iii JN, Lloyd‐Jones DM, et al. Health behaviors, nocturnal hypertension, and non‐dipping blood pressure: the coronary artery risk development in young adults and Jackson heart study. Am J Hypertens. 2019;32(8):759-768. 10.1093/ajh/hpz017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Mokwatsi GG, Schutte AE, Mels C, Kruger R. Morning blood pressure surge in young black and white adults: the African‐PREDICT study. J Hum Hypertens. 2019;33(1):22‐33. [DOI] [PubMed] [Google Scholar]
- 27. Canapari CA, Hoppin AG, Kinane TB, et al. Relationship between sleep apnea, fat distribution, and insulin resistance in obese children. J Clin Sleep Med. 2011;7(3):268‐273. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Mokhlesi B, Hagen EW, Finn LA, Hla KM, Carter JR, Peppard PE. Obstructive sleep apnoea during REM sleep and incident non‐dipping of nocturnal blood pressure: a longitudinal analysis of the Wisconsin Sleep Cohort. Thorax. 2015;70(11):1062‐1069. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Bloomfield D, Park A. Night time blood pressure dip. World J Cardiol. 2015;7(7):373‐376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Chobanian AV, Bakris GL, Black HR, et al. Seventh report of the joint national committee on prevention, detection, evaluation, and treatment of high blood pressure. Hypertension. 2003;42:1206‐1252. [DOI] [PubMed] [Google Scholar]
- 31. Yan B, Sun LU, Gao YA, et al. Blood pressure reverse dipping may associate with stable coronary artery disease in patients with essential hypertension: a cross‐sectional study. Sci Rep. 2016;6:25410. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Siti Suhaila MY, Juwita S, Harmy MY, et al. Circadian blood pressure profile and associated cardiovascular risk factors in non‐dippers. Int Med J Mal. 2016;12(2):23‐31. [Google Scholar]
- 33. Birkenhäger AM, van den Meiracker AH. Causes and consequences of a non‐dipping blood pressure profile. Neth J Med. 2007;65(4):127‐131. [PubMed] [Google Scholar]
- 34. Gruzdeva O, Borodkina D, Uchasova E, Dyleva Y, Barbarash O. Localization of fat depots and cardiovascular risk. Lipids Health Dis. 2018;17(1):218. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Britton KA, Fox CS. Ectopic fat depots and cardiovascular disease. Circulation. 2011;124:e837‐841. [DOI] [PubMed] [Google Scholar]
- 36. Liu B, Sun W, Kong XQ. Perirenal fat: a unique fat pad and potential target for cardiovascular disease. Angiology. 2018;9:3319718799967. [DOI] [PubMed] [Google Scholar]
- 37. Cuspidi C, Meani S, Valerio C, et al. Reproducibility of dipping/nondipping pattern in untreated essential hypertensive patients: impact of sex and age. Blood Press Monit. 2007;12:101‐106. [DOI] [PubMed] [Google Scholar]
- 38. Monte M, Cambão M, Mesquita Bastos J, Polónia J. Reproducibility of ambulatory blood pressure values and circadian blood pressure patterns in untreated subjects in a 1–11 month interval. Rev Port Cardiol. 2015;34(11):643‐650. [DOI] [PubMed] [Google Scholar]