Abstract
Visceral fat accumulation is a key factor in the onset of cardiometabolic diseases, including hypertension. Metabolic Score for Visceral Fat (METS-VF) is an innovative, non-invasive metric developed to estimate visceral fat based on commonly accessible clinical parameters. A total of 13,822 participants were included in this cross-sectional analysis. METS-VF was calculated using a validated formula incorporating age, sex, metabolic score for insulin resistance, waist-to-height ratio. Hypertension was defined based on measured blood pressure or self-reported physician diagnosis. Logistic regression models were used to estimate the association between METS-VF and hypertension, adjusting for sociodemographic, clinical, and dietary covariates. Subgroup, threshold effect, and ROC analyses were conducted to assess robustness and predictive ability. External validation was conducted using 8400 fasting participants from the China Health and Retirement Longitudinal Study (CHARLS) 2011 baseline cohort. METS-VF was positively associated with hypertension. In model 3, participants in the highest METS-VF quartile had substantially higher odds of hypertension (OR: 5.82, 95% CI 4.76–7.11, P < 0.001). A threshold effect was observed at a METS-VF value of 6.42. Subgroup analyses confirmed the consistency of associations across various demographic and clinical strata. Notably, high intake of protein and unsaturated fatty acids attenuated this association. ROC analysis showed METS-VF had the best discriminatory power for hypertension (AUC = 0.749) compared to BMI, WHtR and METS-IR. In the CHARLS external validation cohort, METS-VF also showed the highest AUC among the evaluated indices (AUC = 0.668), and logistic regression confirmed a consistent positive association with hypertension. METS-VF was significantly associated with prevalent hypertension and showed better discriminatory performance than traditional adiposity indices. External validation in CHARLS supports the robustness of these findings, although prospective validation is still warranted.
Keywords: METS-VF, Hypertension, Dietary factors, NHANES, Lipids
Subject terms: Cardiovascular diseases, Metabolic disorders
Introduction
With changes in lifestyle patterns, obesity has become a major global public health concern1. Numerous studies have demonstrated that the accumulation of adipose tissue, particularly visceral adipose tissue (VAT), is linked to a heightened risk of insulin resistance (IR), metabolic syndrome (MetS), and various cardiovascular diseases (CVD)2–4. Given that magnetic resonance imaging (MRI) is considered the gold standard for assessing VAT, previous studies have relied on imaging tools such as computed tomography (CT) and dual-energy X-ray absorptiometry (DXA), and their clinical application has been limited by a variety of factors such as equipment, cost, and professional interpretation. Bello et al. created a novel alternative index for VAT assessment, metabolic score for visceral fat (METS-VF), which utilizes the metabolic score for IR (METS-IR), age, sex, waist-to-height ratio (WHtR), and other simple indicators to indirectly reflect the content and distribution of visceral fat, which greatly improves the feasibility of clinical application5.
Hypertension is a highly prevalent chronic condition and serves as an important risk factor for many diseases, including stroke, myocardial infarction, and chronic kidney disease (CKD)6. Despite advances in treatment and the critical importance of early intervention for hypertension, many people remain without effective diagnosis or treatment given that pre-hypertension and first-degree hypertension often go unnoticed. Emerging evidence has increasingly underscored the pivotal role of visceral adiposity in the pathogenesis of hypertension. Unlike subcutaneous fat, visceral fat is metabolically active and is closely associated with IR, systemic inflammation, and dysregulation of the renin–angiotensin–aldosterone system (RAAS), all of which are associated with elevated blood pressure7–10. Therefore, the discovery of novel metabolism-related biomarkers is imperative for a more comprehensive understanding of the intricate pathophysiology of hypertension and for the formulation of multidimensional diagnostic and therapeutic strategies.
Previous studies systematically evaluating the association between METS-VF and hypertension remain scarce and have mainly focused on Asian populations11,12, the aim of research was to apply the National Health and Nutrition Examination Survey (NHANES) database to examine the relationship between METS-VF and hypertension in different populations in the United States. In addition to variable demographic, socioeconomic, and clinical subgroups, this study adjusted for the effects of dietary factors such as carbohydrates, proteins, and fatty acids, with the aim of providing partial dietary guidance for the hypertensive population.
Methods
This study used data from the NHANES database related to the 2007–2018 survey cycle. The study design and data collection procedures are thoroughly documented on the Centers for Disease Control and Prevention (CDC) NHANES website (CDC NHANES).
Study population and eligibility criteria
A total of 59,846 participants (adults aged 20 years or older) were involved in the original data (Fig. 1). According to the study objectives, 42,522 missing values for the exposure factor METS-VF were excluded, because METS-VF could not be calculated without fasting glucose, triglycerides (TG), body mass index (BMI), WHtR, and other required components. The large reduction in sample size was mainly due to the availability of fasting laboratory measurements in the NHANES fasting subsample. Participants with missing data on relevant covariates were further excluded: coronary heart disease (CHD) (n = 3089), diabetes mellitus (n = 358), heart failure (HF) (n = 18), stroke (n = 10), smoking (n = 12), education (n = 12), and missing marital data (n = 3). Finally, 13,822 participants were included in the study analysis.
Fig. 1.
Flowchart of participant selection. NHANES, National Health and Nutrition Examination Survey.
Definition of METS-VF and hypertension
METS-VF was calculated as 4.466 + 0.011*(Ln(METS-IR))3 + 3.239*(Ln(WHtR))3 + 0.319*(Sex) + 0.594*(Ln(Age). Sex was coded as 1 for men and 0 for women. METS-VF levels were categorized into quartiles as follows: Q1 (< 6.47), Q2 (≥ 6.47 to < 6.99), Q3 (≥ 6.99 to < 7.37), and Q4 (≥ 7.37). Hypertension was defined based on either self-reported physician diagnosis or the mean value of repeated measured blood pressure. Blood pressure was measured by trained personnel according to standardized NHANES protocols, and the mean systolic blood pressure (SBP) and diastolic blood pressure (DBP) values from consecutive measurements were used for analysis. Participants were classified as having hypertension if they met at least one of the following criteria: self-reported physician-diagnosed hypertension, mean SBP ≥ 130 mmHg, or mean DBP ≥ 80 mmHg. Antihypertensive medication use was not included as a separate diagnostic criterion in the primary definition of hypertension because medication information was not uniformly incorporated into the operational outcome definition across the analytical datasets.
Covariates
The following covariates were included in this study for analysis: age (years), sex (male/female), race, education level, BMI (kg/m2), total cholesterol (TC, mg/dL), high-density lipoprotein cholesterol (HDL-C, mg/dL), low-density lipoprotein cholesterol (LDL-C, mg/dL), fasting glucose (mg/dL), serum creatinine (μmol/L), protein intake (g/day), energy intake (kcal/day), carbohydrate intake (CARB, g/day), total fat intake (TFAT, g/day), saturated fatty acid intake (SFAT, g/day), monounsaturated fatty acid intake (MFAT, g/day), and polyunsaturated fatty acid intake (PFAT , g/day). The following disease states were also included: diabetes, stroke, CHD, and HF were identified based on self-reported physician diagnosis. Smoking status was defined by self-reported lifetime consumption of at least 100 cigarettes, and alcohol use (based on the results of the survey, “In the past year, how many times on average have you consumed alcoholic beverages on drinking days?”). Nutrient intake variables—including total energy (kcal), macronutrients (protein, carbohydrate, total fat), fatty acid subtypes (saturated, monounsaturated, polyunsaturated). The data were collected by trained personnel using standardized 24-h dietary recall interviews, and nutrient intakes were estimated from the United States Department of Agriculture food composition database. For participants with two available dietary recalls, the average intake across the two recall days was used; when only one valid recall was available, the available recall value was used. Physical activity (PA) was assessed using the NHANES physical Activity questionnaire. Weekly hours of moderate-intensity physical activity were calculated as the sum of hours of moderate-intensity work, transportation-related walking/cycling, and moderate-intensity recreational activity, plus twice the weekly hours of vigorous-intensity work and vigorous-intensity recreational activity. This calculation is consistent with that based on the NHANES/ Global Physical Activity Questionnaire (GPAQ), which found 4.0 METs (metabolic equivalent of task) for moderate and transport activities, the vigorous activity was 8.0 METs. According to the current adult PA recommendations, participants were divided into three groups: Inactive (0 min/week), Insufficiently active (1–149 min/week), or Active (≥ 150 min/week). The threshold of ≥ 150 moderate-equivalent minutes/week is equivalent to achieving the GPAQ recommended level of aerobic physical activity and is equivalent to ≥ 600 MET-min/week. Missing PA-related variables were handled using multiple imputation, after which the derived PA category variable was reconstructed and included in the fully adjusted model.
External validation analysis
To further evaluate the robustness of the ROC findings, we performed an independent external validation analysis using the 2011 baseline data from the China Health and Retirement Longitudinal Study (CHARLS). CHARLS is independent from NHANES and includes middle-aged and older Chinese adults. Participants with complete data on age, sex, height, weight, waist circumference, glucose, TG, HDL-C, measured blood pressure, and self-reported hypertension status were included in the validation analysis. Because METS-VF calculation requires fasting glucose and triglycerides, the validation cohort was restricted to participants with fasting blood samples.
BMI, WHtR, METS-IR, and METS-VF were calculated using the same formulas as in the NHANES analysis. Hypertension was also defined using the same operational definition. ROC curves and AUCs were calculated for METS-VF, WHtR, METS-IR, and BMI in the CHARLS validation cohort. In addition, logistic regression analyses were performed to examine whether the association between METS-VF and hypertension was directionally consistent in the external cohort. The simplified adjusted model included education level, marital status, smoking status, drinking status, heart disease, and stroke.
Statistical analysis
All analyses accounted for the complex multistage probability sampling design of NHANES, including sampling weights, strata, and primary sampling units. The NHANES Mobile Examination Center examination weight variable WTMEC2YR was used as the original 2-year sampling weight. Because six consecutive 2-year NHANES cycles from 2007–2008 to 2017–2018 were combined, a pooled 12-year weight was calculated as WEIGHT = WTMEC2YR/6 according to the NHANES analytic guidance for combining multiple cycles. The recalculated weight variable, together with SDMVSTRA as the stratification variable and SDMVPSU as the primary sampling unit variable, was used in all weighted analyses.
Continuous variables were expressed as mean ± standard deviation (SD) for normally distributed data. Categorical variables were presented as percentages. Baseline characteristics were compared using one-way ANOVA, given the continuous nature of the variables. Linear and non-linear regression models were applied to explore both linear and potential non-linear associations between METS-VF and hypertension risk. Sensitivity analyses were performed to assess the robustness and consistency of the results.
All statistical analyses were performed using EmpowerStats statistical software, version 5.0 (www.empowerstats.com; X&Y Solutions, Inc., Boston, MA, USA), and R software version 4.5.3. Weighted analyses accounting for the complex NHANES survey design were conducted using the survey package. Multiple imputation of PA-related variables was performed using the mice package. ROC curves and AUCs were generated using the pROC package, and figures were produced using ggplot2. Data management was performed using base R and dplyr. NHANES datasets were downloaded directly from the official CDC NHANES website and merged by the authors using the unique participant identifier SEQN. No NHANES-specific extraction or processing package was used in this study. A two-sided P value < 0.05 was considered statistically significant.
Results
Baseline characteristics
The weighted baseline characteristics of the 13,822 participants are summarized in Table 1, stratified by hypertension status. Individuals with hypertension (n = 6721) were significantly older than those without hypertension (n = 7101), with a higher proportion aged over 60 years (38.69% vs. 12.25%, P < 0.001). The proportion of males was significantly higher in the hypertensive group compared to the non-hypertensive group (51.89% vs. 45.51%, P < 0.001). Racial/ethnic distribution differed significantly between groups, with non-Hispanic White participants being the most prevalent in both groups. Educational attainment was generally lower among those with hypertension; only 25.40% had completed college or above, compared to 34.53% in the normotensive group (P < 0.001). Regarding marital status, widowed, divorced, or separated individuals were more common among hypertensives, whereas those who had never married were more prevalent in the non-hypertensive group (P < 0.001).
Table 1.
Weighted baseline characteristics of participants with and without hypertension.
| Variable | Overall (n = 13,822) | Hypertension (n = 6721) | Non-hypertension (n = 7101) | P value |
|---|---|---|---|---|
| Age (years), n (%) | < 0.001 | |||
| ≤ 40 | 5186 (37.52) | 1432 (18.51) | 3754 (52.84) | |
| > 40 to ≤ 60 | 5312 (38.43) | 2834 (42.80) | 2478 (34.91) | |
| > 60 | 3324 (24.05) | 2455 (38.69) | 869 (12.25) | |
| Sex, n (%) | < 0.001 | |||
| Male | 6684 (48.36) | 3470 (51.89) | 3214 (45.51) | |
| Female | 7138 (51.64) | 3251 (48.11) | 3887 (54.49) | |
| Race, n (%) | < 0.001 | |||
| Mexican American | 1193 (8.71) | 649 (10.45) | 544 (6.55) | |
| Other Hispanic | 874 (6.04) | 409 (6.74) | 465 (5.17) | |
| Non-Hispanic White | 9242 (66.93) | 4482 (68.10) | 4760 (65.99) | |
| Non-Hispanic Black | 1432 (10.43) | 754 (12.74) | 678 (8.57) | |
| Other race—including multi-racial | 1081 (7.89) | 427 (7.43) | 654 (8.26) | |
| Education level, n (%) | < 0.001 | |||
| Less than 9th grade | 758 (5.45) | 340 (4.97) | 418 (6.05) | |
| 9–11th grade (Includes 12th grade with no diploma) | 1485 (10.70) | 719 (10.15) | 766 (11.39) | |
| High school graduate/GED or equivalent | 3143 (22.51) | 1665 (20.75) | 1478 (24.70) | |
| Some college or AA degree | 4281 (30.88) | 2014 (32.47) | 2267 (29.61) | |
| College graduate or above | 4155 (30.46) | 2363 (25.40) | 1792 (34.53) | |
| Marital status, n (%) | < 0.001 | |||
| Married | 7780 (56.12) | 3928 (58.44) | 3852 (54.25) | |
| Widowed | 724 (5.00) | 561 (8.35) | 163 (2.3) | |
| Divorced | 1475 (10.50) | 852 (12.67) | 623 (8.77) | |
| Separated | 316 (2.29) | 159 (2.37) | 157 (2.21) | |
| Never married | 2380 (17.66) | 773 (11.50) | 1607 (22.63) | |
| Living with partner | 1147 (8.43) | 448 (6.67) | 699 (9.84) | |
| PIR | 2.92 ± 0.04 | 2.91 ± 0.05 | 2.93 ± 0.04 | 0.698 |
| BMI (kg/m2) | < 0.001 | |||
| ≤ 30 | 8674 (63.63) | 3468 (51.61) | 5206 (73.32) | |
| > 30 | 5148 (36.37) | 3253 (48.40) | 1895 (26.68) | |
| Height (cm) | 168.76 ± 0.13 | 168.75 ± 0.18 | 168.78 ± 0.17 | 0.897 |
| Waist (cm) | 99.21 ± 0.27 | 105.16 ± 0.29 | 94.43 ± 0.33 | < 0.001 |
| WHtR | 0.59 ± 0.00 | 0.62 ± 0.00 | 0.56 ± 0.00 | < 0.001 |
| METS-VF | 6.80 ± 0.01 | 7.13 ± 0.01 | 6.52 ± 0.02 | < 0.001 |
| SBP (mmHg) | 120.41 ± 0.23 | 129.84 ± 0.33 | 112.81 ± 0.16 | < 0.001 |
| DBP (mmHg) | 69.49 ± 0.23 | 73.11 ± 0.34 | 65.57 ± 0.17 | < 0.001 |
| HDL-C (mg/dL) | 54.41 ± 0.26 | 53.03 ± 0.32 | 55.52 ± 0.31 | < 0.001 |
| TG (mg/dL) | 123.27 ± 1.29 | 139.13 ± 1.99 | 110.51 ± 1.35 | < 0.001 |
| Fasting glucose (mg/dL) | 106.68 ± 0.37 | 114.54 ± 0.65 | 100.34 ± 0.36 | < 0.001 |
| Serum creatinine (μmol/L) | 124.91 ± 1.14 | 123.79 ± 1.48 | 125.82 ± 1.53 | 0.309 |
| TC (mg/dL) | 192.62 ± 0.19 | 192.34 ± 0.21 | 192.84 ± 0.25 | 0.063 |
| LDL-C (mg/dL) | 113.97 ± 0.44 | 114.86 ± 0.63 | 113.25 ± 0.55 | 0.045 |
| Energy (kcal/day) | 2167.98 ± 11.13 | 2138.07 ± 16.36 | 2195.67 ± 14.90 | 0.010 |
| Protein (gm/day) | 83.46 ± 0.52 | 82.12 ± 0.74 | 84.47 ± 0.66 | 0.017 |
| Carbohydrate (gm/day) | 256.48 ± 1.34 | 249.93 ± 1.93 | 261.76 ± 1.78 | < 0.001 |
| Total fat (gm/day) | 84.39 ± 0.58 | 83.93 ± 0.85 | 84.76 ± 0.77 | 0.463 |
| Total saturated fatty acids (gm/day) | 27.52 ± 0.22 | 27.43 ± 0.32 | 27.61 ± 0.28 | 0.678 |
| Total monounsaturated fatty acids (gm/day) | 29.94 ± 0.22 | 29.77 ± 0.32 | 30.08 ± 0.28 | 0.469 |
| Total polyunsaturated fatty acids (gm/day) | 19.25 ± 0.15 | 19.13 ± 0.21 | 19.35 ± 0.21 | 0.431 |
| Diabetes, n (%) | < 0.001 | |||
| Yes | 1441 (9.87) | 1180 (17.55) | 261 (3.67) | |
| No | 12,381 (90.13) | 5541 (82.45) | 6840 (96.33) | |
| CHD, n (%) | < 0.001 | |||
| Yes | 489 (3.35) | 400 (5.95) | 89 (1.26) | |
| No | 13,333 (96.65) | 6321 (94.05) | 7012 (98.74) | |
| Stroke, n (%) | < 0.001 | |||
| Yes | 399 (2.73) | 327 (4.87) | 72 (1.01) | |
| No | 13,243 (97.27) | 6394 (95.13) | 7029 (98.99) | |
| Heart failure, n (%) | < 0.001 | |||
| Yes | 338 (2.30) | 296 (4.41) | 42 (0.59) | |
| No | 13,484 (97.70) | 6425 (95.59) | 7059 (99.41) | |
| Smoking, n (%) | < 0.001 | |||
| Yes | 6193 (44.48) | 3290 (48.95) | 2903 (40.88) | |
| No | 7629 (55.52) | 3431 (51.05) | 4198 (59.12) | |
| Alcohol consumption (drinks/day), n (%) | 0.583 | |||
| ≤ 5 | 12,941 (93.61) | 6304 (93.79) | 6637 (93.47) | |
| > 5 | 881 (6.39) | 417 (6.21) | 464 (6.53) | |
| Physical activity, n (%) | < 0.001 | |||
| Active | 8838 (63.95) | 3950 (57.89) | 4888 (68.83) | |
| Insufficiently active | 2080 (15.05) | 1085 (16.32) | 995 (14.02) | |
| Inactive | 2904 (21.01) | 1686 (25.79) | 1218 (17.16) | |
This table presents the baseline characteristics of the study population divided into hypertensive and non-hypertensive groups. Continuous variables are described as weighted mean ± standard deviation (SD), while categorical variables are presented as unweighted n (%). METS-VF, Metabolic score for visceral fat; BMI, Body mass index; HDL-C, High-density lipoprotein cholesterol; TG, Triglycerides; LDL-C, Low-density lipoprotein cholesterol; TC, Total cholesterol; PIR, Poverty-income ratio; SBP, Systolic blood pressure; DBP, Diastolic blood pressure; CHD, Coronary heart disease; gm, Gram.
Participants with hypertension had significantly higher levels of BMI (≥ 30 kg/m2), waist circumference, WHtR, METS-VF, SBP, and DBP (all P < 0.001). In contrast, HDL-C levels were lower, while TG and fasting blood glucose were higher in the hypertensive group. In terms of dietary intake, individuals with hypertension had significantly lower average intake of total energy, carbohydrates, protein, and total fat compared to those without hypertension (P < 0.001 for all). Additionally, the intake of specific fatty acid subtypes—SFAT, MFAT and PFAT—was also reduced in the hypertensive group. The distribution of physical activity differed significantly between participants with and without hypertension (P < 0.001). Compared with the non-hypertension group, the hypertension group had a lower proportion of active individuals (57.89% vs. 68.83%) and a higher proportion of inactive individuals (25.79% vs. 17.16%).
Additionally, the prevalence of comorbidities—including diabetes (17.55% vs. 3.67%), CHD, stroke, and HF—was substantially higher among individuals with hypertension (P < 0.001 for all). Regarding lifestyle factors, current smoking was more common among hypertensive participants (P < 0.001), while there was no statistically significant variation in alcohol consumption between the two groups (P = 0.583).
Association between METS-VF and hypertension
Table 2 presents the weighted association between METS-VF and hypertension across the three models. When treated as a continuous variable, each one-unit increase in METS-VF was significantly associated with increased odds of hypertension, with odds ratios (ORs) progressively attenuated across the models: OR = 2.37 (95% CI 2.25–2.49) in Model 1, OR = 1.87 (95% CI 1.76–1.98) in Model 2, and OR = 1.79 (95% CI 1.68–1.90) in Model 3 (all P < 0.001). Compared to participants in the lowest quartile (Q1), the risk of hypertension increased significantly with each ascending quartile. After adjusting for a wide range of demographic, clinical, biochemical, and lifestyle covariates, the ORs were 2.05 (95% CI 1.74–2.41) for Q2, 3.56 (95% CI 3.03–4.18) for Q3, and 5.82 (95% CI 4.76–7.11) for Q4 (all P < 0.001) in Model 3. The test for trend remained statistically significant across all models (P for trend < 0.001).
Table 2.
Weighted association between METS-VF and hypertension in all participants.
| Exposure | Model 1 | Model 2 | Model 3 |
|---|---|---|---|
| OR (95% CI), P value | OR (95% CI), P value | OR (95% CI), P value | |
| Continuous | 2.37 (2.25–2.49), < 0.001 | 1.87 (1.76–1.98), < 0.001 | 1.79 (1.68–1.90), < 0.001 |
| METS-VF quartile | |||
| Q1 | Reference | Reference | Reference |
| Q2 | 2.75 (2.33–3.24), < 0.001 | 2.07 (1.75–2.44), < 0.001 | 2.05 (1.74–2.41), < 0.001 |
| Q3 | 5.86 (5.05–6.81), < 0.001 | 3.72 (3.16–4.37), < 0.001 | 3.56 (3.03–4.18), < 0.001 |
| Q4 | 13.93 (11.74–16.54), < 0.001 | 6.67 (5.51–8.07), < 0.001 | 5.82 (4.76–7.11), < 0.001 |
| P for trend | 3.10 (2.78–3.46), < 0.001 | 3.10 (2.78–3.46), < 0.001 | 2.88 (2.57–3.23), < 0.001 |
For quartile analyses, METS-VF was categorized into quartiles, with Q1 as the reference group.
Model 1: no covariates adjusted.
Model 2: adjusted for age, sex, and race.
Model 3: adjusted for the covariates in Model 2 as well as education level, marital status, poverty-to-income ratio, serum creatinine, total cholesterol, low-density lipoprotein cholesterol, energy, protein, carbohydrate, total fat, total saturated fatty acids, total monounsaturated fatty acids, total polyunsaturated fatty acids ,smoking, alcohol consumption, physical activity, stroke, coronary heart disease, diabetes and heart failure.
CI, Confidence interval; METS-VF, Metabolic score for visceral fat; OR, Odds ratio.
Nonlinearity and threshold effect analysis
Threshold effect analysis was conducted to assess the nonlinear relationship between METS-VF and hypertension, comparing a standard linear regression model with a two-piecewise regression model (Table 3). In the standard linear model, a one-unit increase in METS-VF was associated with significantly increased odds of hypertension, with an adjusted OR of 2.67 (95% CI 2.47–2.89, P < 0.001). For METS-VF values below 6.42, the risk of hypertension increased moderately (OR: 1.84, 95% CI 1.56–2.18, P < 0.001). For METS-VF values equal to or above 6.42, the association was stronger, with an OR of 3.26 (95% CI 2.90–3.66, P < 0.001). The log-likelihood ratio test (P < 0.001) indicated that the two-piecewise regression model fitted the data significantly better than the linear model (Fig. 2).
Table 3.
Threshold effect analysis of the association between METS-VF and hypertension.
| METS-VF | Adjusted OR (95% CI), P value |
|---|---|
| Fitting by the standard linear model | 2.67 (2.47–2.89), < 0.001 |
| Fitting by the two-piecewise model | |
| Inflection point | 6.42 |
| METS-VF < 6.42 | 1.84 (1.56–2.18), < 0.001 |
| METS-VF ≥ 6.42 | 3.26 (2.90–3.66), < 0.001 |
| P for log-likelihood ratio | < 0.001 |
The data were adjusted for age, sex, race, educational level, marital status, poverty-to-income ratio, serum creatinine, total cholesterol, low-density lipoprotein cholesterol, smoking, alcohol consumption, coronary heart disease, physical activity, stroke, diabetes and heart failure.
CI, Confidence interval; METS-VF, Metabolic score for visceral fat; OR, Odds ratio.
Fig. 2.
The association between METS-VF and hypertension. Note The solid red line represents the fit between the variables. The blue bands represent 95% confidence. The smooth curve was adjusted for covariates including age, sex, race, education level, marital status, PIR status, serum creatinine, total cholesterol (TC), low-density lipoprotein cholesterol (LDL), smoking status, drinking status, stroke, coronary heart disease (CHD), heart failure (HF) and diabetes status.
Subgroup analysis
Table 4 summarizes the weighted stratified analyses exploring the association between METS-VF and hypertension across various demographic, socioeconomic, and clinical subgroups. Overall, the positive association remained consistent in all subgroups, though the strength of the association varied. No significant interactions were observed between METS-VF and factors such as race, marital status, diabetes, stroke, HF, PA and alcohol consumption, indicating the robustness of the association (P > 0.05) (Table 4). However, significant associations between METS-VF and hypertension were observed across subgroups defined by age, sex, poverty income ratio (PIR), education level, coronary heart disease (CHD), and smoking status (P < 0.05).
Table 4.
Weighted subgroup analysis of the association between METS-VF and hypertension.
| Variable | OR (95% CI), P value | P for interaction |
|---|---|---|
| Age (years) | < 0.001 | |
| ≤ 40 | 3.84 (3.26–4.51), < 0.001 | |
| > 40 to ≤ 60 | 3.26 (2.68–3.95), < 0.001 | |
| > 60 | 1.89 (1.57–2.27), < 0.001 | |
| Sex | 0.009 | |
| Male | 2.53 (2.18–2.92), < 0.001 | |
| Female | 3.30 (2.82–3.85), < 0.001 | |
| Race | 0.078 | |
| Mexican–American | 4.24 (3.04–5.90), < 0.001 | |
| Other Hispanic | 3.05 (2.29–4.07), < 0.001 | |
| Non-Hispanic White | 2.86 (2.44–3.35), < 0.001 | |
| Non-Hispanic Black | 2.49 (2.09–2.96), < 0.001 | |
| Other, including multiple | 3.04 (2.25–4.10), < 0.001 | |
| Education level | < 0.001 | |
| < 9th grade | 3.27 (2.25–4.73), < 0.001 | |
| 9–11th grade (Includes 12th grade without diploma) | 2.00 (1.57–2.54), < 0.001 | |
| High school graduate/GED | 2.97 (2.42–3.64), < 0.001 | |
| Some college or 2-year degree | 2.53 (2.08–3.07), < 0.001 | |
| College degree or beyond | 3.70 (3.06–4.47), < 0.001 | |
| Marital status | 0.419 | |
| Married | 2.92 (2.48–3.45), < 0.001 | |
| Widowed | 2.04 (1.45–2.88), < 0.001 | |
| Divorced | 3.28 (2.41–4.46), < 0.001 | |
| Separated | 2.84 (1.66–4.86), < 0.001 | |
| Never married | 2.88 (2.39–3.48), < 0.001 | |
| Living with partner | 2.62 (1.95–3.53), < 0.001 | |
| Poverty-to-income ratio | 0.036 | |
| ≤ 1 | 2.69 (2.26–3.20), < 0.001 | |
| > 1 to ≤ 3 | 2.64 (2.23–3.13), < 0.001 | |
| > 3 | 3.37 (2.87–3.95), < 0.001 | |
| Diabetes | 0.972 | |
| Yes | 2.86 (1.88–4.33), < 0.001 | |
| No | 2.88 (2.56–3.23), < 0.001 | |
| Coronary heart disease | 0.015 | |
| Yes | 1.32 (0.68–2.54), 0.409 | |
| No | 2.92 (2.61–3.27), < 0.001 | |
| Stroke | 0.056 | |
| Yes | 1.65 (0.96–2.84), 0.076 | |
| No | 2.91 (2.58–3.28), < 0.001 | |
| Heart failure | 0.208 | |
| Yes | 2.90 (2.56–3.25), < 0.001 | |
| No | 1.91 (1.01–3.62), 0.053 | |
| Smoking | < 0.001 | |
| Yes | 2.39 (2.08–2.74), < 0.001 | |
| No | 3.44 (2.97–3.98), < 0.001 | |
| Alcohol consumption, drinks/day | 0.605 | |
| ≤ 5 | 2.89 (2.56–3.27), < 0.001 | |
| > 5 | 2.66 (1.98–3.57), < 0.001 | |
| Physical activity | 0.49 | |
| Active | 2.90 (2.52–3.35), < 0.001 | |
| Insufficiently active | 2.49 (1.96–3.17), < 0.001 | |
| Inactive | 3.12 (2.37–4.13), < 0.001 | |
Significant values are in [bold].
The subgroup analysis was adjusted for the same covariates as in Model 3, except that the stratification variable itself was not adjusted for in the corresponding subgroup model. P for interaction was calculated to assess whether the association between METS-VF and hypertension differed across subgroups.
CI, Confidence interval; METS-VF, Metabolic score for visceral fat; OR, Odds ratio.
In addition to better explore the potential nonlinear associations of each nutrient intake with the relationship between METS-VF and hypertension, as well as to reduce the interference of extreme values in the analysis of the results, all dietary variables were grouped into quartiles and then subgrouped for subgroup analyses in the present study (Fig. 3). The results showed that the association was significantly attenuated in the higher protein, polyunsaturated fatty acid and monounsaturated fatty acid intake groups (P < 0.05).
Fig. 3.
Forest plot of the association between METS-VF and hypertension across quartiles of nutrient intake. Note This figure illustrates adjusted odds ratios (ORs) and 95% confidence intervals for the association between METS-VF and hypertension across quartiles (Q1–Q4) of various dietary factors. The ORs represent the association between METS-VF and hypertension within each nutrient-intake quartile.
ROC analysis of METS-VF and other predictors of hypertension
The paper displays the ROC (Receiver Operating Characteristic) curves for four indicators—METS-VF, BMI, WHtR, and METS-IR —in predicting the presence of hypertension (Fig. 4). Among all the indicators, METS-VF demonstrated the highest predictive performance with an AUC (area under the curve) of 0.749, indicating good discrimination. In comparison, the AUCs for WHtR, METS-IR, and BMI were 0.681, 0.644, and 0.638, respectively. The relatively lower AUC values of BMI and METS-IR highlight their limited effectiveness in hypertension risk prediction compared to METS-VF.
Fig. 4.
ROC curves of METS-VF, WHtR, METS-IR, and BMI for predicting hypertension. Note ROC, Receiver operating characteristic; AUC, Area under the curve; METS-VF, Metabolic score for visceral fat; WHtR, Waist-to-height ratio; METS-IR, Metabolic score for insulin resistance; BMI, Body mass index. The ROC curves compare the discriminatory power of different adiposity and metabolic indices in predicting hypertension. A higher AUC indicates better predictive performance. Among them, METS-VF had the highest AUC value (0.749).
External validation in the CHARLS cohort
To further validate the discriminatory performance of METS-VF, we performed an independent external validation analysis using the CHARLS 2011 baseline cohort. After restricting the analysis to participants with complete data required for METS-VF calculation, hypertension status, and fasting blood samples, 8400 participants were included in the external validation cohort.
In this independent validation cohort, METS-VF retained the highest discriminatory ability for hypertension among the evaluated indices. The AUC for METS-VF was 0.668, followed by WHtR, METS-IR, and BMI, with AUCs of 0.640, 0.631, and 0.625, respectively (Fig. 5). In addition, logistic regression analysis showed that METS-VF remained positively associated with hypertension in the CHARLS cohort. Each one-unit increase in METS-VF was associated with higher odds of hypertension in the unadjusted model (OR = 2.564, 95% CI 2.203–2.983, P < 0.001) and in the simplified adjusted model (OR = 2.601, 95% CI 2.227–3.037, P < 0.001) (Table 5). These findings support the robustness of the ROC results observed in the NHANES analysis.
Fig. 5.
ROC curves of METS-VF, WHtR, METS-IR, and BMI for hypertension in the CHARLS external validation cohort. Note ROC, Receiver operating characteristic; AUC, Area under the curve; METS-VF, Metabolic score for visceral fat; WHtR, Waist-to-height ratio; METS-IR, Metabolic score for insulin resistance; BMI, Body mass index; CHARLS, China Health and Retirement Longitudinal Study. The ROC curves show the discriminatory performance of METS-VF, WHtR, METS-IR, and BMI for hypertension in the independent CHARLS 2011 baseline validation cohort. METS-VF showed the highest AUC among the evaluated indices (0.668).
Table 5.
External validation of the association between METS-VF and hypertension in CHARLS.
| Model | Exposure | OR (95% CI), P value |
|---|---|---|
| Unadjusted model | METS-VF | 2.56 (2.20–2.98), < 0.001 |
| Adjusted model | METS-VF | 2.60 (2.23–3.04), < 0.001 |
The table shows the association between METS-VF and hypertension in the CHARLS external validation cohort. The unadjusted model included METS-VF as the only independent variable.
Unadjusted model: no covariates adjusted.
Adjusted model: adjusted for education level, marital status, smoking, alcohol consumption, stroke, coronary heart disease, diabetes and heart failure.
CHARLS, China Health and Retirement Longitudinal Study; CI, Confidence interval; METS-VF, Metabolic score for visceral fat; OR, Odds ratio.
Discussion
This study systematically assessed the association between METS-VF and hypertension using a nationally representative sample of 13,822 U.S. adults. The findings indicated that elevated METS-VF levels were significantly correlated with a higher chance of hypertension, and the association remained robust after adjusting for multiple confounders including race, marital status, PA, diabetes, stroke, HF and drinking. Further stratified analyses showed that the association was generally consistent across gender, age and other prespecified subgroups.
The differences observed across race, marital status, and education level may partly reflect the influence of social determinants of health on both hypertension and metabolic dysfunction. Recent evidence suggests that racial disparities in hypertension are closely related to structural and social factors, including socioeconomic inequalities, access to health care, neighborhood conditions, chronic stress, and discrimination, rather than race itself as a purely biological construct. Higher educational attainment has generally been associated with lower hypertension prevalence, better blood pressure control, and lower odds of metabolic syndrome, possibly through improved health literacy, healthier dietary and lifestyle patterns, and greater access to preventive health resources. Marital status may also be relevant, as partnership can affect cardiometabolic health through social support, treatment adherence, shared lifestyle behaviors, and psychosocial stress. Therefore, the associations observed in our study may not only reflect biological metabolic risk, but also the combined contribution of behavioral, psychosocial, and socioeconomic factors. However, given the cross-sectional design of NHANES, these findings should be interpreted as associative rather than causal13–15.
The results of this study are consistent with findings from several recent studies examining the association between visceral fat and hypertension. Several studies have confirmed a significant positive association with hypertension using visceral fat volume measured by imaging modalities 16,17. And a study by Asghari et al. found that fat in people with elevated systolic blood pressure was predominantly distributed in the abdomen and visceral fat 18. However, the high cost and high radiation of current imaging modalities limit their widespread use in the general population. In contrast, METS-VF is based on available clinical indicators (e.g., waist circumference, BMI, TG, HDL-C, etc.), and has good operability and generalization value 5,19.
In addition, several studies have conducted preliminary investigations into the relationship between METS-VF and MetS, diabetes, atherosclerosis, CKD20–22. In a large prospective study of 41,756 participants, Liu et al. found that a significant association between high levels METS-VF and CVD and all-cause mortality–hazard ratio (HR): 2.78 and 4.90 in Q4 group, and the association was stronger with longer duration of exposure to high METS-VF 23. In a retrospective study conducted by Qian et al., METS-VF was found to have a positive relationship with the risk of carotid atherosclerosis when METS-VF was lower than 8.09 (OR:1.874, P < 0.001)24. In addition, Yu et al. confirmed that the positive relationship between METS-VF and the incidence of CKD, with ORs and 95% CIs of 3.585 (1.585–8.109) in men and 4.627 (2.485–8.616) in women25. These findings imply that METS-VF may be a practical, reliable, and cost-effective marker for identifying individuals at high risk for mortality.
The fully adjusted model indicated that elevated METS-VF levels were associated with an increased risk of hypertension, which may involve several mechanisms. On the one hand METS-VF is based on METS-IR derivation and reflects IR status, which can lead to increased sympathetic excitability, sodium retention and elevated blood pressure26–28. In addition, VAT is metabolically active and secretes a variety of pro-inflammatory factors, and chronic inflammation promotes vascular endothelial dysfunction, which leads to a decrease in the elasticity of the wall. Moreover, excess VAT activates the RAAS system, increasing blood volume and vasoconstrictor responsiveness29–31. At the same time, excessive fat accumulation in the human body can directly compress the kidneys, affecting their sodium excretion and pressure regulation function32. Multiple metabolic mechanisms act synergistically and ultimately lead to hypertension. He et al. showed that dietary intake of all types was a modifiable risk factor for hypertension at any stage of life33, high intake of SFAT and non-esterified fatty acids may activate pro-inflammatory pathways and increase oxidative stress, thereby favoring endothelial dysfunction in such subjects34. Therefore, the present study model simultaneously adjusted for multiple dietary factors to enhance the robustness of the relationship between hypertension and METS-VF.
Moreover, threshold effects were further analyzed for the association between hypertension and METS-VF. The positive correlation increased significantly when METS-VF > 6.42, which may be due to the following reasons: First, there is a metabolic compensation phase, in which the body may resist blood pressure increase through regulatory mechanisms (e.g., vasodilatation, metabolic buffering) at low or moderate levels of METS-VF, which may be manifested as a weaker effect. Once a certain “metabolic load threshold” is exceeded, regulatory mechanisms fail and blood pressure rises rapidly35. Second, individual differences in fat distribution can influence the expression of hypertension. Third, high METS-VF levels often imply a combination of multiple metabolic abnormalities (hyperglycemia, hypertriglycerides, insulin resistance)36,37. The combined effect of these indicators on blood pressure produces a synergistic boosting effect, resulting in a “non-linear enhancement” of the curves. At the same time, we conducted ROC analysis of the four indexes in predicting hypertension, and the AUC value of METS-VF was significantly better than the other three indexes compared with WHtR, METS-IR, and BMI, further emphasizing its advantage in predicting the risk of hypertension. Importantly, the discriminatory performance of METS-VF was further supported by an independent external validation analysis using the CHARLS cohort. Although the AUC of METS-VF in CHARLS was lower than that observed in NHANES, METS-VF still showed the highest AUC among the evaluated indices, including WHtR, METS-IR, and BMI. This difference in absolute AUC values may be partly explained by differences in age distribution, ethnicity, population structure, and healthcare context between NHANES and CHARLS. Nevertheless, the consistent ranking of METS-VF across the two independent cohorts suggests that METS-VF may provide more comprehensive information on hypertension-related metabolic risk than conventional anthropometric indicators alone. Therefore, the METS-VF may provide an important basis for evaluating the risk of hypertension in the early stages.
Strengths and limitations
This research possesses several key strengths. First, the primary analysis was based on NHANES, a large nationally representative survey of the US population, and all analyses accounted for the complex multistage sampling design. Second, the study comprehensively adjusted for demographic, socioeconomic, clinical, lifestyle, PA, and dietary covariates, thereby reducing potential confounding as much as possible. Third, subgroup, threshold effect, and ROC analyses were performed to assess the robustness and discriminatory performance of METS-VF. Importantly, we further conducted an independent external validation analysis using the CHARLS cohort. In this external validation cohort, METS-VF retained the highest discriminatory ability for hypertension among the evaluated indices, and logistic regression analyses showed a directionally consistent positive association between METS-VF and hypertension. These findings strengthen the robustness and external validity of the association between METS-VF and hypertension.
Several limitations should be acknowledged. First,Although we added an independent external validation analysis using the CHARLS cohort, several limitations should be acknowledged. CHARLS mainly includes middle-aged and older Chinese adults, whereas NHANES represents a broader US adult population. Therefore, differences in age structure, ethnicity, lifestyle, and healthcare context may influence the absolute AUC estimates. Second, the sample data are cross-sectional, and the results are not causally related. Third, the METS-VF is an indirect formula to estimate visceral fat load and is not a direct imaging measurement, which is susceptible to measurement error and could influence the accuracy of the findings. Fourthly, the data on physical activity and related past medical history are evaluated through self-reported questionnaire data, which may be affected by recall bias and incorrect classification. Antihypertensive medication use was not included as a separate diagnostic criterion in the primary definition of hypertension. Fifth, although self-reported physician-diagnosed hypertension may capture most clinically diagnosed cases, some participants with medication-controlled blood pressure may have been misclassified. Therefore, potential outcome misclassification cannot be fully excluded. Meanwhile, because the final analytical cohort was restricted to participants with complete fasting laboratory data and complete covariate information, potential selection bias cannot be fully excluded. Participants included in the fasting subsample or complete-case analysis may differ from excluded participants in demographic characteristics, metabolic status, health behaviors, or comorbidity profiles. Therefore, although NHANES sampling weights were applied to improve national representativeness, the findings should be interpreted as representative of the eligible fasting analytical population rather than the entire original NHANES population. Finally, although we have made additional adjustments to the physical activity situation, residual confounding factors related to cardiopulmonary function and muscle strength still cannot be completely excluded.
Conclusions
The study identified a significant association between higher METS-VF levels and an increased risk of hypertension among US adults, which remained robust across multiple confounders. Compared with traditional obesity indicators, METS-VF has a better predictive value for hypertension, suggesting that it may serve as a practical and accessible marker for hypertension risk stratification in population-based settings.
Acknowledgements
We are grateful to all the organizations and people who provided support and assistance for this study.
Abbreviations
- VAT
Visceral adipose tissue
- IR
Insulin resistance
- MetS
Metabolic syndrome
- MET
Metabolic equivalent of task
- T2D
Type 2 diabetes
- MRI
Magnetic resonance imaging
- CT
Computed tomography
- DXA
Dual-energy-X-ray absorptiometry
- METS-VF
Metabolic score for visceral fat
- METS-IR
Metabolic score for insulin resistance
- GPAQ
Global physical activity questionnaire
- WHtR
Waist-to-height ratio
- HF
Heart failure
- CKD
Chronic kidney disease
- PIR
Poverty-income ratio
- CVD
Cardiovascular disease
- CHD
Coronary heart disease
- SBP
Systolic blood pressure
- DBP
Diastolic blood pressure
- BMI
Body mass index
- TG
Triglycerides
- TC
Total cholesterol
- HDL-C
High-density lipoprotein cholesterol
- LDL-C
Low-density lipoprotein cholesterol
- CARB
Carbohydrate
- TFAT
Total fat
- SFAT
Total saturated fatty acids
- MFAT
Total monounsaturated fatty acids
- PFAT
Total polyunsaturated fatty acids
- PA
Physical activity
- CHARLS
China Health and Retirement Longitudinal Study
- CI
Confidence interval
- OR
Odds ratio
- SD
Standard deviations
- ROC
Receiver operating characteristic
- AUC
Area under the curve
- NHANES
National Health and Nutrition Examination Survey
- CDC
Centers for Disease Control and Prevention
- RAAS
Renin–angiotensin–aldosterone system
Author contributions
ZG and YW: Responsible for the research’s major methodology and formal analysis. ZZ, YT, and XG contributed to data curation and validation. HC, HX, LG and SW contributed to the review and editing of the manuscript. All authors participated in the final manuscript.
Funding
The work was supported by Jilin Provincial Department of Education [Grant Number: JJKH20250188KJ] and Jilin Province Higher Education Research Project [Grant Number: JGJX24D0050].
Data availability
The datasets analyzed in this study are publicly available. The NHANES data can be obtained from the National Health and Nutrition Examination Survey website (https://www.cdc.gov/nchs/nhanes/), and the CHARLS data used for external validation can be obtained from the China Health and Retirement Longitudinal Study website (https://charls.charlsdata.com/). All data were de-identified before public release.
Declarations
Competing interests
The authors declare no competing interests.
Ethical approval
This study was based on publicly available, de-identified data from the National Health and Nutrition Examination Survey (NHANES) and the China Health and Retirement Longitudinal Study (CHARLS). NHANES is conducted by the National Center for Health Statistics (NCHS), and all NHANES study protocols were approved by the NCHS Ethics Review Board. Written informed consent was obtained from all NHANES participants. The external validation analysis was conducted using the 2011 baseline data from CHARLS. The CHARLS study was approved by the Biomedical Ethics Review Committee of Peking University (IRB00001052-11015), and written informed consent was obtained from all participants in the original survey. Because the present study involved secondary analysis of publicly available and de-identified data, no additional ethical approval was required. All methods were performed in accordance with relevant guidelines and regulations.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Caballero, B. Humans against obesity: Who will win?. Adv. Nutr.10, S4–S9 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Després, J.-P. et al. Abdominal obesity and the metabolic syndrome: Contribution to global cardiometabolic risk. Arterioscler. Thromb. Vasc. Biol.28, 1039–1049 (2008). [DOI] [PubMed] [Google Scholar]
- 3.Berg, A. H. & Scherer, P. E. Adipose tissue, inflammation, and cardiovascular disease. Circ. Res.96, 939–949 (2005). [DOI] [PubMed] [Google Scholar]
- 4.Moerland, M. A., De Leeuw, A. A., Van Der Der Voort Van Zyp, J. R. & Jürgenliemk-Schulz, I. M. University Medical Center Utrecht, Utrecht, The Netherlands. In Emerging Technologies in Brachytherapy, Series: Series in Medical Physics and Biomedical Engineering (eds Song, W. Y. et al.) 285–294 (CRC Press Taylor & Francis Group, Boca Raton, 2017). 10.1201/9781315120966-20. [Google Scholar]
- 5.Bello-Chavolla, O. Y. et al. Metabolic score for visceral fat (METS-VF), a novel estimator of intra-abdominal fat content and cardio-metabolic health. Clin. Nutr.39, 1613–1621 (2020). [DOI] [PubMed] [Google Scholar]
- 6.Zhao, M., Bovet, P. & Xi, B. Weight status change from adolescence to young adulthood and the risk of hypertension and diabetes mellitus. Hypertension76, 583–588 (2020). [DOI] [PubMed] [Google Scholar]
- 7.Cesaro, A. et al. Visceral adipose tissue and residual cardiovascular risk: A pathological link and new therapeutic options. Front. Cardiovasc. Med.10, 1187735 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Potenza, M. A., Addabbo, F. & Montagnani, M. Vascular actions of insulin with implications for endothelial dysfunction. Am. J. Physiol. Endocrinol. Metab.10.1152/ajpendo.00297.2009 (2009). [DOI] [PubMed] [Google Scholar]
- 9.McEniery, C. M. et al. Nondiabetic glucometabolic status and progression of aortic stiffness: The Whitehall II study. Diabetes Care40, 599–606 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Anfossi, G., Russo, I., Doronzo, G. & Trovati, M. Contribution of insulin resistance to vascular dysfunction. Arch. Physiol. Biochem.115, 199–217 (2009). [DOI] [PubMed] [Google Scholar]
- 11.Feng, Y. et al. Metabolic score for visceral fat: A reliable indicator of visceral obesity for predicting risk for hypertension. Nutrition93, 111443 (2022). [DOI] [PubMed] [Google Scholar]
- 12.Ji, L. et al. Association of metabolic score for visceral fat (METS-VF) with gout risk in patients with hypertension and hyperuricemia: A multicenter study based on the Chinese population. Diabetes Metab. Syndr. Obes.18, 761–779 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Fan, Y. et al. The impact of marital status on hypertension: SEM analysis of a Chinese general population cohort. BMC Public Health25, 2981 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Chaturvedi, A., Zhu, A., Gadela, N. V., Prabhakaran, D. & Jafar, T. H. Social determinants of health and disparities in hypertension and cardiovascular diseases. Hypertension81, 387–399 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Metlock, F. E. et al. Impact of social determinants of health on hypertension outcomes: A systematic review. Hypertension81, 1675–1700 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wang, Z. et al. Association of visceral and total body fat with hypertension and prehypertension in a middle-aged chinese population. J. Hypertens.33, 1555–1562 (2015). [DOI] [PubMed] [Google Scholar]
- 17.Navarro-Prado, S., Schmidt-RioValle, J., Montero-Alonso, M. A., Fernández-Aparicio, Á. & González-Jiménez, E. Stricter adherence to dietary approaches to stop hypertension (DASH) and its association with lower blood pressure, visceral fat, and waist circumference in university students. Nutrients12, 740 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Asghari, G. et al. Dietary approaches to stop hypertension (DASH) dietary pattern is associated with reduced incidence of metabolic syndrome in children and adolescents. J. Pediatr.174, 178-184.e1 (2016). [DOI] [PubMed] [Google Scholar]
- 19.Torun, C. Prediction of visceral adipose tissue magnitude using a new model based on simple clinical measurements. [DOI] [PMC free article] [PubMed]
- 20.Huang, J.-C. et al. Exploring the relationship between visceral fat and coronary artery calcification risk using metabolic score for visceral fat (METS-VF). Life14, 1399 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zhu, Y. et al. Associations between metabolic score for visceral fat and the risk of cardiovascular disease and all-cause mortality among populations with different glucose tolerance statuses. Diabetes Res. Clin. Pract.203, 110842 (2023). [DOI] [PubMed] [Google Scholar]
- 22.Yang, R. et al. Assessing the usefulness of a newly proposed metabolic score for visceral fat in predicting future diabetes: Results from the NAGALA cohort study. Front. Endocrinol.14, 1172323 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Liu, Q., Cui, H., Si, F., Wu, Y. & Yu, J. Association of cumulative exposure to metabolic score for visceral fat with the risk of cardiovascular disease and all-cause mortality: A prospective cohort study. J. Cachexia Sarcopenia Muscle16, e13702 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Qian, J., Huang, G. & Mao, Y. Association between metabolic score of visceral fat and carotid atherosclerosis in Chinese health screening population: A cross-sectional study. BMC Public Health24, 1723 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Yu, P., Meng, X., Kan, R., Wang, Z. & Yu, X. Association between metabolic scores for visceral fat and chronic kidney disease: A cross-sectional study. Front. Endocrinol.13, 1052736 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Aroor, A. R., Whaley‐Connell, A. & Sowers, J. R. Utility of obesity and metabolic dyslipidemia (a non‐insulin based determinate of the metabolic syndrome and insulin resistance) in predicting arterial stiffness. J. Clin. Hypertens.21, 1071–1074 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.DeMarco, V. G., Aroor, A. R. & Sowers, J. R. The pathophysiology of hypertension in patients with obesity. Nat. Rev. Endocrinol.10, 364–376 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Lastra, G. & Manrique, C. Perivascular adipose tissue, inflammation and insulin resistance: Link to vascular dysfunction and cardiovascular disease. Horm. Mol. Biol. Clin. Investig.22, 19–26 (2015). [DOI] [PubMed] [Google Scholar]
- 29.Gao, T. et al. Causality of visceral adipose tissue on chronic kidney disease and renal function measure indicators, and mediation role of hypertension: Mendelian randomization study. Obes. Rev.26, e13845 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Koenen, M., Hill, M. A., Cohen, P. & Sowers, J. R. Obesity, adipose tissue and vascular dysfunction. Circ. Res.128, 951–968 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Kralova Lesna, I. et al. Cardiovascular disease predictors and adipose tissue macrophage polarization: Is there a link?. Eur. J. Prev. Cardiol.25, 328–334 (2018). [DOI] [PubMed] [Google Scholar]
- 32.Hall, J. E., Do Carmo, J. M., Da Silva, A. A., Wang, Z. & Hall, M. E. Obesity-induced hypertension: Interaction of neurohumoral and renal mechanisms. Circ. Res.116, 991–1006 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.He, F. J., Li, J. & MacGregor, G. A. Effect of longer term modest salt reduction on blood pressure: Cochrane systematic review and meta-analysis of randomised trials. BMJ346, f1325–f1325 (2013). [DOI] [PubMed] [Google Scholar]
- 34.Hall, W. L. Dietary saturated and unsaturated fats as determinants of blood pressure and vascular function. Nutr. Res. Rev.22, 18–38 (2009). [DOI] [PubMed] [Google Scholar]
- 35.Guzik, T. J., Nosalski, R., Maffia, P. & Drummond, G. R. Immune and inflammatory mechanisms in hypertension. Nat. Rev. Cardiol.21, 396–416 (2024). [DOI] [PubMed] [Google Scholar]
- 36.U-Din, M. et al. Characteristics of abdominal visceral adipose tissue, metabolic health and the gut microbiome in adults. J. Clin. Endocrinol. Metab.109, 680–690 (2024). [DOI] [PubMed] [Google Scholar]
- 37.Kubota, S. & Yabe, D. Visceral adipose tissue quality and its impact on metabolic health. J. Clin. Endocrinol. Metab.109, e1665–e1666 (2024). [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets analyzed in this study are publicly available. The NHANES data can be obtained from the National Health and Nutrition Examination Survey website (https://www.cdc.gov/nchs/nhanes/), and the CHARLS data used for external validation can be obtained from the China Health and Retirement Longitudinal Study website (https://charls.charlsdata.com/). All data were de-identified before public release.