Associations of Arm‐to‐Waist Ratio With Hypertension and Blood Pressure: The Role of Peripheral‐Central Body Size Distribution in U.S. Individuals Authors: Zhang Fang, Zhanyang Zhou, Jian Zhang, Yuqing Zhang Published: 2026 ABSTRACT ABSTRACTThe arm‐to‐waist ratio (AWR) reflects peripheral‐central body size distribution. This study investigated the associations of AWR with hypertension and blood pressure. A total of 50,656 participants from NHANES 1999–2018 were included. Pearson correlation analyses were performed to evaluate the correlations among anthropometric indices and blood pressure. We utilized multivariable logistic regression for hypertension and linear regression for systolic and diastolic blood pressure (SBP and DBP). Overall, 15,211 (30%) individuals had hypertension. The prevalence of hypertension decreased progressively across increasing quartiles of AWR (47.5%, 32.2%, 23.6%, and 16.9%, respectively; P for trend < 0.001). AWR was correlated with BMI (r = −0.16), WC (r = −0.41), WHtR (r = −0.46), and AC (r = 0.16; all P < 0.001). Each 1% increment in AWR was associated with lower odds of hypertension (adjusted odds ratio [OR] 0.97, 95% CI 0.96–0.97; P < 0.001), whereas BMI, WC, WHtR, and AC were all positively associated with hypertension (all P < 0.001). Significant interactions were observed between AWR and both age and sex (both P for interaction < 0.001). After multivariable adjustment, nonlinear associations were observed between AWR and hypertension (men: P for overall < 0.001, P for nonlinearity < 0.001; women: P for overall = 0.051, P for nonlinearity = 0.021). Among those without prior hypertension, increased AWR was independently associated with lower SBP and DBP (both P < 0.001). AWR is independently and inversely associated with hypertension and blood pressure in U.S. individuals, exhibiting notable sex differences. Our study highlights the potential importance of peripheral‐central body size distribution in hypertension development.Keywords: arm‐to‐waist ratio, anthropometric indices, blood pressure, body composition, hypertension Abbreviations arm circumference arm‐to‐waist ratio body mass index confidence interval diastolic blood pressure odds ratio systolic blood pressure waist circumference waist‐to‐height ratio Introduction Obesity is a well‐established risk factor for hypertension [1]. However, individuals with comparable levels of overall adiposity often exhibit markedly different cardiometabolic risks, underscoring the heterogeneous nature of obesity [2]. Increasing evidence suggests that body fat distribution, rather than obesity severity alone, plays a central role in hypertension development [3, 4, 5]. A pooled analysis including more than 7 million participants demonstrated that both general and abdominal adiposity were independently associated with hypertension across diverse populations [1]. Accordingly, anthropometric measures reflecting body fat distribution, such as waist circumference (WC) and waist‐to‐height ratio (WHtR), have gained increasing attention because they provide additional information beyond body mass index (BMI) for hypertension risk assessment [3, 4, 5]. Arm circumference (AC) is a measure of peripheral body composition reflecting skeletal muscle mass, subcutaneous adipose tissue, and nutritional reserve [6]. Previous studies have shown that larger AC was associated with a more favorable metabolic profile independent of BMI and WC [6, 7]. However, existing anthropometric indices share a fundamental limitation: they characterize either central adiposity or peripheral body size in isolation and therefore cannot capture the relative balance between these two physiologically distinct compartments. This distinction matters because the cardiometabolic consequences of a given WC may differ substantially depending on how much peripheral lean tissue and subcutaneous fat buffer the effects of central adiposity. The arm‐to‐waist ratio (AWR), calculated as AC divided by WC, integrates peripheral body size and central adiposity into a single anthropometric index. Compared with BMI, which does not distinguish fat mass from lean mass, and WC or WHtR, which primarily characterize abdominal adiposity, AWR captures whether an individual's peripheral body reserves are proportionate to their central fat accumulation. Evidence showed higher AWR was associated with better metabolism and prognosis in individuals with diabetes [8]. A recently proposed limb‐to‐waist ratio was inversely associated with hypertension and other cardiometabolic outcomes [9], providing indirect support for the clinical relevance of AWR. However, AWR has not been systematically evaluated as the primary exposure in a dedicated hypertension study. Therefore, this study aimed to examine the associations of AWR with hypertension prevalence and blood pressure in U.S. individuals. Methods Study Design Data were derived from the National Health and Nutrition Examination Survey (NHANES) cycles spanning 1999–2018. Conducted by the National Center for Health Statistics under the Centers for Disease Control and Prevention, NHANES is a biennial survey designed to monitor health and nutritional status and identify disease risk factors in the U.S. population. This report adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. Population A total of 110,578 participants from NHANES 1999–2018 were initially screened. Participants under 16 years of age (n = 40,760), pregnant women (n = 1,892) and participants with missing blood pressure or anthropometric measurements (n = 17,270) were excluded (Figure 1). Study flowchart. Measurement of Body Composition Indices Anthropometric measurements were obtained by trained health technicians during standardized physical examinations at NHANES mobile examination centers, following the NHANES Anthropometry Procedures Manual. Body weight was measured in kilograms and standing height in centimeters. BMI was calculated as weight divided by height squared (kg/m2). WC (cm) was measured at the high point of the iliac crest at minimal respiration, to the nearest 0.1 cm. AC (cm) was measured on the right arm at the midpoint between the posterior border of the acromion process and the tip of the olecranon process, with the arm hanging freely at the side, to the nearest 0.1 cm. WHtR was calculated as (WC/height) × 100%. AWR was calculated as AC divided by WC, multiplied by 100 and expressed as a percentage (AC/WC × 100%). Hypertension Definition and Blood Pressure Measurement Hypertension was defined based on self‐reported physician diagnosis. Specifically, participants who answered “Yes” to the questionnaire item (“Have you ever been told by a doctor or other health professional that you have hypertension?”) were classified as having hypertension. This approach prioritizes clinically recognized, prevalent disease over single‐visit physiological measurements, aligning with survey‐based epidemiological objectives. Regarding blood pressure levels, while three measurements were taken after a 5‐min rest, we utilized the first systolic and diastolic readings to maximize sample size, given that subsequent measurements had higher rates of missing data due to participant dropout. For blood pressure analyses, nonhypertensive individuals were defined as those without a self‐reported diagnosis of hypertension and without current use of antihypertensive medications. Covariates Demographic data, including key covariates (age, gender, race, educational attainment, and income level), were collected via questionnaire. Educational attainment comprised five categories: <9th grade, 9th‐12th grade (non‐graduate), high school graduate, some college/associate degree, and bachelor's or higher. Income was indexed by the family income‐to‐poverty ratio, where 1 denotes the poverty line and higher values indicate greater affluence. Smoking status (never, former, or current) was determined based on self‐reported cumulative cigarette consumption and current use. Alcohol consumption was defined as consuming ≥12 drinks in the past year. Laboratory tests included lipid profiles and plasma protein levels. Information on eight self‐reported comorbidities was collected. Only chronic kidney disease was included as a covariate. Stroke, coronary heart disease, and myocardial infarction are established downstream cardiovascular consequences of both hypertension and adiposity. Diabetes shares substantial etiological overlap with central adiposity and hypertension, and its inclusion alongside anthropometric indices would risk over‐adjustment of the adiposity‐hypertension pathway. Cancer and chronic bronchitis have no established direct causal relationship with hypertension and were therefore excluded to maintain model parsimony. Statistical Analyses Continuous variables are presented as means±standard deviations (SDs), and categorical variables as numbers (percentages). Comparisons across AWR quartiles were performed using appropriate trend tests. Missing values in covariates were imputed using multiple imputation by chained equations, generating five imputed datasets. The estimates from each dataset were combined according to Rubin's rules [10]. The correlation between variables following a normal distribution was analyzed using Pearson correlation analysis. Logistic regression models were constructed to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for the association between anthropometric indices and hypertension. To further investigate the relationships between anthropometric indices and blood pressure independent of antihypertensive treatment, we performed multivariable linear regression analyses among nonhypertensive participants. To address potential non‐linear associations, restricted cubic splines (RCS) with four knots (placed at the 5th, 35th, 65th, and 95th percentiles) were utilized [11]. Potential interactions between AWR and prespecified variables, including age, sex, and BMI, were assessed by introducing multiplicative interaction terms into the regression models. All statistical analyses were conducted using R software (version 4.1.2), with statistical significance defined as a two‐sided P < 0.05. Results Population Characteristics A total of 50,656 participants were finally included, with a mean age of 45.2 ± 19.9 years, and 50.5% were men (Table 1). Overall, 15,211 (30.0%) participants had hypertension. Individuals in the higher AWR quartile were younger, more likely to be male, and had higher educational attainment and serum albumin levels than those in the lowest quartile (all P for trend < 0.001). In contrast, participants with lower AWR showed a less favorable cardiometabolic profile, including higher concentrations of total cholesterol and triglycerides, and a higher prevalence of diabetes mellitus, stroke, coronary artery disease, myocardial infarction, chronic bronchitis, cancer, and chronic kidney disease (all P for trend < 0.001). TABLE 1Baseline characteristics of participants according to quartiles of arm‐to‐waist ratio.Characteristics ALL (N = 50656) Q1 (N = 12664) Q2 (N = 12664) Q3 (N = 12664) Q4 (N = 12664) P for trendAge45.2 (19.9)59.0 (18.3)47.4 (18.6)40.4 (17.4)33.9 (15.8)< 0.001Gender:< 0.001Men25571 (50.5%)6129 (48.4%)5837 (46.1%)6060 (47.9%)7545 (59.6%)Women25085 (49.5%)6535 (51.6%)6827 (53.9%)6604 (52.1%)5119 (40.4%)Race:< 0.001Mexican American9118 (18.0%)2285 (18.0%)2526 (19.9%)2443 (19.3%)1864 (14.7%)Other Hispanic4054 (8.00%)993 (7.84%)1079 (8.52%)1052 (8.31%)930 (7.34%)Non‐Hispanic White21769 (43.0%)6708 (53.0%)5544 (43.8%)5114 (40.4%)4403 (34.8%)Non‐Hispanic Black11073 (21.9%)1747 (13.8%)2275 (18.0%)2760 (21.8%)4291 (33.9%)Others4642 (9.16%)931 (7.35%)1240 (9.79%)1295 (10.2%)1176 (9.29%)Education:< 0.001Less than nineth grade5327 (10.5%)1963 (15.5%)1425 (11.3%)1142 (9.02%)797 (6.29%)9–11th grade7630 (15.1%)1895 (15.0%)1893 (14.9%)1921 (15.2%)1921 (15.2%)High school graduate11771 (23.2%)3054 (24.1%)2908 (23.0%)2868 (22.6%)2941 (23.2%)Some college14907 (29.4%)3351 (26.5%)3607 (28.5%)3819 (30.2%)4130 (32.6%)College and above11021 (21.8%)2401 (19.0%)2831 (22.4%)2914 (23.0%)2875 (22.7%)Income‐to‐poverty ratio2.50 (1.62)2.38 (1.57)2.54 (1.64)2.55 (1.64)2.54 (1.64)< 0.001Serum albumin, g/L42.6 (3.34)41.6 (3.31)42.4 (3.23)43.0 (3.21)43.5 (3.33)< 0.001Total cholesterol, mmol/L4.94 (1.10)5.04 (1.15)5.06 (1.10)4.93 (1.08)4.73 (1.03)< 0.001Triglycerides, mmol/L1.62 (1.46)1.87 (1.53)1.74 (1.60)1.57 (1.51)1.30 (1.11)< 0.001Smoking status:< 0.001Never27839 (55.0%)6057 (47.8%)6859 (54.2%)7271 (57.4%)7652 (60.4%)Former11282 (22.3%)4112 (32.5%)2999 (23.7%)2354 (18.6%)1817 (14.3%)Current11535 (22.8%)2495 (19.7%)2806 (22.2%)3039 (24.0%)3195 (25.2%)Alcohol consumption:< 0.001No14406 (28.4%)3976 (31.4%)3695 (29.2%)3559 (28.1%)3176 (25.1%)Yes36250 (71.6%)8688 (68.6%)8969 (70.8%)9105 (71.9%)9488 (74.9%)Diabetes mellitus:< 0.001No45439 (89.7%)10227 (80.8%)11242 (88.8%)11823 (93.4%)12147 (95.9%)Yes5217 (10.3%)2437 (19.2%)1422 (11.2%)841 (6.64%)517 (4.08%)Stroke:< 0.001No49175 (97.1%)11916 (94.1%)12309 (97.2%)12425 (98.1%)12525 (98.9%)Yes1481 (2.92%)748 (5.91%)355 (2.80%)239 (1.89%)139 (1.10%)Coronary artery disease:< 0.001No48189 (95.1%)11393 (90.0%)12043 (95.1%)12300 (97.1%)12453 (98.3%)Yes2467 (4.87%)1271 (10.0%)621 (4.90%)364 (2.87%)211 (1.67%)Myocardial infarction:< 0.001No48791 (96.3%)11668 (92.1%)12183 (96.2%)12422 (98.1%)12518 (98.8%)Yes1865 (3.68%)996 (7.86%)481 (3.80%)242 (1.91%)146 (1.15%)Chronic bronchitis:< 0.001No48798 (96.3%)11741 (92.7%)12237 (96.6%)12367 (97.7%)12453 (98.3%)Yes1858 (3.67%)923 (7.29%)427 (3.37%)297 (2.35%)211 (1.67%)Cancer:< 0.001No46507 (91.8%)10776 (85.1%)11532 (91.1%)11955 (94.4%)12244 (96.7%)Yes4149 (8.19%)1888 (14.9%)1132 (8.94%)709 (5.60%)420 (3.32%)Chronic kidney disease:< 0.001No49347 (97.4%)12107 (95.6%)12358 (97.6%)12431 (98.2%)12451 (98.3%)Yes1309 (2.58%)557 (4.40%)306 (2.42%)233 (1.84%)213 (1.68%)Hypertension:< 0.001No35445 (70.0%)6654 (52.5%)8587 (67.8%)9678 (76.4%)10526 (83.1%)Yes15211 (30.0%)6010 (47.5%)4077 (32.2%)2986 (23.6%)2138 (16.9%)SBP, mmHg124 (19.3)131 (21.7)125 (19.6)121 (17.5)119 (15.5)< 0.001DBP, mmHg69.7 (13.3)69.7 (14.7)70.9 (13.1)69.8 (12.6)68.3 (12.7)< 0.001AWR, %34.0 (3.31)30.0 (1.47)32.9 (0.61)34.9 (0.63)38.3 (1.97)< 0.001BMI, kg/m2 28.3 (6.62)29.8 (7.01)28.6 (6.55)27.8 (6.44)26.9 (6.09)< 0.001WC, cm96.7 (16.5)106 (16.5)98.9 (15.4)93.8 (14.8)87.7 (13.4)< 0.001WHtR, %57.8 (9.95)63.9 (9.49)59.2 (8.98)56.0 (8.79)52.0 (8.49)< 0.001AC, cm32.6 (5.09)31.8 (4.92)32.5 (5.05)32.8 (5.15)33.6 (5.09)< 0.001 Note: Data are presented as mean (standard deviation) for continuous variables and n (%) for categorical variables. Trend across quartiles was assessed using the Jonckheere‐Terpstra test for continuous variables and the Cochran‐Armitage trend test for categorical variables. AWR was calculated as (AC / WC) ×100%.Abbreviations: AC, arm circumference; AWR, arm‐to‐waist ratio; BMI, body mass index; DBP, diastolic blood pressure; SBP, systolic blood pressure; WC, waist circumference; WHtR, waist‐to‐height ratio; Baseline characteristics of participants according to quartiles of arm‐to‐waist ratio. ALL (N = 50656) Q1 (N = 12664) Q2 (N = 12664) Q3 (N = 12664) Q4 (N = 12664) Note: Data are presented as mean (standard deviation) for continuous variables and n (%) for categorical variables. Trend across quartiles was assessed using the Jonckheere‐Terpstra test for continuous variables and the Cochran‐Armitage trend test for categorical variables. AWR was calculated as (AC / WC) ×100%. Abbreviations: AC, arm circumference; AWR, arm‐to‐waist ratio; BMI, body mass index; DBP, diastolic blood pressure; SBP, systolic blood pressure; WC, waist circumference; WHtR, waist‐to‐height ratio; The prevalence of hypertension decreased progressively across increasing quartiles of AWR, from 47.5% in Q1 to 32.2% in Q2, 23.6% in Q3, and 16.9% in Q4 (P for trend < 0.001; Table 1). Similarly, systolic blood pressure (SBP) declined steadily across AWR quartiles (131, 125, 121, and 119 mmHg for Q1–Q4, respectively), and diastolic blood pressure (DBP) was overall lower in participants with higher AWR (69.7, 70.9, 69.8, and 68.3 mmHg, respectively; P for trend < 0.001). Regarding anthropometric indices, BMI, WC, and WHtR decreased progressively with increasing AWR, whereas AC increased across quartiles (31.8, 32.5, 32.8, and 33.6 cm for Q1–Q4, respectively; p < 0.001). Correlations Among Anthropometric Indices Correlations among anthropometric indices were calculated in the overall population, and correlations with blood pressure were assessed only among participants without self‐reported hypertension to avoid confounding by antihypertensive treatment (Figure 2). AWR was weakly correlated with BMI (r = −0.16; P < 0.001) and AC (r = 0.16; P < 0.001), and was moderately negatively correlated with WC (r = −0.41; P < 0.001) and WHtR (r = −0.46; P < 0.001). In contrast, BMI, WC, WHtR, and AC were positively correlated with each other (P < 0.001). AWR showed weaker correlations with conventional obesity indices compared with the correlations among BMI, WC, WHtR, and AC (Figure 2). Regarding blood pressure, conventional anthropometric indices (BMI, WC, WHtR, and AC) were positively associated with SBP and DBP, whereas AWR showed a weak negative association with both SBP (r = −0.19; P < 0.001) and DBP (r = −0.08; P < 0.001). Pearson correlation matrix among anthropometric indices and blood pressure Pearson correlation coefficients among anthropometric indices are presented. Color intensity and numbers indicate the strength and direction of the correlations; ***, P < 0.001. Anthropometric indices were calculated based on overall population, and correlations involving blood pressure levels were analyzed exclusively in participants without a self‐reported history of hypertension to avoid confounding by antihypertensive treatment. AC, arm circumference; AWR, arm‐to‐waist ratio; BMI, body mass index; WC, waist circumference; WHtR, waist‐to‐height ratio; SBP, systolic blood pressure; DBP, diastolic blood pressure. Associations Between Anthropometric Indices and Hypertension In the unadjusted model (Model 1; Table 2), AWR was inversely associated with hypertension, with each 1% increase in AWR associated with a 17% lower odds of hypertension (OR, 0.83; 95% CI, 0.83–0.84; P < 0.001). After adjustment for demographic factors, including age, sex, race, education, and income (Model 2; Table 2), the inverse association remained significant (OR, 0.95; 95% CI, 0.95–0.96; P < 0.001). This association persisted after further adjustment for alcohol consumption, smoking, serum albumin, total cholesterol, triglycerides, and chronic kidney disease (Model 3; Table 2), with each 1% increase in AWR associated with a 3% lower odds of hypertension (OR, 0.97; 95% CI, 0.96–0.97; P < 0.001). TABLE 2Associations of body anthropometric indices with hypertension.CharacteristicModel 1Model 2Model 3OR (95% CI) p‐valueOR (95% CI) p‐valueOR (95% CI) p‐valueAWR, %0.83 (0.83–0.84)< 0.0010.95 (0.95–0.96)< 0.0010.97 (0.96–0.97)< 0.001BMI, kg/m2 1.08 (1.08–1.09)< 0.0011.09 (1.08–1.09)< 0.0011.08 (1.08–1.09)< 0.001WC, cm1.04 (1.04–1.05)< 0.0011.04 (1.04–1.04)< 0.0011.04 (1.04–1.04)< 0.001WHtR, %1.08 (1.08–1.08)< 0.0011.06 (1.06–1.06)< 0.0011.06 (1.06–1.06)< 0.001AC, cm1.08 (1.08–1.09)< 0.0011.12 (1.11–1.12)< 0.0011.11 (1.11–1.12)< 0.001Abbreviations: CI, confidence interval; OR, odds ratio.Model 1: Unadjusted analysis;.Model 2: Adjusted for age, sex, race, education, income;.Model 3: Further adjusted for alcohol consumption, smoking, serum albumin, total cholesterol, triglycerides, and chronic kidney disease. Associations of body anthropometric indices with hypertension. Abbreviations: CI, confidence interval; OR, odds ratio. Model 1: Unadjusted analysis;. Model 2: Adjusted for age, sex, race, education, income;. Model 3: Further adjusted for alcohol consumption, smoking, serum albumin, total cholesterol, triglycerides, and chronic kidney disease. In contrast, BMI, WC, WHtR, and AC were all positively associated with hypertension across all models. In the fully adjusted model (Model 3), the ORs per one‐unit increase were 1.08 (95% CI, 1.08–1.09) for BMI, 1.04 (95% CI, 1.04–1.04) for WC, 1.06 (95% CI, 1.06–1.06) for WHtR, and 1.11 (95% CI, 1.11–1.12) for AC (all P < 0.001). Interaction and Subgroup Analyses A significant interaction was observed between AWR and prespecified variables, including age and sex (both P for interaction < 0.001). Compared with participants aged ≥65 years, the inverse association of AWR was more pronounced among those aged <65 years. Similar trends were observed across sex strata. In contrast, BMI, WC, WHtR, and AC consistently showed positive associations with hypertension in all subgroups. Among the five anthropometric indices examined, AWR was the only anthropometric index showing a consistent inverse association with hypertension across age subgroups and in men; in women, the association was attenuated and not statistically significant (Figure 3). Associations of anthropometric indices with hypertension across age and sex subgroups. P for interaction values are shown for each stratification variable and represent the significance of the multiplicative interaction term between the anthropometric index and the respective subgroup variable. A significant interaction between AWR and BMI was observed (P for interaction < 0.001). Among participants with BMI ≤25 kg/m2, each 1% increment in AWR was associated with lower odds of hypertension (OR 0.98, 95% CI 0.96–0.998; p = 0.027). Significant inverse associations were similarly observed among those with BMI 25–30 kg/m2 (OR 0.98, 95% CI 0.96–0.99; P < 0.001) and BMI 30–35 kg/m2 (OR 0.97, 95% CI 0.96–0.99; p = 0.002). Among participants with BMI ≥35 kg/m2, the association was not statistically significant (OR 1.00, 95% CI 0.98–1.02; p = 0.957). Dose–Response Relationship Between AWR and Hypertension RCS analyses were performed to explore the dose‐response relationship between AWR and hypertension. After adjustment for age, race, education, income, alcohol consumption, smoking, serum albumin, total cholesterol, triglycerides, and chronic kidney disease, significant nonlinearity between AWR and hypertension was observed (P for overall < 0.001, P for nonlinearity < 0.001), whereas the association differed by sex (men: P for overall < 0.001, P for nonlinearity < 0.001; women: P for overall = 0.051, P for nonlinearity = 0.021; Figure 4). The inverse association was more evident among men, whereas the association in women was weaker and of borderline statistical significance. Restricted cubic spline analysis of the association between AWR and hypertension in men (A) and women (B). Adjusted for age, race, education, income, alcohol consumption, smoking, serum albumin, total cholesterol, triglycerides, and chronic kidney disease. The solid line represents the adjusted odds ratio (OR), and the shaded area represents the 95% confidence interval (CI). The horizontal dashed line indicates an OR of 1.0. Associations of Anthropometric Indices With Blood Pressure Among Population With no Prior Diagnosis of Hypertension Among individuals without previously diagnosed hypertension, 1% increment in AWR was associated with lower SBP (β = −0.16, 95% CI −0.21 to −0.11; P < 0.001) and DBP (β = −0.08, 95% CI −0.12 to −0.04; P < 0.001) after adjustment for potential confounders. This relationship was consistent between males and females, except for diastolic blood pressure where no significant difference was observed in females (Figure 5). In contrast, BMI, WC, WHtR, and AC were all positively associated with both SBP and DBP (Figure 5). Associations of anthropometric indices with systolic and diastolic blood pressure among nonhypertensive individuals. A. For systolic blood pressure; B. For diastolic blood pressure. Adjusted regression coefficients (β) and 95% CIs for the associations. Analyses were performed among individuals without a self‐reported hypertension diagnosis (N = 35,445). Each anthropometric index was examined in a separate regression model. Independent Associations of AWR Beyond BMI Analysis was performed by further adjusting for BMI in addition to the covariate‐adjusted model (Model 3, Table 3). In the overall population, AWR was associated with lower odds of hypertension (OR 0.98, 95% CI 0.97–0.99; P < 0.001). Among participants without prior hypertension diagnosis, AWR showed a significant inverse association with SBP (β = −0.10, 95% CI −0.15 to −0.05; P < 0.001) but no significant association with DBP (β = −0.03, 95% CI −0.08 to 0.01; p = 0.122). TABLE 3Multivariable associations of AWR and BMI with hypertension and blood pressure.OutcomeCharacteristicOR or β (95% CI) p‐valueHypertensionAWR, %0.98 (0.97–0.99)< 0.001BMI, kg/m2 1.08 (1.08–1.09)< 0.001SBPAWR, %−0.10 (−0.15 to –0.05)< 0.001BMI, kg/m2 0.31 (0.29–0.34)< 0.001DBPAWR, %−0.03 (−0.08–0.01)0.122BMI, kg/m2 0.25 (0.22–0.27)< 0.001 Note: Multivariable models were adjusted for potential confounders listed in model 3, table 2. BMI was included in the model simultaneously with AWR. Associations with hypertension were assessed in all participants (N = 50,656). Associations with SBP and DBP were assessed among individuals without a self‐reported diagnosis of hypertension or antihypertensive medication use (N = 35,445). Multivariable associations of AWR and BMI with hypertension and blood pressure. Note: Multivariable models were adjusted for potential confounders listed in model 3, table 2. BMI was included in the model simultaneously with AWR. Associations with hypertension were assessed in all participants (N = 50,656). Associations with SBP and DBP were assessed among individuals without a self‐reported diagnosis of hypertension or antihypertensive medication use (N = 35,445). Discussion Our study demonstrated that AWR exhibited a distinct relationship with hypertension compared with conventional anthropometric indices. Unlike BMI, WC, WHtR, and AC, all of which were positively associated with hypertension, higher AWR was consistently associated with lower hypertension prevalence. This association differed by sex, with a more pronounced nonlinear relationship in men than in women. Among participants without diagnosed hypertension, higher AWR remained independently associated with lower systolic blood pressure even after adjustment for BMI, suggesting that AWR captures information beyond overall adiposity. Obesity is one of the most important modifiable risk factors for hypertension, but anthropometric indices differ substantially in their ability to characterize body composition and fat distribution. BMI cannot differentiate adipose tissue from lean mass or capture regional fat distribution. In contrast, WC and WHtR more directly reflect central adiposity and are more closely related to visceral fat accumulation and cardiometabolic risk [3, 5, 12, 13]. A pooled analysis of 837 population‐based studies involving more than 7.5 million participants confirmed that both general and abdominal adiposity are strongly associated with hypertension across diverse world regions [1], highlighting the importance of body fat distribution in cardiovascular risk assessment. Consistent with these observations, BMI, WC, and WHtR were all positively associated with hypertension in the present study. Unlike conventional anthropometric indices, AWR was inversely associated with hypertension, although both AC and WC were individually associated with higher hypertension odds. AWR appears to capture the relative balance between peripheral body size and central adiposity rather than reflecting the absolute size of either body compartment [5]. This interpretation is supported by the correlation analyses, which showed only weak correlations between AWR and BMI and moderate correlations with WC and WHtR, whereas conventional obesity indices were strongly intercorrelated. These findings indicate that AWR represents a distinct anthropometric construct that is not simply another measure of overall obesity. The biological basis underlying this association may be multifactorial. A lower AWR is primarily driven by a disproportionately larger WC relative to AC. Given that WC is closely correlated with visceral adiposity, individuals with lower AWR may tend to have greater visceral fat accumulation, which has been linked in other studies to insulin resistance, chronic low‐grade inflammation, sympathetic nervous system activation, and renin‐angiotensin‐aldosterone system stimulation, all of which are implicated in blood pressure elevation [14]. Conversely, AC reflects not only peripheral subcutaneous fat but also skeletal muscle mass, the principal site of insulin‐mediated glucose uptake [6]. A relatively larger AC in relation to WC may therefore reflect preserved peripheral tissue mass together with lower central adiposity and represent a more favorable metabolic phenotype. Consistent with this interpretation, prior observational data have shown that larger AC is independently associated with lower insulin resistance after adjustment for BMI and WC [7], suggesting that the peripheral component of AWR may carry metabolically favorable information. Collectively, these observations raise the possibility that AWR captures aspects of body composition distribution relevant to blood pressure regulation, though the mechanistic pathways require confirmation in longitudinal and experimental studies. Notably, AWR and BMI showed a significant interaction on hypertension risk, indicating that the association between AWR and hypertension varies according to obesity status. In severe obesity, the adverse cardiometabolic consequences of excessive adiposity may partially overwhelm the positive influence associated with a favorable peripheral‐to‐central body size distribution [15]. We also observed marked sex differences in the dose‐response relationship. In men, hypertension risk declined substantially once AWR exceeded approximately 32%, whereas no significant overall association was identified in women, whose dose‐response curve exhibited a U‐shaped pattern. This sex‐specific pattern is biologically plausible given the well‐recognized differences in fat distribution between men and women [16]. Men generally accumulate more visceral adipose tissue, whereas women tend to store a greater proportion of fat subcutaneously in peripheral regions [16]. Consequently, a higher AWR in women may not necessarily reflect a more favorable lean‐to‐central body composition because AC includes a larger proportion of subcutaneous fat relative to skeletal muscle. Nevertheless, the U‐shaped association observed in women should be interpreted cautiously because the overall association did not reach statistical significance. Several limitations of this study should be acknowledged. First, hypertension was ascertained by self‐reported physician diagnosis, which may be subject to recall bias and does not capture individuals with undiagnosed hypertension. Consequently, some participants classified as nonhypertensive may have had unrecognized elevated blood pressure, potentially attenuating the observed associations with blood pressure levels. Second, AWR is an indirect anthropometric measure and cannot distinguish skeletal muscle mass from subcutaneous adipose tissue. Third, although we adjusted for a broad range of potential confounders, residual confounding and selection bias resulting from complete‐case analysis cannot be completely excluded. Forth, because of the cross‐sectional design of NHANES, causal relationships between AWR and hypertension cannot be established. Future prospective studies should determine whether AWR improves cardiovascular risk prediction and whether interventions targeting central adiposity while preserving peripheral lean mass can modify AWR and reduce hypertension incidence. Finally, survey sampling weights were not incorporated in the present analyses; therefore, results reflect associations within the study sample rather than population‐level weighted estimates. Conclusion Higher AWR was independently associated with lower hypertension prevalence and lower blood pressure, with evidence of sex‐specific differences. As a simple and readily obtainable anthropometric measure, AWR may provide complementary information beyond traditional obesity indices for hypertension risk assessment. Author Contributions ZJ and ZYQ: conceptualization, review, and editing. FZ and ZZY: data analysis and interpretation, and draft writing. Funding National Natural Science Foundation of China (Grant No. 82470412). Ethics Statement The NHANES protocol was approved by the National Center for Health Statistics Research Ethics Review Board. This study involved secondary analysis of publicly available, de‐identified data and was exempted from further review. Consent Written informed consent was obtained from all original NHANES participants by NCHS. No additional consent was required for this study. Conflicts of Interest The authors declare no conflicts of interest. Acknowledgments We thank the National Center for Health Statistics of the Centers for Disease Control and Prevention for the design and administration of the National Health and Nutrition Examination Survey. Contributor Information Jian Zhang, Email: fwzhangjian62@126.com. Yuqing Zhang, Email: yqzhang9988@163.com. Data Availability Statement All data is publicly available (https://wwwn.cdc.gov/nchs/nhanes/ and https://www.cdc.gov/nchs/data‐linkage/mortality.htm). References Associated Data Data Availability Statement All data is publicly available (https://wwwn.cdc.gov/nchs/nhanes/ and https://www.cdc.gov/nchs/data‐linkage/mortality.htm).