CLINICAL PERSPECTIVE
WHAT IS NEW?
- In patients with normal weight (defined as a body mass index of 18.5–24.9 kg/m2), individuals with higher waist circumference had a higher prevalence of hypertension and a less favorable cardiometabolic risk profile.
- In the fully adjusted model, waist circumference was positively correlated with hypertension with an odds ratio (95% confidence interval) of 1.27 (1.12–1.44) in normal-weight patients.
WHAT ARE THE CLINICAL IMPLICATIONS?
- Our study highlighted that waist circumference is a useful anthropometric index in normal-weight individuals, and it should be measured regardless of body mass index.
- A more in-depth understanding of waist circumference could assist cardiologists to improve and manage cardiometabolic risk profiles, while neglecting waist circumference may lead to failure to counsel certain patients regarding their high-risk obesity phenotype.
- Measuring waist circumference may improve the evaluation of the risk of hypertension and help to manage cardiometabolic risk in normal-weight individuals.
Introduction
Obesity has been a worldwide health problem over the past 4 decades, with an estimated age-standardized prevalence of 30% in males and 35% in females.[1,2] Body mass index (BMI) is the most commonly used body measurement parameter for defining obesity. It is associated with multiple cardiovascular diseases in a “dose-dependent” manner, with a J-shaped curve.[3,4] However, recent studies have revealed that BMI-defined obesity is a heterogeneous condition in which body fat distribution is closely correlated with metabolic perturbations.[5,6]
The failure of BMI to fully capture abdominal adiposity is a major limitation, as BMI only provides a crude measure of general obesity without taking into account the regional body fat distribution, especially in individuals with normal BMI (18.5–24.9 kg/m2).[7,8] In contrast, waist circumference is strongly related to the absolute amount of abdominal fat.[9] Increased waist circumference is significantly associated with a higher risk of cardiovascular diseases and related death, with or without adjusting for BMI.[10–13] Accumulating evidence also supports the superiority of waist circumference over BMI regarding evaluating cardiovascular risk in both males and females.[14]
Recent consensus statements have highlighted waist circumference as a vital sign to assess abdominal adiposity and estimate cardiovascular risk alongside BMI in clinical practice.[15–17] However, it is generally recommended to measure waist circumference only in overweight or obese individuals, and little is known about the role of waist circumference in the normal-weight population. This study investigated the influence of waist circumference on the prevalence of hypertension and cardiometabolic dysregulation in normal-weight individuals.
Methods
Study population
All the data used in this study were publicly available from the National Health and Nutrition Examination Survey (NHANES) website, which contains health and nutrition information on representative samples of the noninstitutionalized civilian US population in 2-year cycles. We examined data from 5 consecutive NHANES cycles from 2009 to 2018 and included non-pregnant participants aged 20 to 79 years with normal weight (defined as a BMI of 18.5–24.9 kg/m2) who underwent physical and laboratory examinations.[18,19] In addition, we excluded the participants who had missing data on body measurements, blood pressure, cardiometabolic risk factors, dietary interview, standard biochemistry profile, or family poverty-income ratio (PIR). Our study population comprised a total of 8795 individuals. National Center for Health Statistics Research Ethics Review Board has approved this study, and informed consent was obtained from all participants.
Waist circumference measurement
To measure waist circumference, a trained examiner would stand on the right side of a participant, palpate the hip area, locate the right ilium of the pelvis, and mark the most superior lateral border. Subsequently, the examiner would extend a steel tape around the waist at the mark level and measure the waist circumference directly against the skin after the subject exhaled 1 normal breath. A detailed description is provided in the NHANES anthropometry procedures manual.
Hypertension definition
Blood pressure was measured by certified examiners using a mercury sphygmomanometer following the standardized protocol of the American Heart Association. After resting in a seated position for 5 minutes, 3 consecutive blood pressure measurements were obtained and averaged to determine the mean blood pressure. Hypertension was defined as (1) mean systolic blood pressure ≥130 mmHg or mean diastolic blood pressure ≥80 mmHg, (2) self-reported hypertension, and/or (3) self-reported use of anti-hypertensive medications.[20–22]
Demographic variables
Demographic variables included age, sex, race/ethnicity, education, and income. The self-reported race/ethnicity was categorized as non-Hispanic white, non-Hispanic black, Mexican American, other Hispanic, and Other. We used an established PIR formula (PIR = family income/federal poverty level) to evaluate household poverty. PIR was categorized as <1.33, 1.33–<3.50, and ≥3.50, following the qualification criterion for the US federal Supplemental Nutrition Assistance Program.[23]
Cardiometabolic risk factors
We included body measurements, history of cardiovascular diseases, chronic kidney disease (CKD), lipid profiles, blood glucose, total energy intake, sodium intake, physical activity, alcohol consumption, and smoking status as cardiometabolic risk factors. BMI was calculated as weight in kilograms divided by the square of height in meters (kg/m2). We used serum total and high-density lipoprotein (HDL) cholesterol levels to calculate the total-to-HDL cholesterol ratio, and ≥5.9 was defined as high cholesterol.[24] Diabetes was defined as fasting plasma glucose ≥126 mg/dL, hemoglobin A1c ≥6.5%, and/or self-reported diabetes. Behavioral risk factors included smoking status (smoked ≥100 cigarettes in a lifetime or not), alcohol consumption (≥12 alcoholic drinks/year or not), sodium intake, total energy intake, and insufficient leisure-time physical activity. Physical activity was defined as self-reported moderate- or vigorous-intensity leisure time activity, and physical activity <150 minutes/week was considered insufficient. The estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation,[25] and CKD was defined as eGFR ≤60 mL/(min·1.73 m2). Participants were considered to have cardiovascular diseases if they were ever diagnosed with angina pectoris, heart attack, coronary heart disease, congestive heart failure, and/or stroke. Methods and protocols for the questionnaires, laboratory tests, and examination are detailed on the NHANES website.
Statistical analysis
Normally distributed continuous variables are presented as mean ± standard deviation, and non-normally distributed continuous variables are presented as median (Q1, Q3), based on the Kolmogorov-Smirnov normality test. Categorical variables are presented as frequencies with percentages. The demographic characteristics and cardiometabolic risk factors were compared among waist circumference quartiles by one-way analysis of variance (normally distributed continuous variables), the Kruskal-Wallis test (non-normally distributed continuous variables), or the chi-square test (categorical variables). We imputed missing covariates using multivariate multiple imputation strategies (based on 5 replications), to maximize the statistical power and minimize the selection bias.[26,27]
We used boxplots to illustrate the distribution of waist circumference across 10 BMI categories (deciles) by sex, and the waist circumference distributions are also shown based on kernel density estimation with Gaussian kernels. Moreover, we used multivariate logistic regression to evaluate the association of waist circumference with hypertension, and the odds ratios (ORs) with 95% confidence intervals (CIs) were calculated accordingly. In the fully adjusted model, we adjusted for BMI, age categories (in decades), sex, race/ethnicity, diabetes, weight, height, smoking status, alcohol consumption, education level, PIR level, total-to-HDL cholesterol, triglyceride level, energy intake, sodium intake, and physical activity. Furthermore, we illustrated the associations between waist circumference and hypertension by sex using a restricted cubic spline with 5 knots (located at the following percentiles: 5%, 27.5%, 50%, 72.5%, and 95%).[28] The median waist circumference (85.0 cm for males and 80.6 cm for females) was used as the reference. Subsequently, in a sensitivity analysis, we assessed the associations in different subgroups, including sex (male or female), age (<45 or ≥45 years), total-to-HDL cholesterol ratio (below median or above/equal to median), physical activity (yes or no), and energy intake (below median or above/equal to median) categories.
Furthermore, we used the random forest supervised machine learning method, as well as least absolute shrinkage and selection operator (LASSO) regression,[29] to select hypertension-related features with high importance among the demographic variables, body measurements, comorbidities, and health behavior variables. In the random forest method, mean decrease accuracy and mean decrease Gini index are fundamental outcomes. For each variable, the higher values indicate a more important role in classifying the data. To be more specific, mean decrease accuracy indicates the decrease in model accuracy when excluding each variable. The loss in accuracy suggests the importance of the excluded variable for successful classification. The mean decrease Gini index quantifies the contribution of each variable to the homogeneity of the nodes and leaves in the random forest. Similar to mean decrease accuracy, a higher mean decrease Gini index indicates the importance of the variable in the model.
The variables with the top 12 mean decrease accuracy or Gini index were further screened by LASSO regression, with an optimal value for the penalization coefficient lambda.[30] Thereafter, the identified variables were used to create a logistic regression-based model to predict hypertension in individuals with normal weight. All individuals were randomly divided into a training set and a testing set at a ratio of 7:3, and no participant was in more than one dataset. We used the training set to create a multivariate logistic regression model, whereas the testing set was used to evaluate the model performance. The performance was evaluated based on the area under the curve (AUC), sensitivity, specificity, positive predictive value, and negative predictive value. Finally, we created a simple-to-use nomogram website based on the proposed model. All statistical analysis was performed in R software (version 3.6.1). P < 0.05 was considered statistically significant.
Results
Characteristics of participants
Although all the participants had normal BMI (22.6 (21.2–23.9) kg/m2), their waist circumferences ranged from 61.3 to 117.6 cm, with a median (Q1, Q3) of 82.5 (77.4, 87.8) cm [Figure 1A]. Hypertension was observed in 2963 (33.7%) individuals. Table 1 summarizes the demographic characteristics by waist circumference quartiles, and Supplementary Table 1 (https://links.lww.com/CD9/A16) summarizes the cardiometabolic risk factors by waist circumference quartiles. The normal-weight individuals with high waist circumference were older and had a lower education level, and more were males, compared to those with low waist circumference. However, no significant difference was observed in the PIR level. Moreover, individuals with higher waist circumference had a higher prevalence of hypertension and a less favorable cardiometabolic risk profile. They were also more likely to have elevated blood pressure, high total-to-HDL cholesterol ratio and triglyceride level, poor blood glucose control, and more self-reported cardiovascular diseases. The distribution of waist circumference stratified by sex is shown in Figure 1B. The age-adjusted mean waist circumference was 85.0 cm in males and 80.8 cm in females. Waist circumference was positively and significantly correlated with systolic and diastolic blood pressure, with Spearman rank correlation coefficients of 0.29 and 0.16, respectively.
;)
Figure 1: Distribution of waist circumference in normal-weight individuals stratified by (A) BMI categories and (B) sex. BMI: Body mass index.
Table 1 -
Demographic characteristics by waist circumference quartiles.
| Variables |
Q1 [61.3,77.4] (n = 2200) |
Q2 (77.4, 82.5] (n = 2245) |
Q3 (82.5, 87.8] (n = 2152) |
Q4 (87.8, 118.0] (n = 2198) |
P
|
| Age (years), median (Q1, Q3) |
29.0 (22.0, 43.0) |
35.0 (25.0, 50.0) |
43.0 (30.0, 59.0) |
55.0 (41.0, 67.0) |
<0.001 |
| Sex (Female/Male), n (%) |
1478/722 (67.2/32.8) |
1309/936 (58.3/41.7) |
1045/1107 (48.6/51.4) |
704/1494 (32.0/68.0) |
<0.001 |
| Race/ethnicity, n (%) |
<0.001 |
| Non-Hispanic White |
811 (36.9) |
933 (41.6) |
954 (44.3) |
1131 (51.5) |
|
| Non-Hispanic Black |
547 (24.9) |
412 (18.4) |
336 (15.6) |
337 (15.3) |
|
| Mexican American |
212 (9.6) |
267 (11.9) |
266 (12.4) |
247 (11.2) |
|
| Other Hispanic |
176 (8.0) |
187 (8.3) |
203 (9.4) |
151 (6.9) |
|
| Other |
454 (20.6) |
446 (19.9) |
393 (18.3) |
332 (15.1) |
|
| PIR level, n (%) |
0.222 |
| <1.33 |
663 (30.1) |
679 (30.2) |
592 (27.5) |
612 (27.8) |
|
| 1.33–<3.5 |
682 (31.0) |
717 (31.9) |
705 (32.8) |
736 (33.5) |
|
| ≥3.5 |
855 (38.9) |
849 (37.8) |
855 (39.7) |
850 (38.7) |
|
| Education, n (%) |
<0.001 |
| Below high school |
360 (16.4) |
432 (19.2) |
474 (22.0) |
555 (25.3) |
|
| High school |
454 (20.6) |
484 (21.6) |
468 (21.7) |
491 (22.3) |
|
| Above high school |
1386 (63.0) |
1329 (59.2) |
1210 (56.2) |
1152 (52.4) |
|
PIR: Poverty-income ratio.
Association of waist circumference with hypertension
When analyzed as a continuous variable, waist circumference was positively and significantly associated with hypertension in the non-adjusted, minimally adjusted, and fully adjusted models, with ORs (95% CI) of 2.28 (2.14–2.44), 1.27 (1.12–1.44), and 1.27 (1.12–1.44), respectively [Table 2]. When fully adjusted for BMI, age categories (in decades), sex, race/ethnicity, diabetes, weight, height, smoking status, alcohol consumption, education level, PIR level, total-to-HDL cholesterol, triglyceride level, energy intake, sodium intake, and physical activity, individuals in the highest quartile had a 3.87-fold increased risk of hypertension compared to those in the lowest quartile. Figure 2A–B visualize the relationship between waist circumference and hypertension using restricted cubic splines in males and females, respectively, and we observed a consistent and significant association after adjusting for BMI, race/ethnicity, education level, diabetes, smoking status, alcohol consumption, PIR level, total-to-HDL cholesterol, triglyceride level, energy intake, sodium intake, and physical activity.
Table 2 -
Association of waist circumference with hypertension using logistic regression models.
|
Non-adjusted model |
Minimally adjusted model |
Fully adjusted model |
|
|
|
|
| Items |
Odds ratio (95% CI) |
P
|
Odds ratio (95% CI) |
P
|
Odds ratio (95% CI) |
P
|
| Waist circumference (per 10 cm) |
2.28 (2.14–2.44) |
<0.001 |
1.27 (1.12–1.44) |
<0.001 |
1.27 (1.121.44) |
<0.001 |
| Categories |
| Q1 [61.3,77.4] |
Reference |
|
Reference |
|
Reference |
|
| Q2 (77.4–82.5] |
1.60 (1.39–1.84) |
<0.001 |
1.29 (1.08–1.53) |
0.004 |
1.86 (0.82–4.65) |
0.157 |
| Q3 (82.5–87.8] |
2.56 (2.23–2.94) |
<0.001 |
1.54 (1.26–1.87) |
<0.001 |
2.40 (1.05–6.09) |
0.049 |
| Q4 (87.8–118.0] |
4.67 (4.08–5.35) |
<0.001 |
1.64 (1.29–2.07) |
<0.001 |
3.87 (1.59–10.34) |
0.004 |
Minimally adjusted model: adjusted for age (in decades), sex, race/ethnicity, diabetes, weight, height, smoking status, alcohol consumption, PIR level, total-to-HDL cholesterol, triglyceride level, energy intake, sodium intake, and physical activity. Fully adjusted model: adjusted for body mass index, age (in decades), sex, race/ethnicity, diabetes, weight, height, smoking status, alcohol consumption, PIR level, total-to-HDL cholesterol, triglyceride level, energy intake, sodium intake, and physical activity. CI: Confidence interval; HDL: High-density lipoprotein; PIR: Poverty-income ratio.
Figure 2: Restricted cubic spline plots of the association between waist circumference and hypertension in normal-weight (A) males and (B) females. The associations were adjusted for BMI, race/ethnicity, education level, diabetes, smoking status, alcohol consumption, PIR level, total-to-HDL cholesterol, triglyceride level, energy intake, sodium intake, and physical activity. The median waist circumference (85.0 cm for males and 80.6 cm for females) was used as the reference. BMI: Body mass index; CI: Confidence interval; HDL: High-density lipoprotein; OR: Odds ratio; PIR: Poverty-income ratio.
Sensitivity analysis
As shown in Figure 3, the positive association between waist circumference and prevalence of hypertension remained robust across sex (male or female), age (<45 or ≥45 years), total-to-HDL cholesterol ratio (below median or above/equal to median), and physical activity (yes or no) categories. Interestingly, in the low energy intake group, no significant hypertension-inducing effect was observed, with an OR (95% CI) of 1.11 (0.95–1.30), whereas the OR (95% CI) was 1.35 (1.14–1.59) in the subgroup with high energy intake.
Figure 3: Forest plot of sensitivity analysis of the associations between waist circumference and hypertension across sex (male or female), age (<45 or ≥45 years), total-to-HDL cholesterol ratio (below median or above/equal to median), physical activity (yes or no), and energy intake (below median or above/equal to median) categories. The associations were adjusted for BMI, age categories (in decades), sex, race/ethnicity, diabetes, weight, height, smoking status, alcohol consumption, PIR level, total-to-HDL cholesterol, triglyceride level, energy intake, sodium intake, and physical activity. Importantly, when analyzing each subgroup, we did not adjust for the specific variable, except age. BMI: Body mass index; CI: Confidence interval; HDL: High-density lipoprotein; OR: Odds ratio; PIR: Poverty-income ratio.
Predictive model creation and validation
Figure 4 summarizes the mean decrease accuracy or Gini index of the input variables, and the variables with the top 12 mean decrease accuracy or Gini index were selected for further screening (including age, waist circumference, weight, height, CKD, diabetes, BMI, race/ethnicity, energy intake, sodium intake, sex, coronary artery disease, PIR level, education level, and smoking status). Importantly, waist circumference was the second most critical feature (after age), with a mean decrease accuracy of 41 and a mean decrease Gini index of 437. The variables were further assessed using the LASSO method, with an optimal value for the penalization coefficient lambda. A total of 13 variables (age, waist circumference, weight, CKD, diabetes, BMI, race/ethnicity, sodium intake, sex, coronary artery disease, PIR level, education level, and smoking status) were finally selected to create the predictive model. The visualized nomogram can be easily accessed at https://data15651725761.shinyapps.io/Predict-hypertension-in-normal-weight-individuals/. In the testing set, the model performance was good, with an AUC of 0.803, sensitivity of 0.72, specificity of 0.76, positive predictive value of 0.61, and negative predictive value of 0.84 [Figure 5].
Figure 4: (A) Mean decrease accuracy and (B) mean decrease Gini index of the input variables when using the random forest supervised machine learning method. Variables are presented by descending importance. We inputted multiple variables (including demographic variables, body measurements, comorbidities, and health behavior variables) into the random forest model to select hypertension-related features. Mean decrease accuracy and mean decrease Gini index are fundamental outcomes of the random forest that are positively associated with the importance of the input variables. Importantly, waist circumference was the second most critical feature (after age), with a mean decrease accuracy of 41 and a mean decrease Gini index of 437. BMI: Body mass index; CAD: Coronary artery disease; CKD: Chronic kidney disease; PIR: Poverty-income ratio.
Figure 5: Receiver operating characteristic curves of the model performance in the (A) training set and (B) testing set. All individuals were randomly divided into a training set and testing set at a ratio of 7:3. We used the training set to create a multivariate logistic regression model, whereas the testing set was used to evaluate the model performance. The AUCs were 0.816 and 0.803 in the training set and testing set, respectively. AUC: Area under the curve.
Discussion
BMI is currently the most widely used anthropometric index to diagnose obesity in clinical practice. However, obesity is a heterogeneous condition in which fat distribution strongly influences metabolic perturbations and cardiovascular risk: visceral fat is more likely to lead to metabolic abnormalities than subcutaneous fat.[31,32] BMI alone is insufficient to identify differences in body fat content, lean mass, and fat distribution, failing to capture the full view of obesity-induced health risks. Therefore, applying the current BMI-based cut-points to a diverse population can presumably lead to misclassification, as individuals with normal BMI can have excessive adipose tissue accumulation. In contrast, waist circumference, as a recommended central obesity measure, is probably the best anthropometric index to evaluate the abdominal adipose tissue accumulation in both males and females.[9,33] Additionally, it has been reported that height is only marginally associated with waist circumference.[34]
Based on nationally representative US samples, we demonstrated that normal-weight abdominal obesity identified by waist circumference is a risk factor for hypertension and cardiometabolic dysregulation. Individuals in the highest quartile of waist circumference (87.8–118.0 cm), more than half of whom had hypertension, had a 3.87-fold increased risk compared to those in the lowest quartile (61.3–77.4 cm). When analyzed as a continuous variable, waist circumference was consistently and significantly associated with increased prevalence of hypertension in both males and females. In the further analysis using machine learning, waist circumference was also identified as the second most important variable (after age) among the many demographic variables, body measurements, comorbidities, and health behavior variables. Moreover, we observed elevated blood pressure, poorer lipid profiles, and worse blood glucose control in individuals with a high waist circumference. These results suggest that a specific fat distribution might be the clue to the “obesity paradox” (which is more of a “BMI paradox”).[3,4,35]
It has been argued that waist circumference and BMI are interchangeable adiposity indexes and that it is unnecessary to assess waist circumference once BMI has been measured. In this study, although all participants had normal weight, which is generally considered to be associated with a low risk of cardiovascular disease, the prevalence of hypertension was 33.7%. Also, these individuals who had various waist circumferences show distinct hypertension risks. Our data indicated that waist circumference and BMI are not interchangeable at the individual level. We highlighted the additional advantage of waist circumference beyond BMI for evaluating abdominal obesity-induced cardiovascular risk. Therefore, cardiologists should focus on the substantial individual variation in fat distribution among patients with similar BMI in clinical practice.
Normal weight but high waist circumference indicates a distinct metabolic phenotype characterized by low muscle mass, poor cardiorespiratory fitness, and an excess of intra-abdominal adipose tissue.[36,37] The mechanisms underlying the association between abdominal obesity and blood pressure are probably complex, with numerous environmental and genetic factors involved. It should be pointed out that the primary risk attributed to obesity is mediated by concomitant cardiometabolic abnormalities, including dyslipidemia, insulin resistance, diabetes, and inflammation.[35,38,39] Notably, among the many risk factors, adiposity is a well-established one.[40,41] Adipose tissue, once thought to be a simple energy warehouse, is now recognized to interact with systemic inflammatory responses, the sympathetic nervous system, and the renin-angiotensin system, facilitating obesity-related cardiometabolic dysregulation and contributing to the development of hypertension. Recent studies have pointed out that upregulation of the sympathetic nervous system is an important mechanism underlying obesity-related hypertension.[38,39,42,43] Also, leptin released by adipose tissue disturbs the hypothalamic neuronal systems, which can subsequently reduce food consumption, upregulate energy expenditure, and stimulate sympathetic activity.
The waist circumference cut-points for abdominal obesity are 102 cm for males and 88 cm for females, which were originally converted from the BMI threshold for obesity (30.0 kg/m2) as an alternative anthropometric index of obesity.[9] However, waist circumference is strongly linked to abdominal obesity-related risks beyond BMI. The current waist circumference cut-points do not consider the unique advantage of using waist circumference to evaluate the risk attributed to excessive abdominal fat accumulation.[44] A meta-analysis also suggested that the waist circumference cut-points of 102/88 cm might not be suitable to estimate the risk of hypertension.[45] Therefore, this study analyzed waist circumference, as both a continuous variable and in terms of quartiles, to attempt to capture a full view of the association between waist circumference and the risk of hypertension. A novel waist circumference threshold value should be proposed and validated based on abdominal obesity-related risks rather than BMI.[46]
Despite its advantage beyond BMI, waist circumference is not routinely measured in clinical practice and is generally recommended only for the overweight or obese population. Our study highlighted that waist circumference is a useful anthropometric index in normal-weight individuals, and it should be measured regardless of BMI. Additionally, our waist circumference-based predictive model exhibited good performance, with a high AUC and negative predictive value. A more in-depth understanding of waist circumference could assist cardiologists to improve and manage cardiometabolic risk profiles, while neglecting waist circumference may lead to failure to counsel certain patients regarding their high-risk obesity phenotype. Furthermore, therapeutic interventions to decrease waist circumference might reduce the risk of hypertension and cardiometabolic dysregulation.
Last but not least, it is important to understand that waist circumference should not replace BMI as the single adiposity index in cardiology, as combining them would provide a fuller view of obesity-related cardiovascular risk. It should now be recognized that BMI is to anthropometry what “total cholesterol” is to lipidology.[8]
Limitations
First, we did not distinguish between primary and secondary high blood pressure in this study. Abnormal fat distribution might have distinct roles in these 2 subtypes of hypertension. Second, body fat distribution differs considerably by ethnicity. This study involved nationally representative US samples, but representative data from other regions such as Europe and China were lacking. The generalizability of our results should be further validated. Third, although we controlled for many potential confounders, residual confounding and unknown confounders may still exist. Fourth, the cross-sectional nature of NHANES precluded longitudinal analysis, and we could not determine causality.
Conclusion
Measuring waist circumference may improve the evaluation of the risk of hypertension, and reducing waist circumference may improve cardiometabolic risk in normal-weight individuals.
Acknowledgments
We sincerely acknowledge the US National Center for Health Statistics for conducting the survey. Jinyu Sun sincerely acknowledges Dr. Jing Wang for her generous help during resident training.
Funding
This study was supported by the Postgraduate Research & Practice Innovation Program of Jiangsu Province (SJCX21_0626).
Author contributions
Jinyu Sun, Guozhen Sun, Yue Yuan, Wei Sun, and Xiangqing Kong conceived and designed the study; Jinyu Sun and Qiang Qu analyzed the data; Jinyu Sun, Qiang Qu, and Yue Yuan wrote the paper. All authors provided critical revisions of the manuscript and approved the final manuscript.
Conflicts of interest
None.
References
[1]. GBD 2015 Obesity Collaborators, Afshin A, Forouzanfar MH, et al. Health effects of overweight and obesity in 195 countries over 25 years. N Engl J Med 2017;377(1):13–27. doi: 10.1056/NEJMoa1614362.
[2]. Liu B, Du Y, Wu Y, et al. Trends in obesity and adiposity measures by race or ethnicity among adults in the United States 2011–18: population based study. BMJ 2021;372:n365. doi: 10.1136/bmj.n365.
[3]. Shihab HM, Meoni LA, Chu AY, et al. Body mass index and risk of incident hypertension over the life course: the Johns Hopkins Precursors Study. Circulation 2012;126(25):2983–2989. doi: 10.1161/CIRCULATIONAHA.112.117333.
[4]. Pandey A, Patel KV, Bahnson JL, et al. Association of intensive lifestyle intervention, fitness, and body mass index with risk of heart failure in overweight or obese adults with type 2 diabetes mellitus: an analysis from the look AHEAD trial. Circulation 2020;141(16):1295–1306. doi: 10.1161/CIRCULATIONAHA.119.044865.
[5]. Tchernof A, Després JP. Pathophysiology of human visceral obesity: an update. Physiol Rev 2013;93(1):359–404. doi: 10.1152/physrev.00033.2011.
[6]. Eckel N, Meidtner K, Kalle-Uhlmann T, et al. Metabolically healthy obesity and cardiovascular events: a systematic review and meta-analysis. Eur J Prev Cardiol 2016;23(9):956–966. doi: 10.1177/2047487315623884.
[7]. Franzosi MG. Should we continue to use BMI as a cardiovascular risk factor? Lancet 2006;368(9536):624–625. doi: 10.1016/S0140-6736(06)69222-2.
[8]. Poirier P. Adiposity and cardiovascular disease: are we using the right definition of obesity. Eur Heart J 2007;28(17):2047–2048. doi: 10.1093/eurheartj/ehm321.
[9]. Snijder MB, van Dam RM, Visser M, et al. What aspects of body fat are particularly hazardous and how do we measure them. Int J Epidemiol 2006;35(1):83–92. doi: 10.1093/ije/dyi253.
[10]. Cerhan JR, Moore SC, Jacobs EJ, et al. A pooled analysis of waist circumference and mortality in 650,000 adults. Mayo Clin Proc 2014;89(3):335–345. doi: 10.1016/j.mayocp.2013.11.011.
[11]. Katzmarzyk PT, Hu G, Cefalu WT, et al. The importance of waist circumference and BMI for mortality risk in diabetic adults. Diabetes Care 2013;36(10):3128–3130. doi: 10.2337/dc13-0219.
[12]. Chen Q, Li L, Yi J, et al. Waist circumference increases risk of coronary heart disease: evidence from a Mendelian randomization study. Mol Genet Genomic Med 2020;8(4):e1186. doi: 10.1002/mgg3.1186.
[13]. Levine DA, Calhoun DA, Prineas RJ, et al. Moderate waist circumference and hypertension prevalence: the REGARDS Study. Am J Hypertens 2011;24(4):482–488. doi: 10.1038/ajh.2010.258.
[14]. Lee CM, Huxley RR, Wildman RP, et al. Indices of abdominal obesity are better discriminators of cardiovascular risk factors than BMI: a meta-analysis. J Clin Epidemiol 2008;61(7):646–653. doi: 10.1016/j.jclinepi.2007.08.012.
[15]. Ross R, Neeland IJ, Yamashita S, et al. Waist circumference as a vital sign in clinical practice: a Consensus Statement from the IAS and ICCR Working Group on Visceral Obesity. Nat Rev Endocrinol 2020;16(3):177–189. doi: 10.1038/s41574-019-0310-7.
[16]. Klein S, Allison DB, Heymsfield SB, et al. Waist circumference and cardiometabolic risk: a consensus statement from shaping America's health: Association for Weight Management and Obesity Prevention; NAASO, the Obesity Society; the American Society for Nutrition; and the American Diabetes Association. Diabetes Care 2007;30(6):1647–1652. doi: 10.2337/dc07-9921.
[17]. Rao G, Powell-Wiley TM, Ancheta I, et al. Identification of obesity and cardiovascular risk in ethnically and racially diverse populations: a scientific statement from the American Heart Association. Circulation 2015;132(5):457–472. doi: 10.1161/CIR.0000000000000223.
[18]. Skinner AC, Perrin EM, Moss LA, et al. Cardiometabolic risks and severity of obesity in children and young adults. N Engl J Med 2015;373(14):1307–1317. doi: 10.1056/NEJMoa1502821.
[19]. Bhaskaran K, Dos-Santos-Silva I, Leon DA, et al. Association of BMI with overall and cause-specific mortality: a population-based cohort study of 3.6 million adults in the UK. Lancet Diabetes Endocrinol 2018;6(12):944–953. doi: 10.1016/S2213-8587(18)30288-2.
[20]. Liao S, Yao W, Cheang I, et al. Association between perfluoroalkyl acids and the prevalence of hypertension among US adults. Ecotoxicol Environ Saf 2020;196:110589. doi: 10.1016/j.ecoenv.2020.110589.
[21]. Bakris G, Ali W, Parati G. ACC/AHA versus ESC/ESH on hypertension guidelines: JACC guideline comparison. J Am Coll Cardiol 2019;73(23):3018–3026. doi: 10.1016/j.jacc.2019.03.507.
[22]. Whelton PK, Carey RM, Aronow WS, et al. 2017ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol 2018;71(19):e127–127e248. doi: 10.1016/j.jacc.2017.11.006.
[23]. Saydah SH, Siegel KR, Imperatore G, et al. The cardiometabolic risk profile of young adults with diabetes in the U.S. Diabetes Care 2019;42(10):1895–1902. doi: 10.2337/dc19-0707.
[24]. Christianson TJ, Bryant SC, Weymiller AJ, et al. A pen-and-paper coronary risk estimator for office use with patients with type 2 diabetes. Mayo Clin Proc 2006;81(5):632–636. doi: 10.4065/81.5.632.
[25]. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150(9):604–612. doi: 10.7326/0003-4819-150-9-200905050-00006.
[26]. Sterne JA, White IR, Carlin JB, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ 2009;338:b2393. doi: 10.1136/bmj.b2393.
[27]. Jakobsen JC, Gluud C, Wetterslev J, et al. When and how should multiple imputation be used for handling missing data in randomised clinical trials – a practical guide with flowcharts. BMC Med Res Methodol 2017;17(1):162. doi: 10.1186/s12874-017-0442-1.
[28]. Gauthier J, Wu QV, Gooley TA. Cubic splines to model relationships between continuous variables and outcomes: a guide for clinicians. Bone Marrow Transplant 2020;55(4):675–680. doi: 10.1038/s41409-019-0679-x.
[29]. Friedman J, Hastie T, Tibshirani R. Regularization paths for generalized linear models via coordinate descent. J Stat Softw 2010;33(1):1–22.
[30]. Liu F, Long Q, He H, et al. Combining the fecal immunochemical test with a logistic regression model for screening colorectal neoplasia. Front Pharmacol 2021;12:635481. doi: 10.3389/fphar.2021.635481.
[31]. Nakamura T, Tokunaga K, Shimomura I, et al. Contribution of visceral fat accumulation to the development of coronary artery disease in non-obese men. Atherosclerosis 1994;107(2):239–246. doi: 10.1016/0021-9150(94)90025-6.
[32]. Coutinho T, Goel K, Corrêa de Sá D, et al. Central obesity and survival in subjects with coronary artery disease: a systematic review of the literature and collaborative analysis with individual subject data. J Am Coll Cardiol 2011;57(19):1877–1886. doi: 10.1016/j.jacc.2010.11.058.
[33]. Pouliot MC, Després JP, Lemieux S, et al. Waist circumference and abdominal sagittal diameter: best simple anthropometric indexes of abdominal visceral adipose tissue accumulation and related cardiovascular risk in men and women. Am J Cardiol 1994;73(7):460–468. doi: 10.1016/0002-9149(94)90676-9.
[34]. Han TS, Seidell JC, Currall JE, et al. The influences of height and age on waist circumference as an index of adiposity in adults. Int J Obes Relat Metab Disord 1997;21(1):83–89. doi: 10.1038/sj.ijo.0800371.
[35]. Després JP. Excess visceral adipose tissue/ectopic fat the missing link in the obesity paradox. J Am Coll Cardiol 2011;57(19):1887–1889. doi: 10.1016/j.jacc.2010.10.063.
[36]. Despres JP. Body fat distribution and risk of cardiovascular disease: an update. Circulation 2012;126(10):1301–1313. doi: 10.1161/CIRCULATIONAHA.111.067264.
[37]. Murase T, Hattori T, Ohtake M, et al. Cardiac remodeling and diastolic dysfunction in DahlS.Z-Lepr(fa)/Lepr(fa) rats: a new animal model of metabolic syndrome. Hypertens Res 2012;35(2):186–193. doi: 10.1038/hr.2011.157.
[38]. Liu BX, Sun W, Kong XQ. Perirenal fat: a unique fat pad and potential target for cardiovascular disease. Angiology 2019;70(7):584–593. doi: 10.1177/0003319718799967.
[39]. Liu BX, Qiu M, Zong PY, et al. Distribution, morphological characterization, and resiniferatoxin-susceptibility of sensory neurons that innervate rat perirenal adipose tissue. Front Neuroanat 2019;13:29. doi: 10.3389/fnana.2019.00029.
[40]. Timpson NJ, Harbord R, Davey Smith G, et al. Does greater adiposity increase blood pressure and hypertension risk?: Mendelian randomization using the FTO/MC4R genotype. Hypertension 2009;54(1):84–90. doi: 10.1161/HYPERTENSIONAHA.109.130005.
[41]. Hall JE, do Carmo JM, da Silva AA, et al. Obesity-induced hypertension: interaction of neurohumoral and renal mechanisms. Circ Res 2015;116(6):991–1006. doi: 10.1161/CIRCRESAHA.116.305697.
[42]. Grassi G, Quarti-Trevano F, Seravalle G, et al. Sympathetic neural mechanisms underlying attended and unattended blood pressure measurement. Hypertension 2021;78(4):1126–1133. doi: 10.1161/HYPERTENSIONAHA.121.17657.
[43]. Hooper JS, Taylor-Clark TE. Irritant inhalation evokes P wave morphological changes in spontaneously hypertensive rats via reflex modulation of the autonomic nervous system. Front Physiol 2021;12:642299. doi: 10.3389/fphys.2021.642299.
[44]. Lemieux S, Prud’homme D, Bouchard C, et al. A single threshold value of waist girth identifies normal-weight and overweight subjects with excess visceral adipose tissue. Am J Clin Nutr 1996;64(5):685–693. doi: 10.1093/ajcn/64.5.685.
[45]. Seo DC, Choe S, Torabi MR. Is waist circumference ≥102/88 cm better than body mass index ≥30 to predict hypertension and diabetes development regardless of gender, age group, and race/ethnicity? Meta-analysis. Prev Med 2017;97:100–108. doi: 10.1016/j.ypmed.2017.01.012.
[46]. Zhu S, Wang Z, Heshka S, et al. Waist circumference and obesity-associated risk factors among whites in the third National Health and Nutrition Examination Survey: clinical action thresholds. Am J Clin Nutr 2002;76(4):743–749. doi: 10.1093/ajcn/76.4.743.
