Abstract
OBJECTIVE
Various organizations have highlighted the need to examine whether abdominal obesity cut points are appropriate for identification of cardiovascular risk among ethnic minority adults, particularly Hispanic/Latino adults living in Western societies. This study aimed 1) to establish optimal definitions for abdominal obesity among Hispanics/Latinos and 2) to determine the level of agreement between the presence of metabolic syndrome diagnosed by the current Joint Interim Statement (JIS) definition and an updated definition with optimal abdominal obesity cut points.
RESEARCH DESIGN AND METHODS
The sample included 16,289 adults who self-identified as Hispanic/Latino ages 18–74 years enrolled in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL). Receiver operating characteristic curves were used to derive sensitivity and specificity values. The largest sum of sensitivity plus specificity was used to determine appropriate cut points.
RESULTS
Among U.S. Hispanic/Latino adults, waist circumference cut points of >102 cm in men (in line with current JIS criteria) and >97 cm (9 points higher than JIS criteria) in women provide optimal discrimination for cardiovascular risk as judged by the presence of coronary heart disease. When using these cut points to create an updated metabolic syndrome definition among women, we found disagreement between our updated definition and the current JIS criteria. The prevalence of the metabolic syndrome was overestimated by ∼5 percentage points among women based on JIS criteria in comparison with our definition.
CONCLUSIONS
Our results suggest that the current recommendations for waist circumference cut points may not be appropriate for U.S. Hispanic/Latino women.
Introduction
The metabolic syndrome is a cluster of anthropometric, hemodynamic, and metabolic disturbances (1). The Joint Interim Statement (JIS) issued by several organizations established the criteria for clinical diagnosis of the metabolic syndrome in 2009 (2). Within this statement, central adiposity, elevated blood pressure, abnormal glucose regulation, elevated triglycerides, and lowered HDL cholesterol (HDL-C) were recognized as central components of the syndrome and recommendations regarding categorical cut points of each component were proposed. In addition to harmonization of the diagnostic criteria for the metabolic syndrome for use around the world, the need for population- and ethnicity-specific abdominal obesity cut points was emphasized. However, since the publication of this document, no major changes have been made to either the definition of the syndrome or the cut point outlined for specific ethnic groups.
Appropriate identification of patients with metabolic syndrome is crucial, particularly because the syndrome is widely and conveniently used in both clinical practice and research to “visualize” not only cardiovascular disease (CVD) risk but also risk for type 2 diabetes mellitus (T2DM) development (3,4). The metabolic syndrome is considered a useful tool for the prevention of T2DM. By virtue of a diagnosis of metabolic syndrome, patients might be motivated to actively engage in lifestyle modification and physicians can begin to implement targeted strategies to decrease T2DM risk (3). Furthermore, early and adequate identification of patients with metabolic syndrome among those already diagnosed with T2DM is also important because its timely management may help mitigate the known cardiovascular complications of T2DM (3). Both the prevention of T2DM and the mitigation of its associated complications are particularly important for individuals with known vulnerability, such as the Hispanic/Latino population (5).
It is now well established that pronounced differences exist in abdominal obesity across sex and ethnic groups. Hispanics/Latinos, in particular, have a markedly distinct prevalence of abdominal obesity compared with other ethnic populations (4). Furthermore, the prevalence of abdominal obesity is significantly greater in Hispanics/Latinos residing in the U.S. compared with those residing in their heritage countries. However, the prevalence of CVD among Hispanics/Latinos is not in concordance with these high abdominal obesity rates. Therefore, the 2013 American Heart Association/American College of Cardiology/The Obesity Society obesity guidelines highlighted the need to examine whether overall abdominal obesity cut points were appropriate for use in ethnic minorities within Western countries, particularly emphasizing the need for research among Hispanic Americans (6).
Previous groups attempted to provide optimal cut points for waist circumference among Hispanics/Latinos outside the U.S. However, the generalizability of these results is limited due to the inclusion of only one Hispanic/Latino heritage group (7–10) or the use of relatively small samples (8,11). Therefore, at the moment and until more specific data are available, the JIS recommended the use of the cutoff values proposed for South Asian populations, which defined abdominal obesity as waist circumference ≥90 cm in men and ≥80 cm in women for ethnic South and Central Americans regardless of country of residence (12). The use of waist circumference cut points applicable to populations that are likely to be genetically, clinically, and culturally different may result in the misclassification of risk among Hispanics/Latinos in the U.S. and around the world.
The Hispanic Community Health Study/Study of Latinos (HCHS/SOL) comprises the largest cohort of Hispanics/Latinos in the U.S. With the inclusion of >16,000 U.S. Hispanic/Latino adults, it provides an ideal opportunity to identify optimal cut points for waist circumference applicable to Hispanics/Latinos in the U.S. and perhaps even those residing in other countries. In addition, the inclusion of various Hispanic/Latino heritage groups, such as Central/South Americans, Cubans, Dominicans, Mexicans, and Puerto Ricans, enhances the generalizability of results. In this study, we take advantage of this large sample to 1) establish optimal definitions for abdominal obesity among Hispanic/Latino adults and 2) determine the level of agreement as to the presence of the metabolic syndrome between diagnosis by the current JIS definition and diagnosis by an updated definition with optimal abdominal obesity cut points.
Research Design and Methods
Study Sample
The HCHS/SOL is a population-based cohort study supported by the National Institutes of Health that aimed to characterize the health of U.S. Hispanics/Latinos (13). A total of 16,415 self-identified Hispanic/Latino adults between the ages of 18 and 74 years were recruited in four cities: Miami, FL; San Diego, CA; Chicago, IL; and the Bronx, NY. The study was designed to include Hispanics/Latinos of various backgrounds enrolling participants from Cuban, Dominican, Mexican, Puerto Rican, Central American, and South American descent (14) using a multistage, stratified, probabilistic sample design.
Census block groups were randomly selected in specified geographic areas of each study site (stratified by cross classification of high/low categories of % of Hispanic/Latino population and high/low socioeconomic status [SES] measured by % of population with at least a high school education), and households were randomly selected in each sample block group. Households were screened for eligibility, and self-identified Hispanic/Latino persons aged 18–74 years were selected in each household. The study was approved by institutional review boards at each participating institution, and all participants provided written informed consent. The baseline examination included interviewer-administered questionnaires in the participant’s language of preference, anthropometry, blood draw, and other measures. Further details on the design, sampling strategy, and baseline examination of the HCHS/SOL study have previously been published (13,14).
Of the 16,415 participants, only those with complete data on variables of interest (metabolic syndrome components and coronary heart disease [CHD]) were included in the analyses sample. The final analytic sample was comprised of 9,763 women and 6,526 men (total n = 16,289).
Measures
Metabolic Syndrome Components
Waist circumference was measured to the nearest 0.1 cm at the uppermost lateral border of the right ilium with a measuring tape. After 5 min in the seated position, systolic (SBP) and diastolic (DBP) blood pressure was measured three times at 1-min intervals using an automatic sphygmomanometer (model HEM-907XL; Omron Healthcare, Inc., Bannockburn, IL), and the average of the three readings was used. Measurements of HDL-C, triglycerides, and glucose were obtained from collected fasting blood samples. Blood samples were obtained following a nontraumatic venipuncture protocol. Fresh as well as frozen specimens were shipped to the HCHS/SOL central laboratory for assays and long-term storage. HDL-C was measured by a magnesium/dextran sulfate method, and plasma glucose was measured using a hexokinase enzymatic method (Roche Diagnostics, Indianapolis, IN). Triglycerides were measured in serum on a Roche Modular P chemistry analyzer, using a glycerol blanking enzymatic method (Roche Diagnostics). The assay methodologies and their performance are described in HCHS/SOL Manual 7 (15).
Metabolic Syndrome JIS Criteria
As specified in the JIS (2), participants were classified as having the metabolic syndrome if they met three or more of the following criteria: 1) waist circumference ≥102 cm in men and ≥88 cm in women; 2) triglycerides ≥150 mg/dL; 3) HDL-C <40 mg/dL in men and <50 mg/dL in women; 4) SBP ≥130 mmHg, DBP ≥85 mmHg, or on blood pressure medication; and 5) fasting glucose ≥100 mg/dL and/or on medication.
CHD
The presence of CHD was used as a marker of CVD. Each participant received a standard digital 12-lead electrocardiogram (GEMSIT MAC 1200 portable electrocardiograph), and readings were electronically transmitted to a central electrocardiogram reading center (the Epidemiological Cardiology Research Center [EPICARE] of Wake Forest School of Medicine). The Minnesota Code system of classification was used to ascertain possible old myocardial infarction. Self-reported information on angina, heart attack, and coronary procedures (angioplasty, stent, or bypass surgery to the arteries of the heart) was collected via standard questionnaire and interview. Prevalent CHD was specified as a dichotomous variable that combined information from electrocardiogram reports of possible old myocardial infarction as well as self-report of heart attack, coronary procedures, and angina.
Statistical Analysis
Preliminary statistical analyses included descriptive statistics and assessment of distributions. All analyses accounted for the complex sampling design with the use of sampling weights, probability, and cluster units. SPSS, version 22.0, was used for data preparation and descriptive analysis. SAS, version 9.3, was used for receiver operating characteristic (ROC) analyses. The t test was used to examine differences by sex in demographic and biological continuous variables. The χ2 test of independence was used to test differences on categorical variables. ROC analyses were constructed to identify optimal waist circumference cut points (16,17). Separate analyses were conducted for men and women. Sensitivity and specificity values (weighted) were examined to determine optimal waist circumference cut points for each sex. The waist circumference value (cm) resulting in the largest sum of sensitivity and specificity for the presence of CHD was selected as an optimal cut point. Sensitivity analysis were conducted with two alternative outcomes including CHD without the presence of angina and CHD with a history of stroke. The agreement between the presence of metabolic syndrome diagnosed by the JIS definition and an updated definition was calculated using the Cohen κ coefficient.
Results
Descriptive Characteristics of the Study Target Population
Approximately 7.4% of individuals identified as Central American, 20.0% as Cuban, 10.0% as Dominican, 37.4% as Mexican, 16.1% as Puerto Rican, and 5.0% as South American. Mean waist circumference was 98.2 cm for men and 96.6 cm for women. The percentage of individuals with prevalent CHD was 6.8% men and 5.4% women. Further descriptive characteristics are presented in Table 1.
Table 1.
Sociodemographic and health characteristics: HCHS/SOL (2008–2011)
| All (n = 16,289) | Men (n = 6,526) | Women (n = 9,763) | |
|---|---|---|---|
| Age, years | 41.0 (0.3) | 40.2 (0.3) | 41.8 (0.3) |
| Hispanic/Latino heritage group (%) | |||
| Central American | 7.4 | 7.3 | 7.5 |
| Cuban | 20.0 | 21.8 | 18.3 |
| Dominican | 10.0 | 8.2 | 11.6 |
| Mexican | 37.4 | 36.6 | 38.1 |
| Puerto Rican | 16.1 | 17.0 | 15.3 |
| South American | 5.0 | 4.8 | 5.2 |
| Other | 4.1 | 4.3 | 4.0 |
| GED or high school or higher (%) | 67.6 | 68.2 | 67.1 |
| Income (%)€ | |||
| <$10,000 | 14.6 | 11.5 | 17.6 |
| $10,000–$20,000 | 31.6 | 29.5 | 33.7 |
| $20,000–$40,000 | 33.3 | 34.4 | 32.2 |
| $40,000–$75,000 | 14.5 | 16.8 | 12.4 |
| >$75,000 | 5.9 | 7.9 | 4.0 |
| Current smoking (%) | 21.4 | 26.8 | 16.4 |
| BMI, kg/m2 | 29.7 (0.1) | 28.9 (0.1) | 29.8 (0.1) |
| Waist circumference, cm | 97.4 (0.2) | 98.2 (0.3) | 96.6 (0.3) |
| SBP, mmHg | 119.9 (0.3) | 123.4 (0.3) | 116.7 (0.3) |
| DBP, mmHg | 72.2 (0.2) | 73.5 (0.2) | 70.9 (0.2) |
| Total cholesterol, mg/dL | 194.3 (0.6) | 194.4 (0.8) | 194.2 (0.7) |
| HDL cholesterol, mg/dL | 48.5 (0.2) | 44.9 (0.2) | 51.9 (0.2) |
| LDL cholesterol, mg/dL | 119.7 (0.5) | 121.1 (0.7) | 118.5 (0.6) |
| Triglycerides, mg/dL | 133.1 (1.3) | 148.1 (2.3) | 119.4 (1.1) |
| Glucose, g/dL | 101.8 (0.4) | 104.3 (0.5) | 99.6 (0.5) |
| Stroke (%) | 2.3 | 2.4 | 2.2 |
| Diabetes (%) | 15.5 | 14.7 | 16.2 |
| CHD (%) | 6.0 | 6.8 | 5.4 |
Data are means (SE) unless otherwise indicated.
GED, general equivalence diploma.
Income data were available among only 14,927 participants.
Optimal Waist Circumference Cut Points
Table 2 presents sensitivity and specificity values across different waist circumference values for both men and women. The value with the largest sum of sensitivity and specificity was selected as optimal. According to this selection criterion, the optimal waist circumference value for men with a sensitivity of 50.6% and a specificity of 64.0% was 102 cm. This value is in line with the current recommendations for waist circumference in men outlined in the JIS metabolic syndrome definition.
Table 2.
Sensitivity and specificity of waist circumference measurements in association with prevalent CHD: HCHS/SOL (2008–2011)
| Sex | WC (cm) | Sensitivity (%) | Specificity (%) | Sum of sensitivity and specificity |
|---|---|---|---|---|
| Men (N = 6,526) | 95 | 72.8 | 39.4 | 112.2 |
| 96 | 69.7 | 42.6 | 112.3 | |
| 97 | 67.2 | 46.1 | 113.3 | |
| 98 | 64.8 | 49.6 | 114.4 | |
| 99 | 61.2 | 53.1 | 114.3 | |
| 100 | 56.4 | 57.0 | 113.4 | |
| 101 | 53.4 | 61.0 | 114.4 | |
| 102* | 50.6* | 64.0* | 114.6* | |
| 103 | 47.0 | 66.6 | 113.6 | |
| 104 | 42.5 | 69.3 | 111.8 | |
| 105 | 38.8 | 72.0 | 110.8 | |
| 106 | 37.5 | 74.3 | 111.8 | |
| 107 | 36.3 | 76.5 | 112.8 | |
| Women (N = 9,763) | 87 | 90.0 | 21.5 | 111.5 |
| 88 | 87.7 | 24.2 | 111.9 | |
| 89 | 85.5 | 27.0 | 112.5 | |
| 90 | 84.4 | 29.7 | 114.1 | |
| 91 | 81.9 | 33.0 | 114.9 | |
| 92 | 78.7 | 36.0 | 114.7 | |
| 93 | 76.4 | 39.0 | 115.4 | |
| 94 | 73.1 | 42.1 | 115.2 | |
| 95 | 70.4 | 45.2 | 115.6 | |
| 96 | 68.1 | 48.3 | 116.4 | |
| 97* | 65.4* | 51.3* | 116.7* | |
| 98 | 62.3 | 54.1 | 116.4 | |
| 99 | 58.5 | 57.1 | 115.6 | |
| 100 | 56.0 | 60.1 | 116.1 |
WC, waist circumference.
Value was chosen as cutpoint.
For women, however, the optimal waist circumference value was 97 cm, which yielded a sensitivity of 65.4% and a specificity of 51.3%. This value is higher than the current JIS recommendation for waist circumference in women, which is 88 cm. Our analyses indicated that in identification of the presence of CHD among Hispanic/Latino women, a cut point of 88 cm has a great level of sensitivity at 87.7% but significantly compromises specificity, which was found to be at only 24.2%.
Additional ROC analyses were conducted to identify optimal waist circumference cut points associated with other CVD outcomes. These outcomes included CHD without the presence of angina as well as CHD with a history of stroke. Sensitivity and specificity values for each outcome are presented in Supplementary Table 1. Results were consistent with our primary analyses and supported a cut point of 102 cm for Hispanic/Latino men and 97 cm for Hispanic/Latino women.
Prevalence of Metabolic Syndrome by Definition
Based on the previous results, and using a waist circumference cut point of 97 cm for women, we updated the metabolic syndrome criteria and estimated new age-adjusted prevalence estimates for metabolic syndrome in the HCHS/SOL target population. This updated definition generated lower prevalence estimates than the original JIS metabolic syndrome definition. The original JIS definition and our updated definition differed in classifying 5.1% of women in our overall sample. Furthermore, the Cohen κ coefficient, which measures the level of agreement between definitions, was 0.89, indicating some disagreement between the original JIS definition and our updated definition of the metabolic syndrome.
In addition to the overall results, we estimated prevalence estimates and level of agreement among women across the different Hispanic/Latino heritage groups (Table 3). The lowest level of agreement between definitions was among Dominicans and Mexicans (κ = 0.89), and the highest level of agreement was found among South Americans (κ = 0.90).
Table 3.
Age-standardized prevalence of the metabolic syndrome according to JIS and updated definition in women across Hispanic/Latino heritage groups
| Women | Men | |||
|---|---|---|---|---|
| JIS definition | Updated definition | κ statistic | JIS definition* | |
| Overall | 0.353 (0.338–0.367) | 0.301 (0.288–0.315) | 0.8859 | 0.330 (0.315–0.345) |
| Central American | 0.379 (0.348–0.409) | 0.317 (0.287–0.348) | 0.8696 | 0.321 (0.279–0.362) |
| Cuban | 0.337 (0.308–0.366) | 0.280 (0.254–0.306) | 0.8626 | 0.343 (0.316–0.369) |
| Dominican | 0.317 (0.283–0.350) | 0.255 (0.224–0.287) | 0.8551 | 0.289 (0.242–0.335) |
| Mexican | 0.355 (0.328–0.381) | 0.303 (0.278–0.328) | 0.8925 | 0.332 (0.307–0.356) |
| Puerto Rican | 0.405 (0.370–0.441) | 0.356 (0.322–0.390) | 0.8961 | 0.315 (0.277–0.352) |
| South American | 0.260 (0.221–0.298) | 0.227 (0.190–0.265) | 0.9088 | 0.265 (0.214–0.315) |
| Mix/other | 0.384 (0.289–0.479) | 0.361 (0.259–0.462) | 0.9497 | 0.370 (0.285–0.456) |
Our results indicated that the waist circumference cut point outlined in the JIS definition is appropriate for men; therefore, no updated definition was calculated for men.
JIS Metabolic Syndrome Definition Versus Updated Definition
In an effort to illustrate the difference between individuals meeting JIS criteria for the metabolic syndrome and those meeting our updated criteria, we estimated mean cardiovascular risk factor values among individuals in these two groups as well as among individuals not meeting metabolic syndrome criteria under either definition. These results are presented in Table 4. Of note is the fact that mean values for most metabolic syndrome components, with the exception of SBP and triglycerides, were comparable between women with metabolic syndrome according to the JIS criteria and both men and women without the metabolic syndrome.
Table 4.
Mean values of metabolic syndrome components among women and men by each metabolic syndrome definition
| Women | Men | ||||
|---|---|---|---|---|---|
| No MetS | JIS definition | Updated definition | No MetS | JIS definition | |
| Waist circumference | 92.382 (0.358) | 92.287 (0.180) | 106.963 (0.383) | 93.882 (0.294) | 108.197 (0.364) |
| SBP | 111.458 (0.296) | 124.106 (1.165) | 127.395 (0.58) | 120.222 (0.272) | 130.66 (0.533) |
| DBP | 68.291 (0.204) | 73.001 (0.627) | 76.615 (0.328) | 70.993 (0.242) | 79.339 (0.333) |
| HDL cholesterol | 55.243 (0.281) | 48.434 (0.892) | 44.698 (0.263) | 47.754 (0.236) | 38.295 (0.302) |
| Triglycerides | 90.886 (0.731) | 155.457 (6.559) | 178.892 (2.307) | 113.457 (1.472) | 226.198 (5.411) |
| Fasting glucose | 91.088 (0.286) | 106.152 (2.443) | 117.874 (1.366) | 97.395 (0.401) | 119.841 (1.256) |
Data are means (SE). MetS, metabolic syndrome.
Conclusions
In this article, we aimed to establish optimal definitions for abdominal obesity among U.S. Hispanic/Latino men and women in association with prevalent CHD. For the first time, we provide empirically derived recommendations for waist circumference cut points based on data from a large epidemiological sample of Hispanics/Latinos living in the U.S. Our results indicate than among U.S. Hispanic/Latino adults, waist circumference cut points of >102 cm in men and >97 cm in women provide optimal discrimination for cardiovascular risk as judged by the presence of CHD. When using these cut points to create an updated metabolic syndrome definition among women, we found disagreement between our updated definition and the current JIS criteria for metabolic syndrome. The prevalence of the metabolic syndrome was estimated to be ∼5 percentage points higher among women based on JIS criteria compared with our updated criteria.
The optimal waist circumference value in the association with the presence of CHD among Hispanic/Latino men in our target population was 102 cm. This value is in line with current recommendations for waist circumference definitions for non-Hispanic white males as part of the JIS metabolic syndrome criteria. Our results suggest, however, that the optimal waist circumference cut point associated with the presence of CHD among U.S. Hispanic/Latino women is 97 cm. This value is in disagreement with current recommendation outlined in the JIS metabolic syndrome definition. In fact, current recommendations for waist circumference cut point for non-Hispanic white women, 88 cm, yielded high levels of sensitivity at 87.7% but compromised specificity at only 24.2%.
While our study is the first to estimate empirically validated cut points of abdominal obesity for Hispanic/Latino men and women in the U.S., other groups have attempted to provide ethnicity-specific cut points to predict the presence of cardiometabolic risk among Central and South American Hispanics outside the U.S. Several outcome variables have been used across these studies, including cardiovascular end points such as abnormal carotid intima-media thickness and CVD, as well as cardiometabolic risk factors such as hypertension, abnormal lipid profile, and obesity. Groups using cardiovascular end points, such as the Peruvian Study of Cardiovascular Disease (PREVENCION) study, a population-based study of 1,439 Peruvian adults, recommended that cut points of 97 cm in men and 87 cm in women were optimal in determining the presence of abnormal carotid intima-media thickness and manifest CVD (7). With regard to cardiometabolic risk factors, the Mexican National Health and Nutrition Survey, which recruited 11,730 men and 26,647 women (10), established that waist circumference cut points for predicting the presence of T2DM and hypertension were 98 cm for men and 96 cm for women. Other estimates have been provided by groups in South America based on smaller samples. For example, studies in Colombia (8) and Brazil (9) have determine that 84 cm is an optimal cut point to detect the presence of obesity (defined by elevated BMI), and 88 cm to detect abnormal lipid profiles.
It is important to note that comparability across all of these studies is challenging given differences in methodological designs (population-based vs. convenience samples) and, most importantly, the choice of a diagnostic outcome or reference standard (hypertension, obesity, abnormal lipid profiles, abnormal carotid intima-media thickness, T2DM, manifest CVD). A consideration of ROC analysis, the method used across all of these different studies, is that the choice of an optimal cut point depends entirely on the outcome against which the ROC curve is to be constructed (18). Therefore, the choice of reference standard should be based upon the purpose to which the categorical classification is to be put. Similarly, the outcome should reflect a true state or diagnosis given that arbitrary dichotomization of outcomes may potentially introduce inaccuracy to the ROC analysis.
Our results indicate there is considerable disagreement in the prevalence of metabolic syndrome defined according to the JIS criteria and by our updated definition. In general, it has been the case that the misclassification of obesity in certain ethnic groups has resulted in missing obesity and cardiovascular risk in large numbers of people. This has been especially relevant for populations of East Asian and South Asian origin (19). Our data suggest, however, that the misclassification of abdominal obesity among Hispanic/Latino adults results in an overdiagnosis of this component and, therefore, in the overall prevalence of the metabolic syndrome. There was a 5.2% difference in the prevalence of the metabolic syndrome among women diagnosed with our updated definition (30.1%) compared with the JIS criteria (35.3%). Although still significantly higher than the prevalence among other ethnic groups, these estimates are closer to the rates of metabolic syndrome observed among non-Hispanic whites and blacks (20). The latest prevalence estimated among U.S. adults based on data from the 2010 National Health and Nutrition Examination Survey (NHANES) indicated that the metabolic syndrome is prevalent among 20.3% and 24.5% of non-Hispanic white and black women, respectively.
The substantial difference in the prevalence estimates of metabolic syndrome under each definition is not surprising given the high proportion of women who met the metabolic syndrome JIS criterion of three or more factors by virtue of exceeding the threshold value for abdominal girth (88 cm). In fact, ∼96% of women in the overall sample had abdominal obesity compared with only 73% of men. The prevalence of this component was remarkably high regardless of Hispanic/Latino background or age-group, particularly when compared with the prevalence of this component among other ethnic groups. Data from NHANES 2010 indicated that abdominal obesity is prevalent only among 62.4% of non-Hispanic women and among 79.3% of black women (20). This evidence supports the need for ethnicity-specific waist circumference cut points for U.S. Hispanic/Latino women.
Overdiagnosis of obesity has also occurred among other ethnic minority groups. For example, using NHANES III data, a study compared BMI with total body fat and percentage body fat (%BF) measured through bioelectrical impedance among black and non-Hispanic white adults (21). They showed that, in spite of having a significantly higher BMI, blacks had between 1.3 kg (men) and 3.2 kg (women) greater fat-free mass than non-Hispanic whites. When %BF, instead of BMI, was used to define obesity, the race/ethnicity gap in obesity prevalence decreased significantly (particularly among women). Overdiagnosis of obesity is not limited to blacks. It has also been demonstrated that for the same level of body fat, Polynesians have a 4.5 kg/m2 higher BMI than non-Hispanic whites (22). Although waist circumference measures are less susceptive to the contribution of lean body mass than BMI, it is possible that the difference in waist circumference cut points among Hispanic/Latino and other groups is a result of differences in %BF. This highlights the need for research on nonanthropometric imaging techniques to measure adiposity and predict cardiovascular risk among ethnic minority populations. These include imaging techniques such as computer tomography, MRI, and ultrasonography, as well as other modalities such as DEXA, bioelectrical impedance analysis, and air displacement plethysmography. Their performance among many ethnic groups including Hispanics/Latinos has not yet been studied (19).
Important strengths of our study include our large sample size and our population-based approach. The HCHS/SOL cohort was selected through a stratified multistage area probability sample (14), which allows us to estimate prevalence of diseases and baseline risk factors among noninstitutionalized Hispanic/Latino adults aged 18–74 years residing in four defined community areas (Miami, FL; Chicago, IL; San Diego, CA; and the Bronx, NY). Although the target population is limited to the four communities rather than the entire nation, HCHS/SOL’s hybrid design, which uses probability sampling within preselected diverse regions, is superior to the convenience samples, which are typically exploited in epidemiological cohort studies. The HCHS/SOL field centers are in cities with large Hispanic/Latino populations. The ranking of these cities among the metropolitan areas in the U.S. with the largest Hispanic population are as follows: New York #1, Chicago #5, San Diego #9, and Miami #11 (23). Furthermore, our study aimed to include Hispanic/Latino adults of various heritage groups such as Central and South Americans, Cubans, Dominicans, Mexicans, and Puerto Ricans. This ensured the representation of U.S. Hispanic/Latino adults of various ancestries enhancing the generalizability of our results.
Our study is not without limitations. This particular study aimed to recruit and represent Hispanic/Latino adults of various subgroups living in the U.S. As such, it is an ideal cohort for the estimation of ethnicity-specific definitions of abdominal obesity among U.S. Hispanics/Latinos. However, due to the fact that it did not include individuals of other ethnic groups, such as African Americans or non-Hispanic whites, it did not allow for comparisons across ethnicities. Another limitation of our study is the fact that waist circumference was only measured once. While the HCHS/SOL followed standardized protocols commonly used in major epidemiological studies, more rigorous approaches, such as repeated measurements, would have increased accuracy. An important limitation of our study is its cross-sectional design. The metabolic syndrome was identified with the purpose of serving as an early risk (preclinical) indicator of future CHD incidence. Therefore, risk classification is ideally performed in the context of longitudinal incidence data. Given the fact that abdominal girth and presence of CHD were assessed at the same time, we are unable to infer temporal precedence or draw conclusions regarding CHD incidence. Thus, our findings need to be validated externally with the use of longitudinal designs. A prospective study examining the incidence and extent of cardiovascular problems between individuals with and without the metabolic syndrome according to our proposed cut points is needed to further ascertain the accuracy of our abdominal obesity definition in classifying cardiovascular risk. Nevertheless, pending prospective data, our study provides important insights on the need to use ethnicity-specific abdominal obesity definitions among U.S. Hispanic/Latino adults.
Conclusion
Taken together, our results indicate that among Hispanics/Latino adults living in the U.S., waist circumference cut points of 102 cm in men and 97 cm in women provide optimal discrimination for the presence of CHD. Using these cut points in the context of an updated metabolic syndrome definition, we determined there was considerable disagreement between our definition and the current JIS metabolic syndrome definition. Current JIS criteria estimated the prevalence of metabolic syndrome to be ∼5% higher among women in our sample compared with our updated definition. Future reports should examine our recommended waist circumference definition cut points and the performance of our updated metabolic syndrome definition as a predictor of cardiovascular risk among U.S. Hispanics/Latinos in prospective designs.
Article Information
Funding. The HCHS/SOL was carried out as a collaborative study supported by contracts from the National Heart, Lung, and Blood Institute (NHLBI) to the University of North Carolina (N01-HC65233), University of Miami (N01-HC65234), Albert Einstein College of Medicine (N01-HC65235), Northwestern University (N01-HC65236), and San Diego State University (N01-HC65237). The following institutes/centers/offices contribute to the HCHS/SOL through a transfer of funds to the NHLBI: the National Institute on Minority Health and Health Disparities, National Institute of Deafness and Other Communication Disorders, National Institute of Dental and Craniofacial Research, National Institute of Diabetes and Digestive and Kidney Diseases, National Institute of Neurological Disorders and Stroke, and Office of Dietary Supplements.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. D.A.C. conducted data analysis and wrote the manuscript. M.M.L. oversaw data analysis and provided critical feedback and edited the manuscript. R.G., A.M., M.G., and N.S. contributed to discussion and reviewed and edited the manuscript. J.C., D.S.-A., M.D., and L.C.G. reviewed and edited the manuscript. D.A.C. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Footnotes
This article contains supplementary material online at https://doi.org/10.2337/figshare.12241073.
References
- 1.Palmer MK, Toth PP. Trends in lipids, obesity, metabolic syndrome, and diabetes mellitus in the United States: an NHANES analysis (2003-2004 to 2013-2014). Obesity (Silver Spring) 2019;27:309–314 [DOI] [PubMed] [Google Scholar]
- 2.Alberti KG, Eckel RH, Grundy SM, et al.; International Diabetes Federation Task Force on Epidemiology and Prevention; Hational Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; International Association for the Study of Obesity . Harmonizing the metabolic syndrome: a Joint Interim Statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation 2009;120:1640–1645 [DOI] [PubMed] [Google Scholar]
- 3.Shin J-A, Lee J-H, Lim S-Y, et al. . Metabolic syndrome as a predictor of type 2 diabetes, and its clinical interpretations and usefulness. J Diabetes Investig 2013;4:334–343 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Shin D, Kongpakpaisarn K, Bohra C. Trends in the prevalence of metabolic syndrome and its components in the United States 2007-2014. Int J Cardiol 2018;259:216–219 [DOI] [PubMed] [Google Scholar]
- 5.Schneiderman N, Llabre M, Cowie CC, et al. . Prevalence of diabetes among Hispanics/Latinos from diverse backgrounds: the Hispanic Community Health Study/Study of Latinos (HCHS/SOL). Diabetes Care 2014;37:2233–2239 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Jensen MD, Ryan DH, Apovian CM, et al.; American College of Cardiology/American Heart Association Task Force on Practice Guidelines; Obesity Society . 2013 AHA/ACC/TOS guideline for the management of overweight and obesity in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the Obesity Society [published correction appears in Circulation 2014;129(Suppl. 2):S139–S140]. Circulation 2014;129(Suppl. 2):S102–S138 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Medina-Lezama J, Pastorius CA, Zea-Diaz H, et al.; PREVENCION Investigators . Optimal definitions for abdominal obesity and the metabolic syndrome in Andean Hispanics: the PREVENCION study. Diabetes Care 2010;33:1385–1388 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Perez M, Casas JP, Cubillos-Garzon LA, et al. . Using waist circumference as a screening tool to identify Colombian subjects at cardiovascular risk. Eur J Cardiovasc Prev Rehabil 2003;10:328–335 [DOI] [PubMed] [Google Scholar]
- 9.Barbosa PJ, Lessa I, de Almeida Filho N, Magalhães LB, Araújo J. Criteria for central obesity in a Brazilian population: impact on metabolic syndrome. Arq Bras Cardiol 2006;87:407–414 [in English, Portuguese] [DOI] [PubMed] [Google Scholar]
- 10.Sánchez-Castillo CP, Velázquez-Monroy O, Berber A, Lara-Esqueda A, Tapia-Conyer R, James WP; Encuesta Nacional de Salud (ENSA) 2000 Working Group . Anthropometric cutoff points for predicting chronic diseases in the Mexican National Health Survey 2000. Obes Res 2003;11:442–451 [DOI] [PubMed] [Google Scholar]
- 11.Aschner P, Buendía R, Brajkovich I, et al. . Determination of the cutoff point for waist circumference that establishes the presence of abdominal obesity in Latin American men and women. Diabetes Res Clin Pract 2011;93:243–247 [DOI] [PubMed] [Google Scholar]
- 12.The IDF consensus worldwide definition of the metabolic syndrome [Internet] Available from https://www.idf.org/e-library/consensus-statements/60-idfconsensus-worldwide-definitionof-the-metabolic-syndrome.html. Accessed 1 June 2020
- 13.Sorlie PD, Avilés-Santa LM, Wassertheil-Smoller S, et al. . Design and implementation of the Hispanic Community Health Study/Study of Latinos. Ann Epidemiol 2010;20:629–641 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Lavange LM, Kalsbeek WD, Sorlie PD, et al. . Sample design and cohort selection in the Hispanic Community Health Study/Study of Latinos. Ann Epidemiol 2010;20:642–649 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Biospecimen Collection & Processing [Internet] Available from https://sites.cscc.unc.edu/hchs/node/4055. Accessed 1 June 2020.
- 16.Eng J. Receiver operating characteristic analysis: a primer. Acad Radiol 2005;12:909–916 [DOI] [PubMed] [Google Scholar]
- 17.Alemayehu D, Zou KH. Applications of ROC analysis in medical research: recent developments and future directions. Acad Radiol 2012;19:1457–1464 [DOI] [PubMed] [Google Scholar]
- 18.Kraemer HC, Noda A, O’Hara R. Categorical versus dimensional approaches to diagnosis: methodological challenges. J Psychiatr Res 2004;38:17–25 [DOI] [PubMed] [Google Scholar]
- 19.Rao G, Powell-Wiley TM, Ancheta I, et al.; American Heart Association Obesity Committee of the Council on Lifestyle and Cardiometabolic Health . Identification of obesity and cardiovascular risk in ethnically and racially diverse populations: a scientific statement from the American Heart Association. Circulation 2015;132:457–472 [DOI] [PubMed] [Google Scholar]
- 20.Beltrán-Sánchez H, Harhay MO, Harhay MM, McElligott S. Prevalence and trends of metabolic syndrome in the adult U.S. population, 1999-2010. J Am Coll Cardiol 2013;62:697–703 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Burkhauser RV, Cawley J. Beyond BMI: the value of more accurate measures of fatness and obesity in social science research. J Health Econ 2008;27:519–529 [DOI] [PubMed] [Google Scholar]
- 22.Deurenberg P, Yap M, van Staveren WA. Body mass index and percent body fat: a meta analysis among different ethnic groups. Int J Obes Relat Metab disord 1998;22:1164–1171 [DOI] [PubMed] [Google Scholar]
- 23.The Hispanic Population Census 2010 [Internet], 2010 Available from https://factfinder2.census.gov/faces/nav/jsf/pages/index.xhtml. Accessed 28 October 2013.