Obesity is associated with an increased risk of developing comorbidities, such as impaired glucose tolerance, hypertension, type 2 diabetes mellitus (T2D), dyslipidemia, hypertension, heart failure, coronary artery disease (CAD), stroke, and several cancers [1–5]. Various studies define metabolic phenotypes based on different criteria; therefore, it is difficult to determine their prevalence . According to Blüher , the metabolically healthy overweight/obese (MHO) are generally defined as the absence of insulin resistance, T2D, dyslipidemia, and hypertension in patients with BMIs greater than 30 kg/m2. However, there are those who are considered normal weight but are at risk for metabolic disorders, and these patients are classified as metabolically unhealthy normal weight (MUHNW) . MUHNW are defined as hyperinsulinemic, insulin resistant, hypertriglyceremic, and predisposed to the development of T2D and CAD [8,9]. Although they are also known to have BMIs less than 25 kg/m2, they still have metabolic abnormalities, such as abdominal fat distribution and elevated blood pressure (BP) .
MHO individuals have been shown to be insulin sensitive, however, have a low cardiovascular risk profile despite excessive body fat . Longitudinal data demonstrate that some MHO individuals do not remain metabolically healthy over time, suggesting that up to 30% convert to metabolically unhealthy overweight/obese (MUHO) with cardiometabolic complications during a 5- to 10-year period [6,11,12]. Aging and weight gain are two risk factors that may be associated with this conversion, whereas increased physical activity and weight loss may help maintain the MHO status .
According to Peppa et al. , MUHO and MUHNW are common in postmenopausal women, representing 63.3% and 75.6%, respectively. The postmenopausal state is associated with increased cardiometabolic risk regardless of body weight . The hormonal changes during menopause are associated with an increased cardiometabolic risk in all postmenopausal women, whether they are obese or nonobese . It has been suggested that body composition may be a major determinant of cardiometabolic risk after menopause , with excess of central-to-peripheral fat being the most prominent .
While cardiometabolic risk burden is significant in general among postmenopausal women, data are lacking on the magnitude of difference in key cardiometabolic risk factors according to women who are metabolically healthy vs. unhealthy according to obesity status. A better understanding of this would likely help drive the clinical importance of these categorizations. The purpose of this study was to identify sociodemographic and metabolic correlates of metabolic weight categories in Women’s Health Initiative (WHI) postmenopausal women. Our hypothesis was that those MUHO, MUHNW, and MHO postmenopausal women were more likely to have adverse risk factors as compared to metabolically healthy normal weight (MHNW) postmenopausal women.
The WHI consisted of two major segments: a group of randomized clinical trials (N = 68 132) and an observation cohort study (N = 93 676) . The clinical trials comprised three concurrent, randomized controlled trials among postmenopausal women aged 50–79 years: hormone therapy trials, dietary modification trial, and the calcium and vitamin D trial . WHI recruitment was conducted from 1993 to 1998 by 40 Clinical Centers in 24 states and the District of Columbia . Recruitment took place locally, at Clinical Centers, and nationally at the National Institutes of Health, the Clinical Coordinating Center, and various study-wide committees, including various recruitment strategies of mass mailings, community presentations, local newspaper ads, public service announcements (TV and radio), and health fairs . The WHI is monitored by the Data and Safety Monitoring Board, an independent board of medical researchers, which ensures the participants’ safety and protocol procedures are followed . Also, each clinical center has their own independent institutional review board to keep their participants informed and to protect their safety (see Supplementary Appendix, Supplemental digital content 1, http://links.lww.com/CAEN/A21) .
A baseline cardiovascular disease (CVD) risk factor and biomarker subset sample of approximately 25 000 participants was derived from three studies: (1) clinical trials 6% subsample, clinic complete blood counts, quality control pools; (2) CVD biomarkers for 2010–2015 single nucleotide polymorphisms health association resource cohort only; and (3) CVD, diabetes mellitus, and renal biomarkers in the estrogen alone hormone therapy trial cohort were included as variables . The measures consisted of glucose, total cholesterol, high-density lipoprotein cholesterol (HDL-C), triglycerides, insulin, and C-reactive protein (CRP) . We excluded women with prevalent CVD at baseline (no prior myocardial infarction, stroke, percutaneous intervention, heart failure, or peripheral arterial disease).
Overweight/obesity was defined by a BMI ≥25 kg/m2 or elevated waist circumference (≥88 cm) measured by trained staff during clinic visits [18–20]. MUHO was based on having at least two of four metabolic traits, including: high triglycerides (≥150 mg/dl); elevated systolic BP (≥130 mmHg) or diastolic BP (≥85 mmHg) were measured by averaging two baseline measurements (if only one BP measurement was known, the single value was used)  or antihypertensive drugs (including diuretics); high fasting glucose (≥100 mg/dl) or medications for diabetes (insulin and oral medications); and low HDL-C (<50 mg/dl) [18–20]. MHO required less than two of the above metabolic traits [18–20]. Normal body weight was defined as a BMI ≥18.5 and <25 kg/m2 and without elevated waist circumference (<88 cm) (measured by trained staff during clinic visits) [18–20]. Metabolically unhealthy normal (MUHNW) required at least two of the four above metabolic traits whereas metabolically healthy normal (MHNW) required less than two of the above metabolic traits [18–20].
Prevalent diabetes was defined as those with known diabetes at baseline (self-report, fasting glucose ≥ 126 mg/dl or insulin now or oral medications for diabetes) were included in the analyses.
CVD risk factor and biomarker assays consisted of glucose, total cholesterol, HDL-C, triglycerides, insulin, and CRP . Glucose was measured in serum via the hexokinase method on the Hitachi 747 (Boehringer Mannheim Diagnostics, Indianapolis, Indiana, USA) and the Gluco-quant Glucose/hexokinase reagent (Roche Diagnostics, Indianapolis, IN, USA) on the Roche Modular P Chemistry analyzer (Roche Diagnostics Corporation) . Total cholesterol was analyzed by enzymatic methods on a Hitachi 747 analyzer (Boehringer Mannheim Diagnostics) . In 2006, the instrument changed from the Hitachi 747 to the Roche Modular for both the total cholesterol and HDL-C biomarkers . It was also measured in serum using a cholesterol oxidase method (Roche Diagnostics Corporation) on the Roche Modular P Chemistry analyzer (Roche Diagnostics Corporation) . HDL-C was isolated using heparin manganese chloride with the supernate measured enzymatically on the Hitachi 747 . HDL-C was also measured in serum using the HDL-C plus third generation direct method (Roche Diagnostics Corporation) on the Roche Modular P Chemistry analyzer (Roche Diagnostics Corporation) . Triglycerides were analyzed by enzymatic methods on a Hitachi 747 analyzer (Boehringer Mannheim Diagnostics) and in serum using Triglyceride GB reagent (Roche Diagnostics) on the Roche Modular P Chemistry analyzer (Roche Diagnostics Corporation) . Insulin was analyzed via sandwich immunoassay (Roche Diagnostics) on Roche Elecsys 2010 Analyzer . Also, it was measured with the Pharmacia RIA method, then, starting on 8 October 1998, they were measured in a step-wise sandwich ELISA procedure on an ES 300 (Boehringer Mannheim Diagnostics) . Insulin was also measured in serum by Roche 2010 Electrochemiluminescence . In 2009, the instrument changed from the Roche 2010 to the ADVIA Centaur system . CRP was measured in serum via immunoassay on a Roche Modular P Chemistry analyzer .
Covariates: In addition to age, variables hypothesized to be associated with MHO, MUHO, MHNW, and MUHNW in postmenopausal women were included as self-reported covariates, including age, race/ethnicity, income, education, cigarette smoking (at least 100 cigarettes ever), alcohol consumption (drank 12 alcoholic drinks ever), family history (relatives had heart attack), and total energy expenditure from recreational physical activity (MET-hours/week). The biomarkers (total cholesterol, log CRP, and insulin) were measured by multiple laboratories utilizing different techniques and various instruments (see CVD risk factor and biomarker assays section above) . The biomarkers were collected as blood samples with participants fasting for a minimum of 12 h before draws . The residual blood samples were stored at 4°C for up to 1 h until plasma or serum was separated from the cells .
Baseline sociodemographic and additional metabolic risk factor characteristics within the WHI postmenopausal populations were compared across the four groups of interest (MHO, MUHO, MHNW, MUHNW) using Pearson’s Chi-square test of proportions for categorical variables and analysis of variance for the continuous variables. Polytomous multinomial logistic regression with generalized link logit function provided the odds of membership among the four groups with the MHNW as the reference group. Estimates were adjusted for age, race/ethnicity, income, education, cigarette smoking (at least 100 cigarettes ever), alcohol consumption (drank 12 alcoholic drinks ever), family history (relatives had heart attack), total energy expenditure from recreational physical activity (MET-hours/week), total cholesterol, log CRP, and insulin. The CVD biomarkers were log transformed and standardized to account for multiple laboratories utilizing different techniques and various instruments . All statistical tests were two-sided, and all statistical analyses were performed using SAS software Version 9.4 .
Our study included 19 412 postmenopausal women, with 2369 (12.2%) participants having prevalent diabetes. The data indicate that the majority of participants was non-Hispanic White (47.0%), aged 50–79 years, had some college, vocational training, or associates degree (39.4%), and earned an income between $20 000 and $34 999 (25.9%) (Table 1).
Table 2 reports the distributions of the general characteristics of the analysis sample as follows: MHNW (reference group) (16.8%), MUHNW (5.9%), MHO (35.4%), and MUHO (41.9%). For those with MHNW, the mean age was 63.8 years, ethnicity primarily White (not of Hispanic origin), education consisting of a baccalaureate degree or higher, and a reported annual income of $20 000–$34 999. For those with MUHNW, the mean age was 65.9 years, ethnicity mainly White (not of Hispanic origin), education being some college, vocational training, or associates degree, and an annual income of less than $19 999. In contrast, those with MHO averaged 62.4 years old, and were mostly Black or African American, had some college, vocational training, or associates degree, and reported an annual income of $20 000–$34 999. Among those with MUHO, the mean age was 63.5 years, ethnicity largely White (not of Hispanic origin), and education involved some college, vocational training, or associates degree, with a reported annual income of less than $19 999 (Table 2).
To determine the odds of membership, all demographic and behavioral factors as well as clinical biomarkers were included in the polytomous multinomial logistic regression (Table 3). With the MHNW as the reference group, the odds ratio (OR) for age was statistically significant for the MUHNW group [OR 1.04, 95% confidence interval (CI): 1.02–1.05, P < 0.0001] after adjusting for covariates, indicating MUHNW was 4% more likely for each year greater in age. The OR for age as compared to MHNW (reference group) also was statistically significant (inversely) for the MHO (OR 0.98, 95% CI: 0.98–0.99, P < 0.0001) after adjusting for covariates, indicating a 2% lower likelihood of MHO for every year increase in age. The odds ratio for age as compared to MHNW was not statistically significant for the MUHO (OR 1.00, 95% CI: 0.99–1.01, P = 0.60) after adjusting covariates (Table 3).
The ORs for race/ethnicity as compared with White (not of Hispanic origin) MHNW (reference group) were significant for Black/African American MUHNW (OR 0.64, 95% CI: 0.51–0.79, P < 0.0001), Black/African American MHO (OR 1.50, 95% CI: 1.31–1.72, P < 0.0001), Black/African American MUHO (OR 0.77, 95% CI: 0.66–0.89, P = 0.0004), Hispanic/Latino MHO (OR 0.76, 95% CI: 0.64–0.89, P = 0.001), and Hispanic/Latino MUHO (OR 0.65, 95% CI: 0.54–0.77, P < 0.0001) after adjusting for covariates. The Black/African American MUHNW, Black/African American MUHO, Hispanic/Latino MHO, and Hispanic/Latino MUHO reported lower odds of 36, 23, 24, and 35% of having MUHNW, MUHO, or MHO (compared with being metabolically healthy normal weight) after adjusting for covariates. The Black/African American MHO reported a higher odds of 50% of having MHO (compared with being metabolically healthy normal weight) after adjusting for covariates. The OR for race/ethnicity as compared with White (not of Hispanic origin) MHNW (reference group) was not statistically significant for Hispanic/Latino MUHNW (OR 0.95, 95% CI: 0.74–1.21, P = 0.67) after adjusting for covariates (Table 3).
In addition, lower education was associated with a greater likelihood of being MUHNW in postmenopausal women. Also, increased physical activity was associated with a lower likelihood of being MUHO. Family history was associated with a greater likelihood of being MUHO, while total cholesterol was associated with a greater likelihood of being MUHNW and MUHO. Last, insulin and log CRP were associated with a greater likelihood of being MUHNW, MHO, and MUHO (Table 3).
Our study showed that different metabolic risk patterns of all these sociodemographic, behavioral, and clinical characteristics across four weight groups were observed. Prevalent diabetes was most common among those with MUHO. In addition, those with MUHO were the least physically active, had the largest waist circumferences and BMIs, highest diastolic BPs, and the highest biomarkers of glucose, triglyceride, insulin, and CRP.
These findings provide us with additional information to better understand factors that may relate to WHI postmenopausal women transitioning from MHO to MUHO. It has been reported that up to 30% of MHO convert to MUHO with cardiometabolic complications within a 5- to 10-year period [6,11,12]. It is possible that physical activity and other lifestyle interventions might help improve such risk factors in those with MHO. As discussed by Chang et al. , aging and weight gain may be associated with this shift, whereas increased physical activity and weight loss may benefit them by assisting them to maintain their ‘healthy’ status. Ultimately, this may help reduce their risk of developing comorbidities, such as nonalcoholic fatty liver disease, impaired glucose tolerance, hypertension, T2D, dyslipidemia, hypertension, heart failure, CAD, stroke, and several cancers [1–5].
Advanced age was positively associated with the MUHNW as compared with MHNW (reference group) after adjusting for covariates. It is widely reported in the literature that aging is associated with higher prevelances of overweight and obesity in older adults . This outcome is consistent with the current literature, although age was not associated with the likelihood of having MUHO in our study. According to Alam et al. , obesity and its comorbidities are associated with chronic inflammation. Aging also further influences an increase in inflammatory factors, specifically cytokines . However, once aging and obesity are both present, they exacerbate inflammation, eventually intensifying chronic diseases .
Black/African American postmenopausal women were associated with a decreased likelihood of becoming MUHNW and MUHO as compared with White (not of Hispanic origin) MHNW (reference group) after adjusting for covariates, although they were associated with a greater likelihood of being MHO. However, Hispanic/Latino women were associated with a lower likelihood of being MHO and MUHO as compared with White (not of Hispanic origin) MHNW (reference group) after adjusting for covariates. This conclusion is very important to note that the Black/African American MUHNW and MUHO and Hispanic/Latino MHO and MUHO postmenopausal women were less likely to be part of MUHNW, MHO, and MUHO after adjusting for covariates as compared with the reference group. Coleman et al.  reported that Latino and African Americans had the highest prevalence of lean body weight with diabetes, which may suggest higher levels of MUHNW and cardiometabolic abnormalities regardless of body weight.
Our study had important strengths and limitations. A strength was that the CVD biomarkers were measured by standardized methods among a large sample of diverse demographics. A proposed limitation is that study participants may not be representative of the US population, thus limiting generalizability. Although the participants were recruited nationally and from various geographic and socioeconomic backgrounds, the majority were non-Hispanic White. The WHI population also consists of only postmenopausal women, and therefore, may not be generalizable to younger women and men. Also, as a secondary data analysis, not all potential variables of interest may have been collected during the original study sample data collection to address particular issues or hypotheses. Importantly, we did not have adequate repeated measures to assess the degree of transition between groups, for example, transition to MUHO from MHO over time.
In summary, findings from this work may help in developing pertinent health promotion methods to assist postmenopausal women in the prevention of obesity and diabetes. By making lifestyle changes, risk factors can be reduced to help promote healthier, longer, and an improved quality of life for postmenopausal women . Further research efforts are needed to perform studies with added emphasis on policy and environmental factors to reduce obesity and diabetes .
The WHI program is funded by the National Heart, Lung, and Blood Institute, National Institutes of Health, US Department of Health and Human Services through contracts HHSN268201600018C, HHSN268201600001C, HHSN268201600002C, HHSN268201600003C, and HHSN268201600004C.
A.R.C.H., H.A.-C., and N.D.W. were involved in study concept and design. A.R.C.H., S.L.A., B.X., D.V.P., H.A.-C., and N.D.W. were involved in acquisition, analysis, or interpretation of data and administrative, technical, or material support. A.R.C.H., B.X., and N.D.W. were involved in statistical analysis. S.L.A., B.X., D.V.P., H.A.-C., and N.D.W. were involved in study supervision. All authors involved in drafting of the manuscript and critical revision of the manuscript for important intellectual content.
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent was obtained from all individual participants included in the study.
Abstract was accepted/presented at the American Heart Association 2020 Epi Lifestyle Scientific Sessions.
Conflicts of interest
There are no conflicts of interest.
1. Mulhall BP, Ong JP, Younossi ZM.. Non-alcoholic fatty liver disease: an overview. J Gastroenterol Hepatol. 2002; 17:1136–1143
2. Haslam DW, James WP.. Obesity
. Lancet. 2005; 366:1197–1209
3. Lau DC, Douketis JD, Morrison KM, Hramiak IM, Sharma AM, Douketis JD, et al. 2006 Canadian clinical practice guidelines on the management and prevention of obesity
in adults and children [summary]. CMAJ. 2007; 176:S1–S13
4. Mathew H, Farr OM, Mantzoros CS.. Metabolic health and weight: understanding metabolically unhealthy normal weight or metabolically healthy obese patients. Metabolism. 2016; 65:73–80
5. National Heart, Lung, and Blood Institute (NHLBI). Managing overweight and obesity
in adults: systematic evidence review from the obesity
expert panel. 2013
6. Chang Y, Ryu S, Suh BS, Yun KE, Kim CW, Cho SI.. Impact of BMI on the incidence of metabolic abnormalities in metabolically healthy men. Int J Obes (Lond). 2012; 36:1187–1194
7. Blüher M.. Are metabolically healthy obese individuals really healthy?. Eur J Endocrinol. 2014; 171:R209–R219
8. Ruderman NB, Schneider SH, Berchtold P.. The ‘metabolically-obese’, normal-weight individual. Am J Clin Nutr. 1981; 34:1617–1621
9. Ruderman N, Chisholm D, Pi-Sunyer X, Schneider S.. The metabolically obese, normal-weight individual revisited. Diabetes
. 1998; 47:699–713
10. Sims EA.. Are there persons who are obese, but metabolically healthy?. Metabolism. 2001; 50:1499–1504
11. Appleton SL, Seaborn CJ, Visvanathan R, Hill CL, Gill TK, Taylor AW, Adams RJ; North West Adelaide Health Study Team. Diabetes
and cardiovascular disease outcomes in the metabolically healthy obese phenotype: a cohort study. Diabetes
Care. 2013; 36:2388–2394
12. Soriguer F, Gutiérrez-Repiso C, Rubio-Martín E, García-Fuentes E, Almaraz MC, Colomo N, et al. Metabolically healthy but obese, a matter of time? Findings from the prospective pizarra study. J Clin Endocrinol Metab. 2013; 98:2318–2325
13. Peppa M, Koliaki C, Papaefstathiou A, Garoflos E, Katsilambros N, Raptis SA, et al. Body composition determinants of metabolic phenotypes of obesity
in nonobese and obese postmenopausal women
(Silver Spring). 2013; 21:1807–1814
14. Peppa M, Koliaki C, Raptis SA.. Body composition as an important determinant of metabolic syndrome in postmenopausal women
. Endocrinol Metab Syndrome. 2012S1–009
15. Carr MC.. The emergence of the metabolic syndrome with menopause. J Clin Endocrinol Metab. 2003; 88:2404–2411
16. The Women’s Health Initiative
Study Group. Design of the Women’s Health Initiative
clinical trial and observational study. The Women’s Health Initiative
Study Group. Control Clin Trials. 1998; 19:61–109
17. Women’s Health Initiative
(WHI). About WHI. WHI website. https://www.whi.org/about/SitePages/About%20WHI.aspx
. [Accessed 10 October 2017]
: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser. 2000; 894:1–253
19. Alberti KG, Eckel RH, Grundy SM, Zimmet PZ, Cleeman JI, Donato KA, 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
20. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of The National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). JAMA. 2001; 285:2486–2497
21. SAS Institute Inc(2002-2012). SAS version 9.4. Cary, NC
22. Alam I, Ng TP, Larbi A.. Does inflammation determine whether obesity
is metabolically healthy or unhealthy? The aging perspective. Mediators Inflamm. 2012; 2012:456456
23. Coleman NJ, Miernik J, Philipson L, Fogelfeld L.. Lean versus obese diabetes
mellitus patients in the United States minority population. J Diabetes
Complications. 2014; 28:500–505
24. Remington PL, Brownson RC, Wegner MV.. Chronic Disease Epidemiology and Control. 2010Ed. 3, Washington, DC: American Public Health Association