Abstract
Objective:
Carotid artery intima media thickness (cIMT) and adventitial diameter (AD) are subclinical atherosclerosis indicators. Metabolic syndrome (MetS) and obesity are risk factors for atherosclerosis, but their combined impact on atherosclerosis risk is unknown. This study sought to examine the effect of the co-occurrence of MetS with obesity on cIMT and AD.
Methods:
The Study of Women’s Health Across the Nation (SWAN) is a multi-center, multi-ethnic study. Carotid ultrasound assessments and concurrent physiologic measurements were undertaken between 2009–2013. This cross-sectional analysis included 1,433 women with body mass index ≥18.5kg/m2 and free of prevalent clinical cardiovascular disease. Multivariable linear regression models were used to relate maximum cIMT and AD (dependent variables) with obesity, MetS and their interaction.
Results:
The average age was 60.1 years (standard deviation (SD):2.7). The prevalence of obesity and MetS was 44% and 35%, respectively. Women with obesity had a 0.051mm larger mean cIMT and women with MetS had a 0.057mm larger cIMT versus women without the respective conditions (both p<0.001). There was a statistically significant interaction between obesity and MetS (p=0.011); women with both had a model-adjusted predicted mean cIMT of 0.955mm (95%CI, 0.897–1.013), higher than those with MetS alone (0.946mm; 95%CI, 0.887–1.005), obesity alone (0.930mm; 95%CI, 0.873–0.988), or neither condition (0.878mm; 95%CI, 0.821–0.935). Adventitial diameter results were similar.
Conclusions:
Early detection and treatment of atherosclerotic changes may prevent significant disease. This study suggests there is a minimal impact of obesity on carotid artery thickness beyond MetS alone. All individuals with metabolic dysfunction, regardless of obesity status, should be considered at increased risk for atherosclerotic changes.
Keywords: carotid intima media thickness, obesity, metabolic syndrome, adventitial diameter
INTRODUCTION
Carotid artery intima media thickness (cIMT) and adventitial diameter (AD) are early indicators of subclinical atherosclerosis, a thickening and stiffening of the artery walls due to plaque build-up, which can lead to stroke, heart attack, and death1–4. Early detection of atherosclerotic cardiovascular disease (ASCVD) can lead to prompt identification of ASCVD risk factors with management of these risk factors leading to reductions in risk for events.3 There are many known risk factors for atherosclerosis including smoking, unhealthy diet, older age, obesity, hypertension, and diabetes1,3,4. However, little is known about the combined impact of obesity and MetS on atherosclerosis. Understanding the combined effect of multiple risk factors is essential given that individuals rarely exhibit only one factor in isolation.
Although MetS is often considered a condition of obesity, obesity itself, based purely upon body size, is not a required part of the definition5. Instead, MetS definitions include a measure of body fat distribution (waist circumference), given evidence that fat location is more associated with metabolic dysfunction than is body size alone6–8. In fact, despite the known co-occurrence of obesity with diseases such as diabetes, hypertension, and hyperlipidemia7, there remain a number of individuals with obesity who are otherwise healthy – a condition known as “metabolically healthy obese” (MHO)7. The rate of MHO varies from 10% to more than 25%7, and these individuals with MHO often have higher insulin sensitivity and cardiovascular fitness than those who are “metabolically unhealthy obese”7. Importantly, the prevalence of MHO decreases with age7, suggesting that those who have obesity for extended periods of time may be at risk to develop additional comorbidities despite an initial “healthy” obese presentation. It remains unknown if those with metabolically healthy obesity are at decreased risk for cardiovascular sequalae while in this state compared to those with metabolic disease.
Both metabolic syndrome and higher body mass index have been linked to increased risk of atherosclerosis, though the effect has been shown to be greater in women than in men9–12. Notably, for women the menopausal transition has been identified as a critical window of risk for adverse metabolic function. Both the prevalence and severity of MetS and its constituent components increase during this time13. To our knowledge, while individual risk has been assessed no studies have evaluated the joint effects of having obesity and MetS on the outcome of cIMT and adventitial diameter. Given the high rate of MHO, it is important to distinguish risk between these two groups so prophylactic treatment can be appropriately administered.
This study sought to examine the effect of the co-occurrence of metabolic syndrome with obesity on cIMT and adventitial diameter, both surrogate markers of subclinical atherosclerosis and early arterial remodeling. We hypothesized that cIMT and AD measures would be higher in individuals with both obesity and MetS compared to those without both conditions or in individuals with only one of the two conditions, and that there would be a synergistic joint effect of obesity and MetS.
MATERIAL AND METHODS
Study Population
The Study of Women’s Health Across the Nation (SWAN) is a community based, multi-center, multi-ethnic longitudinal study designed to investigate physiologic and psychosocial changes that occur during the menopausal transition, midlife and early late adulthood14. In 1996, 3302 women were enrolled from seven study sites, and each site recruited White women and site-specific non-White sample including Black women from Boston, MA, Pittsburgh, PA, southeast Michigan, MI, and Chicago, IL; Hispanic women from Newark, NJ; Chinese women from Oakland, CA; and Japanese women from Los Angeles, CA. Eligibility criteria for enrolment into the longitudinal SWAN follow-up included: age 42–52 years at study baseline in 1996, an intact uterus and at least one ovary, no use of exogenous hormones affecting ovarian function in the past 3 months, and at least one menstrual period in the previous 3 months. The institutional review board at each participating site approved the study protocol and all participants provided written, signed informed consent.
SWAN women were invited to participate in a carotid ultrasound assessment at either the SWAN follow-up visit 12 (2009–2011) or 13 (2011–2013). Women from the Los Angeles SWAN site did not participate in these measurements. Of the 2,466 women who participated in SWAN follow-up visit 12 or 13 at one of the six sites administering the ultrasound protocol, 1,606 women underwent the carotid ultrasound (Figure 1). Those who received an ultrasound were more likely to have obesity, to have metabolic syndrome, and more likely to have a family history of heart disease as compared to those who did not participate in the ultrasound protocol. There were no significant differences in age or diabetes status. In this analytic sample, we excluded women with a body mass index < 18.5 kg (n=45), unclassified metabolic syndrome at the time of the carotid ultrasound (n=45), or history of cardiovascular disease including myocardial infarction, prior transient ischemic attack/stroke, or congestive heart failure (n=83), thereby leaving a final analytic sample of 1,433 women.
Figure 1. Study participant flow diagram.
A visual description of the exclusion criteria for our study.
Measures of Arterial Remodeling
At each site, digitized carotid artery ultrasound images were taken by centrally trained and certified sonographers using a Terason t3000 Ultrasound System (Teratech Corp, Burlington, MA, USA) equipped with a variable frequency 5–12 MHz linear array transducer15. Two digitized images were obtained bilaterally by the site sonographer for later reading; images were obtained of the left and right distal common carotid artery (CCA) at end-diastole (upstroke of R wave) by automatic cine-loops detection. At each site most scans were performed by one sonographer (100% at 4 sites and 91% at 1 site). Trained sonographers at the SWAN Ultrasound Reading Center, blinded to the health status of SWAN participants, read each of these four images, using a semi-automated edge detection software [(Artery Measurement System (AMS)]16, near and far wall CCA-IMT measures were obtained by electronically tracing the lumen-intima interface and the media-adventitia interface across a 1-centimeter segment proximal to the carotid bulb. Common carotid artery IAD was measured directly as the distance from the adventitial-medial interface on the near wall to the medial-adventitial interface on the far wall across the same CCA segments used for CCA-IMT measurement. For CCA-IMT, the mean of the maximum IMT and for IAD, the mean of the average IAD across the four images were used in analysis. SWAN images were read by 2 readers (61% by reader 1 and 39% by reader 2). Scanning reproducibility was evaluated at each site. Reproducibility of IMT measures was good to excellent with an intraclass correlation coefficient between sonographers of ≥ 0.77 and between and within readers of > 0.90. Carotid ultrasound data were reviewed quarterly during data collection. Potential outliers were identified as observations with values outside mean ± 2SD of site-specific quarterly data, and were assessed again once data collection was completed. Random data checks were also performed throughout data collection and anomalies were queried.
Metabolic Syndrome
Metabolic syndrome was defined using the National Cholesterol Education Program Adult Treatment Panel III5 which requires the presence of three or more of the following: high waist circumference (≥80 cm for Chinese women, otherwise >88cm), high fasting blood glucose (>100 mg/dL), high blood pressure (systolic pressure ≥130mmHg or diastolic pressure ≥85mmHg or on blood pressure lowering medication), low high density lipoprotein cholesterol (HDL, <50mg/dL), and high fasted triglyceride values (≥150mg/dL). Waist circumference was measured in centimeters at the narrowest part of the torso from the anterior aspect. Blood pressure was measured by a trained technician following a five-minute rest; the average of two measurements was used in analysis. Assays for glucose, HDL and triglycerides were conducted on fasted serum samples and have been described previously17,18. Briefly, glucose was measured using a hexokinase-coupled reaction (Boehringer Mannheim Diagnostics), HDL cholesterol was measured using heparin-2 M manganese chloride, and triglycerides were measured using enzymatic methods on a Hitachi 747 analyzer (Boehringer Mannheim Diagnostics). Those missing data for any element of MetS were considered to have unknown MetS status. Missing MetS information was usually due to lack of participation in the blood collection protocol.
Height and weight were measured by a trained interviewer using a stadiometer and balance beam scale, respectively, and used to calculate body mass index (kg/m2). Obesity was defined as a BMI >= 30 kg/m2, and the same BMI threshold was used for all races.
Covariates
With the exception of race/ethnicity which was self-identified at the start of the study, all other measures, including hormone use and menopausal status were collected at the time of the carotid ultrasound assessment. Age was calculated as date of visit minus date of birth. Socioeconomic strain (very hard, somewhat hard, not at all hard) was characterized based upon participant report of difficulty paying for basic necessities. Family history of heart disease was defined from self-report. Smoking status was defined as never, former, or current smoker. Menopausal status (post-menopausal, peri-menopausal, pre-menopausal, and unknown status) was based upon bleeding patterns and levels of sex steroid hormones. Use of lipid-lowering medications and exogenous hormones were collected via a full medication inventory and coded using Iowa Drug Information Service coding.
While this is a cross-sectional analysis, the analytic sample is seated within the larger SWAN study, an ongoing longitudinal study with near-annual study visits dating back to 1996. Thus, we are able to examine longitudinal use of lipid-lowering medications. In our analytic sample, approximately 33% (n=127) of participants reported their first use of a lipid-lowering medication at the time of the visit associated with their carotid ultrasound. Of those that commenced use at a prior visit, there was no statistically significant difference in length of use of lipid-lowering agents by obesity and/or MetS status (p=0.354). Thus, we did not adjust for length of use of lipid-lowering medications in our analysis.
Statistical Analysis
Descriptive statistics were presented for all continuous variables as means with standard deviations (SD) or medians with 25th and 75th percentiles, as appropriate, and for categorical variables as frequencies with percentages. Continuous data were assessed for normality using histograms and the Kolmogorov-Smirnov test.
Bivariate analyses were conducted to evaluate differences in key covariates between those with and without MetS, and those with and without obesity using two-sample independent t-tests or Wilcoxon rank-sum tests for continuous variables and Chi-square or Fisher’s exact tests for categorical variables, as appropriate. Percent differences and percent changes, as appropriate, were additionally computed to quantify differences between groups.
To determine if the interaction of obesity and MetS was associated with a significant increase in the outcomes of interest, a series of multivariable linear models were constructed for each of the outcomes: maximum IMT and adventitial diameter. The interaction was operationalized as a 4-level variable: no obesity/no MetS, obesity with no MetS, no obesity with MetS, and both obesity and MetS. Prior to model entry, all variables were assessed for collinearity using a Pearson correlation matrix and the variance inflation factor. If the correlation between two variables is > 0.7 or the variance inflation factor >10 then the variable pair was determined to have high collinearity and the variable with the smaller bivariate effect size was removed from the model. Due to the known collinearity between race/ethnicity and study site, race was chosen a priori to remain in the models given the clinical significance of race in the context of cardiometabolic dysfunction.
A set of models was first run for the exposure of obesity: first, only the exposure against the outcome of mean maximum cIMT; second, obesity adjusting for age, race, hormonal therapy use, and menopausal status; and third obesity adjusting for age, race, hormonal therapy use, menopausal status, smoking status, family history of heart disease, and lipid-lowering medication use. Similar models were constructed for MetS, MetS and obesity in the same model, and MetS and obesity with their interaction. This same set of models was constructed for the outcome of adventitial diameter (AD). Model residuals were assessed for outliers and leverage points using Cook’s D. All variables included in the models were selected based on clinical relevance with the outcome. Model effect sizes were reported as beta coefficients with 95% confidence intervals.
Additional sensitivity analyses included: using an obesity BMI threshold of 27.5 kg/m2 for Chinese women, excluding overweight individuals from the non-obese cohort, limiting to only those post-menopausal women, stratifying the fully adjusted model by race, and excluding participants from the New Jersey site due to inconsistencies in data collection.
Analyses were conducted using SAS v. 9.4 (SAS Institute, Cary, NC) and a p <0.05 was considered to be statistically significant for all analyses conducted.
RESULTS
The average age of participants was 60.1 years (SD 2.7) with an overall obesity prevalence of 44.1% and an overall metabolic syndrome prevalence of 34.8%. African American women and Hispanic women, those with greater economic strain, and those peri-menopausal or with unknown menopausal status were more likely to have obesity and metabolic syndrome whereas women who were never smokers were less likely to have obesity or MetS (Table 1).
Table 1.
Study participant characteristics by metabolic syndrome and obesity status at time of carotid ultrasound.
| No Obesity or Metabolic Syndrome (N = 654) |
Obesity with no Metabolic Syndrome (N = 281) |
No Obesity with Metabolic Syndrome (N = 147) |
Obesity with Metabolic Syndrome (N = 351) |
p-value | |
|---|---|---|---|---|---|
|
| |||||
| DEMOGRAPHICS | |||||
| Age, years | 60.0 ± 2.7 | 60.1 ± 2.6 | 60.5 ± 2.7 | 60.1 ± 2.9 | p=0.314 |
| Body Mass Index | 24.6 ± 2.9 | 35.7 ± 5.1 | 27.1 ± 2.3 | 37.7 ± 5.7 | p<0.001abcdef |
| Race/Ethnicity | p<0.001abcdf | ||||
| White | 371 (56.7) | 129 (45.9) | 59 (40.1) | 167 (47.6) | |
| Black | 128 (19.6) | 130 (46.3) | 43 (29.3) | 143 (40.7) | |
| Chinese | 130 (19.9) | 5 (1.8) | 33 (22.5) | 11 (3.1) | |
| Hispanic | 25 (3.8) | 17 (6.1) | 12 (8.2) | 30 (8.6) | |
| Socioeconomic Strain | p<0.001abce | ||||
| Very Hard | 17 (2.7) | 16 (5.8) | 8 (5.5) | 31 (9.2) | |
| Somewhat Hard | 115 (18.0) | 74 (26.9) | 51 (35.2) | 111 (32.9) | |
| Not Hard at All | 507 (79.3) | 185 (67.3) | 86 (59.3) | 195 (57.9) | |
| Current Hormone Use | 48 (7.3) | 18 (6.4) | 8 (5.4) | 24 (6.8) | p =0.852 |
| Menopausal Status | p<0.001abcf | ||||
| Post-Menopausal | 634 (96.9) | 260 (92.5) | 137 (93.2) | 318 (90.6) | |
| Peri-Menopausal | 9(1.4) | 7 (2.5) | 4 (2.7) | 16 (4.6) | |
| Pre-Menopausal | 0 (0.0) | 1 (0.4) | 0 (0.0) | 0 (0.0) | |
| Unknown due to hormone therapy | 1 (0.2) | 3 (1.1) | 3 (2.0) | 0 (0.0) | |
| Unknown due to hysterectomy | 10 (1.5) | 10 (3.6) | 3 (2.0) | 17 (4.8) | |
| METABOLIC SYNDROME CHARACTERISTICS | |||||
| Waist Circumference (cm) | 80.8 ± 8.0 | 102.4 ± 10.9 | 91.4 ± 6.3 | 109.5 ± 11.2 | p<0.001abcdef |
| Triglycerides ≥ 150 (mg/dL) | 45 (6.9) | 15 (5.3) | 83 (56.5) | 127 (36.2) | p <0.001bcdef |
| HDL ≤ 45 (mg/dL) | 16 (2.5) | 4 (1.4) | 34 (23.1) | 80 (22.8) | p<0.001bcde |
| Hypertension | 251 (38.4) | 177 (63.0) | 125 (85.0) | 306 (87.2) | p<0.001abcde |
| Diabetes | 16 (2.5) | 4 (1.4) | 36 (24.5) | 146 (41.6) | p<0.001bcdef |
| ADDITIONAL COMORBIDITIES | |||||
| Smoking status | p<0.001bcde | ||||
| Never | 412 (64.0) | 171 (62.4) | 84 (58.7) | 177 (52.1) | |
| Former | 182 (28.3) | 89 (32.5) | 37 (25.9) | 121 (35.6) | |
| Current | 50 (7.8) | 14 (5.1) | 22 (15.4) | 42 (12.4) | |
| Family history of heart disease | 367 (56.1) | 168 (59.8) | 95 (64.6) | 244 (69.5) | p<0.001ce |
| Lipid-lowering medication | 102 (15.6) | 56 (19.9) | 69 (46.9) | 163 (46.4) | p<0.001bcde |
| OUTCOMES | |||||
| Maximum Carotid IMT | 0.880 ± 0.114 | 0.947 ± 0.115 | 0.959 ± 0.149 | 0.975 ± 0.152 | p<0.001abce |
| Adventitial Diameter | 6.993 ± 0.561 | 7.248 ± 0.657 | 7.367 ± 0.654 | 7.456 ± 0.701 | p<0.001abce |
Data are presented as frequency (percentage of non-missing data) or mean ± standard deviation. P-values were computed using chi-square, Fisher’s exact test, ANOVA, and independent t-tests, as appropriate.
P-value <0.05 comparing “no obesity or MetS” to “obesity with no MetS”
P-value <0.05 comparing “no obesity or MetS” to “no obesity with MetS”
P-value <0.05 comparing “no obesity or MetS” to “obesity with MetS”
P-value <0.05 comparing “obesity with no MetS” to “no obesity with MetS”
P-value <0.05 comparing “obesity with no MetS” to “obesity with MetS”
P-value <0.05 comparing “no obesity with MetS” to “obesity with MetS”
Abbreviations: MetS – metabolic syndrome
Overall, 654 women (45.6%) had neither obesity nor MetS, and 351 women (24.5%) had both conditions. We observed a high prevalence of metabolically healthy obesity. Of the 632 women with obesity by BMI criteria, 281 (44.5%) did not meet criteria for MetS. Furthermore, of the 498 women with MetS, 147 (29.5%) were not obese by BMI criteria; the majority of women in this group met race-specific waist circumference criteria (n=130, 88%), but had a median BMI in the middle of the overweight range (median 27.7 kg/m2, 25th percentile 26.4 kg/m2 and 75th percentile 29.1 kg/m2). Hypertension was the most prevalent component of MetS in the cohort, with a prevalence ranging from 38.4% in the non-obese, non-MetS group to 87.2% in the obese with MetS group.
The mean maximum cIMT in the full sample was 0.924mm (SD 0.135mm); women with obesity and women with MetS had 7.3% and 8.6%, respectively, higher mean maximum cIMT values as compared to women without obesity or MetS (both p<0.001; Table 1). The mean AD in the full sample was 7.191 (SD 0.655mm); women with obesity and women with MetS had 3.6% and 5.2%, respectively, higher AD as compared to women without (both p < 0.001).
Unadjusted models for the outcome of cIMT showed significantly higher mean cIMT values for obese compared to non-obese individuals, for those with MetS compared to those without MetS, and for those with either condition when both variables were included (Table 2). When both obesity and metabolic syndrome were included in the model without an interaction, there was a 31.8% attenuation of the association of obesity with maximum cIMT and a 28.0% attenuation of the association of MetS with maximum cIMT compared to the models without both covariates, though both remained statistically significant (p < 0.001). In the fully adjusted model, there was a statistically significant though not synergistic interaction between obesity and Mets. Women with both conditions had a model-adjusted predicted mean cIMT of 0.955mm (95%CI, 0.897–1.013; Figure 2). This value is 0.009mm higher than the model-adjusted predicted mean of those with MetS alone (mean 0.946mm; 95%CI, 0.887–1.005) and 0.025mm higher than the model-adjusted predicted mean of those with obesity alone (mean 0.930mm; 95%CI, 0.873–0.988). Those with neither obesity nor MetS had a model-adjusted mean cIMT of 0.878mm (95%CI, 0.821–0.935) when all other variables are held constant. While there were no statistically significant differences in cIMT by menopausal status, the vast majority of women (94%) were postmenopausal at the time of data collection. Black women had a 0.062mm higher cIMT than White women (95%CI, 0.046–0.078; p<0.001) in the fully adjusted model, though there were no statistically significant differences for Chinese or Hispanic women compared to White women.
Table 2.
Multivariable linear regression for the outcomes of maximum carotid intima media thickness and adventitial diameter
| Unadjusted Model | p-Value | Covariate Set 1 | p-Value | Covariate Set 2 | p-Value | |
|---|---|---|---|---|---|---|
|
| ||||||
| MAXIMUM CAROTID INTIMA MEDIA THICKNESS | ||||||
| Model Set 1 | ||||||
| Obesity | 0.069 (0.055 to 0.082) | p <0.001 | 0.055 (0.041 to 0.069) | p<0.001 | 0.051 (0.037 to 0.066) | p<0.001 |
| Model Set 2 | ||||||
| MetS | 0.071 (0.056 to 0.085) | p<0.001 | 0.063 (0.048 to 0.077) | p<0.001 | 0.057 (0.041 to 0.072) | p<0.001 |
| Model Set 3 | ||||||
| MetS | 0.051 (0.035 to 0.066) | p<0.001 | 0.049 (0.034 to 0.064) | p<0.001 | 0.043 (0.027 to 0.060) | p<0.001 |
| Obesity | 0.050 (0.035 to 0.065) | p<0.001 | 0.037 (0.022 to 0.052) | p<0.001 | 0.037 (0.022 to 0.053) | p<0.001 |
| Model Set 4 | ||||||
| No Obesity or MetS | Reference | NA | Reference | NA | Reference | NA |
| Obesity with no MetS | 0.068 (0.050 to 0.086) | <0.001 | 0.052 (0.033 to 0.070) | <0.001 | 0.052 (0.033 to 0.071) | <0.001 |
| No Obesity with MetS | 0.080 (0.056 to 0.103) | <0.001 | 0.073 (0.050 to 0.096) | <0.001 | 0.067 (0.044 to 0.091) | <0.001 |
| Obesity with MetS | 0.096 (0.079 to 0.113) | <0.001 | 0.082 (0.065 to 0.099) | <0.001 | 0.077 (0.059 to 0.095) | <0.001 |
| ADVENTITIAL DIAMETER OF CAROTID ARTERY | ||||||
| Model Set 1 | ||||||
| Obesity | 0.301 (0.234, 0.369) | p<0.001 | 0.300 (0.230 to 0.371) | p<0.001 | 0.295 (0.223 to 0.367) | p<0.001 |
| Model Set 2 | ||||||
| MetS | 0.361 (0.290, 0.431) | p<0.001 | 0.354 (0.285 to 0.424) | p<0.001 | 0.340 (0.265 to 0.415) | p<0.001 |
| Model Set 3 | ||||||
| MetS | 0.281 (0.206, 0.356) | p<0.001 | 0.282 (0.208, 0.356) | p<0.001 | 0.266 (0.187 to 0.345) | p<0.001 |
| Obesity | 0.198 (0.126, 0.270) | p<0.001 | 0.197 (0.124, 0.271) | p<0.001 | 0.210 (0.135 to 0.285) | p<0.001 |
| Model Set 4 | ||||||
| No Obesity or MetS | Reference | NA | Reference | NA | Reference | NA |
| Obesity with no MetS | 0.255 (0.166 to 0.343) | <0.001 | 0.246 (0.156 to 0.336) | <0.001 | 0.256 (0.164 to 0.348) | <0.001 |
| No Obesity with MetS | 0.374 (0.261 to 0.487) | <0.001 | 0.361 (0.249 to 0.472) | <0.001 | 0.341 (0.224 to 0.458) | <0.001 |
| Obesity with MetS | 0.463 (0.380 to 0.545) | <0.001 | 0.467 (0.383 to 0.551) | <0.001 | 0.464 (0.374 to 0.554) | <0.001 |
Data are presented as estimated β-coefficients with 95% confidence intervals. Abbreviations: MetS – metabolic syndrome
Covariate Set 1: adjusted for age, race, menopausal status, and hormonal therapy use
Covariate Set 2: adjusted for age, race, menopausal status, hormonal therapy use, smoking status, family history of heart disease, and lipid-lowering medication use
Figure 2. Adjusted multivariable linear regression model results.
Model-predicted mean and 95 percent confidence intervals for (A) maximum carotid intima media thickness and (B) adventitial diameter, after adjustment for age, menopausal status, hormonal therapy use, smoking status, family history of heart disease, and anti-lipemic medication use. Abbreviation: cIMT - carotid intima media thickness, MetS – metabolic syndrome
Similar to mean cIMT, the exposures of interest and their interaction were statistically significant for all models for the outcome of AD. After adjusting for age, race, hormonal therapy use, and menopausal status, all fixed effects of obesity and MetS remained statistically significant, as did the interaction. The attenuation from adjusting for these variables ranged from 0.9% - 3.6%. When both obesity and metabolic syndrome were included in the model without an interaction, there was a 33.7% attenuation of the effect of obesity on AD and a 24.4% attenuation of the effect of MetS on AD compared to the models without both covariates, though both remained statistically significant (p < 0.001). In the fully adjusted model, there was a statistically significant though not synergistic interaction between obesity and Mets. Women with both conditions had a model-adjusted predicted mean AD of 7.323mm (95%CI, 7.042– 7.605). This value is 0.123mm higher than the model-adjusted predicted mean of those with MetS alone (mean 7.200mm; 95%CI, 6.912–7.488) and 0.208mm higher than the model-adjusted predicted mean of those with obesity alone (mean 7.115mm; 95%CI, 6.834–7.396). Those with neither obesity nor MetS had a model-adjusted mean AD of 6.859mm (95%CI, 6.580–7.139) when all other variables are held constant. Compared to White women, Black women had a 0.212mm larger AD (95%CI, 0.134–0.290, p<0.001) and Chinese women had a 0.229mm larger AD (95%CI, 0.121–0.337, p<0.001) in the fully adjusted model. There were no significant differences in AD for Hispanic women.
Results were largely similar for sensitivity analyses (Table 3). When stratified by race, there were no statistically significant differences in cIMT between those with no obesity/MetS and those with obesity and MetS for White, Black, and Chinese women (Table 4). Similar results were seen in AD. Additionally, there were no statistically significant differences in cIMT or AD between those with no obesity/no MetS and those with no obesity and MetS for Hispanic women. Finally, model results were similar when a race-specific BMI threshold for obesity was used (data not shown).
Table 3.
Results of sensitivity analyses, excluding stratification by race.
| Sensitivity – No Overweight | P-Value | Sensitivity – No New Jersey Site | P-Value | Sensitivity – Only Post-Menopausal | P-Value | |
|---|---|---|---|---|---|---|
|
| ||||||
| MAXIMUM CAROTID INTIMA MEDIA THICKNESS | ||||||
|
| ||||||
| No Obesity or MetS | Reference | NA | Reference | NA | Reference | NA |
| Obesity with no MetS | 0.068 (0.046 to 0.089) | <0.001 | 0.055 (0.035 to 0.074) | <0.001 | 0.056 (0.037 to 0.075) | <0.001 |
| No Obesity with MetS | 0.078 (0.026 to 0.131) | 0.003 | 0.072 (0.047 to 0.097) | <0.001 | 0.073 (0.049 to 0.098) | <0.001 |
| Obesity with MetS | 0.091 (0.069 to 0.113) | <0.001 | 0.074 (0.054 to 0.093) | <0.001 | 0.079 (0.060 to 0.098) | <0.001 |
|
| ||||||
| ADVENTITIAL DIAMETER OF CAROTID ARTERY | ||||||
|
| ||||||
| No Obesity or MetS | Reference | NA | Reference | NA | Reference | NA |
| Obesity with no MetS | 0.328 (0.219 to 0.436) | <0.001 | 0.262 (0.166 to 0.357) | <0.001 | 0.284 (0.191 to 0.378) | <0.001 |
| No Obesity with MetS | 0.342 (0.081 to 0.603) | 0.010 | 0.360 (0.239 to 0.482) | <0.001 | 0.377 (0.258 to 0.496) | <0.001 |
| Obesity with MetS | 0.538 (0.428 to 0.647) | <0.001 | 0.439 (0.345 to 0.533) | <0.001 | 0.469 (0.377 to 0.561) | <0.001 |
Data are presented as estimated β-coefficients with 95% confidence intervals.
Models were run separately for each race and are adjusted for age, menopausal status, hormonal therapy use, smoking status, family history of heart disease, and anti-lipemic medication use
Abbreviations: MetS – metabolic syndrome
Table 4.
Multivariable linear regression stratified by race
| White | p-Value | Black | p-Value | Chinese | p-Value | Hispanic | p-Value | |
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| MAXIMUM CAROTID INTIMA MEDIA THICKNESS | ||||||||
| No Obesity or MetS | Reference | NA | Reference | NA | Reference | NA | Reference | NA |
| Obesity with no MetS | 0.071 (0.047 to 0.095) | <0.001 | 0.033 (−0.003 to 0.069) | 0.073 | 0.001 (−0.110 to 0.113) | 0.980 | −0.010 (−0.097 to 0.076) | 0.813 |
| No Obesity with MetS | 0.041 (0.008 to 0.075) | 0.016 | 0.084 (0.032 to 0.137) | 0.002 | 0.118 (0.066 to 0.170) | <0.001 | 0.017 (−0.074 to 0.108) | 0.703 |
| Obesity with MetS | 0.065 (0.042 to 0.088) | <0.001 | 0.077 (0.039 to 0.114) | <0.001 | 0.134 (0.050 to 0.219) | 0.002 | 0.109 (0.036 to 0.182) | 0.004 |
| ADVENTITIAL DIAMETER OF CAROTID ARTERY | ||||||||
| No Obesity or MetS | Reference | NA | Reference | NA | Reference | NA | Reference | NA |
| Obesity with no MetS | 0.288 (0.169 to 0.408) | <0.001 | 0.244 (0.071 to 0.418) | 0.006 | 0.244 (−0.277 to 0.765) | 0.357 | −0.032 (−0.476 to 0.412) | 0.886 |
| No Obesity with MetS | 0.333 (0.164 to 0.503) | <0.001 | 0.366 (0.113 to 0.620) | 0.005 | 0.372 (0.129 to 0.615) | 0.003 | 0.136 (−0.330 to 0.602) | 0.562 |
| Obesity with MetS | 0.400 (0.285 to 0.515) | <0.001 | 0.515 (0.335 to 0.700) | <0.001 | 0.583 (0.187 to 0.978) | 0.004 | 0.487 (0.114 to 0.860) | 0.012 |
Data are presented as estimated β-coefficients with 95% confidence intervals.
Models were run separately for each outcome and race, and are adjusted for age, menopausal status, hormonal therapy use, smoking status, family history of heart disease, and anti-lipemic medication use
Abbreviations: MetS – metabolic syndrome
DISCUSSION
In this cohort of mid- to early late adulthood women, we confirmed that obesity and metabolic syndrome were strongly associated with carotid IMT and adventitial diameter, markers of atherosclerosis. Importantly, however, we observed a negative interaction between obesity and metabolic syndrome such that the co-occurrence of both conditions did not result in an additive or multiplicative impact on cIMT; instead, for women who had metabolic syndrome the association with cIMT was largely similar regardless of if they concurrently had obesity. There was a similar joint effect of obesity and MetS seen for the outcome of adventitial diameter. This suggests that there is no additional effect of having obesity with MetS on measures of subclinical atherosclerosis; clinicians treating normal-weight individuals with MetS should monitor them for atherosclerotic changes as they would those with obesity and MetS. Given the known risk of developing MetS and constituent components during the menopausal transition, these findings within this population are particularly important.
In our study, we found that 45% of women with obesity did not have MetS and 30% of women with MetS did not have obesity. This prevalence of those with metabolically healthy obesity is higher than what has been reported in the literature7. This may be due in part to the definition of obesity by BMI rather than waist circumference. Waist circumference, part of the definition of metabolic syndrome, is a measure of adiposity rather than a ratio of height to weight. Additionally, there was a large difference in rate of those with obesity without MetS by race, with the lowest number of those with metabolically healthy obesity in the Chinese women (31%) and the highest number in the Black women (48%). Despite the higher reported prevalence of metabolically healthy obesity, our study found that, independently, obesity was associated with higher cIMT and AD, which has been shown frequently in the literature19. After full model adjustment, however, we found that the effect of having both obesity and MetS on both cIMT and AD was not much higher than that of just MetS alone. These results were unexpected but are thought to have been driven by what our outcomes are measuring. Specifically, cIMT and AD reflect arterial remodeling and hypertrophy of medial layer due to aging and hypertension rather than exclusively atherosclerosis20. This is even more the case in the common carotid artery, where atherosclerotic lesions tend to develop with lower frequency21.
Both MetS and obesity are known to have strong systemic inflammatory effects, which can cause long-term damage to the cardiovascular system and increase atherosclerosis risk22,23. The unadjusted and adjusted model results within our study support this. There was not a synergistic joint effect of obesity and MetS on cIMT and AD as was expected. This may be because those with MetS are already in a constant inflammatory state; all of the conditions encompassed within MetS are associated with inflammatory processes within the body, including obesity 24. Adipose tissue can produce a number of proinflammatory adipokines associated with subclinical atherosclerosis, such as adiponectin and leptin22. However, a recent study of these SWAN women found that after adjusting for other cardiovascular risk factors, there was no association between adiponectin and cIMT or AD22. Additionally, neither BMI nor waist circumference fully account for adipose tissue distribution across the body. Visceral adipose tissue produces pro-inflammatory cytokines and is more vascular than subcutaneous adipose tissue8. Interestingly, with the exception of the no obesity/no MetS cohort, all groups had a waist circumference meeting the MetS definition of abdominal obesity (≥80 cm for Chinese women, otherwise >88cm). Additionally, there was a high rate of hypertension (63%) in group of women with obesity but without MetS. Thus, while the majority of these women have hypertension, they lack additional criteria to satisfy the metabolic syndrome definition. These factors could also have affected the minimal additional risk of obesity seen with MetS. . Given the similar effect of MetS and obesity on cIMT, it is important to recognize the increased risk associated with MetS even among those without obesity.
Some of the results of this study may be due to the high prevalence of metabolically healthy obese in our cohort. A large recent study of UK Biobank data suggests these individuals may not be at higher risk of atherosclerotic cardiovascular disease, including myocardial infarction and stroke, compared to those of normal weight25. However, whether these individuals are truly lower risk than those with metabolically unhealthy obesity remains controversial. Part of this is due to the transitory nature of metabolic healthy obesity. Indeed, a recent study within the Multi-Ethnic Study of Atherosclerosis cohort found that transition from metabolically healthy obesity to metabolically unhealthy obesity occurred at a rate of 44% over ten years, with higher baseline BMI and longer duration of obesity showing significant association with this transition26. They additionally found that 26% of those without obesity at baseline transitioned having metabolic syndrome during the same time period26. An additional study within the same cohort found that presence of MetS mediated 62% of the relationship between obesity and cardiovascular disease27. In combination, these findings could explain why there was minimal additional risk of ASCVD seen with obesity over MetS alone.
The current study is one of the few studies to examine the interaction effect of metabolic syndrome and obesity on the carotid atherosclerosis markers of maximum IMT and adventitial diameter. A cIMT greater than 0.6mm has been shown to be associated with at least 2.5-fold higher risk of cardiovascular events, highlighting the significance of early detection1. The physiology of both obesity and metabolic syndrome are complex, and identifying this relationship is the first step to disentangling how these factors interplay with future risk for atherosclerosis. Future research is needed to determine if certain conditions within the metabolic syndrome are driving the associations seen in the current study, and if there might be a differential effect of obesity on carotid artery health in the presence of these conditions. Additionally, the joint effect of MetS and obesity on cardiovascular outcomes including myocardial infarction, stroke, and mortality warrants investigation.
Study Strengths
A major strength of this study is that the cohort is from a well-established, multi-ethnic, multi-center study of women during a critical time of risk – the transition from middle to late adulthood, and across the menopausal transition. The central ultrasound reading and interpretation for all sites ensured consistency of results between data collected from different sites. Additionally, capture of multiple measures of the metabolic syndrome components including treatment medications for these conditions allowed for a comprehensive definition beyond what laboratory and physiologic values alone can show.
Study Limitations
This study has several limitations, most notably that cIMT was only captured at one time point for each participant. Causation therefore cannot be determined because we are unable to establish temporality between the cIMT measures and the MetS and obesity outcomes. Because of this, findings from this study can not be considered as definitive and should instead generate further hypotheses on this topic. Future studies which are able to establish temporality between obesity/MetS and atherosclerotic damage are needed, as are longitudinal studies with repeated assessment of cIMT.
Another potential limitation is the possibility of residual confounding. Although we adjusted for several important covariates in our models including socioeconomic strain, family history of heart disease, smoking status, menopausal status and use of lipid-lowering medications, there is the possibility of residual confounding. For example, although we characterized smoking status as current, former, or never, there may be confounding by cumulative smoking exposure that is not fully captured and may be important. Additionally, the current analysis did not examine circulating inflammatory biomarkers or biomarkers associated with the inflammatory status of the endothelium, which may be important covariates. Finally, the study cohort did not include men, and thus the findings cannot be generalized to males.
CONCLUSION
Atherosclerosis can lead to major clinical sequelae; identification of subclinical ASCVD may allow provider and patients to identify and manage risk factors earlier. The findings from this study suggest that there is only a minimal impact of having obesity on carotid artery thickness over the effect of having metabolic syndrome alone. Clinicians should monitor all individuals with MetS for vascular changes regardless of obesity status, even while waiting for interventions such as weight loss or an insulin regimen to reduce metabolic syndrome symptomology.
ACKNOWLEDGEMENTS
We thank the study staff at each site and all the women who participated in SWAN.
The Study of Women’s Health Across the Nation (SWAN) has grant support from the National Institutes of Health (NIH), Department of Health and Human Services (DHHS), through the National Institute on Aging (NIA), the National Institute of Nursing Research (NINR) and the National Institutes of Health Office of Research on Women’s Health (ORWH) (Grants U01NR004061; U01AG012505, U01AG012535, U01AG012531, U01AG012539, U01AG012546, U01AG012553, U01AG012554, U01AG012495, and U19AG063720). The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Aging, National Institute of Nursing Research, Office of Research on Women’s Health or the National Institutes of Health.
Clinical Centers: University of Michigan, Ann Arbor – Carrie Karvonen-Gutierrez, PI 2021 – present, Siobán Harlow, PI 2011 – 2021, MaryFran Sowers, PI 1994–2011; Massachusetts General Hospital, Boston, MA – Sherri-Ann Burnett-Bowie, PI 2020 – Present; Joel Finkelstein, PI 1999 – 2020; Robert Neer, PI 1994 – 1999; Rush University, Rush University Medical Center, Chicago, IL – Imke Janssen, PI 2020 – Present; Howard Kravitz, PI 2009 – 2020; Lynda Powell, PI 1994 – 2009; University of California, Davis/Kaiser – Elaine Waetjen and Monique Hedderson, PIs 2020 – Present; Ellen Gold, PI 1994 – 2020; University of California, Los Angeles – Arun Karlamangla, PI 2020 – Present; Gail Greendale, PI 1994 – 2020; Albert Einstein College of Medicine, Bronx, NY – Carol Derby, PI 2011 – present, Rachel Wildman, PI 2010 – 2011; Nanette Santoro, PI 2004 – 2010; University of Medicine and Dentistry – New Jersey Medical School, Newark – Gerson Weiss, PI 1994 – 2004; and the University of Pittsburgh, Pittsburgh, PA – Rebecca Thurston, PI 2020 – Present; Karen Matthews, PI 1994 – 2020.
NIH Program Office: National Institute on Aging, Bethesda, MD – Rosaly Correa-de-Araujo 2020 - present; Chhanda Dutta 2016- present; Winifred Rossi 2012–2016; Sherry Sherman 1994 – 2012; Marcia Ory 1994 – 2001; National Institute of Nursing Research, Bethesda, MD – Program Officers.
Central Laboratory: University of Michigan, Ann Arbor – Daniel McConnell (Central Ligand Assay Satellite Services).
Coordinating Center: University of Pittsburgh, Pittsburgh, PA – Maria Mori Brooks, PI 2012 - present; Kim Sutton-Tyrrell, PI 2001 – 2012; New England Research Institutes, Watertown, MA - Sonja McKinlay, PI 1995 – 2001.
Steering Committee: Susan Johnson, Current Chair; Chris Gallagher, Former Chair
Footnotes
Conflicts of Interests: Elizabeth A Jackson receives ongoing institutional funding from the National Institute of Health, AHRQ, and the VA health system. Dr. Jackson is a member of the AHA editorial board, a consultant for ACC and Optum Health, an expert witness for DeBlase, Brown, and Everly LLP, and received past institutional funding from Amgen Research. Carrie A Karvonen-Gutierrez received payment from the University of Jyväskylä Finland for services as expert reviewer of a dissertation. The other authors have nothing to disclose.
An abstract of this work was presented at the 2021 Gerontological Society of America scientific research meeting.
REFERENCES
- 1.Chambless LE, Heiss G, Folsom AR, et al. Association of coronary heart disease incidence with carotid arterial wall thickness and major risk factors: the Atherosclerosis Risk in Communities (ARIC) Study, 1987–1993. Am J Epidemiol. Sep 15 1997;146(6):483–94. doi: 10.1093/oxfordjournals.aje.a009302 [DOI] [PubMed] [Google Scholar]
- 2.Sobiczewski W, Wirtwein M, Trybala E, Gruchala M. Severity of coronary atherosclerosis and stroke incidence in 7-year follow-up. J Neurol. Jul 2013;260(7):1855–8. doi: 10.1007/s00415-013-6892-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Makover ME, Shapiro MD, Toth PP. There is urgent need to treat atherosclerotic cardiovascular disease risk earlier, more intensively, and with greater precision: A review of current practice and recommendations for improved effectiveness. Am J Prev Cardiol. Dec 2022;12:100371. doi: 10.1016/j.ajpc.2022.100371 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.National Heart L, and Blood Institute. What Is Atherosclerosis? Accessed 9 22, 2022, 2022. https://www.nhlbi.nih.gov/health/atherosclerosis [Google Scholar]
- 5.Grundy SM, Cleeman JI, Daniels SR, et al. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation. Oct 25 2005;112(17):2735–52. doi: 10.1161/CIRCULATIONAHA.105.169404 [DOI] [PubMed] [Google Scholar]
- 6.Smitka K, Marešová D. Adipose Tissue as an Endocrine Organ: An Update on Pro-inflammatory and Anti-inflammatory Microenvironment. Prague Med Rep. 2015;116(2):87–111. doi: 10.14712/23362936.2015.49 [DOI] [PubMed] [Google Scholar]
- 7.Blüher M Metabolically Healthy Obesity. Endocr Rev. 05 January 2020;41(3)doi: 10.1210/endrev/bnaa004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ibrahim MM. Subcutaneous and visceral adipose tissue: structural and functional differences. Obes Rev. Jan 2010;11(1):11–8. doi: 10.1111/j.1467-789X.2009.00623.x [DOI] [PubMed] [Google Scholar]
- 9.Iglseder B, Cip P, Malaimare L, Ladurner G, Paulweber B. The metabolic syndrome is a stronger risk factor for early carotid atherosclerosis in women than in men. Stroke. Jun 2005;36(6):1212–7. doi: 10.1161/01.STR.0000166196.31227.91 [DOI] [PubMed] [Google Scholar]
- 10.Giltay EJ, Lambert J, Elbers JM, Gooren LJ, Asscheman H, Stehouwer CD. Arterial compliance and distensibility are modulated by body composition in both men and women but by insulin sensitivity only in women. Diabetologia. Feb 1999;42(2):214–21. doi: 10.1007/s001250051141 [DOI] [PubMed] [Google Scholar]
- 11.Kawamoto R, Tomita H, Inoue A, Ohtsuka N, Kamitani A. Metabolic syndrome may be a risk factor for early carotid atherosclerosis in women but not in men. J Atheroscler Thromb. Feb 2007;14(1):36–43. doi: 10.5551/jat.14.36 [DOI] [PubMed] [Google Scholar]
- 12.Lin HF, Liu CK, Liao YC, Lin RT, Chen CS, Juo SH. The risk of the metabolic syndrome on carotid thickness and stiffness: sex and age specific effects. Atherosclerosis. May 2010;210(1):155–9. doi: 10.1016/j.atherosclerosis.2009.11.027 [DOI] [PubMed] [Google Scholar]
- 13.Napoleone JM, Boudreau RM, Lange-Maia BS, et al. Metabolic Syndrome Trajectories and Objective Physical Performance in Mid-to-Early Late Life: The Study of Women’s Health Across the Nation (SWAN). J Gerontol A Biol Sci Med Sci. Feb 03 2022;77(2):e39–e47. doi: 10.1093/gerona/glab188 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sowers M, Crawford S, Sternfeld B, et al. SWAN: A Multicenter, Multiethnic, Community-Based Cohort Study of Women and the Menopausal Transition. Academic Press; 2000:175–188. [Google Scholar]
- 15.Thurston RC, Chang Y, Derby CA, et al. Abuse and subclinical cardiovascular disease among midlife women: the study of women’s health across the nation. Stroke. Aug 2014;45(8):2246–51. doi: 10.1161/STROKEAHA.114.005928 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wendelhag I, Gustavsson T, Suurküla M, Berglund G, Wikstrand J. Ultrasound measurement of wall thickness in the carotid artery: fundamental principles and description of a computerized analysing system. Clin Physiol. Nov 1991;11(6):565–77. doi: 10.1111/j.1475-097x.1991.tb00676.x [DOI] [PubMed] [Google Scholar]
- 17.Thurston RC, El Khoudary SR, Sutton-Tyrrell K, et al. Are vasomotor symptoms associated with alterations in hemostatic and inflammatory markers? Findings from the Study of Women’s Health Across the Nation. Menopause. Oct 2011;18(10):1044–51. doi: 10.1097/gme.0b013e31821f5d39 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Appelhans BM, Baylin A, Huang MH, et al. Beverage Intake and Metabolic Syndrome Risk Over 14 Years: The Study of Women’s Health Across the Nation. J Acad Nutr Diet. Apr 2017;117(4):554–562. doi: 10.1016/j.jand.2016.10.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kaul A, Kumar A, Baksi A, et al. Impact of bariatric surgery on carotid intima-medial thickness and cardiovascular risk: results of a prospective study. Surg Endosc. November 2021;35(11):6006–6012. doi: 10.1007/s00464-020-08088-0 [DOI] [PubMed] [Google Scholar]
- 20.Xu X, Wang B, Ren C, et al. Age-related Impairment of Vascular Structure and Functions. Aging Dis. Oct 2017;8(5):590–610. doi: 10.14336/AD.2017.0430 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Riles TS, Imparato AM, Posner MP, Eikelboom BC. Common carotid occlusion. Assessment of the distal vessels. Ann Surg. Mar 1984;199(3):363–6. doi: 10.1097/00000658-198403000-00019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Everson-Rose SA, Barinas-Mitchell EJM, El Khoudary SR, et al. Adipokines and Subclinical Cardiovascular Disease in Post-Menopausal Women: Study of Women’s Health Across the Nation. J Am Heart Assoc. 04 June 2021;10(7):e019173. doi: 10.1161/JAHA.120.019173 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Nikolopoulou A, Kadoglou NP. Obesity and metabolic syndrome as related to cardiovascular disease. Expert Rev Cardiovasc Ther. Jul 2012;10(7):933–9. doi: 10.1586/erc.12.74 [DOI] [PubMed] [Google Scholar]
- 24.Marsland AL, McCaffery JM, Muldoon MF, Manuck SB. Systemic inflammation and the metabolic syndrome among middle-aged community volunteers. Metabolism. Dec 2010;59(12):1801–8. doi: 10.1016/j.metabol.2010.05.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Zhou Z, Macpherson J, Gray SR, et al. Are people with metabolically healthy obesity really healthy? A prospective cohort study of 381,363 UK Biobank participants. Diabetologia. Sep 2021;64(9):1963–1972. doi: 10.1007/s00125-021-05484-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Mongraw-Chaffin M, Foster MC, Kalyani RR, et al. Obesity Severity and Duration Are Associated With Incident Metabolic Syndrome: Evidence Against Metabolically Healthy Obesity From the Multi-Ethnic Study of Atherosclerosis. J Clin Endocrinol Metab. November 2016;101(11):4117–4124. doi: 10.1210/jc.2016-2460 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Mongraw-Chaffin M, Foster MC, Anderson CAM, et al. Metabolically Healthy Obesity, Transition to Metabolic Syndrome, and Cardiovascular Risk. J Am Coll Cardiol. 05 January 2018;71(17):1857–1865. doi: 10.1016/j.jacc.2018.02.055 [DOI] [PMC free article] [PubMed] [Google Scholar]