Characteristics of Metabolic Syndrome in Morbidly Obese Subjects

Abstract

Background:

Metabolic syndrome (MetSyn) magnifies risks of cardiovascular disease (CVD) and type 2 diabetes, but its expression varies within the obese population. We examined body mass index (BMI), metabolic traits, and fat distribution in morbidly obese individuals.

Methods:

Lipids and inflammatory, oxidative stress and hepatic biomarkers in 346 women and 203 men (BMI ≥35 kg/m² and co-morbidity or ≥40 kg/m²) were stratified by MetSyn components (1–5, excluding diabetes). Age- and smoking-adjusted partial correlations were calculated. Dual-energy X-ray absorptiometry was measured in 206 participants.

Results:

Apolipoprotein B, ferritin, uric acid, and alanine aminotransferase (ALT) concentrations worsened with increasing MetSyn components (P ≤ 0.0001), while BMI and LDL-cholesterol showed no association. BMI correlated inversely with triglycerides (r = −0.16, P = 0.03) and positively with HDL-cholesterol in men (r = 0.16, P = 0.02), but not in women. BMI correlated with C-reactive protein (CRP) (r = 0.32, P < 0.0001; r = 0.24, P < 0.0001 in men and women, respectively) and white blood cell count (r = 0.24, P = 0.001 in men; r = 0.15, P = 0.008 in women). Truncal fat percentage correlated to CRP (r = 0.31, P = 0.03; r = 0.20, P = 0.02 in men and women, respectively). In women, number of MetSyn components was inversely related to truncal and peripheral fat (r = −0.20, P = 0.02; r = −0.42, P < 0.0001, respectively) as was ALT (r = −0.21, P = 0.009; r = −0.38, P < 0.0001, respectively) and triglycerides with peripheral fat (r = −0.38, P < 0.0001), while HDL cholesterol was positively associated with truncal and peripheral fat (r = 0.26; P = 0.001).

Conclusions:

BMI and fat distribution showed expected associations to inflammation biomarkers, but paradoxical relations between fat indices, and MetSyn components and biomarkers were seen. This suggests a need for better markers of CVD risk in morbid obesity.

Introduction

Obesity represents one of the most complex public health challenges for our societies. Cardiometabolic disorders (metabolic syndrome [MetSyn] and type 2 diabetes) follow in the wake of obesity adding to health burdens on the population. Of all diseases caused by obesity, relative risks of type 2 diabetes are the highest.¹ Type 2 diabetes is characterized by insulin resistance that is tightly related to abdominal obesity and is accompanied by changes in metabolic traits and subclinical inflammation that lead to cardiovascular disease (CVD).²

Clinically the presence of three or more of five risk factors, namely abdominal obesity, low HDL cholesterol, high triglycerides, high fasting glucose, and high blood pressure characterizes MetSyn.³

Persons with morbid obesity, defined as body mass index (BMI) ≥35 kg/m² with co-morbidities or ≥40 kg/m² regardless of co-morbidities, typically carry increased burdens of psychological and physical discomfort and disease. While the risk of CVD is clearly magnified by the presence of MetSyn or type 2 diabetes, CVD risk factors and metabolic traits may not increase linearly with increasing BMI and may differ by gender.⁴ While lipid disturbances are common, they are not uniformly related to total body fat mass.⁵ Subgroups of morbid obesity have been identified, which appear metabolically healthy and exhibit normal lipids or high insulin sensitivity.^6,7 Furthermore, waist circumference is uniformly above standard cutoff levels that are used to define MetSyn and thus of lesser utility in determining risks in the morbidly obese population.

The purpose of the study was to examine the relationships of MetSyn components and CVD risk factors to BMI and fat distribution in morbidly obese men and women. We examined subjects with morbid obesity grouped according to their degree of metabolic disturbance, with the goal of better understanding the expression and utility of CV risk markers in this population.

Materials and Methods

Patients referred consecutively to the Preventive Cardiology Clinic at Oslo University Hospital, Oslo, Norway, participated between April 2005 and December 2010. The study conformed to the Helsinki Declaration and was evaluated by the Ethics committee for region 1 in Norway.

Participants with BMI ≥35 kg/m² and co-morbidity, or related disorders (including hypertension, sleep apnea, dyspnea, polycystic ovarian syndrome, asthma, hypercholesterolemia, gout, musculoskeletal symptoms, gall bladder symptoms, esophageal reflux, pulmonary or deep vein embolism, intermittent claudication, angina pectoris, depression, or eating disorder) or BMI ≥40 kg/m² regardless of co-morbidity, but without diabetes were included (n = 549). Subjects with type 2 diabetes (n = 209) were not included in current analyses.

After written informed consent, participants completed a health questionnaire and underwent anthropometric measurements. A constant tension body tape measure was used to determine waist and hip circumferences. Waist circumference was measured at midpoint between the inferior costal margin and the highest point of the iliac crest, and hip circumference was measured at the widest point around the hips. Height was measured using a stadiometer and recorded to the nearest cm. Patients were weighed to the nearest 1.0 kg using a calibrated mobile electronic scale (Seca 720; Medical Scales and Measuring Systems). BMI was calculated in accordance to the Quetelet's formula: Body weight in kilograms divided by the square of body height in meters (kg/m²). Body total and regional fat percentages were analyzed by using dual X-ray absorptiometry (DXA) (Lunar DPX-L, Lunar) in 206 individuals. Only individuals with body weights below 140 kg could be examined by DXA. Blood pressure was measured automatically using an automatic blood pressure monitor (52000 Series Vital Signs Monitor; Welch Ally) following a 5-min rest.

Subjects were stratified by number of MetSyn components (from 1 to 5). Definition of MetSyn components was as follows: waist circumference ≥102 cm for men and ≥88 cm for women; systolic blood pressure ≥130 mmHg and/or diastolic ≥85 mmHg, or drug treatment for hypertension; triglycerides ≥1.7 mM; HDL-cholesterol ≤1.0 for men or ≤1.3 for women; and fasting glucose ≥ 5.6 mM.

Laboratory analyses

Participants were instructed to fast overnight for at least 10 hr, before providing blood samples between 8:00 and 11:00 a.m. Analyses of blood samples were performed at Oslo University Hospital (Clinical Chemistry Laboratory at Ullevål and Endocrine Laboratory at Aker). Total cholesterol, HDL-cholesterol, triglycerides, glucose, alanine aminotransferase (ALT), uric acid, creatinine, and high-sensitivity C-reactive protein (CRP) concentrations were measured on an automated analyzer Cobas Integra 800 (Roche Diagnostics). LDL-cholesterol was calculated using Friedewald's formula. Apolipoprotein B and lipoprotein (a) were determined with an immunoturbidimetric assay on an automated analyzer (Cobas Tinaquant 917, Roche/Hitachi; Roche Diagnostics). White blood cells were analyzed using Sysmex XE 2100 (Sysmex). Serum ferritin was determined by an ADIVA Centaur analysis (ADIVA Centaur; Siemens Healthcare Diagnostics, Inc.).

Statistical analyses

Statistical analyses were performed using SPSS 21 (SPSS, Inc.). Categorical and continuous data are presented as counts and percentages, or mean ± standard deviation (SD), respectively. Variables were tested for normality and logarithmically transformed values for lipoprotein (a), CRP, ferritin, white blood cell count, and ALT were used where indicated. One-way analysis of variance (ANOVA) was performed to compare components of MetSyn, other lipids, and inflammatory biomarkers. Partial correlation coefficients (corrected for age and smoking) were calculated to analyze correlations between BMI and waist circumference and MetSyn components, other lipids, and inflammatory biomarkers. Two-sided P values of <0.05 were considered statistically significant.

Results

A total of 346 women and 203 men aged between 18 and 78 years and BMI between 35 kg/m² and 74 kg/m² participated. Sample characteristics and MetSyn components are presented in Table 1. Two-thirds (373 of 547) met criteria of MetSyn (three or more components). Age and the proportion of males increased with increasing numbers of components of MetSyn, and the level of each component increased with increasing numbers of components. BMI was not associated with number of MetSyn components. As shown in Table 2, there was no relationship between LDL-cholesterol concentrations and MetSyn components, while Lp(a) concentrations decreased as the number of components increased. Total cholesterol, apolipoprotein B, ferritin, uric acid, and ALT levels increased as number of components of MetSyn increased.

Table 1.

Age, Gender, Anthropometrics, Smoking Status, Components of Metabolic Syndrome, and Medication Use Stratified by Number of Metabolic Syndrome Components (N = 547)

	Number of components
Parameter	One n = 43	Two n = 131	Three n = 200	Four n = 121	Five n = 52	P
Age, years	33.5 (11.6)	39.9 (12.9)	40.6 (12.5)	40.6 (11.2)	44.9 (9.6)	<0.0001
Male, n (percentage)	6 (14.0)	42 (32.1)	74 (37.0)	54 (44.6)	27 (51.9)	0.001
BMI, kg/m²	42.1 (5.1)	42.5 (5.3)	42.4 (5.4)	42.3 (5.4)	41.6 (5.2)	0.88
Smokers, n (percentage)	13 (30.2)	25 (19.1)	45 (22.5)	31 (25.6)	17 (32.7)	0.26
Metabolic syndrome components
Waist circumference, cm	116.5 (11.6)	122.6 (14.6)	124.9 (15.0)	127.0 (14.3)	125.3 (14.0)	0.001
Systolic BP, mm Hg	117 (8)	130 (14)	135 (18)	135 (14)	138 (15)	<0.0001
Diastolic BP, mmHg	77 (6)	83 (9)	86 (11)	88 (10)	90 (10)	<0.0001
HDL-cholesterol, mM	1.5 (0.3)	1.3 (0.3)	1.2 (0.3)	1.1 (0.2)	0.9 (0.2)	<0.0001
Triglycerides, mM	1.1 (0.3)	1.2 (0.4)	1.8 (1.4)	2.4 (1.1)	3.2 (1.6)	<0.0001
Glucose, mM	4.9 (0.3)	5.1 (0,4)	5.4 (0.7)	5.5 (0.6)	6.0 (0.4)	<0.0001
Use of statins (percentage)	0 (0)	9 (6.9)	22 (11.0)	13 (10.7)	6 (11.5)	0.15
Use of antihypertensive drugs (percentage)	0 (0)	27 (20.6)	57 (28.5)	42 (34.7)	21 (40.4)	<0.0001

Mean (SD) shown except for percentages. P value indicates ANOVA comparison across groups.

BMI, body mass index; BP, blood pressure.

Table 2.

Other Lipids and Lipoproteins, Biomarkers of Inflammation, and Other Markers of Metabolic Syndrome Stratified by Number of Metabolic Syndrome Components (N = 547)

	Number of components
Parameter	One n = 43	Two n = 131	Three n = 200	Four n = 121	Five n = 52	P^a
Other lipids
Total cholesterol, mM	5.2 (1.1)	5.0 (0.9)	5.3 (1.2)	5.5 (1.1)	5.4 (0.9)	0.009
LDL-cholesterol, mM	3.2 (1.0)	3.2 (1.2)	3.3 (1.0)	3.4 (0.9)	3.1 (1.0)	0.56
Lipoprotein (a), nM	319 (122, 782)	222 (84, 508)	144 (63, 354)	180 (87, 471)	161 (63, 280)	0.03
Apolipoprotein B, g/L	0.9 (0.2)	0.9 (0.2)	1.0 (0.2)	1.1 (0.2)	1.1 (0.2)	<0.0001
Biomarkers
C-reactive protein, mg/L	8.5 (4.0, 16.0)	6.0 (3.0, 11.0)	6.0 (3.0, 11.0)	6.0 (3.3, 10.8)	5.0 (3.0, 8.0)	0.09
White blood cells, × 10⁹/L	6.7 (5.6, 8.3)	6.5 (5.4, 8.1)	6.8 (5.8, 8.0)	6.8 (5.8, 7.9)	6.9 (5.8, 8.2)	0.24
Ferritin, μg/L	54 (38, 102)	73 (36, 150)	92 (30, 150)	118 (51, 211)	162 (77, 264)	<0.0001
Uric acid, μM	336 (74)	363 (75)	385 (91)	405 (81)	431 (85)	<0.0001
ALT, U/L	20 (17, 31)	30 (20, 47)	31 (22, 45)	34 (24, 47)	36 (24, 57)	<0.0001

Mean (SD) shown except for lipoprotein (a), C-reactive protein (CRP), ferritin, white blood cells, and alanine aminotransferase (ALT) for which median (25th, 75th percentiles) are shown.

For total cholesterol, 1 missing; for LDL-cholesterol, 25 missing (could not be estimated because high level of triglycerides); for lipoprotein (a), 12 missing; for apolipoprotein B, 14 missing; for CRP, 10 missing; for ferritin, 14 missing; for white blood cells, 18 missing; for uric acid, 13 missing; and for ALT, 12 missing.

P value indicates ANOVA comparison across groups.

BMI and waist circumference were highly correlated (r = 0.67, P < 0.0001 in women; r = 0.85, P < 0.0001 in men). Partial correlation coefficients between BMI and waist circumference and metabolic traits and biomarkers are shown in Table 3. Number of MetSyn components did not correlate with BMI, and weakly with waist circumference (only in women). In men, all components of MetSyn correlated with BMI, with the exception of fasting glucose. However, waist circumference correlated only with blood pressure. The correlations of HDL-cholesterol and triglycerides to BMI and waist circumference were opposite to expected (positively for HDL-cholesterol and inversely for triglycerides). In women, only systolic blood pressure correlated with BMI and waist circumference.

Table 3.

BMI and Waist Circumference Correlations with Number of Metabolic Syndrome Components, Components of Metabolic Syndrome, Other Lipids and Lipoproteins, Biomarkers of Inflammation, and Other Markers of Metabolic Syndrome, Adjusted for Age and Smoking Status (N = 547)

	Male (n = 203)				Female (n = 344)
	BMI		Waist circumference		BMI		Waist circumference
	Coefficient	P	Coefficient	P	Coefficient	P	Coefficient	P
#MetSyn components	−0.07	0.34	0.01	0.88	0.05	0.38	0.11	0.04
Systolic BP (all)	0.27	<0.0001	0.29	<0.0001	0.19	<0.0001	0.19	<0.0001
Diastolic BP (all)	0.25	<0.0001	0.32	<0.0001	0.01	0.79	0.05	0.34
Systolic BP (no medication)	0.32	<0.0001	0.35	<0.0001	0.21	<0.0001	0.23	<0.0001
Diastolic BP (no medication)	0.25	0.004	0.29	<0.0001	0.05	0.46	0.09	0.14
HDL-cholesterol	0.16	0.02	0.08	0.24	0.02	0.69	0.01	0.83
Triglycerides	−0.16	0.03	−0.11	0.12	−0.02	0.71	0.03	0.59
Glucose	0.13	0.06	0.07	0.34	0.02	0.79	0.07	0.21
Total cholesterol	0.07	0.33	0.01	0.89	−0.06	0.24	0.03	0.56
LDL-cholesterol	0.13	0.07	0.04	0.60	−0.04	0.44	0.04	0.45
Lipoprotein (a)	−0.05	0.45	−0.08	0.29	−0.07	0.20	−0.10	0.07
Apolipoprotein B	0.11	0.14	0.07	0.30	−0.06	0.30	0.03	0.59
C-reactive protein	0.32	<0.0001	0.27	<0.0001	0.24	<0.0001	0.24	<0.0001
White blood cells	0.24	0.001	0.20	0.005	0.15	0.008	0.19	0.001
Ferritin	0.03	0.72	0.09	0.21	−0.01	0.83	0.04	0.49
Uric acid	0.04	0.57	0.02	0.75	0.19	<0.0001	0.10	0.07
ALT	0.05	0.47	0.06	0.38	0.07	0.24	0.14	0.01

CRP concentrations and white blood cell count correlated with both BMI and waist circumference in both genders. Uric acid concentrations correlated with BMI, while ALT correlated with waist circumference in women.

The mean (SD) BMI of the group with DXA measurements was 40.5 (3.6) kg/m², versus 43.4 (5.9) kg/m² in the group that did not take part in DXA (P < 0.0001). As was the case in the total sample, over two-thirds had three or more criteria of MetSyn (142 of 206). As shown in Table 4, number of MetSyn components in men and women was inversely associated with fat percentages of total body mass and truncal region, and also inversely associated with fat percentage of peripheral region in women. In Table 5, age- and smoking-adjusted partial correlations between MetSyn components, other lipids, and biomarkers with fat percentage of total body mass and truncal and peripheral fat are shown. In men, CRP concentrations correlated with increasing fat percentage of total body mass and central fat. These relationships were also seen in women. However, in women, paradoxical relationships between number of MetSyn components and HDL-cholesterol and triglyceride concentrations to total fat percentage and to peripheral fat were seen (lesser number of components and lower triglycerides, but higher HDL cholesterol with increasing fat). HDL-cholesterol correlated positively to truncal fat percentage. Also, ALT concentrations were inversely associated with all fat percentages in women, and uric acid was inversely associated with total and peripheral fat percentages.

Table 4.

Fat Distribution Stratified by Number of Metabolic Syndrome Components (N = 206)

	Number of components
Parameter	One n = 14	Two n = 50	Three n = 77	Four n = 43	Five n = 22	P^a
Men	n = 0	n = 10	n = 20	n = 12	n = 10
Fat percentage of total body mass	—	44.3 (3.3)	39.9 (3.2)	41.5 (3.7)	41.2 (4.0)	0.02
Fat percentage of truncal region	—	54.6 (2.4)	50.6 (3.8)	52.0 (4.0)	49.9 (4.5)	0.02
Fat percentage of peripheral region	—	45.8 (4.8)	41.4 (4.9)	42.4 (4.6)	42.8 (5.5)	0.16
Women	n = 14	n = 40	n = 57	n = 31	n = 12
Fat percentage of total body mass	54.3 (2.9)	53.7 (3.0)	51.2 (3.4)	50.9 (4.2)	49.5 (3.6)	<0.0001
Fat percentage of truncal region	59.3 (2.2)	59.1 (4.3)	57.2 (4.0)	58.3 (4.4)	55.5 (4.0)	0.03
Fat percentage of peripheral region	57.8 (4.1)	57.6 (3.7)	55.2 (3.8)	53.4 (4.6)	52.3 (3.5)	<0.0001

Mean (SD) shown.

P indicates ANOVA comparison across groups.

Table 5.

Partial Correlations Between Total, Truncal, and Peripheral Fat Percentages in Men and Women with Components of Metabolic Syndrome, Other Lipids and Lipoproteins, and Biomarkers Adjusted for Age and Smoking (N = 206) ^a

	Men						Women
	Total		Truncal		Peripheral		Total		Truncal		Peripheral
	r	P	r	P	r	P	r	P	r	P	r	P
#MetSyn Components	−0.13	0.37	−0.19	0.18	−0.01	0.54	−0.38	<0.0001	−0.20	0.02	−0.42	<0.0001
Metabolic syndrome components
Systolic BP	−0.24	0.10	−0.19	0.18	−0.23	0.12	−0.15	0.06	−0.08	0.35	−0.10	0.24
Diastolic BP	−0.14	0.32	−0.03	0.85	−0.24	0.10	−0.13	0.13	−0.11	0.18	−0.10	0.25
Systolic BP (no meds)	−0.26	0.18	−0.31	0.10	−0.24	0.21	−0.16	0.10	−0.02	0.87	−0.12	0.22
Diastolic BP (no meds)	−0.18	0.34	−0.14	0.47	−0.30	0.12	−0.18	0.06	−0.13	0.16	−0.17	0.07
HDL-cholesterol	−0.02	0.90	0.01	0.97	0.02	0.87	0.31	<0.0001	0.26	0.001	0.26	0.001
Triglycerides	−0.17	0.24	−0.17	0.24	−0.21	0.15	−0.31	<0.0001	−0.13	0.10	−0.38	<0.0001
Glucose	−0.02	0.88	−0.03	0.85	−0.01	0.94	−0.11	0.19	−0.03	0.73	−0.12	0.13
Other lipids and lipoproteins
Total cholesterol	0.03	0.83	0.01	0.96	0.02	0.91	−0.04	0.63	0.06	0.47	−0.10	0.21
LDL-cholesterol	0.15	0.34	0.08	0.61	0.12	0.44	−0.03	0.69	0.02	0.78	−0.06	0.48
Lipoprotein (a)	0.15	0.30	0.15	0.31	0.08	0.59	0.12	0.16	0.06	0.45	0.10	0.21
Apolipoprotein B	0.07	0.64	0.02	0.88	−0.02	0.89	−0.15	0.08	−0.06	0.48	−0.19	0.02
Biomarkers
C-reactive protein	0.42	0.003	0.31	0.03	0.27	0.06	0.28	0.001	0.20	0.02	0.27	0.001
White blood cell count	0.10	0.49	0.13	0.36	−0.04	0.79	−0.06	0.48	−0.03	0.73	−0.08	0.34
Ferritin	0.09	0.55	−0.01	0.97	0.11	0.44	−0.11	0.20	0.04	0.66	−0.14	0.08
Uric acid	−0.08	0.57	−0.20	0.18	0.01	0.95	−0.20	0.01	−0.07	0.42	−0.23	0.006
ALT	−0.10	0.48	−0.14	0.33	−0.15	0.31	−0.37	<0.0001	−0.21	0.009	−0.38	<0.0001

r = partial correlation coefficient.

For LDL-cholesterol, 8 missing (could not be estimated because high level of triglycerides); for Lp(a), 2 missing; for apolipoprotein B, 3 missing; and for C-reactive protein, 3 missing.

Discussion

Our main findings were that degree of obesity, as measured by BMI or waist circumference, did not predict MetSyn components, with the exception of systolic blood pressure in women with morbid obesity. In men with morbid obesity, systolic and diastolic blood pressures were positively related to BMI and waist circumference, but paradoxical relationships between BMI and lipids were observed. Furthermore, increasing number of components of MetSyn was associated with decreasing total and truncal fat in women and men and decreasing peripheral fat in women. In women, paradoxical relationships were seen between total, truncal, and peripheral fat percentages and HDL-cholesterol, triglyceride, and ALT concentrations. Positive associations between BMI, waist circumferences, fat percentages, and inflammatory biomarkers were noted. Associations were controlled for age and smoking, given the effects of smoking on MetSyn.⁸

The strength of this study is the inclusion of a large sample of consecutive patients of which over one-third consisted of men. Notably, there was no association between BMI and number of MetSyn components, suggesting that other factors than BMI determine MetSyn in morbid obesity, as suggested previously.⁹ Contrastingly, some studies have reported that metabolic traits remain associated with traditional anthropometric measures in severe obesity;^10
–12 however, these studies have included limited samples of severely obese individuals. Furthermore, we found only a weak association between waist circumference and number of MetSyn components in women and none in men. This is in line with a previous study of 100 morbidly obese individuals, in which waist circumference did not correlate with the prevalence of MetSyn, its severity, or with visceral fat area measured by CT.⁹ However, CT-assessed visceral fat did show an association with MetSyn in this study.⁹ These findings put together suggest that waist circumference is less useful as a tool to assess MetSyn in the morbidly obese than in nonmorbidly obese populations, and more specific tools may be needed, including visualization by CT or other methods.

Of our total sample, 68% met the harmonized definition of MetSyn, while the remaining one-third may be characterized as metabolically healthy with high waist circumference, but exhibiting none or only one other MetSyn component above the proposed cutoff levels. This proportion is similarly to that previously reported in morbidly obese individuals.^7
–9 The metabolically healthy participants were more likely to be female and were younger than those exhibiting a larger number of components of MetSyn. The predominance of younger individuals and women is in line with the notion that metabolically healthy obese individuals are unlikely to be permanently protected from metabolic disturbances.⁶ Furthermore, compared with metabolically healthy normal weight individuals, obese persons are at an increased risk for adverse outcomes, even in the absence of metabolic abnormalities.¹³

A key indicator of metabolic risks of obesity is the presence of subclinical inflammation, as evidenced by an elevated high-sensitivity CRP concentration, and increased concentrations of other inflammatory markers, including white blood cell count.^14,15 Elevated CRP is prevalent even among individuals with metabolically healthy obesity,¹⁴ and may underlie the risk of adverse outcomes in this group.¹³ We observed markedly elevated high-sensitivity CRP concentrations across all metabolic risk categories, and CRP concentrations were correlated with BMI and waist circumference. Likewise, white blood cell counts, although not grossly elevated, correlated with BMI and waist circumference in both genders. While these biomarkers of inflammation did not associate with severity of MetSyn, as indicated by the number of components, CRP concentrations showed statistically significant associations with all fat depots, while white blood cell count showed no associations.

Other biomarker indicators of obesity-related stress include oxidative stress and liver dysfunction.¹⁵ Both ferritin and uric acid concentrations were associated with severity of MetSyn, as seen in many obese populations,¹⁵ suggesting that they may be useful markers of metabolic damage in the morbidly obese, although causality cannot be determined in a cross-sectional study. These biomarkers were not associated with waist circumference. Uric acid concentrations were positively associated with BMI only in women, but a novel finding was the inverse relation between uric acid in relation to total and peripheral fat depots in women. While uric acid acts as an antioxidant in response to increased oxidative stress in normal weight individuals, it does not act as an antioxidant in patients with MetSyn, as the system appears to be saturated and unable to respond appropriately to increased oxidative stress.¹⁶

ALT concentrations indicating hepatic dysfunction were classically associated with severity of MetSyn and waist circumference in women. This finding is in line with a study of liver fat in morbidly obese patients where MetSyn was found to be independently related to fatty liver measured by CT.⁹ However, inverse associations were found between ALT and total, truncal, and peripheral fat in women, a surprising finding that questions the suitability of DXA in the extremely obese.

One of the best established links between obesity and CVD risk is obesity-related dyslipidemia,⁵ which is characterized by reduced lipolysis of triglyceride-rich lipoproteins, and the development of small dense LDL particles, small dense HDL particles, and relatively high apolipoprotein B levels. Notably, these changes do not appear to depend on total body fat mass.⁵ Previously, studies have revealed lower free fatty acids and triglyceride concentrations before and after fat overloads in patients with MetSyn with or without morbid obesity.¹⁶ In line with this, we found no relationship between BMI or waist circumference and lipids in women, and in men, relationships were paradoxical, with positive relations between BMI and HDL-cholesterol and an inverse relationship between BMI and triglycerides. We suggest that divergent findings regarding risk factor expression may be due to the sequestering of fat in adipose tissue in morbid obesity. In elegant experiments, Tinahones et al. showed that in the face of increased lipolysis, the net release of free fatty acids into the bloodstream is counteracted by increased reuptake and cycling in the morbidly obese.¹⁷ This is facilitated by increased expression of PPARγ and other enzymes involved in lipid storage, thus ensuring the storage and relocalization of the excess triglyceride.¹⁷

A trend toward lower lipoprotein (a) concentrations with increasing number of MetSyn components was noted in our study. This observation has been made previously in a predominantly healthy Asian occupational cohort and appeared to be independent of confounders.¹⁸ Also, in a cohort of Italian hypertensive patients, lipoprotein (a) levels were significantly, independently, and progressively lower with increasing insulin resistance.¹⁹ Lipoprotein (a) concentrations are strongly linked to apolipoprotein (a) allele size and expression, while dietary, lifestyle, and environmental factors have less influence. Thus, these data suggest that shared genetic mechanisms may underlie the associations, although other nongenetic explanations are possible.

In the subgroup that completed the DXA procedure, severity of MetSyn appeared to be associated with lower total, truncal, and peripheral fat depots (Table 4). Given that BMI did not differ according to MetSyn severity, this finding underscores the role of large depots in “absorbing” metabolic risks. The protective associations between lower body fat and cardiovascular risk in nonobese populations have been long known.²⁰ We previously reported that leg fat mass protected against metabolic risks in obese women, although not in obese men.²¹ A study of severely obese premenopausal women, with BMI >40 found that leg fat correlated negatively with cardiovascular risk factors.²² These findings extend these observations to women and men with morbid obesity and suggest that adequate fat depots may be advantageous, although ectopic fat would not be expected to protect against metabolic risks even in this population. This idea would need in depth and further investigation with CT- or MR-based methodology. Furthermore, HDL-cholesterol concentrations were positively associated with all fat depots, and triglyceride concentrations were inversely associated with total and peripheral depots in women. Notably, a large study of South Koreans found that metabolically healthy obese persons had an attenuated risk of preclinical atherosclerosis in contrast to metabolically abnormal obese and nonobese persons.²³ While not directly comparable to the current findings, this suggests that obesity in itself is less important than the metabolic abnormalities.

Study limitations

We lacked data on dietary patterns, alcohol consumption, and physical activity of the participants. Only 38% participated in the DXA due, in part, to limitations in performing the procedure in very heavy individuals. Thus, the findings regarding fat distribution may not apply to individuals with extreme obesity. Furthermore, hepatic fat was not quantified through ultrasound or other methods. We did not adjust for use of statins, but findings excluding users of statins did not differ substantially from shown associations (data not shown).

Conclusion

MetSyn severity does not seem to linearly increase with BMI, suggesting a need for better markers of CVD risk in the morbidly obese, particularly women. The paradoxical inverse correlation between BMI and fat distribution with lipids and other biomarkers suggests protective effects of fat depots in the morbidly obese, but prospective studies are needed to establish causality.

Footnotes

Author Disclosure Statement

No competing financial interests exist. All authors agree with the content of the text and tables.

References

Knight

. Diseases and disorders associated with excess body weight. Ann Clin Lab Sci, 2011; 41:107–121.

Eckel

, Grundy

, Zimmet

. The metabolic syndrome. Lancet, 2005; 365:1415–1428.

Alberti

, Eckel

, Grundy

, et al. 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.

Morais

, Faintuch

, Leal

AAD

, et al. Inflammation and biochemical features of bariatric candidates: Does gender matter?. Obes Surg, 2011; 21:71–77.

Klop

, Elte

JWF

, Cabezas

. Dyslipidemia in obesity: Mechanisms and potential targets. Nutrients, 2013; 5:1218–1240.

Plourde

, Karelis

. Current issues in the identification and treatment of metabolically healthy but obese individuals. Nutr Metab Cardiovasc Dis, 2014; 24:455–459.

Soverini

, Moscatiello

, Villanova

, et al. Metabolic syndrome and insulin resistance in subjects with morbid obesity. Obes Surg, 2010; 20:295–301.

Sun

, Liu

, Ning

. Active smoking and risk of metabolic syndrome: A meta-analysis of prospective studies. PLoS One, 2012; 7:e47791.

Faria

, Gonçalves

, Cunha

, et al. Beyond central adiposity: Liver fat and visceral area are associated with metabolic syndrome in morbidly obese patients. Int J Surg, 2015; 14:75–79.

10.

Ledoux

, Coupaye

, Essig

, et al. Traditional anthropometric parameters still predict metabolic disorders in women with severe obesity. Obesity, 2010; 18:1026–1032.

11.

Zazai

, Wilms

, Ernst

, et al. Waist circumference and related anthropometric indices are associated with metabolic traits in severely obese subjects. Obes Surg, 2014; 24:777–782.

12.

Ebron

, Andersen

, Aguilar

, et al. A larger body mass index is associated with increased atherogenic dyslipidemia, insulin resistance, and low-grade inflammation in individuals with metabolic syndrome. Metab Syndr Relat Dis, 2015; 13:458–464.

13.

Kramer

, Zinman

, Fetnakaran

. Are metabolically healthy overweight and obesity benign conditions? A systematic review and meta-analysis. Ann Intern Med, 2013; 159:758–769.

14.

Gonçalves

, Glade

, Meguid

. Metabolically healthy obese individuals: Key protective factors. Nutrition, 2016; 32:14–20.

15.

Robberecht

, Hermans

. Biomarkers of metabolic syndrome: Biochemical background and clinical significance. Metab Syndr Relat Dis, 2016; 14:47–93.

16.

Tinahones

, Garrido-Sánchez

, Murri

. Particular characteristics of the metabolic syndrome in patients with morbid obesity. Endocrinol Nutr, 2013; 60:127–135.

17.

Tinahones

, et al. Obesity and insulin resistance-related changes in the expression of lipogenic and lipolytic genes in morbidly obese subjects. Obes Surg, 2010; 20:1559–1567.

18.

Sung

K-C

, Wild

, Byrne

. Lipoprotein (a), metabolic syndrome and coronary calcium score in a large occupational cohort. Nutr Metab Cardiovasc Dis, 2013; 23:1239–1246.

19.

Marzano

, Colussi

, De Torre

, et al. Relationships of plasma lipoprotein (a) levels with insulin resistance in hypertensive patients. Metabolism, 2014; 63:1439–1446.

20.

Van Pelt

, Evans

, Schechtman

, et al. Contributions of total and regional fat mass to risk for cardiovascular disease in older women. Am J Physiol Endocrinol Metab, 2002; 282:E1023–E1028.

21.

Aasen

, Fagertun

, Tonstad

, et al. Leg fat mass as measured by dual X-ray absorptiometry (DXA) impacts insulin resistance differently in obese women versus men. Scand J Clin Lab Invest, 2009; 69:181–189.

22.

Faliola

, Tirabassi

, Canibus

, et al. Protective effect of leg fat against cardiovascular risk factors in obese premenopausal women. Nutr Metab Cardiovasc Dis, 2009; 19:39–44.

23.

Sung

K-C

, Cha

S-C

, Sung

J-W

, et al. Metabolically healthy obese subjects are at risk of fatty liver but not of pre-clinical atherosclerosis. Nutr Metab Cardiovasc Dis, 2014; 24:256–262.