- Email: [email protected]
Body Composition and Cardiorespiratory Fitness Between Metabolically Healthy Versus Metabolically Unhealthy Obese Black and White Adolescents
ARTICLE IN PRESS Journal of Adolescent Health 000 (2018) 1−6
www.jahonline.org Original article
Body Composition and Cardiorespiratory Fitness Between Metabolically Healthy Versus Metabolically Unhealthy Obese Black and White Adolescents D1X XSoJung Lee, D2X XPh.D.a,*, and D3X XSilva Arslanian, D4X XM.D.b,c a
Division of Sports Medicine, Graduate School of Physical Education, Kyung Hee University, Yongin, Republic of Korea Center for Pediatric Research in Obesity and Metabolism, Children’s Hospital of Pittsburgh of UPMC, School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania c Division of Pediatric Endocrinology, Metabolism and Diabetes Mellitus, Children’s Hospital of Pittsburgh of UPMC, School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania b
Article History: Received April 25, 2018; Accepted August 22, 2018 Keywords: Metabolically healthy obesity; Liver fat; Cardiorespiratory fitness; Childhood obesity
A B S T R A C T
Purpose: We compared body composition and cardiorespiratory fitness (CRF) between metabolically healthy overweight/obese (MHO) versus metabolically unhealthy overweight/obese (MUO) adolescents in 189 black and white adolescents (BMI ≥ 85th percentile, 12−18 years of age). Methods: Participants were defined as MHO or MUO if their insulin-stimulated glucose disposal, measured by a 3-hour hyperinsulinemic-euglycemic clamp, was in the upper quartile or in the lower three quartiles. Total fat was measured by dual-energy X-ray absorptiometry, and visceral adiposity and liver fat were measured by magnetic resonance imaging and proton magnetic resonance spectroscopy, respectively. CRF was measured by a graded maximal treadmill test. Results: Black MHO adolescents had lower (p < .05) 2-hour oral glucose tolerance test glucose, triglycerides, very-low-density lipoprotein cholesterol, and higher high-density lipoprotein cholesterol, with a lower prevalence of impaired fasting glucose and impaired glucose tolerance compared with black MUO adolescents. White MHO adolescents had lower (p < .05) triglycerides and very-low-density lipoprotein cholesterol, with a lower prevalence of impaired fasting glucose compared with white MUO adolescents. Independent of race, CRF was higher in MHO versus MUO adolescents. After accounting for gender, Tanner stage, and BMI, there were no differences in total fat (kg, %) between MHO versus MUO in both races. MHO adolescents had significantly lower trunk fat, waist circumference, and visceral fat compared with MUO adolescents in both races. In whites, MHO adolescents had lower (p = .055) liver fat compared with MUO adolescents. Conclusions: Independent of race, the MHO phenotype is characterized by high CRF, lower waist circumference and visceral fat, and lower rates of dysglycemia in youth. © 2018 Published by Elsevier Inc. on behalf of Society for Adolescent Health and Medicine.
Although obesity is often associated with metabolic complications, a subgroup of obese individuals appears to be protected Conflicts of interest: All authors have no conflicts of interest to declare. * Address correspondence to: SoJung Lee, Ph.D., Division of Sports Medicine, Graduate School of Physical Education, Kyung Hee University, Yongin 17104, Republic of Korea. E-mail address: [email protected] (S. Lee).
IMPLICATIONS AND CONTRIBUTION
In both black and white adolescents, the MHO phenotype is characterized by lower abdominal adiposity and lower rates of dysglycemia and higher aerobic fitness.
against the development of obesity-related comorbidities. These subjects, referred to as metabolically healthy obese (MHO) individuals, have favorable metabolic profiles such as high insulin sensitivity and do not have any traditional cardiometabolic risk factors (e.g., high blood pressure and dyslipidemia) despite their high BMI or total adiposity [1]. Previous studies have shown that in adults, MHO individuals account for 20%−40% of individuals with obesity
1054-139X/© 2018 Published by Elsevier Inc. on behalf of Society for Adolescent Health and Medicine. https://doi.org/10.1016/j.jadohealth.2018.08.024
ARTICLE IN PRESS 2
S. Lee and S. Arslanian / Journal of Adolescent Health 00 (2018) 1−6
[2,3]. Similarly, in the pediatric populations, there is wide variation in the prevalence of MHO (10%−40%) depending on the definition used to classify MHO [4−6]. Although there has been increased awareness of the MHO phenotype, it remains to be determined as to which factors contribute the MHO versus MUO characteristics in children and adolescents. Recently, Cadenas-Sanchez et al. [7] examined the associations between MHO phenotype and physical activity level in 237 European adolescents. In that study [7], youth were categorized as MHO if they had 0 or 1 of the traditional cardiometabolic risk factors (e.g., high triglycerides, high fasting glucose, low high-density lipoprotein cholesterol, and high systolic or diastolic blood pressure) [8] and leisure-time physical activity level was assessed using an accelerometer. They found that compared with MUO youth, MHO youth are less sedentary and spend more time in moderate-vigorous physical activity [7]. Currently, there is no established definition to define MHO. Studies in adults [3,9−11] have identified MHO individuals based on insulin sensitivity since insulin resistance (i.e., reduced insulin sensitivity) is the primary link between obesity and the cardiometabolic abnormalities [1,12]. Therefore, in this study, we identified MHO adolescents based on in vivo insulin sensitivity using a 3-hour hyperinsulinemic-euglycemic clamp technique, which is considered the gold-standard method. Then, we compared cardiorespiratory fitness (CRF) as measured by peak VO2 during a maximal graded treadmill test, and total and regional body fat distribution between MHO versus MUO adolescents using gold-standard imaging modalities such as magnetic resonance (MR) imaging and proton-MR spectroscopy. We also examined whether racial differences (black vs. white) exist in the cardiometabolic and body composition characteristics between MHO versus MUO adolescents.
Anthropometric measurement, and total and regional body fat distribution Body weight was measured to the nearest .1 kg and height was measured to the nearest .1 cm using standardized equipment. Waist circumference was measured at the last rib and the average of two measurements was used in the analyses. Total body fat (kg and %), trunk fat, and fat free mass (FFM) were measured using lunar iDXA (GE Healthcare, Madison, WI), and abdominal fat was measured at L4−L5 with the use of a 3.0 Tesla MR scanner (Siemens, Magnetom TIM Trio, Erlangen, Germany) [13,14]. For liver fat content, proton magnetic resonance spectroscopy (1H-MRS) was performed in the same MR scanner (Siemens, Tim Trio, Erlangen, Germany) using a body matrix coil and a spine matrix using the following parameters (voxel size = 30 £ 30 £ 20 mm3, TR = 4,000 ms, TE = 30 ms) as prescribed previously [13,14]. Oral glucose tolerance and insulin sensitivity tests All participants reported to the PCTRC after an overnight fast (minimum 10 hours) for a 2-hour oral glucose tolerance test (OGTT, 1.75 g/kg, max 75 g) to assess glucose tolerance status [15]. Following the OGTT, participants were admitted and stayed overnight in the hospital for a 3-hour 80 mU/m2/min hyperinsulinemic-euglycemic clamp test the next morning as described previously [16]. Participants were fed a standardized lunch and dinner and fasted overnight. Insulin-stimulated glucose disposal (Rd) was calculated as the average infusion rate of glucose during the last 30 minutes of the clamp and was expressed as milligrams per min per kg of FFM (RdFFM). Participants were fed a standardized lunch after the clamp test. Fasting lipids and blood pressure measurements
Methods Subjects The study sample included 189 overweight and obese black and white adolescents (BMI ≥ 85th percentile, 12−18 years of age) who were recruited via public advertisements and from the Weight Management and Wellness Center at Children’s Hospital of Pittsburgh (CHP) of the University of Pittsburgh Medical Center. Data from some of these participants have been reported previously [13,14]. Participants were nonsmokers, nondiabetic, physically inactive, and not taking medications known to affect study outcomes. Exclusion criteria included significant weight change (BMI > 2−3 kg/m2) prior to study participation, endocrine disorder (e.g., type 2 diabetes and polycystic ovary syndrome), or syndromic obesity. None of the participants had liver diseases and consumed alcohol. A complete medical history and physical examination, with evaluation of pubertal staging according to Tanner criteria, were conducted by a certified nurse practitioner. Participants self-identified as black or white during physical examination. Participants with mixed race were not included in the current study analyses. Routine hematological and biochemical testing were performed in all participants at the Pediatric Clinical and Translational Research Center (PCTRC) at CHP. Ethical approval was granted by the University of Pittsburgh institutional review board and written parental informed consent and child assent was obtained from all participants before any research participation. Evaluations were completed during an overnight inpatient admission at the PCTRC at CHP.
Fasting plasma lipids were determined in the Nutrition Laboratory of the University of Pittsburgh certified by the National Heart, Lung, and Blood Institute standardization program. Blood pressure was measured during inpatient admission when the patients were resting in the supine position with an automated sphygmomanometer every 10 minutes between 22:00 and 23:00 P. M. on the evening of admission, and between 06:00 and 07:00 A. M. before awakening next morning. The average value of seven measurements during each hour was used in the analysis. CRF Peak VO2 was measured during a graded maximal treadmill test with the use of standard open-circuit spirometry techniques until volitional fatigue [13,14]. A graded maximal treadmill test was conducted at least 2−3 hours after the hyperinsulinemiceuglycemic clamp test and after obtaining normalized glucose levels. For those who were unable to do the treadmill test, the test was conducted on a separate day. Statistical analyses Insulin-stimulated glucose disposal (RdFFM) did not differ (p > .1) between black versus white (11.4 § 3.5 mg/kgFFM/min vs. 11.9 § 3.6 mg/kgFFM/min, respectively) obese adolescents. Given that insulin sensitivity is a continuous trait and that there is no objective definition of insulin resistance, 189 black and white adolescents were combined and stratified into quartiles based on
ARTICLE IN PRESS S. Lee and S. Arslanian / Journal of Adolescent Health 00 (2018) 1−6
their RdFFM. Participants were defined as having high insulin sensitivity and MHO if their RdFFM value was in the upper quartile ( > 13.8 mg/kgFFM/min, black n = 26; white n = 22). Participants were defined as having low insulin sensitivity and MUO if their RdFFM value was in the lower three quartiles ( ≤ 13.8 mg/kgFFM/ min, black n = 87; white n = 54). Physical and metabolic characteristics and CRF between MHO versus MUO adolescents within each race were compared using independent t tests and chi-square tests. General linear modeling with least squared adjusted means was used to examine MHO versus MUO differences in total and regional body fat distribution adjusting for Tanner stage, gender, and BMI. Statistical procedures were performed using SPSS (Version 24; SPSS, Inc., Chicago, IL) and statistical significance was set at p ≤ .05. Results The study sample was composed of 60% black (113 of 189) and 40% white (76 of 189) overweight and obese adolescents. Of the 189 adolescents studied, 26 of 113 (23%) black and 22 of 76 (29%) white adolescents were classified as MHO (Table 1). There were no significant gender differences in the prevalence of MHO within each race. Despite similar age, MHO adolescents had lower BMI, body weight, and waist circumference compared with MUO adolescents in both races. In whites but not in blacks, Tanner stage was lower (p < .01) in MHO compared with MUO adolescents. Cardiometabolic characteristics of MHO versus MUO adolescents are shown in Table 2. By design, insulin-stimulated glucose disposal rate (RdFFM) was significantly different between MHO versus MUO adolescents in both races. In blacks, MHO adolescents had lower (p < .05) 2-hour OGTT glucose, triglycerides, very-lowdensity lipoprotein (VLDL)-cholesterol, and higher high-density lipoprotein cholesterol levels compared with MUO adolescents. Additionally, the prevalence of impaired fasting glucose (IFG, 30.1% vs. 4.4%, p = .06) and impaired glucose tolerance (IGT, 15.0% vs. .8%, p = .055) trended to be lower in MHO adolescents compared with MUO adolescents. In whites, MHO adolescents had significantly lower triglycerides and VLDL cholesterol levels, with a lower prevalence of IFG (17.1% vs. 1.3%, p = .046) compared with MUO adolescents. In both black and white adolescents with MHO phenotypes, CRF was significantly higher (p = .05) compared with adolescents with MUO. Total and regional body fat distribution between MHO versus MUO adolescents are shown in Figure 1. After adjusting for confounding factors (gender, Tanner stage, and BMI), there were no significant differences in the amount of total body fat (kg and %) between MHO versus MUO groups in both black and white adolescents. Despite similar total body fat (kg, %), MHO adolescents had
3
significantly lower trunk fat mass, waist circumference, and visceral fat compared with MUO adolescents independent of race. In whites but not in blacks, MHO adolescents also had lower liver fat when compared with MUO adolescents (4.0 § .5% vs. 1.9 § .9%, p = .055). Discussion The present study demonstrated that in both black and white youth, MHO adolescents have higher CRF measured as peak VO2, and lower triglycerides and VLDL cholesterol, with a lower prevalence of IFG as compared with MUO adolescents. After accounting for confounding factors such as gender, Tanner stage, and BMI, MHO adolescents have significantly lower waist circumference, trunk fat mass, and visceral fat with no differences in subcutaneous adipose tissue as compared with MUO adolescents despite similar total body fat. Further, in whites but not in blacks, MHO adolescents have lower liver fat compared with MUO adolescents. These findings suggest that in adolescents, the MHO phenotype is associated with high CRF, and lower abdominal obesity and visceral fat, and lower rates of dysglycemia in both black and white adolescents. To date, the comparisons of body fat distribution between MHO versus MUO have been mostly limited to adults [9,10,17]. Lower levels of visceral fat [9,10] or DXA-measured central adiposity [18] have been reported in MHO versus MUO postmenopausal women. In a mixed sample of black and white adults (18−68 years), Camhi et al. [17] have shown that MHO adults have significantly lower total adiposity, trunk fat, and visceral adiposity compared with MUO adults, independent of gender. In children and adolescents, lower waist circumference and insulin resistance measured by HOMA-IR have been consistent predictors of MHO phenotype [4−6], whereas dietary component and leisure-time physical activity level did not predict MHO [6]. To our knowledge, this is the first study that examined comprehensively the cardiometabolic and body composition characteristics, and objectively measured CRF between MHO versus MUO in both black and white adolescents. Recently, Cadenas-Sanchez et al. [7] demonstrated that although moderate-vigorous physical activity level was higher in MHO than in MUO adolescents, CRF did not differ between MHO versus MUO adolescents. However, it is important to note that in that study [7], CRF was evaluated by a 20-m shuttle run test and maximal O2 consumption was estimated using an equation. Our findings using the gold-standard methods of measuring CRF, body fat distribution, and insulin sensitivity extend previous findings by Cadenas-Sanchez et al. [7] and others [4−6] and provide evidence that the MHO and MUO phenotypes are characterized by differences in abdominal fat and CRF in both black and white obese adolescents. However, our findings differ
Table 1 Anthropometric characteristics of MHO versus MUO adolescents Black obese adolescents (n = 113)
n (%) Male/female, n Age, years Tanner stage*, II−III/IV−V BMI, kg/m2 Body weight, kg Waist circumference, cm
MHO
MUO
26 (23%) 10/16 14.9 § 1.9 4/21 32.3 § 4.5 87.4 § 17.1 90.6 § 11.2
87 (77%) 29/58 14.5 § 1.7 9/78 34.8 § 3.8 96.2 § 15.7 98.6 § 9.5
Data presented as mean § SD. *Tanner stage is missing in one subject in the black MHO group.
White obese adolescents (n = 76) p
MHO
MUO
p
.64 .38 .48 .005 .02 .001
22 (29%) 13/9 14.3 § 1.6 7/15 31.8 § 4.6 86.1 § 15.0 93.6 § 8.6
54 (71%) 26/28 14.7 § 1.6 4/50 34.0 § 4.1 97.9 § 15.4 100.7 § 9.6
.45 .37 .01 .04 .003 .004
ARTICLE IN PRESS S. Lee and S. Arslanian / Journal of Adolescent Health 00 (2018) 1−6
4
Table 2 Cardiometabolic characteristics of MHO versus MUO adolescents Black obese adolescents (n = 113)
Rd, mg/kgFFM/min Fasting glucose, mg/dl OGTT 2-hour glucose, mg/dl IFG, n (%) IGT, n (%) Total cholesterol, mg/dl Triglycerides, mg/dl HDL cholesterol, mg/dl LDL cholesterol, mg/dl VLDL cholesterol, mg/dl AM systolic BP, mm Hg AM diastolic BP, mm Hg PM systolic BP, mm Hg PM diastolic BP, mm Hg CRF, ml/kg/min RER
White obese adolescents (n = 76)
MHO
MUO
p
MHO
MUO
p
16.1 § 2.1 95.6 § 7.0 111.6 § 15.0 5 (4.4%) 1 (.9%) 141.5 § 22.7 56.8 § 19.1 47.6 § 10.2 82.5 § 19.7 11.4 § 3.8 110.8 § 9.0 59.2 § 6.9 114.4 § 9.6 61.7 § 6.7 25.8 § 4.6 1.07 § .08
10.0 § 2.5 97.0 § 7.2 121.2 § 21.4 34 (30.1%) 17 (15.0%) 144.7 § 24.9 76.0 § 30.7 42.9 § 7.9 86.6 § 23.1 15.2 § 6.1 109.7 § 10.4 59.9 § 6.5 114.0 §11.9 60.7 § 7.2 23.6 § 4.9 1.05 § .08
< .001 .36 .03 .06 .055 .55 .003 .01 .41 .003 .65 .69 .88 .56 .05 .29
16.1 § 2.3 93.6 § 3.9 120.1 §16.9 1 (1.3%) 3 (3.9%) 144.0 § 23.8 88.2 § 38.5 42.5 § 10.0 83.8 § 19.7 17.6 § 7.7 108.8 § 12.1 60.3 § 5.3 115.2 § 12.0 61.7 § 6.6 28.9 § 5.6 1.07 § .06
10.2 § 2.4 95.9 § 5.3 127.0 § 23.4 13 (17.1%) 15 (19.7%) 151.7 § 34.6 123.3 § 80.3 41.4 § 8.3 86.0 § 28.2 24.7 § 16.1 111.6 § 10.2 59.8 § 5.0 114.8 §11.1 60.0 § 5.8 26.4 § 4.4 1.09 § .06
< .001 .07 .21 .046 .19 .35 .055 .62 .74 .055 .31 .69 .88 .27 .046 .19
Data presented as mean § SD. BP = blood pressure; CRF = cardiorespiratory fitness; FFM = fat free mass; HDL = high-density lipoproteins; IFG = impaired fasting glucose; IGT = impaired glucose tolerance; LDL = low-density lipoprotein; NS = not significant; OGTT = oral glucose tolerance test; Rd = rate of glucose disposal; RER = respiratory exchange ratio; VLDL = very-low-density lipoprotein.
from Senechal et al. [4] who reported no differences in visceral fat and CRF between MHO versus MUO adolescents (n = 108). The disparate findings could be due to the differences in the cohort studied and the methodology employed to evaluate CRF (maximal treadmill test vs. cycle ergometer test). In that study [4], the data analyses combined adolescents from various ethnic background
MHO A
C
40 30 20 10 0
25
40
20
30 20 10
*
Visceral fat, cm2
80
*
80
*
60
Whites
† 5 4 3 2 1
40
Blacks
Whites
5
F 100
90
Blacks
10
Whites
E
100
*
0
Blacks
Liver fat, %
*
Whites
*
15
0
Blacks
D
50
Trunk fat, kg
Total adiposity, %
Total fat, kg
MUO
B 50
Waist circumference, cm
(56% white, 11% First Nation, and 32% other ethnic groups) despite the well-known differences in body fat distribution between adolescents of different ethnicities [19]. To our knowledge, this is the first study to compare liver fat content between MHO versus MUO in a biracial sample of adolescents. We found that in whites, MHO adolescents tend to have
0
Blacks
Whites
Blacks
Whites
Figure 1. Total and regional body fat distribution of MHO versus MUO adolescents. Data presented as the least squared means (§SE) adjusted for Tanner stage, gender, and BMI. (A) Total fat (kg), (B) total adiposity (%), (C) trunk fat (kg), (D) waist circumference (cm), (E) visceral fat (cm2), and (F) liver fat (%). *p < .05. yp = .055.
ARTICLE IN PRESS S. Lee and S. Arslanian / Journal of Adolescent Health 00 (2018) 1−6
lower liver fat compared with MUO adolescents, while this was not the case in blacks. The lack of association between the MHO phenotype and liver fat in black adolescents could be explained by substantially lower levels of liver fat in black (average liver fat in 113 adolescents = 1.6%) versus white (average liver fat in 76 adolescents = 3.4%) youth in our study, regardless of the MHO and MUO phenotypes. Our findings provide evidences that in white adolescents, the MHO phenotype imparts a lower risk for increased liver fat. To date, there is no established consensus on the MHO definitions in children and adults. In adults, increased insulin sensitivity has been often used to identify MHO individuals since insulin resistance is suggested to play a key role in the development of cardiometabolic abnormalities in obesity [1,12]. Studies in adults [3,9−11] have indentified MHO individuals based on insulin sensitivity measured using the gold-standard hyperinsulinemiceuglycemic clamp technique. However, the hyperinsulinemiceuglycemic clamp test is costly, time consuming, and invasive, requires technical expertise, and is not readily available at most pediatric centers. Therefore, the most common definitions of MHO in the pediatric literature have utilized simple fasting surrogate index (the homeostasis model assessment [HOMA]) [5,20] and/or the components of metabolic syndrome [5−7,21,22]. Recently, Heinzle et al. [6] examined the prevalence and predictors of MHO in 632 adolescents (12−19 years, BMI ≥ 95th percentile) using the National Health and Nutrition Examination Survey Data (1999−2010). The authors used two different definitions to identify MHO ([1] having ≤ 1 metabolic syndrome risk factors [excluding waist circumference] and free of type 2 diabetes, hypertension, and dyslipidemia, [2] free of all metabolic syndrome criteria, insulin resistance, and inflammation), and demonstrated that the prevalence of MHO in adolescents varies between 7% and 74% depending on the definition used [6]. Further, they found that lower levels of abdominal obesity, measured by waist circumference, and absence of insulin resistance, measured by HOMA-IR, were the most consistent significant predictors of MHO, independent of leisure-time physical activity and diet intake, in both boys and girls [6]. Given the increasing prevalence of comorbidities (e.g., type 2 diabetes and nonalcoholic fatty liver disease) in obese children and adolescents, characterizing MHO versus MUO youth is of importance to help health care professionals to identify highrisk obese youth for early interventions. Further studies are needed to investigate whether different definitions of MHO influence the predictors of MHO phenotype in youth. The strengths and limitations of this study warrant mention. To the best of our knowledge, this is the first study to examine MHO versus MUO phenotypes in a large sample of youth (n = 189), including black adolescents. We employed the state-of-the-art methodologies such as the 3-hour hyperinsulinemic-euglycemic clamp, MRI, 1H-MRS, and a graded maximal treadmill test to comprehensively evaluate cardiometabolic and body composition characteristics between MHO versus MUO adolescents. Whether our findings would remain true in other racial groups such as Asian and Hispanic youth, prepubertal children and severely obese youth are unknown. Further, other lifestyle variables such as sedentary behavior and sleep duration may play a role in MHO phonotype unknown in youth. In conclusion, the MHO phenotype is characterized by lower waist circumference, trunk fat mass and visceral fat, higher CRF, and lower rates of dysglycemia in both black and white adolescents. In whites, the MHO phenotype is also characterized by lower liver fat content.
5
Author Contribution S. Lee designed the study, researched the data, and wrote the manuscript. S. Arslanian researched data, supervised clamp experiment, reviewed clinical data, and provided critical revision of the manuscript. Funding Sources This research was funded by the National Institutes of Health (Grant Numbers 5R01HL114857, 1R21DK083654-01A1, UL1 RR024153, and UL1TR000005), American Diabetes Association (708-JF-27), and Children’s Hospital of Pittsburgh of UPMC (Cochrane-Weber Foundation and Renziehausen Fund) to Lee; the National Center for Advancing Translational Sciences Clinical and Translational Science Award (UL1 RR024153 and UL1TR000005) to the Pediatric Clinical and Translational Research Center, and the Department of Defense (FA7014-02-2-001) and Richard L. Day endowed chair to Arslanian. The American Diabetes Association, Department of Defense, National Institute of Health, and Cochrane-Weber Foundation at Children’s Hospital of Pittsburgh did not have any role in the design and conduct of the study; collection, management, analysis, interpretation of the data, and writing the manuscript. References [1] Karelis AD, St-Pierre DH, Conus F, et al. Metabolic and body composition factors in subgroups of obesity: What do we know? J Clin Endocrinol Metab 2004;89:2569–75. [2] Bonora E, Kiechl S, Willeit J, et al. Prevalence of insulin resistance in metabolic disorders: The Bruneck Study. Diabetes 1998;47:1643–9. [3] Ferrannini E, Natali A, Bell P, et al. Insulin resistance and hypersecretion in obesity. European Group for the Study of Insulin Resistance (EGIR). J Clin Invest 1997;100:1166–73. [4] Senechal M, Wicklow B, Wittmeier K, et al. Cardiorespiratory fitness and adiposity in metabolically healthy overweight and obese youth. Pediatrics 2013;132:e85–92. [5] Prince RL, Kuk JL, Ambler KA, et al. Predictors of metabolically healthy obesity in children. Diabetes Care 2014;37:1462–8. [6] Heinzle S, Ball GD, Kuk JL. Variations in the prevalence and predictors of prevalent metabolically healthy obesity in adolescents. Pediatr Obes 2016;11: 425–33. [7] Cadenas-Sanchez C, Ruiz JR, Labayen I, et al. Prevalence of metabolically healthy but overweight/obese phenotype and its association with sedentary time, physical activity, and fitness. J Adolesc Health 2017;61:107–14. [8] Jolliffe CJ, Janssen I. Development of age-specific adolescent metabolic syndrome criteria that are linked to the Adult Treatment Panel III and International Diabetes Federation criteria. J Am Coll Cardiol 2007;49:891–8. [9] Brochu M, Tchernof A, Dionne IJ, et al. What are the physical characteristics associated with a normal metabolic profile despite a high level of obesity in postmenopausal women? J Clin Endocrinol Metab 2001;86:1020–5. [10] Karelis AD, Faraj M, Bastard JP, et al. The metabolically healthy but obese individual presents a favorable inflammation profile. J Clin Endocrinol Metab 2005;90:4145–50. [11] Marini MA, Succurro E, Frontoni S, et al. Metabolically healthy but obese women have an intermediate cardiovascular risk profile between healthy nonobese women and obese insulin-resistant women. Diabetes Care 2007;30:2145–7. [12] Ferrannini E, Haffner SM, Mitchell BD, Stern MP. Hyperinsulinaemia: The key feature of a cardiovascular and metabolic syndrome. Diabetologia 1991;34:416–22. [13] Lee S, Bacha F, Hannon T, et al. Effects of aerobic versus resistance exercise without caloric restriction on abdominal fat, intrahepatic lipid, and insulin sensitivity in obese adolescent boys: A randomized, controlled trial. Diabetes 2012;61:2787–95. [14] Lee S, Deldin AR, White D, et al. Aerobic exercise but not resistance exercise reduces intrahepatic lipid content and visceral fat and improves insulin sensitivity in obese adolescent girls: A randomized controlled trial. Am J Physiol Endocrinol Metab 2013;305:E1222–9. [15] American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care 2004;27(Suppl 1):S5–10. [16] Lewy VD, Danadian K, Witchel SF, Arslanian S. Early metabolic abnormalities in adolescent girls with polycystic ovarian syndrome. J Pediatr 2001;138:38–44.
ARTICLE IN PRESS 6
S. Lee and S. Arslanian / Journal of Adolescent Health 00 (2018) 1−6
[17] Camhi SM, Katzmarzyk PT. Differences in body composition between metabolically healthy obese and metabolically abnormal obese adults. Int J Obes 2014;38:1142–5. [18] Peppa M, Koliaki C, Papaefstathiou A, et al. Body composition determinants of metabolic phenotypes of obesity in nonobese and obese postmenopausal women. Obesity 2013;21:1807–14. [19] Liska D, Dufour S, Zern TL, et al. Interethnic differences in muscle, liver and abdominal fat partitioning in obese adolescents. PloS One 2007;2: e569.
[20] Vukovic R, Mitrovic K, Milenkovic T, et al. Insulin-sensitive obese children display a favorable metabolic profile. Eur J Pediatr 2013;172:201–6. [21] Margolis-Gil M, Yackobovitz-Gavan M, Phillip M, Shalitin S. Which predictors differentiate between obese children and adolescents with cardiometabolic complications and those with metabolically healthy obesity? Pediatr Diabetes 2018. https://doi.org/10.1111/pedi.12694. [22] Ding W, Cheng H, Chen F, et al. Adipokines are associated with hypertension in metabolically healthy obese (MHO) children and adolescents: A prospective population-based cohort study. J Epidemiol 2018;28:19–26.
www.jahonline.org Original article
Body Composition and Cardiorespiratory Fitness Between Metabolically Healthy Versus Metabolically Unhealthy Obese Black and White Adolescents D1X XSoJung Lee, D2X XPh.D.a,*, and D3X XSilva Arslanian, D4X XM.D.b,c a
Division of Sports Medicine, Graduate School of Physical Education, Kyung Hee University, Yongin, Republic of Korea Center for Pediatric Research in Obesity and Metabolism, Children’s Hospital of Pittsburgh of UPMC, School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania c Division of Pediatric Endocrinology, Metabolism and Diabetes Mellitus, Children’s Hospital of Pittsburgh of UPMC, School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania b
Article History: Received April 25, 2018; Accepted August 22, 2018 Keywords: Metabolically healthy obesity; Liver fat; Cardiorespiratory fitness; Childhood obesity
A B S T R A C T
Purpose: We compared body composition and cardiorespiratory fitness (CRF) between metabolically healthy overweight/obese (MHO) versus metabolically unhealthy overweight/obese (MUO) adolescents in 189 black and white adolescents (BMI ≥ 85th percentile, 12−18 years of age). Methods: Participants were defined as MHO or MUO if their insulin-stimulated glucose disposal, measured by a 3-hour hyperinsulinemic-euglycemic clamp, was in the upper quartile or in the lower three quartiles. Total fat was measured by dual-energy X-ray absorptiometry, and visceral adiposity and liver fat were measured by magnetic resonance imaging and proton magnetic resonance spectroscopy, respectively. CRF was measured by a graded maximal treadmill test. Results: Black MHO adolescents had lower (p < .05) 2-hour oral glucose tolerance test glucose, triglycerides, very-low-density lipoprotein cholesterol, and higher high-density lipoprotein cholesterol, with a lower prevalence of impaired fasting glucose and impaired glucose tolerance compared with black MUO adolescents. White MHO adolescents had lower (p < .05) triglycerides and very-low-density lipoprotein cholesterol, with a lower prevalence of impaired fasting glucose compared with white MUO adolescents. Independent of race, CRF was higher in MHO versus MUO adolescents. After accounting for gender, Tanner stage, and BMI, there were no differences in total fat (kg, %) between MHO versus MUO in both races. MHO adolescents had significantly lower trunk fat, waist circumference, and visceral fat compared with MUO adolescents in both races. In whites, MHO adolescents had lower (p = .055) liver fat compared with MUO adolescents. Conclusions: Independent of race, the MHO phenotype is characterized by high CRF, lower waist circumference and visceral fat, and lower rates of dysglycemia in youth. © 2018 Published by Elsevier Inc. on behalf of Society for Adolescent Health and Medicine.
Although obesity is often associated with metabolic complications, a subgroup of obese individuals appears to be protected Conflicts of interest: All authors have no conflicts of interest to declare. * Address correspondence to: SoJung Lee, Ph.D., Division of Sports Medicine, Graduate School of Physical Education, Kyung Hee University, Yongin 17104, Republic of Korea. E-mail address: [email protected] (S. Lee).
IMPLICATIONS AND CONTRIBUTION
In both black and white adolescents, the MHO phenotype is characterized by lower abdominal adiposity and lower rates of dysglycemia and higher aerobic fitness.
against the development of obesity-related comorbidities. These subjects, referred to as metabolically healthy obese (MHO) individuals, have favorable metabolic profiles such as high insulin sensitivity and do not have any traditional cardiometabolic risk factors (e.g., high blood pressure and dyslipidemia) despite their high BMI or total adiposity [1]. Previous studies have shown that in adults, MHO individuals account for 20%−40% of individuals with obesity
1054-139X/© 2018 Published by Elsevier Inc. on behalf of Society for Adolescent Health and Medicine. https://doi.org/10.1016/j.jadohealth.2018.08.024
ARTICLE IN PRESS 2
S. Lee and S. Arslanian / Journal of Adolescent Health 00 (2018) 1−6
[2,3]. Similarly, in the pediatric populations, there is wide variation in the prevalence of MHO (10%−40%) depending on the definition used to classify MHO [4−6]. Although there has been increased awareness of the MHO phenotype, it remains to be determined as to which factors contribute the MHO versus MUO characteristics in children and adolescents. Recently, Cadenas-Sanchez et al. [7] examined the associations between MHO phenotype and physical activity level in 237 European adolescents. In that study [7], youth were categorized as MHO if they had 0 or 1 of the traditional cardiometabolic risk factors (e.g., high triglycerides, high fasting glucose, low high-density lipoprotein cholesterol, and high systolic or diastolic blood pressure) [8] and leisure-time physical activity level was assessed using an accelerometer. They found that compared with MUO youth, MHO youth are less sedentary and spend more time in moderate-vigorous physical activity [7]. Currently, there is no established definition to define MHO. Studies in adults [3,9−11] have identified MHO individuals based on insulin sensitivity since insulin resistance (i.e., reduced insulin sensitivity) is the primary link between obesity and the cardiometabolic abnormalities [1,12]. Therefore, in this study, we identified MHO adolescents based on in vivo insulin sensitivity using a 3-hour hyperinsulinemic-euglycemic clamp technique, which is considered the gold-standard method. Then, we compared cardiorespiratory fitness (CRF) as measured by peak VO2 during a maximal graded treadmill test, and total and regional body fat distribution between MHO versus MUO adolescents using gold-standard imaging modalities such as magnetic resonance (MR) imaging and proton-MR spectroscopy. We also examined whether racial differences (black vs. white) exist in the cardiometabolic and body composition characteristics between MHO versus MUO adolescents.
Anthropometric measurement, and total and regional body fat distribution Body weight was measured to the nearest .1 kg and height was measured to the nearest .1 cm using standardized equipment. Waist circumference was measured at the last rib and the average of two measurements was used in the analyses. Total body fat (kg and %), trunk fat, and fat free mass (FFM) were measured using lunar iDXA (GE Healthcare, Madison, WI), and abdominal fat was measured at L4−L5 with the use of a 3.0 Tesla MR scanner (Siemens, Magnetom TIM Trio, Erlangen, Germany) [13,14]. For liver fat content, proton magnetic resonance spectroscopy (1H-MRS) was performed in the same MR scanner (Siemens, Tim Trio, Erlangen, Germany) using a body matrix coil and a spine matrix using the following parameters (voxel size = 30 £ 30 £ 20 mm3, TR = 4,000 ms, TE = 30 ms) as prescribed previously [13,14]. Oral glucose tolerance and insulin sensitivity tests All participants reported to the PCTRC after an overnight fast (minimum 10 hours) for a 2-hour oral glucose tolerance test (OGTT, 1.75 g/kg, max 75 g) to assess glucose tolerance status [15]. Following the OGTT, participants were admitted and stayed overnight in the hospital for a 3-hour 80 mU/m2/min hyperinsulinemic-euglycemic clamp test the next morning as described previously [16]. Participants were fed a standardized lunch and dinner and fasted overnight. Insulin-stimulated glucose disposal (Rd) was calculated as the average infusion rate of glucose during the last 30 minutes of the clamp and was expressed as milligrams per min per kg of FFM (RdFFM). Participants were fed a standardized lunch after the clamp test. Fasting lipids and blood pressure measurements
Methods Subjects The study sample included 189 overweight and obese black and white adolescents (BMI ≥ 85th percentile, 12−18 years of age) who were recruited via public advertisements and from the Weight Management and Wellness Center at Children’s Hospital of Pittsburgh (CHP) of the University of Pittsburgh Medical Center. Data from some of these participants have been reported previously [13,14]. Participants were nonsmokers, nondiabetic, physically inactive, and not taking medications known to affect study outcomes. Exclusion criteria included significant weight change (BMI > 2−3 kg/m2) prior to study participation, endocrine disorder (e.g., type 2 diabetes and polycystic ovary syndrome), or syndromic obesity. None of the participants had liver diseases and consumed alcohol. A complete medical history and physical examination, with evaluation of pubertal staging according to Tanner criteria, were conducted by a certified nurse practitioner. Participants self-identified as black or white during physical examination. Participants with mixed race were not included in the current study analyses. Routine hematological and biochemical testing were performed in all participants at the Pediatric Clinical and Translational Research Center (PCTRC) at CHP. Ethical approval was granted by the University of Pittsburgh institutional review board and written parental informed consent and child assent was obtained from all participants before any research participation. Evaluations were completed during an overnight inpatient admission at the PCTRC at CHP.
Fasting plasma lipids were determined in the Nutrition Laboratory of the University of Pittsburgh certified by the National Heart, Lung, and Blood Institute standardization program. Blood pressure was measured during inpatient admission when the patients were resting in the supine position with an automated sphygmomanometer every 10 minutes between 22:00 and 23:00 P. M. on the evening of admission, and between 06:00 and 07:00 A. M. before awakening next morning. The average value of seven measurements during each hour was used in the analysis. CRF Peak VO2 was measured during a graded maximal treadmill test with the use of standard open-circuit spirometry techniques until volitional fatigue [13,14]. A graded maximal treadmill test was conducted at least 2−3 hours after the hyperinsulinemiceuglycemic clamp test and after obtaining normalized glucose levels. For those who were unable to do the treadmill test, the test was conducted on a separate day. Statistical analyses Insulin-stimulated glucose disposal (RdFFM) did not differ (p > .1) between black versus white (11.4 § 3.5 mg/kgFFM/min vs. 11.9 § 3.6 mg/kgFFM/min, respectively) obese adolescents. Given that insulin sensitivity is a continuous trait and that there is no objective definition of insulin resistance, 189 black and white adolescents were combined and stratified into quartiles based on
ARTICLE IN PRESS S. Lee and S. Arslanian / Journal of Adolescent Health 00 (2018) 1−6
their RdFFM. Participants were defined as having high insulin sensitivity and MHO if their RdFFM value was in the upper quartile ( > 13.8 mg/kgFFM/min, black n = 26; white n = 22). Participants were defined as having low insulin sensitivity and MUO if their RdFFM value was in the lower three quartiles ( ≤ 13.8 mg/kgFFM/ min, black n = 87; white n = 54). Physical and metabolic characteristics and CRF between MHO versus MUO adolescents within each race were compared using independent t tests and chi-square tests. General linear modeling with least squared adjusted means was used to examine MHO versus MUO differences in total and regional body fat distribution adjusting for Tanner stage, gender, and BMI. Statistical procedures were performed using SPSS (Version 24; SPSS, Inc., Chicago, IL) and statistical significance was set at p ≤ .05. Results The study sample was composed of 60% black (113 of 189) and 40% white (76 of 189) overweight and obese adolescents. Of the 189 adolescents studied, 26 of 113 (23%) black and 22 of 76 (29%) white adolescents were classified as MHO (Table 1). There were no significant gender differences in the prevalence of MHO within each race. Despite similar age, MHO adolescents had lower BMI, body weight, and waist circumference compared with MUO adolescents in both races. In whites but not in blacks, Tanner stage was lower (p < .01) in MHO compared with MUO adolescents. Cardiometabolic characteristics of MHO versus MUO adolescents are shown in Table 2. By design, insulin-stimulated glucose disposal rate (RdFFM) was significantly different between MHO versus MUO adolescents in both races. In blacks, MHO adolescents had lower (p < .05) 2-hour OGTT glucose, triglycerides, very-lowdensity lipoprotein (VLDL)-cholesterol, and higher high-density lipoprotein cholesterol levels compared with MUO adolescents. Additionally, the prevalence of impaired fasting glucose (IFG, 30.1% vs. 4.4%, p = .06) and impaired glucose tolerance (IGT, 15.0% vs. .8%, p = .055) trended to be lower in MHO adolescents compared with MUO adolescents. In whites, MHO adolescents had significantly lower triglycerides and VLDL cholesterol levels, with a lower prevalence of IFG (17.1% vs. 1.3%, p = .046) compared with MUO adolescents. In both black and white adolescents with MHO phenotypes, CRF was significantly higher (p = .05) compared with adolescents with MUO. Total and regional body fat distribution between MHO versus MUO adolescents are shown in Figure 1. After adjusting for confounding factors (gender, Tanner stage, and BMI), there were no significant differences in the amount of total body fat (kg and %) between MHO versus MUO groups in both black and white adolescents. Despite similar total body fat (kg, %), MHO adolescents had
3
significantly lower trunk fat mass, waist circumference, and visceral fat compared with MUO adolescents independent of race. In whites but not in blacks, MHO adolescents also had lower liver fat when compared with MUO adolescents (4.0 § .5% vs. 1.9 § .9%, p = .055). Discussion The present study demonstrated that in both black and white youth, MHO adolescents have higher CRF measured as peak VO2, and lower triglycerides and VLDL cholesterol, with a lower prevalence of IFG as compared with MUO adolescents. After accounting for confounding factors such as gender, Tanner stage, and BMI, MHO adolescents have significantly lower waist circumference, trunk fat mass, and visceral fat with no differences in subcutaneous adipose tissue as compared with MUO adolescents despite similar total body fat. Further, in whites but not in blacks, MHO adolescents have lower liver fat compared with MUO adolescents. These findings suggest that in adolescents, the MHO phenotype is associated with high CRF, and lower abdominal obesity and visceral fat, and lower rates of dysglycemia in both black and white adolescents. To date, the comparisons of body fat distribution between MHO versus MUO have been mostly limited to adults [9,10,17]. Lower levels of visceral fat [9,10] or DXA-measured central adiposity [18] have been reported in MHO versus MUO postmenopausal women. In a mixed sample of black and white adults (18−68 years), Camhi et al. [17] have shown that MHO adults have significantly lower total adiposity, trunk fat, and visceral adiposity compared with MUO adults, independent of gender. In children and adolescents, lower waist circumference and insulin resistance measured by HOMA-IR have been consistent predictors of MHO phenotype [4−6], whereas dietary component and leisure-time physical activity level did not predict MHO [6]. To our knowledge, this is the first study that examined comprehensively the cardiometabolic and body composition characteristics, and objectively measured CRF between MHO versus MUO in both black and white adolescents. Recently, Cadenas-Sanchez et al. [7] demonstrated that although moderate-vigorous physical activity level was higher in MHO than in MUO adolescents, CRF did not differ between MHO versus MUO adolescents. However, it is important to note that in that study [7], CRF was evaluated by a 20-m shuttle run test and maximal O2 consumption was estimated using an equation. Our findings using the gold-standard methods of measuring CRF, body fat distribution, and insulin sensitivity extend previous findings by Cadenas-Sanchez et al. [7] and others [4−6] and provide evidence that the MHO and MUO phenotypes are characterized by differences in abdominal fat and CRF in both black and white obese adolescents. However, our findings differ
Table 1 Anthropometric characteristics of MHO versus MUO adolescents Black obese adolescents (n = 113)
n (%) Male/female, n Age, years Tanner stage*, II−III/IV−V BMI, kg/m2 Body weight, kg Waist circumference, cm
MHO
MUO
26 (23%) 10/16 14.9 § 1.9 4/21 32.3 § 4.5 87.4 § 17.1 90.6 § 11.2
87 (77%) 29/58 14.5 § 1.7 9/78 34.8 § 3.8 96.2 § 15.7 98.6 § 9.5
Data presented as mean § SD. *Tanner stage is missing in one subject in the black MHO group.
White obese adolescents (n = 76) p
MHO
MUO
p
.64 .38 .48 .005 .02 .001
22 (29%) 13/9 14.3 § 1.6 7/15 31.8 § 4.6 86.1 § 15.0 93.6 § 8.6
54 (71%) 26/28 14.7 § 1.6 4/50 34.0 § 4.1 97.9 § 15.4 100.7 § 9.6
.45 .37 .01 .04 .003 .004
ARTICLE IN PRESS S. Lee and S. Arslanian / Journal of Adolescent Health 00 (2018) 1−6
4
Table 2 Cardiometabolic characteristics of MHO versus MUO adolescents Black obese adolescents (n = 113)
Rd, mg/kgFFM/min Fasting glucose, mg/dl OGTT 2-hour glucose, mg/dl IFG, n (%) IGT, n (%) Total cholesterol, mg/dl Triglycerides, mg/dl HDL cholesterol, mg/dl LDL cholesterol, mg/dl VLDL cholesterol, mg/dl AM systolic BP, mm Hg AM diastolic BP, mm Hg PM systolic BP, mm Hg PM diastolic BP, mm Hg CRF, ml/kg/min RER
White obese adolescents (n = 76)
MHO
MUO
p
MHO
MUO
p
16.1 § 2.1 95.6 § 7.0 111.6 § 15.0 5 (4.4%) 1 (.9%) 141.5 § 22.7 56.8 § 19.1 47.6 § 10.2 82.5 § 19.7 11.4 § 3.8 110.8 § 9.0 59.2 § 6.9 114.4 § 9.6 61.7 § 6.7 25.8 § 4.6 1.07 § .08
10.0 § 2.5 97.0 § 7.2 121.2 § 21.4 34 (30.1%) 17 (15.0%) 144.7 § 24.9 76.0 § 30.7 42.9 § 7.9 86.6 § 23.1 15.2 § 6.1 109.7 § 10.4 59.9 § 6.5 114.0 §11.9 60.7 § 7.2 23.6 § 4.9 1.05 § .08
< .001 .36 .03 .06 .055 .55 .003 .01 .41 .003 .65 .69 .88 .56 .05 .29
16.1 § 2.3 93.6 § 3.9 120.1 §16.9 1 (1.3%) 3 (3.9%) 144.0 § 23.8 88.2 § 38.5 42.5 § 10.0 83.8 § 19.7 17.6 § 7.7 108.8 § 12.1 60.3 § 5.3 115.2 § 12.0 61.7 § 6.6 28.9 § 5.6 1.07 § .06
10.2 § 2.4 95.9 § 5.3 127.0 § 23.4 13 (17.1%) 15 (19.7%) 151.7 § 34.6 123.3 § 80.3 41.4 § 8.3 86.0 § 28.2 24.7 § 16.1 111.6 § 10.2 59.8 § 5.0 114.8 §11.1 60.0 § 5.8 26.4 § 4.4 1.09 § .06
< .001 .07 .21 .046 .19 .35 .055 .62 .74 .055 .31 .69 .88 .27 .046 .19
Data presented as mean § SD. BP = blood pressure; CRF = cardiorespiratory fitness; FFM = fat free mass; HDL = high-density lipoproteins; IFG = impaired fasting glucose; IGT = impaired glucose tolerance; LDL = low-density lipoprotein; NS = not significant; OGTT = oral glucose tolerance test; Rd = rate of glucose disposal; RER = respiratory exchange ratio; VLDL = very-low-density lipoprotein.
from Senechal et al. [4] who reported no differences in visceral fat and CRF between MHO versus MUO adolescents (n = 108). The disparate findings could be due to the differences in the cohort studied and the methodology employed to evaluate CRF (maximal treadmill test vs. cycle ergometer test). In that study [4], the data analyses combined adolescents from various ethnic background
MHO A
C
40 30 20 10 0
25
40
20
30 20 10
*
Visceral fat, cm2
80
*
80
*
60
Whites
† 5 4 3 2 1
40
Blacks
Whites
5
F 100
90
Blacks
10
Whites
E
100
*
0
Blacks
Liver fat, %
*
Whites
*
15
0
Blacks
D
50
Trunk fat, kg
Total adiposity, %
Total fat, kg
MUO
B 50
Waist circumference, cm
(56% white, 11% First Nation, and 32% other ethnic groups) despite the well-known differences in body fat distribution between adolescents of different ethnicities [19]. To our knowledge, this is the first study to compare liver fat content between MHO versus MUO in a biracial sample of adolescents. We found that in whites, MHO adolescents tend to have
0
Blacks
Whites
Blacks
Whites
Figure 1. Total and regional body fat distribution of MHO versus MUO adolescents. Data presented as the least squared means (§SE) adjusted for Tanner stage, gender, and BMI. (A) Total fat (kg), (B) total adiposity (%), (C) trunk fat (kg), (D) waist circumference (cm), (E) visceral fat (cm2), and (F) liver fat (%). *p < .05. yp = .055.
ARTICLE IN PRESS S. Lee and S. Arslanian / Journal of Adolescent Health 00 (2018) 1−6
lower liver fat compared with MUO adolescents, while this was not the case in blacks. The lack of association between the MHO phenotype and liver fat in black adolescents could be explained by substantially lower levels of liver fat in black (average liver fat in 113 adolescents = 1.6%) versus white (average liver fat in 76 adolescents = 3.4%) youth in our study, regardless of the MHO and MUO phenotypes. Our findings provide evidences that in white adolescents, the MHO phenotype imparts a lower risk for increased liver fat. To date, there is no established consensus on the MHO definitions in children and adults. In adults, increased insulin sensitivity has been often used to identify MHO individuals since insulin resistance is suggested to play a key role in the development of cardiometabolic abnormalities in obesity [1,12]. Studies in adults [3,9−11] have indentified MHO individuals based on insulin sensitivity measured using the gold-standard hyperinsulinemiceuglycemic clamp technique. However, the hyperinsulinemiceuglycemic clamp test is costly, time consuming, and invasive, requires technical expertise, and is not readily available at most pediatric centers. Therefore, the most common definitions of MHO in the pediatric literature have utilized simple fasting surrogate index (the homeostasis model assessment [HOMA]) [5,20] and/or the components of metabolic syndrome [5−7,21,22]. Recently, Heinzle et al. [6] examined the prevalence and predictors of MHO in 632 adolescents (12−19 years, BMI ≥ 95th percentile) using the National Health and Nutrition Examination Survey Data (1999−2010). The authors used two different definitions to identify MHO ([1] having ≤ 1 metabolic syndrome risk factors [excluding waist circumference] and free of type 2 diabetes, hypertension, and dyslipidemia, [2] free of all metabolic syndrome criteria, insulin resistance, and inflammation), and demonstrated that the prevalence of MHO in adolescents varies between 7% and 74% depending on the definition used [6]. Further, they found that lower levels of abdominal obesity, measured by waist circumference, and absence of insulin resistance, measured by HOMA-IR, were the most consistent significant predictors of MHO, independent of leisure-time physical activity and diet intake, in both boys and girls [6]. Given the increasing prevalence of comorbidities (e.g., type 2 diabetes and nonalcoholic fatty liver disease) in obese children and adolescents, characterizing MHO versus MUO youth is of importance to help health care professionals to identify highrisk obese youth for early interventions. Further studies are needed to investigate whether different definitions of MHO influence the predictors of MHO phenotype in youth. The strengths and limitations of this study warrant mention. To the best of our knowledge, this is the first study to examine MHO versus MUO phenotypes in a large sample of youth (n = 189), including black adolescents. We employed the state-of-the-art methodologies such as the 3-hour hyperinsulinemic-euglycemic clamp, MRI, 1H-MRS, and a graded maximal treadmill test to comprehensively evaluate cardiometabolic and body composition characteristics between MHO versus MUO adolescents. Whether our findings would remain true in other racial groups such as Asian and Hispanic youth, prepubertal children and severely obese youth are unknown. Further, other lifestyle variables such as sedentary behavior and sleep duration may play a role in MHO phonotype unknown in youth. In conclusion, the MHO phenotype is characterized by lower waist circumference, trunk fat mass and visceral fat, higher CRF, and lower rates of dysglycemia in both black and white adolescents. In whites, the MHO phenotype is also characterized by lower liver fat content.
5
Author Contribution S. Lee designed the study, researched the data, and wrote the manuscript. S. Arslanian researched data, supervised clamp experiment, reviewed clinical data, and provided critical revision of the manuscript. Funding Sources This research was funded by the National Institutes of Health (Grant Numbers 5R01HL114857, 1R21DK083654-01A1, UL1 RR024153, and UL1TR000005), American Diabetes Association (708-JF-27), and Children’s Hospital of Pittsburgh of UPMC (Cochrane-Weber Foundation and Renziehausen Fund) to Lee; the National Center for Advancing Translational Sciences Clinical and Translational Science Award (UL1 RR024153 and UL1TR000005) to the Pediatric Clinical and Translational Research Center, and the Department of Defense (FA7014-02-2-001) and Richard L. Day endowed chair to Arslanian. The American Diabetes Association, Department of Defense, National Institute of Health, and Cochrane-Weber Foundation at Children’s Hospital of Pittsburgh did not have any role in the design and conduct of the study; collection, management, analysis, interpretation of the data, and writing the manuscript. References [1] Karelis AD, St-Pierre DH, Conus F, et al. Metabolic and body composition factors in subgroups of obesity: What do we know? J Clin Endocrinol Metab 2004;89:2569–75. [2] Bonora E, Kiechl S, Willeit J, et al. Prevalence of insulin resistance in metabolic disorders: The Bruneck Study. Diabetes 1998;47:1643–9. [3] Ferrannini E, Natali A, Bell P, et al. Insulin resistance and hypersecretion in obesity. European Group for the Study of Insulin Resistance (EGIR). J Clin Invest 1997;100:1166–73. [4] Senechal M, Wicklow B, Wittmeier K, et al. Cardiorespiratory fitness and adiposity in metabolically healthy overweight and obese youth. Pediatrics 2013;132:e85–92. [5] Prince RL, Kuk JL, Ambler KA, et al. Predictors of metabolically healthy obesity in children. Diabetes Care 2014;37:1462–8. [6] Heinzle S, Ball GD, Kuk JL. Variations in the prevalence and predictors of prevalent metabolically healthy obesity in adolescents. Pediatr Obes 2016;11: 425–33. [7] Cadenas-Sanchez C, Ruiz JR, Labayen I, et al. Prevalence of metabolically healthy but overweight/obese phenotype and its association with sedentary time, physical activity, and fitness. J Adolesc Health 2017;61:107–14. [8] Jolliffe CJ, Janssen I. Development of age-specific adolescent metabolic syndrome criteria that are linked to the Adult Treatment Panel III and International Diabetes Federation criteria. J Am Coll Cardiol 2007;49:891–8. [9] Brochu M, Tchernof A, Dionne IJ, et al. What are the physical characteristics associated with a normal metabolic profile despite a high level of obesity in postmenopausal women? J Clin Endocrinol Metab 2001;86:1020–5. [10] Karelis AD, Faraj M, Bastard JP, et al. The metabolically healthy but obese individual presents a favorable inflammation profile. J Clin Endocrinol Metab 2005;90:4145–50. [11] Marini MA, Succurro E, Frontoni S, et al. Metabolically healthy but obese women have an intermediate cardiovascular risk profile between healthy nonobese women and obese insulin-resistant women. Diabetes Care 2007;30:2145–7. [12] Ferrannini E, Haffner SM, Mitchell BD, Stern MP. Hyperinsulinaemia: The key feature of a cardiovascular and metabolic syndrome. Diabetologia 1991;34:416–22. [13] Lee S, Bacha F, Hannon T, et al. Effects of aerobic versus resistance exercise without caloric restriction on abdominal fat, intrahepatic lipid, and insulin sensitivity in obese adolescent boys: A randomized, controlled trial. Diabetes 2012;61:2787–95. [14] Lee S, Deldin AR, White D, et al. Aerobic exercise but not resistance exercise reduces intrahepatic lipid content and visceral fat and improves insulin sensitivity in obese adolescent girls: A randomized controlled trial. Am J Physiol Endocrinol Metab 2013;305:E1222–9. [15] American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care 2004;27(Suppl 1):S5–10. [16] Lewy VD, Danadian K, Witchel SF, Arslanian S. Early metabolic abnormalities in adolescent girls with polycystic ovarian syndrome. J Pediatr 2001;138:38–44.
ARTICLE IN PRESS 6
S. Lee and S. Arslanian / Journal of Adolescent Health 00 (2018) 1−6
[17] Camhi SM, Katzmarzyk PT. Differences in body composition between metabolically healthy obese and metabolically abnormal obese adults. Int J Obes 2014;38:1142–5. [18] Peppa M, Koliaki C, Papaefstathiou A, et al. Body composition determinants of metabolic phenotypes of obesity in nonobese and obese postmenopausal women. Obesity 2013;21:1807–14. [19] Liska D, Dufour S, Zern TL, et al. Interethnic differences in muscle, liver and abdominal fat partitioning in obese adolescents. PloS One 2007;2: e569.
[20] Vukovic R, Mitrovic K, Milenkovic T, et al. Insulin-sensitive obese children display a favorable metabolic profile. Eur J Pediatr 2013;172:201–6. [21] Margolis-Gil M, Yackobovitz-Gavan M, Phillip M, Shalitin S. Which predictors differentiate between obese children and adolescents with cardiometabolic complications and those with metabolically healthy obesity? Pediatr Diabetes 2018. https://doi.org/10.1111/pedi.12694. [22] Ding W, Cheng H, Chen F, et al. Adipokines are associated with hypertension in metabolically healthy obese (MHO) children and adolescents: A prospective population-based cohort study. J Epidemiol 2018;28:19–26.