Obesity is a global pandemic(Reference Blüher1,Reference Jaacks, Vandevijvere and Pan2) that is associated with comorbidities such as diabetes, hypertension, dyslipidaemia, CVD and some cancers(Reference Garg, Maurer and Reed3,Reference Ortega and Lavie4) . The BMI is a simple index, which considers weight and height, commonly used to classify obesity, but which is not able to differentiate between lean and fat tissue(Reference Group5). In this context, it has been recognised that there are individuals who have a high cardiometabolic risk while maintaining body weight within the normal range by BMI, called metabolically obese normal-weight (MONW)(Reference Ruderman, Schneider and Berchtold6).
Different criteria are used for the diagnosis of this phenotype, and the condition is often defined by the presence of one or more metabolic alterations(Reference Galić, Pavlica and Udicki7,Reference Green, Jacques and Rogers8) . Thus, adults with this phenotype, despite their normal-weight, are described as having an unfavourable lipid and glucose profile, as well as an increased risk of diabetes and CVD(Reference Aung, Lorenzo and Hinojosa9,Reference Kramer, Zinman and Retnakaran10) . Studies with adolescents also show that MONW may have reduced HDL, in addition to elevated leptin, insulin, homoeostasis model assessment – insulin resistance, and high TAG levels(Reference Molero-Conejo, Morales and Fernández11,Reference Kelishadi, Cook and Motlagh12) . Also, MONW individuals may be metabolically affected in a similar way to the BMI’s overweight group(Reference Molero-Conejo, Morales and Fernández11).
The identification of this phenotype in normal-weight adolescents suggests that determinants other than BMI may influence clinical outcomes related to cardiometabolic health. The use of anthropometric measurements and body composition indices, independent of the nutritional status classified by BMI, allow an evaluation of the contribution of lean mass and adiposity in predicting the risk of diseases(Reference Ding, Liu and Shang13) and, thus, could help in the identification of MONW individuals.
Although the researches did already contribute with knowledge about the MONW phenotype, few studies on this subject have been performed with adolescents(Reference Molero-Conejo, Morales and Fernández11–Reference Li, Li and Han16). Besides, the phenotype can often be underdiagnosed in adolescents, due to normal body weight and young age(Reference Karelis, St-Pierre and Conus17). From this perspective, early identification of MONW adolescents, as well as related anthropometric and body composition parameters, becomes important. Therefore, this study aimed to investigate in a sample of normal weight adolescents the anthropometric and body composition parameters associated with the MONW phenotype.
Materials and methods
This study is part of the ‘Comparative study between the three phases of adolescence in relation to excess body fat and cardiovascular risk factors for metabolic syndrome’, already detailed in other publications(Reference Pereira18,Reference Faria19) .
Population, design and sampling
Epidemiological study with a cross-sectional design, population-based, conducted with adolescents between 10 and 19 years old, of both sexes, selected from the rural and urban, public and private school population in the municipality of Viçosa, Minas Gerais (MG), Brazil, between the years 2010 and 2013. The sample size was calculated using the StatCalc of the Epi Info software, version 6.04, from a specific formula for cross-sectional studies, considering a total population of 11 898 adolescents in the city of Viçosa/MG(20), an expected prevalence of 50·0 %(Reference Luiz, Magnanini, Medronho, Carvalho and Block21), acceptable variability of 5 %, and 95 % confidence level, totaling a minimum sample of 372 adolescents. When 20 % was added to this to control for confounding factors, a minimum total of 447 was required. A total of 506 adolescents with an appropriate weight according to BMI/age(22) were evaluated in this study.
The inclusion criteria were no regular use of medications that alter blood glucose, insulinaemia, lipid metabolism, and/or blood pressure levels, no participation in a weight reduction or weight control programme, no regular use of diuretics/laxatives, not pregnant or having ever been pregnant, no neck deformities and no diagnosis of infection, acute inflammation or thyroid disease. According to these criteria, twenty-eight adolescents were excluded from the initial study sample.
All participants and their parents/guardians, in the case of volunteers under the age of 18, signed the Informed Consent Form, under the Declaration of Helsinki. The study was approved by the Ethics Committee on Human Research of the Federal University of Viçosa – Comitê de Ética em Pesquisas com Seres Humanos da Universidade Federal de Viçosa (Of. Ref. No. 0140/2010).
Anthropometry
Weight and height were measured by international standard techniques(Reference Lohman, Roche and Martorell23) using electronic digital scales (LC 200PP, Marte®) and portable stadiometer (Alturexata®).
Waist circumference (WC) was obtained at the end of a normal expiration, at the midpoint between the lower margin of the last rib and the iliac crest(Reference Lohman, Roche and Martorell23), and used continuously. Hip circumference (HC) was measured in the gluteal region, around the largest horizontal circumference between the waist and the knees(Reference Lohman, Roche and Martorell23). The waist:hip ratio (WHR) was obtained by dividing the waist circumference (cm) by the hip circumference (cm), and the waist:height ratio (WHtR) was obtained by the quotient of the waist circumference (cm) by the height measurement (cm), evaluated continuously.
Neck circumference was measured at the midpoint between the spine and the anterior neck, except when the individual had pronounced thyroid cartilage (Adam’s apple). In these cases, the circumference was measured just below the thyroid cartilage(Reference Ben-Noun, Sohar and Laor24). This measure was used on an ongoing basis.
The Paediatric Body Adiposity Index (BAIp), calculated from the measurement of hip circumference and height, was also evaluated continuously, according to the equation: BAIp = ((hip circumference (cm)/(height (m))0·8)–38(Reference El Aarbaoui, Samouda and Zitouni25).
Body composition assessment
For body composition analysis, the dual energy X-ray absorptiometry (DEXA) equipment (Lunar Prodigy Advance DXA System – analysis version: 13.31, GE Healthcare) was used. The participants were barefoot, wearing light clothing, no metal ornaments and fasting for 12 h. The evaluation was performed with each individual in the supine position, through a series of transverse scans from head to toe, with a whole-body scan duration of approximately 10 min. The body composition analysis included android fat, gynoid fat, total fat mass and trunk, arms, and legs fat. Arm and leg fat was defined, respectively, as the sum of the fat in two arms and two legs, both divided by two(Reference Ribeiro, Kogure and Lopes26). The following indices were calculated according to the fat distribution: android:gynoid ratio; trunk:legs ratio; trunk:arms ratio(Reference Ribeiro, Kogure and Lopes26,Reference Guimarães, Guimarães and Kakehasi27) .
Besides, the fat mass index (FMI) and the fat-free mass index, proposed by Van Itallie et al. (1990)(Reference VanItallie, Yang and Heymsfield28), to have a more careful anthropometric evaluation, according to the body compartments, by calculation that considers the amount in kilograms of fat mass and fat-free mass relative to height, as follows: FMI = (body fat (kg)/height (m)2) and fat-free mass index = ((Weight (kg) – body fat (kg))/height (m)²).
The body composition analysis also included the lean mass of the legs and arms, and the following parameters were analysed: appendicular lean mass (ALM), obtained by adding the lean mass of the arms and legs. With the ALM data, the lean mass index relative to height (ALM/height², given in kg/m²)(Reference Baumgartner, Koehler and Gallagher29); the lean mass index relative to weight (ALM/weight × 100, given in %)(Reference Janssen, Heymsfield and Ross30); the lean mass index relative to BMI (lean mass index relative to height BMI: ALM/IMC, given in kg/kg/m²)(Reference Studenski, Peters and Alley31) and the indices of metabolic load and capacity (IMLC), which relate total fat mass to total fat-free mass (TFFM) (total IMLC = total fat mass/TFFM)(Reference Siervo, Prado and Mire32), and trunk fat mass to ALM (regional IMLC = trunk fat mass/ALM) were obtained(Reference Siervo, Prado and Mire32).
Biochemical assessment
Blood samples (12 ml) were collected at the Clinical Analysis Laboratory of the Health Division of the Federal University of Viçosa, after a 12-hour fast, by venipuncture, with disposable syringes. The analyses were carried out at the Nutritional Biochemistry Laboratory of the Department of Nutrition and Health and at the Molecular Immunovirology Laboratory of the Department of General Biology, at the Federal University of Viçosa.
HDL, LDL and TAG analyses were performed on blood serum after the material had been centrifuged in an Excelsa 206 BL centrifuge for 10 min at 3500 rpm. HDL and TAG were measured by the colorimetric enzymatic method, with automation by the Cobas Mira Plus equipment (Roche Corp.), and LDL was calculated by Friedewald’s formula for TAG values lower than 400 mg/dl(Reference Friedewald, Levy and Fredrickson33). Classification of the lipid profile was performed according to the Integrated Guide for cardiovascular health and risk reduction in children and adolescents (Guia integrado para saúde cardiovascular e redução de risco em crianças e adolescentes)(34), in which altered values (mg/dl) for LDL ≥ 130; HDL < 40 and TAG ≥ 130 are considered.
Fasting blood glucose was measured by the enzymatic glucose-oxidase method using the Cobas Mira Plus (Roche Corp.) automation equipment, and it was considered an altered fasting blood glucose ≥ 100 mg/dl(35). Fasting insulin was measured by the electrochemiluminescence method. Insulin resistance was calculated through the mathematical model homoeostasis model assessment – insulin resistance, using insulin and fasting blood glucose levels. Homoeostasis model assessment – insulin resistance values ≥ 3·16 were considered as insulin resistance(36).
Blood pressure assessment
Blood pressure was measured, according to the protocol established by the VI Brazilian Guidelines on Hypertension (I Diretriz Brasileira de Hipertensão Arterial)(Reference de Andrade and Nobre37), using an automatic inflation blood pressure monitor (Omron® Model HEM-741 CINT), recommended by the Brazilian Society of Cardiology (Sociedade Brasileira de Cardiologia). The blood pressure was measured in the right and left arms, and the measurement was repeated twice in the arm with the highest pressure value, with a 1-minute interval between them, and the average of the last two measurements was taken.
Elevated blood pressure levels were defined as systolic or diastolic blood pressure ≥ percentile 90 by age, sex and height for the adolescents under 13 years of age and for the group aged 13 years and older, systolic blood pressure ≥ 120 mmHg or diastolic blood pressure ≥ 80 mmHg were considered elevated pressure levels, according to the recommendations of the American Academy of Pediatrics (2017)(Reference Flynn, Kaelber and Baker-Smith38).
Definition of the metabolically obese normal-weight phenotype
Adolescents who have adequate weight but the presence of at least one metabolic alteration(Reference Green, Jacques and Rogers8) were considered MONW in this study. To evaluate the nutritional status of adolescents and classify them as normal-weight, the BMI was used, obtained by dividing the weight by the square of the height, with values between the percentiles ≥ 3 and < 85, analysed according to sex and age, according to the WHO(22).
Also, it was considered metabolic alterations: high blood pressure(Reference Flynn, Kaelber and Baker-Smith38), high fasting glucose(35), alteration in the lipid profile (low HDL, high LDL or TAG)(34), and insulin resistance by homoeostasis model assessment – insulin resistance(36). Alteration in at least one of these components defined the metabolic abnormality.
Covariates
A questionnaire was applied to evaluate the profile of the adolescents, such as age, sex and type of school where they study (public or private). Besides, the socio-economic condition was investigated through the application of a questionnaire that collects a variety of social and economic issues at the household level, using the same methodology adopted by the Survey on Living Standards (Pesquisa sobre Padrões de Vida – PPV)(39), which was classified as ‘adequate’ or ‘precarious and intermediate’.
The level of physical activity was assessed using the International Physical Activity Questionnaire – short version, validated for this population group(Reference Guedes, Lopes and Guedes40) as a way to classify adolescents as sedentary, irregularly active, active and very active(41). Considering the low prevalence of sedentarism in the studied population, we chose to group the sedentary and irregularly active individuals into insufficiently active, and those considered active and very active were grouped into physically active.
The dietary analysis was performed by applying the qualitative FFQ to know the frequency of consumption of food groups. The FFQ was applied individually, and the adolescents were instructed to report on the frequency of consumption in the month before the date of application of the questionnaire(Reference Serra-Majem, Aracenta-Bartrina, Serra-Majem, Aracenta-Bartrina and Mataix-Verdú42).
The list of foods that made up the FFQ was determined considering the foods that are part of the eating habits of adolescents in the city of Viçosa, MG, based on data from the application of 24-h recall tests on adolescents assisted by the Adolescent Health Care Program (Programa de Atenção à Saúde do Adolescente – PROASA) of UFV. Weekly consumption frequency was categorised as ≥ 4 or < 4 times a week(Reference Olafsdottir, Torfadottir and Arngrimsson43). The consumption of fruits, vegetables and legumes was used as a marker of a healthy diet.
Statistical analysis
The database was prepared by a double entry in Microsoft Office Excel 2007. The statistical analyses were performed in STATA software, version 14. The consistency and distribution of the quantitative variables were evaluated by histograms, coefficient of asymmetry and kurtosis and the Shapiro–Wilk normality test. Categorical variables were expressed as absolute and relative frequency; quantitative variables were expressed as mean and standard deviation or median and interquartile range. The statistical differences of the quantitative variables according to the presence or absence of the MONW phenotype were analysed by Student’s t test or the nonparametric Mann–Whitney test, according to the normality of the variables, as well as the homogeneity of variances. Statistical differences for categorical variables were calculated by the χ 2 test; or Fisher’s exact test for when more than 20 % of the cells had expected counts < 5. In the bivariate analysis, P values < 0·20 were considered for inclusion in the regression models. The WHtR, WHR, android:gynoid ratio and regional IMLC were converted to Z-score in the regression models using the following equation:
Z-score = (individual anthropometric value - mean anthropometric value)/standard deviation(Reference Nascimento-Souza, Lima-Costa and Peixoto44).
To evaluate the associations of anthropometric and body composition parameters (explanatory variables) with the MONW phenotype (outcome variable), Poisson regression with robust variance was employed, and prevalence ratios (PR) with 95 % CI were estimated. Two independent models were built for each explanatory variable, one crude and one adjusted for potential confounding factors defined according to the literature(Reference Ruderman, Schneider and Berchtold6–Reference Karelis, St-Pierre and Conus17) (age, sex, physical activity level, socio-economic condition, and fruit, vegetable and legume consumption). Regression analysis was used in the total sample, and there was also stratification by sex and age (10–13 years old, 14–16 years old and 17–19 years old). The significance level adopted in all analyses was 5 %.
Results
Among the evaluated adolescents, the majority of participants are female (54·5 %), and the prevalence of the MONW phenotype was 29·1 % (n 147) (95 % CI 24·9, 32·9). There were no statistically significant differences in age, sex, school type, socio-economic condition, physical activity level and food intake between the groups with and without the MONW phenotype (Table 1). Regarding the indicative measures of central and total adiposity and the body composition indices, the MONW adolescents presented, compared with the normal-weight adolescents without the phenotype, higher waist circumference (cm), WHtR, WHR, neck circumference (cm), android:gynoid fat ratio, trunk:leg and trunk:arm fat mass ratio and regional IMLC (Table 2).
Characterisation of adolescents according to the absence or presence of the metabolically obese normal-weight phenotype (MONW)
(Numbers and percentages)
The results were expressed as absolute and (relative) frequency for categorical variables. The quantitative variable (age) was expressed as mean and (standard deviation).
* Pearson’s χ 2.
† Student’s t test.
‡ Fisher’s Exact Test.
Anthropometric and body composition parameters, according to the absence or presence of the metabolically obese normal-weight phenotype (MONW)
FM, fat mass; BAIp, paediatric body adiposity index; FMI, fat mass index, FFMI, fat-free mass index; LMI height, lean mass index relative to height; LMI weight, lean mass index relative to body weight; LMI BMI, lean mass index relative to body mass index; IMLC, index of metabolic load and capacity.
The results were expressed as mean and (standard deviation) when analysed by Student’s t test or as median and (interquartile range) when analysed by Mann–Whitney test.
* Mann–Whitney.
† Student’s t test.
As for the main metabolic alterations in the MONW group, it was observed that the frequency of participants with altered HDL among those with the MONW phenotype is 46·9 %; also, 23·8 % of MONW adolescents have altered LDL (Fig. 1).
Distribution of metabolic alterations in adolescents with the metabolically obese normal-weight phenotype.
In the adjusted regression models, the MONW phenotype was positively associated with waist circumference (cm) (PR = 1·05; 95 % CI 1·03, 1·08), WHtR (PR = 1·23; 95 % CI 1·07, 1·41), WHR (PR = 1·25; 95 % CI 1·07, 1·47), android:gynoid fat ratio (PR = 1·34; 95 % CI 1·19, 1·51), trunk:leg fat ratio (PR = 2·67; 95 % CI 1·64, 4·32), trunk:arm fat ratio (PR = 1·12; 95 % CI 1·04, 1·21), regional IMLC (PR = 1·32; 95 % CI 1·15, 1·51), as well as with FMI (PR = 1·10; 95 % CI 1·02, 1·22) (Table 3).
Association between anthropometric and body composition parameters with the MONW phenotype in adolescents
(Prevalence ratios and 95 % confidence intervals, n 506)
FM, fat mass; PR, prevalence ratio; BAIp, paediatric body adiposity index; FMI, fat mass index; FFMI, fat-free mass index; IMLC, index of metabolic load and capacity.
* Statistical significance.
† Adjustment for sex; age; physical activity level; socio-economic condition and consumption of fruits, vegetables,and legumes (healthy diet).
‡ Waist:height ratio, waist:hip ratio, android:gynoid ratio and regional IMLC were converted to Z-score.
In the analysis stratified by sex, there was a positive association between waist circumference (cm) (PR = 1·05; 95 % CI 1·01, 1·09), WHtR (PR = 1·26; 95 % CI 1·07, 1·49), WHR (PR = 1·32; 95 % CI 1·06, 1·65), android:gynoid fat ratio (PR = 1·25; 95 % CI 1·03, 1·51), trunk:arm fat ratio (PR = 1·13; 95 % CI 1·02, 1·24) and the MONW phenotype in males (Table 4). In females, the phenotype was positively associated with waist circumference (cm) (PR = 1·06; 95 % CI 1·02, 1·09), WHtR (PR = 1·29; 95 % CI 1·07, 1·56), android:gynoid fat ratio (PR = 1·39; 95 % CI 1·20, 1·62), trunk:leg fat ratio (PR = 2·84; 95 % CI 1·46, 5·53), BAIp (PR = 1·06; 95 % CI 1·01, 1·12), FMI (PR = 1·24; 95% CI 1·10, 1·41) and regional IMLC (PR = 1·29; 95 % CI 1·09, 1·53) (Table 5).
Association between anthropometric and body composition parameters with the MONW phenotype in male adolescents
(Prevalence ratios and 95 % confidence intervals, n 230)
FM, fat mass; PR, prevalence ratio; BAIp, paediatric body adiposity index; FMI, fat mass index; FFMI, fat-free mass index; IMLC, index of metabolic load and capacity.
* Statistical significance.
† Adjustment for age; physical activity level; socio-economic condition and consumption of fruits, vegetables and legumes (healthy diet).
‡ Waist:height ratio, waist:hip ratio, android:gynoid ratio and regional IMLC were converted to Z-score.
Association between anthropometric and body composition parameters with the MONW phenotype in female adolescents
(Prevalence ratios and 95 % confidence intervals, n 276)
FM, fat mass; PR, prevalence ratio; BAIp, paediatric body adiposity index; FMI, fat mass index; FFMI, fat-free mass index; IMLC, index of metabolic load and capacity.
* Statistical significance.
† Adjustment for age; physical activity level; socio-economic condition and consumption of fruits, vegetables and legumes (healthy diet).
‡ Waist:height ratio, waist:hip ratio, android:gynoid ratio and regional IMLC were converted to Z-score.
Furthermore, to check whether the associations differ by age, we analysed them according to the stages of adolescence (Table 6). In early adolescence (ages 10–13), the phenotype was positively associated with android:gynoid fat ratio (PR = 1·30; 95 % CI 1·10, 1·54), trunk:leg fat ratio (PR = 3·26; 95 % CI 1·47, 7·23), trunk:arm fat ratio (PR = 1·14; 95 % CI 1·02, 1·28) and regional IMLC (PR = 1·31; 95 % CI 1·10, 1·57). In middle adolescence (ages 14–16), there was a positive association with waist circumference (cm) (PR = 1·06; 95 % CI 1·01, 1·12), neck circumference (cm) (PR = 1·25; 95 % CI 1·01, 1·54), android:gynoid fat ratio (PR = 1·46; 95 % CI 1·10, 1·93), trunk:leg fat ratio (PR = 3·63; 95 % CI 1·27, 10·37) and regional IMLC (PR = 1·38; 95 % CI 1·06, 1·80). In late adolescence (ages 17–19), the MONW phenotype was positively associated with waist circumference (cm) (PR = 1·06; 95 % CI 1·02, 1·10), WHtR (PR = 1·28; 95 % CI 1·02, 1·61) and WHR (PR = 1·32; 95 % CI 1·02, 1·72).
Association between anthropometric and body composition parameters with the MONW phenotype according to the stages of adolescence
(Prevalence ratios and 95 % confidence intervals)
FM, fat mass; PR, prevalence ratio; BAIp, paediatric body adiposity index; FMI, fat mass index; FFMI, fat-free mass index; IMLC, index of metabolic load and capacity.
* Statistical significance.
† Adjustment for age; sex; physical activity level; socio-economic condition and consumption of fruits, vegetables and legumes (healthy diet).
‡ Waist:height ratio, waist:hip ratio, android:gynoid ratio and regional IMLC were converted to Z-score.
Discussion
In this study, conducted with normal-weight adolescents, the presence of the MONW phenotype was positively associated with anthropometric and body composition parameters indicative of central fat, such as waist circumference, WHtR, WHR, android/gynoid, trunk/leg and trunk:arm fat ratio, as well as with regional IMLC. The phenotype was also positively associated with FMI, indicative of total body fat. Thus, anthropometric and body composition measurements can be useful in the early diagnosis of normal-weight adolescents with high cardiometabolic risk.
The deposition of fat in the central or abdominal region, found in greater magnitude in the MONW adolescents of this study, is related to a higher risk of metabolic complications, mainly due to the accumulation of visceral fat(Reference Barroso, Marins and Alves45,Reference Bergman, Kim and Hsu46) . This type of predominantly central or abdominal deposition, in the region of the thorax and abdomen, is characteristic of android obesity, also called truncal obesity(Reference Vague47).
Anthropometric measures have been used to identify central fat and cardiometabolic risk. These measures deserve to be highlighted in the context of clinical practice and epidemiological studies for measuring adiposity, given their better applicability, availability, safety, low cost and good correlation with adiposity(Reference Vasques, Priore and Rosado48,Reference Cavalcanti, Carvalho and Barros49) . In this sense, our findings agree with other studies(Reference Molero-Conejo, Morales and Fernández11,Reference Karelis, St-Pierre and Conus17) reporting that individuals with the MONW phenotype, despite having normal-weight by BMI, have higher waist circumference, which is a good marker of metabolic risk(Reference Katzmarzyk, Srinivasan and Chen50–Reference Li, Ford and Mokdad52). Besides, MONW had higher WHtR, an index that can be used to assess abdominal fat(Reference Li, Ford and Mokdad52,Reference Taylor, Jones and Williams53) and identify adolescents with high metabolic and cardiovascular risk(Reference Maffeis, Banzato and Talamini54), as well as higher WHR, used to assess body fat distribution and as an indicator of central obesity(Reference Li, Ford and Mokdad52,Reference Taylor, Jones and Williams53) . When considering the analysis stratified by age, we also found in adolescents aged 14 to 16 years a positive association of the phenotype with another measure indicative of central fat and cardiometabolic risk, the neck circumference.
The MONW phenotype was also positively associated with the regional IMLC in this study, which is an index that relates trunk fat to ALM, referring to visceral adiposity, associated with high morbidity and mortality from CVD(Reference Reis, Loria and Lewis55). The use of the IMLC helps to fill a gap in body composition studies, which is represented by the identification of individuals who present at the same time high adiposity and low muscle mass(Reference Siervo, Prado and Mire32,Reference Powell, Lara and Mocciaro56) . In this way, it is a model that demonstrates the metabolic imbalance in the individual, referring to the tissues that threaten the body’s homoeostasis (fat) and those that maintain it (muscle)(Reference Siervo, Prado and Mire32). To our knowledge, this is the first study to identify this dysfunction of metabolic load and capacity in MONW adolescents.
Moreover, the MONW phenotype was positively associated with the FMI, which is an indicative index of total body fat that allows a more careful anthropometric evaluation of fat mass, according to body compartments, by calculation that considers height. Furthermore, when considering only female participants, we found an association of the MONW phenotype with BAIp, calculated by measuring the hip circumference and height, and which is also indicative of total body fat. Adipose tissue is capable of synthesising substances such as non-esterified fatty acids, hormones and pro-inflammatory cytokines – leptin, TNF and visfatin – that are related to the development of diseases(Reference Van Gaal, Mertens and De Block57,Reference Balagopal, de Ferranti and Cook58) . The secretion of these substances by the adipose tissue, especially in the abdominal region, predicts cardiometabolic alterations, insulin resistance, type 2 diabetes mellitus, metabolic syndrome and, consequently, CVD(Reference Kelishadi, Mirmoghtadaee and Najafi59). Thus, excess body fat is associated with increased cardiometabolic risk, and its early identification, especially in the younger population, is essential for the prevention of chronic diseases in adulthood. Thus, screening for excess body fat may be useful for the early diagnosis of MONW adolescents.
In this study, although we did not find an association between the MONW phenotype and lean mass indices, we emphasise that muscle mass is related to the metabolic profile. Skeletal muscle is the most abundant insulin-sensitive tissue in our body, in addition to being the primary site of glucose utilisation from the insulin-regulated glucose transporter (GLUT4), thus presenting a protective role against resistance to insulin and type 2 diabetes mellitus(Reference Walsh60,Reference Stump, Henriksen and Wei61) . Furthermore, the secretion of myokines by skeletal muscle can positively interfere in the prevention of insulin resistance and inflammation(Reference Walsh60).The fact that we did not find an association between the MONW phenotype and lean mass indices suggests a predominant effect of fat distribution on metabolic risk, exceeding the contribution of other parameters of body composition. However, the absence of observed association, added to the existing heterogeneity in the definition of phenotypes, in addition to the different ways of assessing muscle mass, indicates the need for further studies to clarify the role of this body compartment in predicting cardiometabolic risk in adolescents.
In this context, due to the high prevalence (29·1 %) of adolescents with the MONW phenotype found in this study, our results indicate that using only the BMI seems insufficient to determine cardiometabolic risk. Thus, the inclusion of measures of body fat – especially those indicative of central adiposity – in young people with normal weight will provide additional information and will help us to identify the MOWN phenotype, enabling an early intervention in order to reduce their possible cardiometabolic risk.
This study is not free of limitations. The cross-sectional design makes it impossible for us to ensure the temporality of the observed associations. Furthermore, there is no consensus on the concept of ‘metabolic abnormality’ for the classification of the MONW phenotype. However, it is valid to consider that due to the scarcity of studies with adolescents on this theme, the present study brought important contributions, especially for the nutritional assessment of this group. Moreover, it reinforces the importance of early detection of this phenotype, because for the young age group in question, having an alteration in cardiometabolic parameters, especially for the fact that they have normal-weight, already demonstrates concern about the health status of adolescents. Also noteworthy as positive points are the fact that this is a population-based study, the methodological rigor in data collection and the use of validated methods, such as DEXA. Furthermore, body composition indices adjusted for weight, BMI and height, as well as those that consider body compartments, which have been studied in differentiating the contribution of lean mass and adiposity in predicting disease risk, were evaluated.
Conclusion
The MONW phenotype associates positively with anthropometric and body composition measures indicative of central and total fat. Furthermore, the phenotype was positively associated with regional IMLC, revealing a metabolic imbalance of tissues and a threat to the body homoeostasis of MONW adolescents.
Acknowledgements
The authors thank all adolescents who participated in this work and their parents/guardians and the Coordination for the Improvement of Higher Education Personnel (CAPES, Brazil, funding code 001) for the scholarship granted to B. C. C.
This study was supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) (grant number 485986/2011-6) and Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG) (grant number APQ-01618-10). The CNPq and FAPEMIG had no role in the design, analysis or writing of this article.
No competing financial interests exist.
B. C. C. was responsible for the analysis and interpretation of the data, conducted the literature search as well as wrote the manuscript. S. A. V. R. and L. L. J. participated in the analysis and interpretation of data as well as the critical review of the paper. S. E. P. acted in the conception and design of the study as well as the critical review of the paper. E. R. F. and F. R. F. participated in the conception of the study design, data collection as well as the critical review of the paper. P. F. F. participated in the conception of the study design, data collection and writing of the article. All authors have read and approved the final manuscript.
There are no conflicts of interest.
