Somatotype characteristics of normal-weight and obese women among different metabolic subtypes

ABSTRACT Background Obesity is a well known risk factor for the development of metabolic abnormalities. However, some obese people are healthy and on the other hand some people with normal weight have adverse metabolic profile, therefore it can be assumed that there is a difference in physical characteristics amongst these people. The aim of this study was to establish whether there are somatotype differences between metabolically healthy and metabolically obese women who are obese or of normal weight. Subjects and methods Study included 230 women aged 44.76 ± 11.21y. Metabolic status was assessed according to IDF criteria, while somatotype was obtained using Heath & Carter method. Results Significant somatotype differences were observed in the group of women with normal-weight: metabolically healthy women had significantly lower endomorphy, mesomorphy and higher ectomorphy compared to metabolically obese normal-weight women (5.84-3.97-2.21 vs. 8.69-6.47-0.65). Metabolically healthy obese women had lower values of endomorphy and mesomorphy and higher values of ectomorphy compared to ‘at risk’ obese women but the differences were not statistically significant (7.59-5.76-0.63 vs. 8.51-6.58-0.5). Ectomorphy was shown as an important determinant of the favorable metabolic profile (cutoff point was 0.80). Conclusion We concluded that, in addition to fat mass, metabolic profile could be predicted by the structure of lean body mass, and in particular by body linearity.


INTRODUCTION
S omatotyping provides the quantitative description of the human physique.The most widely used so matotype method was introduced by Heath & Car ter; it is expressed in three components (endomorphy, mesomorphy and ectomorphy) that empirically define different aspects of the body composition: degree of fatness, musculoskeletal development and the linearity of the body (1).Each individual is a unique combina tion of all the above three components in different pro portion.Endomorphy, mesomorphy and ectomorphy correspond with the three primary germ cell layers that give rise to the specific sets of tissues that define body composition.
Obesity, especially central (truncal) type, has been proven to be an independent risk factor for the deve lopment of cardiovascular and metabolic disturbances.However, some phenotypically obese individuals have normal metabolic profile.Some studies indicate that 1025% of obese individuals are actually metabolically healthy (24).On the other hand, it has been estimated that 1318% of normalweight individuals have abnor mal metabolic profile (5,6).The mechanisms under lying the metabolic disturbances in metabolically obese normal weight subjects, as well as those that prevent the development of metabolic abnormalities in metabo lically healthy obese subjects, are poorly understood.It is assumed that the muscle metabolic capacity and also the ability to store fat in the subcutaneous adipose tissue depots instead of in visceral depots could be of great importance in understanding these phenomena (4,7).This would imply certain somototype differences bet ween metabolically healthy and metabolically obese in dividuals of the same nutrition level.The purpose of this study was to analyse somatotype in normalweight and obese women with respect to their metabolic profile.

SUBJECTS AND METHODS
Study group involved 230 women aged 22 to 76 years (average age: 44.76 ± 11.21y) who voluntarily partici pated in the study.This investigation was taken as a part of a larger crosssectional population study of the pre valence of obesity and cardiovascular risk factors among adult population living in the urban and rural areas of Vojvodina province situated in the northern part of Ser bia.Vojvodina represents the most demographically di verse region of Serbia with more than 25 ethnic groups (most prominent ethnic groups are Serbs and Hunga rians).Participants were invited to participate in the study via local media, pamphlets and social networks.Participants underwent thorough evaluation, inclu ding medical and family history, physical examination and blood biochemistry by health professionals; all the tests included were free of charge for all participants.Candidates with any of the following conditions were excluded from the study: history or evidence of cardio vascular diseases, diabetes, malignancies, chronic liver disease, using steroids, hormone displacement therapy, or medication that could affect body composition, car diovascular function or metabolism, pregnancy, curren tly breastfeeding, and large body mass fluctuations in the last 6 months.Participants who had missing data or presented difficulties with measuring were also exclu ded.The study was carried out pursuant to the Decla ration of Helsinki.In order to assess somatotype, nutri tion level and metabolic profile all subjects underwent anthropometric measurements, blood pressure measu rements and biochemical analyses.
Body mass was obtained through body compositi on assessment using the bioelectrical impedance ana lysis (Tanita TBF310 bioimpedance analyzer, Tanita Corporation, Tokyo, Japan).Body height was measu red to the nearest 0.1 cm using GPM anthropometer (Sieber&Hegner, Zürich, Switzerland).Body girths (flexed and tensed upper arm girth, waist girth and calf girth) were measured using Holtain flexible but nonstretchable tape (Holtain Ltd, Croswell, UK) to the nearest 0.1 cm.Upper arm girth was measured as the maximal girth of the upper arm with flexed and tensed elbow.Calf girth was measured as the greatest girth of the calf.Waist circumference was measured at the level midway between the lowest point of the rib margin and the highest point of the iliac crest.Skin fold thicknesses (triceps, subscapular, supraspinale and medial calf) were measured using Harpenden caliper (Holtain Ltd, Croswell, UK) to the nearest 0.2 mm.Triceps skinfold thickness was measured in the vertical direction at the level halfway between the acromion and olecranon.Subscapular skinfold thickness was measured below the inferior angle of the scapula in an oblique direction downwards and laterally at 45 de grees.Supraspinale skinfold thickness was measured above the anterior superior iliac spine on a line to the anterior axillary border and on a diagonal line going downwards and medially at 45 degrees.Medial calf skinfold thickness was measured in the vertical direc tion on the medial side of the leg, at the level of the maximum calf girth.Biepicondylar humeral and femo ral breadth were measured using Holtain bicondylar caliper (Holtain Ltd, Croswell, UK) to the nearest 0.1 cm, between lateral and medial epicondyles of the humerus and femur, compressing the subcutaneous tissue.
Biochemical factors including plasma glucose, tri glycerides and HDLcholesterol were determined in overnight fasting blood sample.Glucose was analyzed using Dialab glucose GOD PAP method, tryglicerides were analyzed using an enzymatic method and HDL cholesterol was analysed using magnesium chloride/ phosphotungstate precipitation technique.Systolic and diastolic blood pressure were measured in the morning using a RivaRocci sphygmomanometer.Using the IDF criteria subjects were defined as having the metabolic syndrome if they had central obesity (defined as waist circumference of ≥ 80 cm) plus any two of the following four factors: blood pressure of ≥ 130/85 mmHg, glu cose of ≥ 5.6 mmol/L, high density lipoprotein choles terol (HDLcholesterol) of < 1.29 mmol/L and trigly ceride of ≥ 1.7 mmol/L (9).
According to BMI and metabolic profile the ex amined group was subdivided into four subgroups: metabolically healthy normalweght (BMI < 25 kg/m 2 and the absence of metabolic syndrome), metabolically obese normalweight (BMI < 25 kg/m 2 and the pre sence of metabolic syndrome), metabolically healthy obese (BMI ≥ 25 kg/m 2 and the absence of metabolic syndrome) and 'at risk' obese (BMI ≥ 25 kg/m 2 and the presence of metabolic syndrome).
Results are presented as mean ± standard devia tion (SD) and percent.The oneway analysis of va Somatotype, nutrition level and metabolic syndrome Arch Endocrinol Metab.2016;60/1 riance (ANOVA) with the Bonferroni posthoc me thod was used to determine whether there are any significant differences between the means of the subgroups.The discrimination abilities (accuracy) of endomorphy, mesomorphy and ectomorphy in the prediction of metabolic syndrome were assessed with the area under the receiveroperating characteristic (ROC) curve.The statistical program used for the calculations was SPSS Statistics 17.0 (IBM SPSS, Chi cago, IL).

RESULTS
Characteristics of the examined subjects are presented in the Table 1.According to the BMI values 35.65% of women were overweight or obese while 64.35% were of normalweight; 9.46% of normalweight subjects were metabolically obese, while 13.41% of obese subjects were metabolically healthy.
Somatotype analysis showed significantly higher values of endomorphy and mesomorphy and lower values of ectomorphy in the overweight and obese compared to normalweight women (Table 2).Consi dering the metabolic profile, metabolically obese sub jects had higher values of endomorphy and mesomor phy and lower values of ectomorphy compared to the metabolically healthy counterparts (Table 3).However, significant differences between metabolically healthy and metabolically obese individuals were found only in the group of normalweight women.Metabolically healthy normalweight women had significantly lower endomorphy and mesomorphy and higher ectomorphy compared to the somatotype of all the other subgroups.Somatotype of metabolically obese normalweight wo men did not differ significantly from the somatotype of 'at risk' obese women, but endomorphy of meta bolically obese normalweight women was significantly higher compared to the metabolically healthy obese women.Receiver operational characteristic (ROC) cur ve analysis revealed the maximum predictive value of endomorphy for metabolic syndrome (AUC: 0.713).The AUC of mesomorphy was 0.673 in identifying metabolic syndrome.Ectomorphy was shown as a best predictor of the favorable metabolic profile (AUC: 0.658) (Table 4, Figure 1).

Somatotype, nutrition level and metabolic syndrome
Arch Endocrinol Metab.2016;60/1 mesomorphy and lower ectomorphy more frequently suffer from arterial hypertension and liver disease.Hi gher values of endomorphy were reported in metabo lic syndrome, type 2 diabetes, hypertension and breast cancer (1013).Baltadjiev found that the diabetic indi viduals mostly present with endomorphic mesomorph or mesomorphendomorph somatotype, pointing also to age, gender, and population dependent somato type differences (14,15).Our results showed higher values of endomorphy and mesomorphy and lower values of ectomorphy in overweight subjects.Howe ver, our results reveal some differences between meta bolically healthy and metabolically obese women of the same nutrition level.In both, normalweight and obese women metabolically obese subjects had higher endo morphy and mesomorphy and lower ectomorphy than the metabolically healthy subjects.Somatotype of me tabolically healthy normalweight women was comple tely distinct showing significantly lower endomorphy and mesomorphy and higher ectomorphy comparing to the other subgroups.Somatotype of the metaboli cally obese normalweight women was similar to the somatotype of those who were overweight and obese, with even significantly higher values of endomorphy compared to the metabolically healthy obese women.The role of adipose tissue in the development of the cardiovascular and metabolic disorders has been well recognized, so the results obtained for endomorphy were not unexpected.
Metabolically obese normalweight individuals are known to demonstrate metabolic disturbances in spite of normal values of BMI.Several studies showed higher body fat (especially visceral depot) and low lean mass in the metabolically obese normalweight individuals (6,1618).According to our results, mesomorphy was higher in metabolically obese women who were both, normalweight and overweight.In order to explain ob tained results the storage capacity of adipose tissue and the muscle metabolic capacity should be considered.It is assumed that the metabolic risk is determined by the capability of subcutaneous adipose tissue to store fat.Fat is initially stored in the subcutaneous adipose tissue, but once the capacity of subcutaneous adipose tissue is reached, storage shifts to visceral depots and ectopic nonadipose sites, including skeletal muscles (19,20).Higher mesomorphy thus could be the result of the deposition of the ectopic fat in skeletal muscles that causes larger girths of extremities.Additionaly, meso morphy reflects the muscular mass but it also contains

DISCUSSION
Our study aimed at revealing somatotype differences between different metabolic subtypes of obesity.The obtained results imply the important role of the non adipose components, presented by mesomorphy and ectomorphy, in the distinction between healthy and risky metabolic profile.Somatotype describes different aspects of body composition.It is used in the assessment of the changes in physique during growth, ageing and physical activity.However, some studies showed that it could be used in the prediction of certain diseases.According to Kole va and cols., mesomorphic endomorphs tend to suffer from digestive system disorders, neurosis, or lumbosa cral radiculitis (10).The same study showed that indi viduals of both genders with higher endomorphy and some measures of peripheral fat which could explain higher mesomorphy in the metabolically obese women registered in our study.Higher ectomorphy in metabolically healthy indi viduals implies the importance of the body linearity.Several studies showed the inverse correlation between body height and cardiometabolic risk (21,22).Our previous results also showed that metabolically healthy obese women are significantly higher than the 'at risk' obese ones (23).Some authors explain this phenome non by the fetal undernutrition which causes the tis sue reprogramming in a way that determines further development of insulin resistance and atherosclerosis and changes of postnatal body composition (24).This could be in line with the embryological aspect of diffe rent components of somatotype.
In analyzing the capability of somatotype to pre dict metabolic risk, endomorphy and mesomorphy were shown as better predictors of metabolic syndrome (cutoffs were 6.89 and 5.35, respectively) while ecto morphy was the best predictor of the favorable metabo lic profile (cutoff was 0.8).
In conclusion, our results clearly show somatotype differences between metabolically healthy and meta bolically obese normalweight women.Endomorphy was the best predictor of metabolic syndrome.Higher mesomorphy in metabolically obese women appears to be controversial since it basically reflects muscularity.An important finding was the higher ectomorphy in metabolically healthy individuals which highlights the protective role of higher body linearity in the develop ment of metabolic syndrome.Despite the minor clini cal relevance of the somatotyping it could help in the explanation of the underlying mechanisms of metabo lically healthy obesity and of metabolic abnormalities in normalweight individuals.Concerning embryonic aspect of the somatotype theory our results may indica te that the susceptibility to metabolic syndrome could be determined by the prenatal environment.Finally, our results support the close relationship between body constitution and metabolic phenotype.Defining opti mal somatotype could help in differentiation between metabolically healthy and metabolically obese indivi duals of the same nutrition level.

Figure 1 .
Figure 1.Receiver operating characteristic (ROC) curve for endomorphy, mesomorphy and ectomorphy in a prediction of metabolic syndrome.

Table 1 .
Physical and metabolic characteristics of examined women

Table 2 .
Somatotype in normal-weight and overweight and obese women * Significantly different from overweight and obese subjects.

Table 3 .
Somatotype in normal-weight and overweight or obese women of different metabolic profiles * Significantly different from metabolically obese normal-weight, metabolically healthy obese and 'at risk' obese subjects; † Significantly different from metabolically healthy obese subjects.

Table 4 .
Performance of somatotype in the prediction of metabolic syndrome