Association of sarcopenic obesity and body composition with metabolically unhealthy overweight/obese phenotypes among Iranian women: A cross-sectional study

Abstract


Results
Of the 241 subjects in this study (average age 35.32 years), 176 (73.03%) were classi ed as MUO phenotype.Based on this study, the prevalence of sarcopenic obesity was 7.88%.We found that high fatfree mass was more strongly and signi cantly associated with MUO phenotype.Furthermore, we found that individuals with high fat-free mass and high skeletal muscle mass had a signi cantly low prevalence of MUO phenotype.A signi cant positive correlation between metabolic phenotypes and sarcopenic obesity was also observed after all potential covariates were adjusted for.

Conclusion
These results of this study suggest that increased adiposity and decreased skeletal muscle mass are associated with unfavorable metabolic traits among overweight and obese Iranian women.SO was also found to be associated with a greater risk of developing MUO phenotype.

Background
Obesity has been de ned according to the World Health Organisation (WHO) as an abnormal accumulation of fat that affects health negatively (1).Obesity is a major risk factor for metabolic and cardiovascular morbidity and mortality, and its prevalence has doubled since 1980 in middle-aged and older adults (2).Individuals who are obese are at greater risk of developing the metabolic disease compared to individuals of normal weight (3), while obesity has also been related to the incidence of various cancers, respiratory disorders, arteriosclerosis, arthritis, and fatty liver (4).According to a study by Fallahzadeh and associates on estimating the prevalence and trends of obesity in the Iranian population, it was reported that 16.5% of Iranian subjects older than 18 years were obese (5).Furthermore, as part of the Jahrom Health Study (JHS), 24.8% of women aged 30 years and above were obese (6).In obese women, it has been reported that a small decline in muscle strength may cause di culties in bearing their excess weight and in moving e ciently (7).Obese subjects have been reported to have heterogeneous phenotypes with different degrees of metabolic risk (8).A sub-group of obese individuals without associated metabolic complications have been phenotypically described as metabolically healthy obese (MHO) individuals (9,10).On the other hand, is the metabolically unhealthy obese individuals characterized by obesity-related metabolic complications (11).Previous studies have reported MHO phenotype to be associated with a healthy metabolic pro le, characterized by a lower amount of visceral adipose tissue (VAT) and liver fat (12), favorable lipid pro le, high insulin sensitivity, and low proin ammatory cytokine levels in plasma (12,13) than metabolically unhealthy obese (MUO) subjects.
Growing evidence however has shown that healthy obese adults without metabolic risk factors may progress to unhealthy obesity over time (14).Schroder et al., in a 10-year follow-up of Spanish obese cohort patients, found half of their subjects with an MHO phenotype at the beginning acquire an unhealthy phenotype (15).There is not a universal criterion to determine MUO and MHO, however, some studies have shown that BMI, blood pressure, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, triglyceride, fasting blood glucose, c-reactive protein, and homeostatic model assessment of insulin resistance (HOMA-IR) can be used as a criterion to diagnose obesity phenotype (3).
Aging is a continuous process characterized by a decline in most physiological functions (16).Major changes associated with aging in body composition include a decline in both fat-free mass and muscle strength, which has been referred to as sarcopenia (17); and, an increase in body fat and a decline in skeletal muscle, referred to as sarcopenic obesity (SO) (18).Sarcopenia and obesity are both related to metabolic disorders, morbidity, and mortality (19).It has therefore been hypothesized that sarcopenic obesity may have a greater impact on metabolic diseases and cardiovascular morbidity and mortality than either sarcopenia or obesity alone (2,20).Sarcopenic obesity has been linked to increased production of in ammatory substances such as leptin, IL-6, tumor necrosis factor [TNF] α, IL-1α known to alter insulin sensitivity, growth hormone secretion, and energy metabolism (18, 21).Furthermore, Chung et al., in a study of elderly Koreans found that the sarcopenic obese group was more closely associated with insulin resistance, metabolic syndrome, and cardiovascular disease risk factors than any other group in both men and women (22).Moreover, numerous studies have also reported a signi cantly higher risk of mortality (23) worse cardiovascular risk pro les including hyperglycemia, hypertension, dyslipidemia, insulin resistance, and lower cardiorespiratory tness (22,24,25) in sarcopenic obese individuals compared to non-sarcopenic or non-obese subjects.
In a study of the Iranian adult population, Rahmanian and colleagues reported a 10.9% and a 7.2% prevalence of MUO and MHO respectively in this group (26).Rasaei and associates also reported a 9.9% prevalence of SO among Tehranian overweight and obese women (27).Despite the high prevalence of SO and MUO in this population, information regarding their association with body composition is scarce.The aim of this study therefore was to determine the association of metabolically healthy and unhealthy obesity with sarcopenic obesity and measures of body composition among overweight and obese Iranian women.

Study sample
This cross-sectional study was conducted among 241 overweight and obese women who were referred to health centers a liated to Tehran University Medical Science in Iran.Subjects were registered by use of the multistage cluster random sampling method.All participants answered a questionnaire developed and administered by well-trained investigators Inclusion criteria were age 18-48 years, being overweight or obese (body mass index (BMI) ≥ 25 kg/m 2 , and not being pregnant.Exclusion criteria included having; an acute or chronic in ammatory disease, regular use of medication, history of hypertension, cardiovascular disease, diabetes mellitus, impaired renal and liver function, intake of alcohol or drug abuse, smoking, thyroid disease, malignancies, pregnancy, and lactation.Also, individuals with chronic diseases affecting their dietary intake as well as those following a special dietary pattern over the past year were excluded from this study and whose daily energy consumption was 800 kcal/day or 4200 kcal/day were registered.All participants were informed about the study aims and procedures and provided written informed consent.The Medical Research Ethics Committee of Tehran University of Medical Sciences approved the study with the following identi cation IR.TUMS.VCR.REC.1398.692.

Measurements Dietary Assessment
A semi-quantitative food frequency questionnaire (FFQ) which contains 147-item food was applied to evaluate usual dietary intake and nutritional status over the past year.Household methods were used to transform portion sizes of the used up foods into grams (28), and a predictable average daily intake of food parameters (including macro and micronutrients) was computed from the FFQ using NUTRITIONIST IV software (version 7.0; N-Squared Computing, Salem, OR), modi ed for Iranian foods.This procedure was designed according to the Willett study including a list of 147 food items along with a standard serving size for each nutrient which its reliability and validity had been approved in Iran (29).

Measurement of blood samples
Blood samples were collected following overnight fasting and the serum was centrifuged, liquated, and stored at a temperature of −80 °C until the analysis was performed.All samples were analyzed using a single assay according to the manufacturer's protocol.The GOD PAP method and the GPO-PAP method were used for measuring the respective fasting blood sugar (FBS) and triglyceride (TG) amounts.Total cholesterol (T-chol) and cholesterol with high-density lipoprotein (HDL) were calculated using the enzymatic endpoint process and the enzymatic clearance assay.All calculations were made with the usage of a package from Randox Laboratories (Hitachi 902).An immunoturbidimetric assay (high sensitivity assay, Hitachi 902) was used to test high-sensitive C-reactive protein (hs-CRP).All measurements were taken at the Endocrinology and Metabolism Research Institute (EMRI) -Bionanotechnology Laboratory of Tehran University of Medical Science.

Body composition and anthropometric measurements assessment
Bioelectrical impedance analyzer (BIA) (InBody 720, South Korea) was used to calculate body composition indicators, including body fat-free mass and fat mass.This bioelectrical impedance analyzer (BIA) is considered to emit a very weak current to assess body impedance (resistance).Anthropometric measurements were performed on all participants, such as weight, BMI, waist circumference (WC), and waist-to-hip ratio (WHR).Weight was measured with digital scales and recorded to the nearest 0.1kg while the subjects were in minimal clothing and without shoes.Height was measured using a tape measure while the subjects were standing erect, without shoes, and with shoulders in a normal position and, head, shoulders, heels, and buttocks are in contact with the wall.BMI was calculated by dividing the weight (kg) by the square of the height (m).Waist circumference (WC) and hip circumference (HC) were recorded to the nearest 0.1 cm in the central point of the iliac crest and rib cage with tape with an accuracy of 0.1 cm.To reduce error, all measurements were taken by the same technician.All measurements were carried out by professionally trained technicians and according to speci c guidelines to reduce the difference between subjects.

Assessment of blood pressure
The blood pressure was assessed after the participants rested for 10 minutes.Hypertension was de ned as blood pressure >130/85 mmHg.This category includes patients taking antihypertensive medicines, even if treatment achieves a blood pressure level that is within the target range.

De nition of metabolic health and sarcopenic obesity
The Karelis criteria were used because all in ammation criteria are tested.The existence of four or more of the following ve is described as metabolic abnormality according to the requirements of Karelis, composition: TG 1.7 mmol/l, HDL 1.3 mmol/l, without treatment; LDL 2.6 mmol/l, without treatment, 3.0 mg/l hs-CRP and 2.7 HOMA-IR (31).According to the de nition, sarcopenic obesity in women is referred to as two lower quintiles of SMM and two highest quintiles of FM (32).In this study, we measured SMM and FM by BIA (Inbody Co., Seoul, Korea) according to guidelines.

Other baseline measurements
The socio-demographic and lifestyle characteristics assessed included age, gender, educational level attained, physical activity status, and smoking habits by questionnaire.A validated International Physical Activity Questionnaire (IPAQ) was used to assess the level of physical activity.For weekly physical activity assessment, a short form (9 items) was applied using MET (metabolic equivalent) scores for each type of activity.For reporting the total physical activity from all activity categories, MET scores across all these sub-components were summed and MET-minutes per week (MET-min/wk) were estimated (33).

Statistical Analysis
All statistical analyses were performed using the IBM SPSS software version 25.0 (SPSS, Chicago, IL, USA), and p-values less than 0.05 were considered statistically signi cant.Normal distribution of data was checked by the Kolmogorov-Smirnov test.Continuous variables information was represented by mean ± standard deviations (SDs) and mean ± standard errors (SEs), and categorical feature information represented by percentage and number.An independent sample t-test was used for assessed differences between groups with the Metabolically Healthy and Metabolically Healthy obese and chi-square (χ 2 ) tests for categorical variables.Analysis of covariance (ANCOVA) was used for adjusting potential confounders such as age, BMI, physical activity, and energy intake.Spearman bivariate correlation between body composition variables and sarcopenic obesity was performed.Association between metabolic healthy statuses with measures of body composition and sarcopenic obesity were assessed by the Binary logistic regression model, adjust for confounders in crude and 2 adjust models.Results were presented as beta and odds ratios (ORs) and 95% con dence intervals (CIs) compared with the metabolic unhealthy status group.

General characteristics of study participants
The general characteristics of the study subjects according to metabolically healthy and unhealthy phenotypes were shown in Table1.In total, 241 women aged 18-48 years, mean± SD age of 35.32±8.68years, participated in our study.Among these, 78.99% were married and 1.26%, 10.50%, 38.66%, and 49.58%, respectively, were illiterate, under diploma education level, diploma education level, and bachelor and a higher academic degree.A signi cant difference was observed among the study participants in terms of their economic status (p=0.02).After controlling for potential confounders (age, physical activity (PA), energy intake, and BMI, this signi cance disappeared.The prevalence of sarcopenic obesity was 7.88%.The aforementioned demographic characteristics and prevalence of sarcopenic obesity were marginally different across the two groups (p=0.06).However, after adjusting for potential confounders, a signi cant difference was observed between the two groups (p=0.04).Subjects in the MUO group had a signi cantly higher body weight, BMI, WC, body fat mass (BFM), and visceral fat area (VFA) compared to subjects in the metabolically healthy (MHO) group (p <0.001).Similar ndings were observed after adjusting for potential confounding variables.The FFM, fat mass index (FMI), and skeletal muscle mass (SMM) were signi cantly higher in the MUO group, p=0.004, p=0.001, and p=0.002 respectively.FBS was seen to be signi cantly higher in the MUO group after potential confounders were controlled for (p=0.03).TG, LDL, and total cholesterol were signi cantly higher (p<0.001) in the unhealthy group, while the levels of HDL were signi cantly higher (p<0.001) in the healthy group compared with the unhealthy group.After controlling for potential confounders, the results remained unchanged.A signi cant difference was also observed among the groups based on HOMA-IR (p<0.001) and insulin level (p=0.001), with the observed outcome remaining unchanged even after confounders were controlled for.While a non-signi cant difference was observed across the groups base on systolic blood pressure (SBP), a signi cantly higher diastolic blood pressure (DBP) was observed among the unhealthy group (p<0.001).(hs-CRP levels in the MUO group was signi cantly higher than in the MHO group (p<0.001).We, however, did not observe a statistically significant difference in physical activity among the two groups (p=0.33).*.Correlation is signi cant at the 0.05 level (2-tailed).

Association between obesity statuses with measures of body composition
With the binary logistic regression analysis, beta (β), odds ratio (OR), and con dence interval (CI) for the association of sarcopenic obesity with measures of body composition in crude and adjusted models are presented in Table 4.In the crude model, participants with MUO had greater odds of increasing FFM (β= 0.76, OR=0.27, 95% CI=0.13-0.56,p=0.001) than the MHO subjects.However, after controlling for age, total energy intake, and physical activity level, the MUO groups had 0.29 folds decreased odds of FFM (β=-0.89,OR=0.29, 95% CI=0.13-0.62,p=0.008) compared to the MHO group.After additionally controlling for weight and economic status, the observed association disappeared (β=-0.55,OR=0.26, 95% CI=0.11-0.59,p=0.07).In the crude model, participants with MUO were observed to be positively associated with greater odds of increasing BFM (β=1.31,OR: 3.73, 95% CI=1.27-10.91,p=0.01) than those with MHO.Also, after controlling for age, total energy intake, and physical activity level, subjects with MUO phenotype had 4.44 greater odds for BFM, compared to those with MHO phenotype (β=1.49,OR=4.44, 95% CI=1.31-15.00,p=0.01).Even after additional control for weight and economic status, the positive association of the MUO score with higher odds for the BFM remained significant (β=0.95,OR=2.60, 95% CI=1.39-17.20,p=0.02).Subjects with MUO phenotype in the crude model were observed to have a 0.87 kg lower SMM compared with the MHO subjects (β=-0.87,OR=0.50, 95% CI=0.27-0.90, p=0.006).After age, total energy intake, and physical activity had been controlled for, the MUO participants had a 0.83 lower SMM compared with the MHO subjects (β=-0.83,OR=0.46, 95% CI=0.22-0.84,p=0.01).Furthermore, after additional adjustment for weight and economic status, the association remained signi cant (β=-0.73,OR=0.47, 95% CI=0.24-0.94,p=0.03).In the crude model, a non-signi cant association was seen between metabolic healthy statuses with sarcopenic obesity scores (β=0.Based on this study, the prevalence of sarcopenic obesity was 7.88%.Our study found that high fat-free mass was more strongly and signi cantly associated with MUO phenotype.Furthermore, we found that individuals with high fatfree mass and high skeletal muscle mass had a signi cantly low prevalence of MUO phenotype.Consistently, the signi cant association between high SMM and MUO phenotype persisted even after all potential confounders (age, total energy intake, physical activity, weight, and economic status) were controlled for.Also, we found a signi cant positive correlation between metabolic phenotypes with sarcopenic obesity after all potential covariates were for.
Previous epidemiological studies have addressed the issue of fat mass, muscle mass, and sarcopenic obesity in relation to metabolic phenotypes in both Asian and Western populations.The body composition analysis of the two phenotype groups (MUO and MHO) showed that BFM, VFM, FFM, SMM, and FMI, as measured by BIA, were signi cantly higher in the MUO group than in the MHO group.
Previous studies have found that MUO women had signi cantly higher body fat mass, visceral fat mass, and muscle mass than MHO women [21] and healthy obese people [29].
Despite the fact that some studies (9,34,35) have observed a non-signi cant difference in the measures of adiposity in "healthy" and "unhealthy" phenotypes, in line with other previous studies (3,10), we observed a signi cant increase in body fat mass in the MUO phenotype even after all potential confounders were controlled for.Body fat mass has been reported as an independent predictor of insulin resistance and dyslipidemia (36) among postmenopausal women.Moreover, higher body fat has been reported as an increased risk factor for type 2 diabetes (37,38), as well as increase metabolic risk and the risk of cardiovascular disease (39,40) in both men and women.Furthermore, researches have shown increasing body fat to be correlated with an increase in fasting blood glucose, triglyceride, LDL cholesterol, and total cholesterol, and a decrease in HDL cholesterol (41,42).
There have been con icting results regarding SMM accumulation among the metabolic phenotype groups.While in some studies, no signi cant differences in SMM indices between metabolic phenotypes of either obese or non-obese postmenopausal women (35) were found, some studies have reported SMM to signi cantly increase in the MUO phenotype of postmenopausal women (11,43), and signi cantly decrease in the metabolic non-obese phenotype of young women (5).Also, Estrella et al., (44) among Hispanic/Latino women, found higher lean body mass to be independently associated with a lower prevalence of the MHO phenotype.In our study, however, each increase of 1-SD in skeletal muscle mass was associated with a lower prevalence of the MUO phenotype.The result from the fourth and fth Korean National Health and Nutrition Examination Survey indicated a protective association of muscle mass with metabolic syndrome, which was attenuated by high-fat mass (45).Additionally, Burrows et al. found in a Chilean birth cohort that low muscle mass is associated with cardiometabolic risk factors independent of dietary intake (46).Furthermore, in a 7-year retrospective cohort study, Kim and colleagues (47) found that an increase in relative skeletal muscle mass over time has a potential preventive effect on developing metabolic syndrome, independently of baseline skeletal muscle mass and glycometabolic parameters.In a nationally representative sample of 4,449 US adults aged 50 years and older from the NHANES surveys, Li and associates (48) found that only participants with low muscle mass but without metabolic syndrome had a signi cantly increased risk of all-cause mortality.Earlier studies have reported maintenance of muscle mass in the elderly as crucial, as it serves as a metabolic reservoir that is needed to effectively withstand disease (49,50).The outcome of our study supports the suggestion from a previous report about the importance of skeletal muscle mass in glucose regulation (51), and thus its protection from the development of cardiometabolic abnormalities (45).
After all potential confounders were adjusted for, each increase of 1-SD in sarcopenic obesity was found to be associated with a higher prevalence of MUO phenotype in the current study.Earlier epidemiological studies have shown that the odds of metabolic syndrome was 6 to 8 times higher in postmenopausal Korean women, elderly Korean men and women, and adult Caucasian subjects with sarcopenic obesity (SO) compared to those without sarcopenic obesity (24,(52)(53)(54).Furthermore, Kim et al., (55) reported an increased risk of metabolic syndrome in subjects with sarcopenic obesity.In their study, compared with normal subjects among women, they found that SO subjects had signi cantly higher values for a number of metabolic syndrome components.Furthermore, they found SO to be independently associated with metabolic syndrome among women after adjustment for age.

Strengths And Limitations
Our study has inherent strengths and limitations.A validated food frequency questionnaire was used to assess participants' usual dietary intake.Also, all measurements and interviews were conducted by trained personnel.
Even though this study provides valuable information regarding metabolic phenotypes with sarcopenic obesity and measures of body composition among Iranian subjects, we acknowledge limitations that need to be addressed.First, the cross-sectional nature of this study does not permit the assessment of causality.Only a prospective study would provide a better understanding of the observed associations.
Another limitation of our study is that we used BIA in order to assess body composition and not dual Xray absorptiometry (DXA) which is considered the gold standard method.However, BIA is also a validated and reliable method for the measurement of body composition in the adult population (56).Furthermore, the result of the current study cannot be generalized since only females were included in the study.

Conclusion
In conclusion, increased adiposity and decreased skeletal muscle mass was associated with unfavorable metabolic traits among overweight and obese Iranian women.SO was also found to be associated with a greater risk of developing MUO phenotype after controlling for potential confounders.

IPAQ
Abbreviations BFM body fat mass

Table 1 .
General characteristics of study participants ‡ Collinear variables did not enter into the model and this P-value obtained from ANCOVA analysis *p-value obtained from independent T-test ** p-value obtained from ANCOVA test, variable adjust for age, physical activity, energy intake, and BMI Chi-square and analysis of variance were used for qualitative and quantitative variables respectively, and the p-value was set to <0.05. a Mean ± SD b Sample size (%)

Table 2 .
Dietary intake of study subjects according to metabolic statuses MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid; SFA, Saturated fatty acid Data are presented as mean ± standard deviation (SD).aCalculated by analysis of variance and p-value<0.05indicates a signi cant level.*
contains the correlation between sarcopenic obesity with the variables of interest in this study.A simple correlation revealed a signi cant inverse associations between FFM and FFMI with correlation coe cient (r=-0.30,p=<0.001) and (r=-0.12,p=0.01) respectively with sarcopenic obesity, while signi cant positive correlations were
35, OR=1.42, 95% CI=0.45-4.45,p=0.54).However, after controlling for age, total energy intake, physical activity level, weight, and economic status, we observed MUO subjects to have 4.21 fold greater odds for sarcopenic obesity compared to the MHO subjects (β=1.43,OR=4.21, 95% CI=1.28-62.26,p=0.02).conducted on overweight and obese Iranian women who were between the ages of 18-50 years, with a BMI greater than or equal to 25 kg/m 2 .The metabolic risk was assessed according to Karelis criteria and the subjects were classi ed as either MHO or MUO.Of the 241 subjects in this study (average age 35.32 years), 176 (73.03%) were classi ed as MUO phenotype.