We assessed the relationship between anthropometric measures with cMetS risk score components in a large, nationally representative sample of children and adolescents using the structural equation modeling. Path analysis is a powerful statistical model in evaluating a complex cluster of dependent variables. This model is more complicated and realistic than multiple regression with its single dependent variable (7). We previously in a liner regression model documented that higher anthropometric indices are associated with higher cMetS risk score in children and adolescents (18). Here, we extend our previous work by evaluating the direct and indirect effect of ZBMI and ZWC, as pivotal indices, on cMets risk score components according to sex. Our study also provided new data regarding the effect of three predictive variables including age, sedentary behaviors and SES on ZBMI and ZWC. In this study, sedentary behaviors and SES in both groups were positively correlated with ZWC but not ZBMI. Age had significant effect on ZBMI and ZWC. The direct effect of age on ZBMI was negative but on ZWC was positive in both genders. It means that in children and adolescents, by increasing age, increasing hours of sedentary behaviors and increasing level of SES, the tendency of accumulating fat centrally increases in spite of not changing or even decreased level of BMI. These results confirm the findings that WC can change independently of BMI (19). Previous studies on children and adolescents revealed that WC is a better marker of metabolic risk factors than BMI and it was shown that children with a high BMI and high WC were twice as likely to have increased levels of TG and insulin and Mets, when compared to those who had a normal WC but high BMI (19–21). Therefore, it seems that BMI by itself does not suffice for monitoring obesity and related metabolic complications in school-aged children and it is suggested paying more attention to WC as predominant underlying risk factor for development of MetS in this age group.
Another point in the current study is the relationship of higher SES with higher ZWC. We previously showed that higher SES was associated with unhealthy diet and inactive lifestyle in Iranian school children (22). During the last decades, because of a rapid nutritional transition in Iran, the trend toward high-calorie diet and sedentary lifestyle has increased (23, 24). It has been shown that in Iran similar to some developing countries, the children in families with high SES have more preference to western diets such as fast foods and they also spend a considerable part of their time with sedentary entertainments, such as watching television, playing computer games, and surfing the Internet which in turn lead to more fat accumulation and overweight in this group (25–27).
In the present study, among different components of cMetS, ZBMI and ZWC showed a significant direct effect on ZMAP and ZTG in both sexes, but the effect of ZWC was stronger than ZBMI on these variables. It should be mentioned that in our models, age, sedentary behaviors and SES impressed indirect effects on cMets risk score components through ZWC.
These results are in consistent with several population studies which documented the close correlation between central obesity and MetS components (28–30). The Bogalusa Heart Study showed that the distribution of central fat determined by WC, was associated with abnormal concentrations of TG, LDL-C, HDL-C, and insulin in children and adolescents at the ages of 5–17 yeas (31). It was shown that the association of BP with WC was stronger than those with BMI implicating that visceral fat is the primary etiological component of excess adiposity underlying the development of adiposity-related hypertension (32).
Since high fat mass during childhood and adolescence is associated with higher blood pressure and unfavorable metabolic profile, childhood nominates as an important period for interventions to manage obesity to minimize long-term metabolic abnormalities. This syndrome may developed by unhealthy diet, lack of exercise, overweight and central obesity (33, 34).
Compared to the previous studies regarding the parameters correlated with MetS in pediatrics, this study has the strength of using the path analysis in a large sample size. However, it had some limitations. First, the cross-sectional nature of study cannot demonstrate a causative relationship. Second, the information on SES and sedentary behaviors was obtained by self-report which may affect the estimates by under- or over-reporting. Third, some effective factors on MetS including dietary pattern, pubertal stage and smoking habits were not assessed in this study.