Association Of Birth Weight, Inflammation And Variables Of The Metabolic Syndrome In Mexican American Girls Aged 6-11 Years
birth weight, child obesity, hispanics, insulin resistance
R Gillum. Association Of Birth Weight, Inflammation And Variables Of The Metabolic Syndrome In Mexican American Girls Aged 6-11 Years. The Internet Journal of Cardiovascular Research. 2004 Volume 2 Number 2.
Low weight at birth has been shown to be a risk factor for cardiovascular disease, non-insulin dependent diabetes and the metabolic syndrome in adults (1, 2). Inflammation has also been associated with cardiovascular risk in adults (3). Few studies have examined the relationship of birth weight (BW) and insulin resistance or abdominal body fat distribution, blood pressure, serum triglycerides and HDL cholesterol, components of the metabolic syndrome, or with inflammation in representative samples of Mexican American children despite the relationship postulated for these variables with atherosclerosis and diabetes, major public health problems in Mexican American adults (1,2,3,4,5,6). In English children aged 10-11 years, fasting and post-load serum insulin levels were inversely associated with BW only after adjusting for current height and ponderal index (7). BW was not associated with central obesity or other components of the metabolic syndrome except blood pressure. In US children aged 7-12, independent associations of BW with subscapular-to-triceps ratio (SFR) and body mass index (BMI), were reported (8). Well established are associations of body circumference measures and skinfold measures of body fat distribution and fatness with insulin sensitivity, non-insulin dependent diabetes, and other cardiovascular risk factors of the metabolic syndrome (4,9, 10). Mechanisms have been advanced to explain the associations of low BW with atherosclerosis and diabetes (1, 2, 7, 8). Since adverse patterns of blood lipids and atherosclerosis itself begin in childhood, studies of population and individual differences in the early onset and progression through adolescence of possible initiating risk factors are important (9, 11,12,13,14). Tracking of blood lipids, obesity, and body fat distribution over long periods has been demonstrated (14,15,16,17).
In order to test the hypothesis that obesity, circumference and skinfold indices of body fat distribution, components of the metabolic syndrome, and inflammation are significantly associated with BW in girls independent of gender, ethnicity, or age, data from the Third National Health and Nutrition Examination Survey (NHANES III) were examined. Mexican Americans were selected for study because of the reported higher prevalence of obesity, diabetes, and insulin resistance compared to non-Hispanic whites or blacks (18, 19). The results of these analyses will show that BW was not strongly or consistently associated with inflammation or components of the metabolic syndrome in children.
Technicians measured height to the nearest 0.1 centimeter, weight to the nearest 0.01 kg, triceps, subscapular, suprailiac and mid-thigh skinfold thickness to the nearest 0.1 millimeter and waist and buttocks circumference to the nearest 0.1 centimeter as described in detail elsewhere (20, 23,24,25,26). The following were computed: waist-to-hip circumference ratio (WHR), waist-to-thigh circumference ratio (WTR), ratio of subscapular to triceps skinfold thickness (SFR), central-peripheral skinfold ratio (CPR=(subscapular skinfold + suprailiac skinfold)/(triceps skinfold + thigh skinfold); sum of the four skinfolds (SSF); body mass index (BMI=weight /height2, kg/m2), and ponderal index (PI=weight/height3, kg/m3). It has been previously reported that 14.3% (SE 2.4) of Mexican American girls aged 6-11 were overweight based on revised NCHS growth charts (23). Therefore in the present study, girls exceeding the 85th percentile of age-, sex-specific BMI were considered overweight. By analogy, girls exceeding the 85th percentile of age-, sex-specific WHR were arbitrarily considered to have elevated WHR. Blood pressure was measured using standardized methods as described elsewhere; the mean of all available readings was used in this analysis (20, 21).
Statistical analysis. Pearson correlation or partial correlation was used for assessing associations of BW with WHR, BMI, SSF and other variables in simple and stratified analyses controlling for age (27). Linear multivariate regression analysis with BW as a continuous variable was used to develop models for assessing overweight, elevated WHR, etc. as correlates of BW controlling for other variables (27). Analyses used techniques that incorporated sampling weights and design features of the survey (28). Population estimates for percentiles of BW and frequencies were produced using weighted SAS or SUDAAN procedures (29). Linear regression results of BW with other variables including statistical testing and variance estimation were performed using the PROC REGRESS procedure for linear regression models in the SUDAAN system (28,29).
Linear regression models were fit with SSF or WHR as the dependent variable and BW as the exposure variable controlling for age, with and without BW interaction terms to determine whether BW was a significant correlate of SSF or WHR and whether, in the case of WHR, BMI was a significant modifier of the effect. To allow for nonlinear effects, dummy variables for quintile of BW were entered with the second quintile as the reference category. A quadratic model was also fit with BW as a continuous variable. In unadjusted analyses, SSF in the fourth and fifth quintile of BW were significantly greater than the second (p<0.05). However, this model explained only 4% of variation in SSF. These coefficients remained significant after controlling for age (p<0.004). Low BW was not significantly associated with SSF. The quadratic model using exact BW and its square was consistent with a nonlinear association, the model minus intercept being significant (p=0.04). However, only 3% of the variation in SSF was explained by this model.
After controlling for age and BMI, BW was not a significant correlate of WHR in the model with dummy variables or in the quadratic model. No significant interaction of BW and BMI was seen.
In Mexican American girls aged 6-11 years, BW was not consistently associated with body fat distribution, obesity, or risk variables of the metabolic syndrome. Significant, independent associations were seen only for elevated SSF in the top two quintiles of BW. Further, body fat distribution or obesity were not significantly independently associated with glycated hemoglobin percent in this population. Unfortunately fasting serum insulin was not available. BW was not associated with level of the marker of inflammation CRP in girls. Despite possible mechanisms for an association with body size at birth and subsequent insulin sensitivity levels, reviewed at length elsewhere, no such associate was evident in this study (1, 2, 7, 8, 30,31,32,33).
Comparisons with previous reports. Few studies have assessed associations of BW and obesity, central or abdominal fat distribution, and insulin resistance or components of the metabolic syndrome in young children (7,8,9, 12, 32,33,34,35,36,37,38). One study of Pima Indian children aged 5-14 reported independent associations of BW with relative weight (increased with high BW), prevalence of type 2 diabetes (increased at high BW at age 5-9 and at high and low BW at 10-14 y (34). BW was negatively associated with fasting and post-load serum insulin when adjusted for current weight and height. In 4-year old Asian Indian children, plasma glucose and insulin 30 minutes post glucose load but not fasting or 120 minutes post load was inversely associated with BW (32). Similar findings were reported in 7-year-old English children (33). In English children aged 10-11 years, fasting and post-load serum insulin levels were inversely associated with BW only after adjusting for current height and ponderal index (7). BW was not associated with central obesity or other components of the metabolic syndrome except blood pressure. In US children aged 7-12, independent associations of BW with subscapular-to-triceps ratio (SFR) and body mass index (BMI), were reported (8). In US children aged 5-11, BW was negatively associated with central adiposity in white, black and Hispanic children (39). The discrepancy with the present results may relate to the earlier report's different age range, lack of weighted analysis, lack of analysis of nonlinear effects and lack of sex-specific analyses (39). In Jamaican prepubertal children, BW was inversely associated with blood pressure, but not with glycated hemoglobin or serum total cholesterol (35). However, crown-heel length was inversely associated and childhood triceps skinfold thickness and directly associated with glycated hemoglobin level. In Danish women aged 18-32, BW was positively associated with insulin sensitivity index and showed a U-shaped association with fasting insulin concentration (36). However, in regression models BW explained only 2% of the variation in insulin sensitivity index. Variables of the metabolic syndrome were not associated with BW in this study. In a matched retrospective cohort study of adolescents with intrauterine growth retardation and controls, cases had lower BMI, height, and weight, but similar SFR as controls (37).
Although WHR may not accurately reflect intra-abdominal fat mass in children (41,42,43), SFR and CPR surely reflect distribution of subcutaneous fat, which is also related to cardiovascular disease occurrence in adults (44,45,46,47,48). Blood collection conditions and laboratory methods in NHANES III were standardized. CRP, an acute phase reactant, is a marker of chronic inflammation, the test characteristics of which are discussed elsewhere (22). Fasting blood samples were not available, so results for triglycerides must be interpreted with caution. Confounding by variables not controlled for cannot be excluded. The number of children with low or high BW was limited and could explain negative findings. Because of small numbers, analyses using the categories <2500 g (low BW, n=34) and >4100 g (high BW, n=32) are not shown, but were generally consistent with results for quintiles.
BW was not strongly or consistently associated with inflammation or components of the metabolic syndrome in children. Future research should include longitudinal studies of BW, body fat distribution, obesity and fasting and post-load serum insulin in Mexican American children and adults to better characterize the development of the relationship (13, 49). Dual-energy X-ray absorptiometry or other techniques for accurate body fat measurement and assessment of regional fat distribution should be used to determine whether adipose tissue in various depots varies in its association with BW in children, adolescents or adults. BW, body fat distribution, obesity and serum insulin should be assessed jointly as risk factors for development of non-invasively measured atherosclerosis (e.g. carotid intima-medial thickness) and non-insulin dependent diabetes.
The author acknowledges the staff and contractors of the Division of Health Examination Statistics of the National Center for Health Statistics, Centers for Disease Control and Prevention, who conducted the survey and prepared the data for analysis.
R. F. R. Gillum, M.D., Centers for Disease Control and Prevention, National Center for Health Statistics, 3311 Toledo Road, 6th Floor, Hyattsville, MD 20782, USA. FAX 301-458-4036; email: email@example.com