Comparison of Various Maternal Anthropometric Indices of Obesity for Identifying Metabolic Syndrome in Offspring

Abstract

Background:

Several maternal anthropometric markers have been associated with the metabolic syndrome (MS) in offspring. The objectives of this study were (1) to determine the association between children's MS and maternal anthropometric markers such as body mass index (BMI), waist circumference (WC), WC/height, weight/sitting height squared, and WC/sitting height and (2) to compare the abilities of these five indices to identify children with MS.

Methods:

Data were collected cross-sectionally from five elementary schools between April 2007 and March 2008. BMI, WC, WC/height, weight/sitting height squared, and WC/sitting height were acquired in mothers and their children. Tanner stage, blood pressure, glucose, lipids, and insulin were measured in children. Criteria analogous to Adult Treatment Panel III for MS were used for children.

Results:

Of 624 children (307 boys) 8.96 ± 1.86 years old, with their mothers being 36.25 ± 7.14 years old, examined, 107 (17.1%) of children were obese (BMI >95^th percentile per Centers for Disease Control and Prevention norms), and 95 (15.2%) were overweight (OW) (85^th percentile ≤BMI < 95^th percentile). Of the mothers, 109 (30.4%) were obese (BMI > 30 kg/m²), and 206 (33.0%) were OW (25 kg/m² < BMI < 30 kg/m²). Approximately 68% of the children were prepubertal. The prevalence of MS was 3.5% overall: 6.7% in OW and 13.9% in obese children. To determine which marker was a better predictor for MS, a receiver operating characteristics (ROC) curve was generated for the five maternal anthropometric measures, with children's MS as the dichotomous variable. The areas under the ROC curves were 0.697 ± 0.07 for BMI, 0.698 ± 0.07 for WC, 0.717 ± 0.07 for WC/height, 0.725 ± 0.07 for WC/sitting height, and 0.704 ± 0.07 for weight/sitting height squared. There was no significant difference between the areas of the five maternal anthropometric markers as predictors of children's MS.

Conclusions:

Measurement of maternal sitting height had no advantages over total height in the prediction of children's MS. All maternal anthropometric measures identified the MS in their children consistent with known familial associations of obesity and type 2 diabetes.

Background

Obesity (OB) has become a major public health problem in developed and developing countries.¹ It is closely associated with diabetes mellitus and cardiovascular diseases.¹ Reports of American children and adolescents (2–19 years old) showed that approximately 32% of the children were at or above the 85^th percentile for body mass index (BMI) according to age and gender.² Argentina, as well, has experienced a marked increase in the prevalence in childhood of overweight (OW) (85^th percentile ≤BMI < 95^th percentile)/OB (BMI ≥ 95^th percentile) over the last few decades with a prevalence very similar to that in the United States.^3,4 High levels of BMI among children are associated with hyperlipidemia, insulin resistance, and hypertension, all components of the metabolic syndrome (MS).^5,6

Parents are in a key position to shape the environments of children, and there is increasing interest in the contribution of parenting behaviors to OB risk.⁷ Parental OB is a strong predictor of offspring OB.^7,8 The common forms of childhood OB seem to result from a predisposition that primarily favors obesogenic behaviors in an enabling environment.⁸

Maternal waist circumference (WC) has been associated with the MS in offspring.³ Other maternal OB indicators such as BMI, WC, and WC/height may serve as predictors of children's MS. Anthropometric measures may need to be adjusted for sitting height as indicators of health risks in special populations.⁹ This study used sitting height as a component of various anthropometric measures and compared these with total height measures to predict children's MS.

The objectives of this study were (1) to determine the association between children's MS and maternal anthropometric markers such as BMI, WC, WC/height, weight/sitting height squared, and WC/sitting height and (2) to compare the abilities of these five maternal indices to identify children with MS.

Subjects and Methods

Study design and participants

Originally 649 children and their mothers were recruited from six elementary schools; however, one child was taking a thyroid medication, eight declined to participate, and BMI information was missing for 16. Therefore, 624 children (307 boys) and their mothers were examined between April 2007 and March 2008. The schools were randomly selected from two poor neighborhoods of the suburbs of Buenos Aires. We calculated the sample size based on the prevalence of OW/OB in children reported in other elementary school studies in Buenos Aires. Given the fact that the prevalence of OW/OB was approximately 33% among children in a similar population,³ the sample size was estimated to achieve that percentage with an error lower than 0.035. The sample size resulting with this error was a figure close to the sample size used.

Although Argentina is a Spanish-speaking country, the population differs greatly from what is usually referred to as Hispanic in the United States. About 85% of the population is of European descent (largely Spanish and Italian), with the remainder of mixed European and Native American (12%) or Native American (3%) descent.¹⁰

Socioeconomic characteristics included age, parental level of education, and the presence or absence of a refrigerator or a dirt floor. Mothers were asked to define their level of formal education as having no formal education, only the first 3 years of elementary school, completed elementary school, completed high school, or a university degree.

Height and weight were measured in the mothers and their children with the subjects wearing light clothing and without shoes. Weight was measured to the nearest 0.1 kg on a medical balance scale. Height was measured to the nearest 0.1 cm with a calibrated wall-mounted stadiometer. Sitting height was measured as the distance from the highest point of the head to the sitting surface.⁹

Exclusion criteria included the following: missing information to calculate BMI, the informed consent not being signed, self-reported pregnancy at the time of the examination, not being in the fasting state for at least 10 h, known diabetes or other chronic diseases, and the use of medication that alters blood pressure or glucose or lipid metabolism. All subjects were examined by the same physician. The study was approved by the Human Rights Committee of Durand Hospital in Buenos Aires. Each parent gave written informed consent, and children gave assent after an explanation of the study and before its initiation.

Children's anthropometric measures, stage of puberty, and blood pressure

Because children's BMI varies according to age and gender we standardized the values for age and sex by converting them to a z-score using the LMS method according to the Centers for Disease Control and Prevention (CDC) growth charts.¹¹ Children were classified as normal weight, OW, or obese by CDC norms. WC was taken at the level of the umbilicus and recorded to 0.1 cm.¹² A nonelastic flexible tape measure was used with the subject standing without clothing covering the waist area. Because WC also varies according to age and gender we standardized values for age and sex by converting them to z-scores. Central OB was defined for children as WC ≥90^th percentile based on 3,000 normal Argentinian children (authors' unpublished data).

Children's puberty was staged according to the criteria of Tanner¹³ from 1, representing immaturity, to 5, for full maturity.¹⁴ Tanner staging was performed in 589 children because 26 children refused to be examined.

Three separate blood pressure measurements were recorded by a trained technician using a random-zero sphygmomanometer after the participant was seated at rest for 5 min. The averages of the last two measurements of systolic and diastolic blood pressures were used. Because normal pediatric blood pressure varies according to age, gender, and height, we used the National Heart, Lung and Blood Institute's recommended cutoff point for these variables.¹⁵

Maternal anthropometric measures

Simple maternal OB indices such as BMI, WC, WC/height, weight/sitting height squared, and WC/sitting height were tested as predictors of children's MS. The World Health Organization recommends measurement of the BMI as a universal criterion of adult OW (25 kg/m² ≤ BMI < 30 kg/m²) and OB (≥30 kg/m²).¹⁶ Measures of abdominal fat distribution such as WC and WC/height were used as indices of central adiposity. The definition of central OB was WC >88 cm.¹⁷ WC has been criticized for not taking into account differences in body height, and the WC/height has been proposed as a better predictor of cardiovascular risk.¹⁸ A cutoff level of 0.5 for WC/height has been recommended for different populations.^19,20 Sitting height was measured to the last completed millimeter using an anthropometer placed against the edge of a chair, and the participant sat on the chair as tall and as straight as possible with his or her back against the vertical stand.⁹ Sitting height was compared with total height to predict MS. Weight/sitting height squared and WC/sitting height were determined as new anthropometric measures.

Biochemistry

Baseline blood samples were obtained from subjects while they were fasting, for measurement of fasting levels of glucose, insulin, and lipids. Plasma glucose was assayed by the glucose oxidase technique, and serum lipids were measured with a Modular P analyzer (Hitachi High Technologies Corp., Tokyo, Japan). Serum insulin levels were determined by radioimmunoassay (Diagnostic Products, Los Angeles, CA), and insulin did not cross-react with proinsulin or C-peptide (percentage coefficient of variance, 5.2–6.8%).

Children's MS

Varying definitions for MS have been applied to children until the International Diabetes Federation recently developed a simple unified definition for children over 10 years of age.²¹ The mean age of the children in this study was approximately 9 years, and as the International Diabetes Federation's definition was proposed only for children over 10 years, we used the definition of Cook et al.²² The latter definition included the presence of three or more of the following five conditions: (1) central OB (WC ≥90^th percentile), (2) fasting triglycerides ≥110 mg/dL, (3) high-density lipoprotein cholesterol (HDL-C) ≤40 mg/dL, (4) blood pressure ≥90^th percentile for age, gender, and height, and (5) fasting glucose >100 mg/dL. The American Diabetes Association changed the glucose categorization for impaired fasting glucose from ≥110 mg/dL to ≥100 mg/dL.²³ Following this new criterion, impaired fasting glucose was defined as fasting glucose concentrations >100 mg/dL. Glucose concentrations ≥126 mg/dL were indicative of overt diabetes.

Insulin resistance

Homeostasis model assessment of insulin resistance (HOMA-IR) was performed. HOMA-IR was validated in children and adolescents and was strongly correlated with insulin resistance.²⁴ The following equation for HOMA-IR was used: fasting insulin (in μU/L) × fasting glucose (in mmol/L)/22.5.^25,26

Data analysis

The χ ² test was used to compare proportions. When more than 20% of the cells had expected frequencies <5, Fisher's exact test was used. The fit to normal distribution of continuous variables was assessed using the Shapiro-Wilks test. When comparing two groups with normally distributed data, Student's t test was performed. When comparing three or more groups and with data that were normally distributed with homogeneity of variances, one-way analysis of variance was used (Student-Newman-Keuls post hoc test). Homogeneity of variances was tested with the Levene test. When this test was rejected the Brown-Forsythe test was performed with Dunnett's post hoc test.

Separate multiple linear regression models were used to determine the association between all the maternal anthropometric measures (z-scores) explored as continuous variables and those of their children adjusted for gender and Tanner stage. Separate multiple logistic regression analyses were done to examine the relationship between children's MS and cardiovascular risk factors as dependent variables and the five maternal anthropometric indices (z-scores), gender, and Tanner stage as the independent variables.

The areas under the receiver operator characteristics (ROC) curves were calculated for maternal BMI, WC, WC/height, weight/sitting height squared, and WC/sitting height, and the difference between these areas was used to determine which marker was the best predictor for children's MS. Comparisons of areas were performed using the nonparametric De Long test. The optimal threshold was determined representing the point on the ROC curve that optimizes specificity and sensitivity. Children's MS was used as the dichotomous variable.

Values of P < 0.05 were considered significant. Data are presented as mean ± SD values. Analyses were done using the SPSS^® (Chicago, IL) statistical software package SPSS version 11.0.

Results

The study population included 624 children (307 boys) 8.96 ± 1.86 years old with 624 mothers 35.19 ± 7.05 years old. All families were in the low socioeconomic class: 7% of the families did not have a refrigerator, 3.1% had a dirt floor, and 1.2% had a dirt floor and no refrigerator. The educational backgrounds of the mothers were as follows: 3.9% (24) had no formal education, 21.5% (132) had only the first 3 years of elementary school, 45.1% (277) (132) had completed elementary school, 21.5% (132) had completed high school, and 8.0% (49) had university and/or advanced degrees. The educational background of six mothers was missing. The social worker's report showed that all of the children in these neighborhoods attended elementary school.

Characteristics of children

Mean levels of various characteristics are presented according to gender (Table 1). Ninety-five (15.2%) of the children were OW, 107 (17.1%) were obese, and 428 (67.6%) were normal. The mean z-score for BMI of these three groups was as follows: OB, 2.02 ± 0.33; OW, 1.31 ± 0.16; and normal, 0.00 ± 0.81. Sixty-seven percent (n = 396), 20.4% (n = 127), 9.0% (n = 56), and 1.6% (n = 10) were at Tanner stage 1, 2, 3, and 4, respectively.

Table 1.

Clinical and Metabolic Characteristics of Children According to Gender

	Boys (n = 307)	Girls (n = 317)
Percentage	49.2%	50.8%
Age (years)	8.88 ± 1.77	9.04 ± 1.93
BMI (kg/m²)	18.32 ± 3.56	18.46 ± 3.60
z-score for BMI (CDC)	0.57 ± 1.08	0.50 ± 1.04
OW/OB	33.8% (28.5–39.1%)	30.9% (25.8–36.0%)
WC (cm)	63.02 ± 9.97	63.49 ± 10.47
Blood pressure (mm Hg)
Systolic	91.83 ± 12.18	90.81 ± 13.40
Diastolic	56.28 ± 8.62	55.44 ± 9.00
Tanner stage 1^**	84.7% (80.8–88.8%)	50.5% (44.8–56.1%)
Cholesterol (mg/dL)	152.62 ± 26.96	154.96 ± 28.03
Triglycerides (mg/dL)^**	67.88 ± 33.54	78.60 ± 35.64
Glucose (mg/dL)	76.25 ± 9.24	75.11 ± 8.43
HDL-C (mg/dL)^*	52.90 ± 12.24	50.13 ± 11.74
Insulin (IU/mL)^*	3.56 ± 3.13	4.52 ± 4.25
HOMA-IR^*	0.68 ± 0.65	0.85 ± 0.86

Data are mean ± SD values, and observed prevalence is in percentage (95% confidence interval).

P values compare levels between boys and girls: *P < 0.05, **P < 0.001.

Girls compared to boys had higher mean triglyceride, HOMA-IR, and insulin levels and lower HDL-C and glucose levels (Table 1). There was no significant difference in BMI, z-score for BMI, WC, systolic blood pressure, and the prevalence of OW and OB between genders. The prevalence of Tanner stage 1 (prepubertal) was significantly lower in girls than in boys, as expected with earlier pubertal maturation in girls (Table 1).

Children were divided into three groups according to the presence of OW or OB (Table 2). Approximately 40% of children had at least one risk factor for cardiovascular disease, and 15% had two or more risk factors, and the prevalence was higher (P < 0.001) in OW/OB than in normal weight. The risk factors of low HDL-C (18%; 14.6–21.7%) and central OB (27%; 23.4–30.5%) were common, whereas hypertension (5.6%; 3.8–7.4%) and impaired fasting glucose (0.9%; 0.02–1.73%) were infrequent. None had diabetes. MS was present in 3.5% (1.8–5.20%) overall, 0.3% with normal weight, 6.7% with OW, and 13.9% with OB (P < 0.01). None had all five risk factors. Mean values of z-score for BMI and HOMA-IR increased as the number of risk factors increased (P < 0.01). There was no significant difference in the prevalence of MS between boys and girls (P = 0.34). There was a significantly higher prevalence of MS in Tanner stages 2 and 3 than in Tanner stage 1 (P = 0.01).

Table 2.

Clinical and Metabolic Characteristics According to the Presence of OW/OB

	Normal weight	OW	OB	Significance
Number	422	95	107
Age (years)	8.3 ± 1.9	8.8 ± 2.0	8.4 ± 1.6	NS
BMI (kg/m²)^*	16.5 ± 1.7	20.3 ± 2.0	24.1 ± 3.1	^a
z-score BMI (CDC)^*	0.0 ± 0.8	1.3 ± 0.2	2.0 ± 0.3	^a
WC (cm)^*	58.7 ± 6.5	68.5 ± 8.6	76.4 ± 10.2	^a
Blood pressure (mm Hg)
Systolic^*	89.1 ± 11.6	94.1 ± 14.0	97.6 ± 13.7	^b
Diastolic^*	54.4 ± 8.3	57.5 ± 9.1	59.9 ± 9.2	^b
Cholesterol (mg/dL)	153.6 ± 28.9	151.4 ± 22.3	156.9 ± 26.2	NS
Triglycerides (mg/dL)^*	67.1 ± 28.1	77.5 ± 29.5	94.8 ± 53.5	^b
Glucose (mg/dL)	75.1 ± 9.2	76.6 ± 8.1	77.1 ± 7.9	NS
HDL-C (mg/dL)^*	53.2 ± 12.2	47.7 ± 9.7	48.3 ± 12.1	^b
Insulin (IU/mL)^*	3.2 ± 2.7	5.4 ± 5.4	7.0 ± 4.9	^b
HOMA-IR^*	0.6 ± 0.6	1.0 ± 1.1	1.4 ± 1.0	^b

Data are mean ± SD values. P values compare mean levels among normal weight, OW, and obese children. The level of significance was *P < 0.001. NS, no significant differences found between any of the groups.

Significance found between each group.

Significance found in comparison of normal weight to OW and OB.

Characteristics of the mothers

One hundred ninety (30.4%) of the mothers were obese, 206 (33.0%) were OW, and 228 (36.5%) were normal weight. The average maternal BMI was 28.2 kg/m², which indicates that, as a group, they were OW. The average maternal WC was 90.8 cm, and WC/height was 0.57, which indicates that, as a group, they had central OB. The mean BMI of these three groups was as follows: OB, 36.07 ± 5.20 kg/m²; OW, 27.42 ± 1.44 kg/m²; and normal, 22.47 ± 1.75 kg/m². The prevalence of mothers with central OB was 56.5% (52.4–60.4%) and of WC/height >0.5 was 81.3% (78.2–84.4%). There was no significant difference in the prevalence of OW/OB in mothers with a lower level of education versus mothers with a higher level (divided into high school or less and more).

Relationship between children and their mothers

Prevalence of children's MS and components of the MS according to maternal OW/OB

There was a lower prevalence of OW/obese children in the group of normal weight mothers than in the group of OW/obese mothers (20.2% vs. 39.4%; P < 0.01). There was a lower prevalence of children's MS in the group of normal weight mothers than in the group of OW/obese mothers (1.4% vs. 8.5%; P = 0.004). There was a lower prevalence of children's MS in the group of mothers without central OB than in the group of mothers with central OB (0.6% vs. 5.5%; P = 0.007). Children of mothers with normal weight had mean values of various characteristics that were significantly lower than those for children with OW/OB mothers: BMI (17.39 vs. 18.97 kg/m²; P < 0.01), WC (60.53 vs. 64.83 cm; P < 0.01), systolic blood pressure (89.91 vs. 92.11 mm Hg; P = 0.031), and insulin (3.50 vs. 4.30 IU/mL; P = 0.046).

Associations between mother and child anthropometric measures

All children's anthropometric measures explored as continuous variables showed a significant linear association with those of their mothers, adjusted for gender and Tanner stage (Table 3). Separate multiple linear regression models showed that maternal BMI and Tanner stage were associated with children's BMI. Maternal WC and Tanner stage were associated with children's WC. Maternal WC/height and Tanner stage were associated with children's WC/height. Maternal WC/sitting height and Tanner stage were associated with children's WC/sitting height. Maternal weight/sitting height squared and Tanner stage were associated with children's weight/sitting height squared.

Table 3.

Associations Between Mother–Child Anthropometric Measures by Separate Multiple Linear Regression Analysis

Children's anthropometrics measure	Maternal anthropometrics measures and Tanner stage	Beta standardized coefficients	P
BMI	BMI	0.31	< 0.01
	Tanner stage	0.35	< 0.01
WC	WC	0.30	< 0.01
	Tanner stage	0.46	< 0.01
WC/height	WC/height	0.35	< 0.01
	Tanner stage	0.05	< 0.01
WC/sitting height	WC/sitting height	0.38	< 0.01
	Tanner stage	0.07	0.03
WC/sitting height squared	Weight/sitting height squared	0.28	< 0.01
	Tanner stage	0.38	0.03

As gender was not significantly associated, it was not included here.

Children's anthropometric measures were used as dependent variables, and maternal anthropometric measures and Tanner stage and gender were entered as independent variables. Gender was excluded from Table 3 because it was not significantly associated in any analysis.

Associations between children's MS and cardiovascular risk and maternal anthropometric measures

In separate multiple logistic regression models, BMI, WC, WC/height WC/sitting height, and weight/sitting height squared were associated with children's MS adjusted for Tanner stage and sex (Table 4). WC and WC/height were the most strongly associated with children's MS. Furthermore, children had 88% higher odds of having MS as the mother's WC/height increased 1 SD and had 81% higher odds of presenting MS as the mother's WC increased 1 SD.

Table 4.

Separate Multiple Logistic Regression Analysis of Children's MS

	Children's MS		Children's central OB		Children's hypertension
Dependent variable	OR	95% CI	OR	95% CI	OR	95% CI
Maternal BMI	1.77	1.18–2.64	2.15	1.73–2.66	1.52	1.21–2.07
Maternal WC	1.81	1.18–2.78	1.99	1.62–2.45	1.58	1.14–2.18
Maternal WC/height	1.88	1.21–2.92	1.88	1.21–2.92	1.55	1.11–2.15
Maternal WC/sitting height	0.79	1.16–2.78	2.09	1.67–2.62	1.64	1.19–2.27
Weight/sitting height squared	1.67	113–2.46	2.13	1.71–2.66	1.59	1.18–2.14

Central OB and hypertension were used as dependent variables, and children's Tanner stage and sex and maternal anthropometric measures were entered as independent variables. CI, confidence interval; OR, odds ratio.

In separate multiple logistic regression models, BMI, WC, WC/height, WC/sitting height, and weight/sitting height squared were associated with children's central OB and children's hypertension, adjusted for Tanner stage and sex.³ However, children's low HDL-C, hyperglycemia, and high triglyceride levels were not associated with any of the maternal anthropometric measures.

ROC curves

A comparison of the ability of maternal anthropometric measures to correctly identify children with risk factors for MS is presented in Table 5. The area under the curve (AUC) is a measure of the degree of separation between affected and nonaffected subjects by a specific test. An AUC of 1 indicates perfect separation between affected and nonaffected subjects, and an AUC of 0.5 indicates no discriminative value of the test used.

Table 5.

AUCs of Maternal Anthropometric Measures and Children's MS, Central OB, Hypertension, Dyslipidemia, and Insulin Resistance

	MS	Central OB	High BP	High TG	Low HDL-C	>III Q HOMA
Maternal WC
AUC	0.70	0.68	0.63	0.61	0.54	0.55
Lower	0.56	0.63	0.54	0.52	0.47	0.46
Upper	0.83	0.73	0.72	0.71	0.62	0.64
Maternal WC/height
AUC	0.72	0.67	0.62	0.60	0.55	0.54
Lower	0.59	0.62	0.52	0.50	0.47	0.45
Upper	0.84	0.72	0.71	0.70	0.63	0.63
Maternal BMI
AUC	0.70	0.67	0.63	0.60	0.50	0.58
Lower	0.57	0.62	0.54	0.50	0.43	0.49
Upper	0.82	0.73	0.73	0.70	0.58	0.66
Maternal WC/sitting height
AUC	0.72	0.67	0.61	0.60	0.56	0.55
Lower	0.61	0.62	0.52	0.50	0.48	0.46
Upper	0.84	0.73	0.71	0.70	0.63	0.64
Weight/sitting height squared
AUC	0.70	0.67	0.63	0.60	0.52	0.60
Lower	0.58	0.62	0.53	0.51	0.44	0.51
Upper	0.83	0.73	0.72	0.70	0.59	0.68

AUC was estimated by ROC analysis. BP, blood pressure; TG, triglycerides. >III Q HOMA-IR, >III quartile HOMA-IR.

Regarding children's MS, the AUCs for all the maternal anthropometric measures were not significantly different. Still, the AUCs, the specificities, and the sensitivities varied across maternal anthropometric measures; maternal WC/height and WC/sitting height seemed to be better predictors.

Regarding children's central OB, hypertension, and hypertriglyceridemia, the AUCs for all the maternal anthropometric measures were not significantly different; however, maternal WC had the larger area. Regarding children's low HDL-C and insulin resistance, the AUCs for all the maternal anthropometric measures were very small, and none was significantly different from 0.5; therefore none of the maternal anthropometric measures could discriminate children with low HDL-C or insulin resistance.

Optimal anthropometric thresholds

Optimal maternal BMI, WC, WC/height, weight/sitting height squared, and WC/sitting height cutoff points for predicting children's MS, central OB, hypertriglyceridemia, and hypertension are shown in Table 6. Thresholds for optimal sensitivity and specificity for maternal WC ranged from 93 to 89 cm, for WC/height from 0.58 to 0.60, for BMI from 27 to 30 kg/m², for weight/sitting height squared from 92.5 to 99.0, and for WC/sitting height from 1.06 to 1.13.

Table 6.

Optimal Maternal Anthropometric Measures

	MS	Central OB	High BP	High TG
Maternal WC
Cutoff	93	89	89	89
Sensitivity	0.71	0.66	0.67	0.56
Specificity	0.60	0.60	0.50	0.52
Maternal WC/height
Cutoff	0.59	0.57	0.58	0.58
Sensitivity	0.71	0.69	0.61	0.59
Specificity	0.62	0.60	0.57	0.55
Maternal BMI
Cutoff	30.14	27.06	27.72	26.63
Sensitivity	0.71	0.71	0.58	0.56
Specificity	0.70	0.59	0.57	0.52
Maternal WC/sitting height
Cutoff	1.13	1.07	1.09	1.06
Sensitivity	0.71	0.67	0.61	0.59
Specificity	0.67	0.60	0.58	0.54
Weight/sitting height squared
Cutoff	98.90	94.65	97.46	92.51
Sensitivity	0.71	0.67	0.58	0.56
Specificity	0.59	0.60	0.58	0.52

BP, blood pressure; TG, triglycerides.

Discussion

We have found little difference in the abilities of the five maternal anthropometric measures to identify schoolchildren with MS. In general, WC/height ratios showed slightly stronger associations with children's MS, whereas maternal WC showed slightly stronger associations with measures of children's central OB, hypertension, and hypertriglyceridemia. These slight differences among the five indices, AUCs, specificities, and sensitivities were similar for each risk factor.

The association between mothers' OB and morbidity and their children's OB justifies looking for various markers of adiposity in all mothers. Parental OB is a strong predictor of OW and OB in children.^27,28 The familial risk ratio for childhood OB when a parent is obese reaches >2.5.⁸ A recent study showed that parental leanness confers significant protection against development of OW in children, whereas parental OB is associated with a prevalence of OW in their children more than double that for children of lean parents.²⁹ Excessive BMI gains of parents during childhood and adulthood were associated with a higher BMI and risk of OB in the offspring.³⁰ Consistent with these studies the results from the linear regression analysis showed that the five maternal anthropometric measures studied were associated with their offspring's anthropometric measures.

Approximately 40% of children had at least one metabolic abnormality, and 3.5% of the children had MS. This could be due to the high prevalence of OW and OB in mothers. Approximately 63% of the mothers were OW/obese. There was a high prevalence of maternal central OB as measured by WC (56%) and by WC/height (81%). However, the WC/height cutoff of 0.5 may have generated a higher prevalence of central OB than that using WC >88 cm as evidenced in our study.

Simple maternal OB indices were used as predictors of cardiovascular disease risk factors and children's MS. BMI¹⁶ has become a universal index of nutritional status. This study found that children had approximately twice the odds of having central OB as the mother's BMI z-scores increased 1 unit. However, BMI is influenced by variation in height caused by variation in the length of the legs. Such variations occur between and within populations.⁹ In the British population, with a sitting height/stature ratio at the mode of population values, adjustments of BMI for sitting height/stature ratio caused BMI to change by ≥1 kg/m² in 33% of women and 10% of men.³¹ However, measurements of maternal sitting height had no advantages over total height in the prediction of children's MS or cardiovascular risk factors in the present study.

Although WC and BMI are highly correlated, they measure different aspects of body fatness. BMI tends to indicate overall fatness but not body fat distribution, whereas WC assesses abdominal adiposity.³² WC was adopted as a measure of OB by The National Cholesterol Education Program.³³ A previous study we conducted in a population similar to that in the present study showed that the mother's WC was a significant independent predictor for her child's MS.³ Consistent with the current study we showed that maternal WC had higher AUCs, specificities, and sensitivities for children's central OB, hypertension, and hypertriglyceridemia, although it did not prove to be the best predictor for children's MS. The multiple logistic regression analysis demonstrated that WC was strongly associated with children's hypertension, whereas BMI best predicted children's central OB. WC/height was first used in the Framingham Study,³⁴ and different studies concluded that it was more strongly associated with cardiovascular risk factors than WC alone.^18,35
–37 This study has shown that maternal WC/height has slightly higher AUCs, specificities, and sensitivities for children's MS. When calculating odds ratios that allow adjusting for further influencing factors, WC/height was still a slightly better predictor for children's MS. There was not a significant difference among the AUCs of the five anthropometric maternal measures as predictors for MS. However, AUCs showed that measures of WC and WC/height had a better performance as predictors for MS and cardiovascular risk factors than BMI, possibly because visceral fat is a greater risk factor for type 2 diabetes.³⁸

The cutoff level for optimal sensitivity and specificity for maternal BMI in predicting children's MS was 30, which is in line with the criterion for maternal OB,¹⁶ whereas the criteria for central OB (WC = 93 cm and WC/height = 0.60) were higher than those published previously.^17,18 Criteria for optimal sensitivity and specificity for maternal BMI in predicting children's central OB, hypertension, and hypertriglyceridemia were from 27 to 28, which match the definition for maternal OW¹⁶; the criteria for central OB, however, were higher than those published.^17,18 However, the influence of these risk factors on the risk of developing type 2 diabetes is a continuous variable, and thus any criterion would be arbitrary.³⁹

A previous study showed a significant increase in odds ratios for OW, abdominal OB, and high triglycerides among offspring of a parent with MS,⁷ whereas high blood pressure, high blood glucose, and low HDL-C were not significant. Consistent with this study, we found that maternal anthropometric measures were associated with central OB, hypertriglyceridemia, and hypertension, whereas low HDL-C and hyperglycemia were not. These findings suggest that the phenotype of OW or central OB appears in children before the appearance of type 2 diabetes or low HDL-C in adulthood. These results indicate that biomarkers for an increased risk of cardiovascular disease are already present in these children. The risk of the MS in children was highly correlated with maternal status of OW/OB. Furthermore, the MS has familial components and suggests the influence of genes and shared environment in the etiology of the MS.

Family is one of the most important components affecting metabolic risk factors in children, in that family displays an interaction between genetic and shared environmental factors.^7,40 Gene–environment interaction studies suggest that the effects of predisposing genes can be enhanced or diminished by exposure to relevant behaviors.⁸ Therefore, dietary behaviors such as high energy intake, consumption of sugar-sweetened beverages,⁴¹ skipping breakfast,⁴² and decreased milk consumption⁴³ could be associated with childhood OB. Several studies have demonstrated that children who watched TV extensively were from families in which parents did the same.^44,45 Furthermore, studies of physical activity have demonstrated that parental support of their children's participation in physical activity produces active children.⁴⁶ Thus, it is reasonable to conclude that the influence or support of parents may promote their children's involvement in a healthy lifestyle and that this could influence their children's BMI and cardiovascular risks.

Conclusions from this report are limited by the fact that it is of a cross-sectional study. Our data only show the association with present risk factor conditions but do not directly predict the future risk of cardiovascular events. To elucidate which anthropometric measure is the predictor of future cardiovascular events, prospective studies are necessary. Second, there is a lack of consensus regarding specific criteria and cutoff values for cardiovascular risk factors and for MS in children. Despite these limitations, our findings of higher levels of cardiovascular risk factors in a large group of schoolchildren are important and timely observations. Third, associations between mother and child anthropometric measures, while significant, were low; therefore the significance could have been a factor of sample size and not of the association. Fourth, other possible confounding factors such as diet, lifestyle, physical activities, and family history were not addressed in the analysis. However, this should not affect the conclusions of the maternal anthropometric measures as predictors of their children's MS. Fifth, the reason why a higher number of schools were not included was due to economic limitations. Lastly, because the sample was drawn from Buenos Aires neighborhoods of a low socioeconomic level, it cannot be seen to represent Argentina as a whole. However, it can be seen as representative of a large portion of the Buenos Aires suburbs.

Conclusions

All maternal anthropometric measurements are simple to use and are obtained in common clinical practice. Measurement of maternal sitting height had no advantages over total height in the prediction of children's MS. This study suggested that even though WC and WC/height tended to show closer associations with children's MS and cardiovascular risk than measures of BMI, the differences were small, suggesting an equal value of the methods. Therefore, any of the five maternal anthropometric markers could identify the MS in their children consistent with known familial associations of OB and type 2 diabetes. As familial transmission is a risk factor for OB, it is important to evaluate familial factors to identify strategies for the prevention and management of the MS.

Footnotes

Acknowledgments

We thank Dr. Arlan Rosenbloom for help in editing this article.

Author Disclosure Statement

No competing financial interests exist.

References

Yusuf

, Hawken

, Ôunpuu

, Bautista

, Franzosi

, Commerford

, Lang

, Rumboldt

, Onen

, Lisheng

, Tanomsup

, Wangai

Jr , Razak

, Sharma

, Anand

. INTERHEART Study Investigators. Obesity and the risk of myocardial infarction in 27000 participants from 52 countries: a case-control study. Lancet, 2005; 366:1640–1649.

Ogden

, Carroll

, Flegal

. High body mass index for age among US children and adolescents, 2003–2006. JAMA, 2008; 299:2401–2405.

Hirschler

, Roque

, Calcagno

, Gonzalez

, Aranda

. Maternal waist circumference and the prediction of children's metabolic syndrome. Arch Pediatr Adolesc Med, 2007; 161:1205–1210.

Hirschler

, Maccallini

, Karam

, Gonzalez

, Aranda

. Are girls more insulin-resistant than boys? Clin Biochem, 2009; 42:1051–1056.

Freedman

, Dietz

, Srinivasan

, Berenson

. The relation of overweight to cardiovascular risk factors among children and adolescents: the Bogalusa Heart Study. Pediatrics, 1999; 103:1175–1182.

Reilly

, Methven

, McDowell

, Hacking

, Alexander

, Stewart

, Kelnar

. Health consequences of obesity. Arch Dis Child, 2003; 88:748–752.

Park

, Park

, Cho

. Familial aggregation of the metabolic syndrome in Korean families with adolescents. Atherosclerosis, 2006; 186:215–221.

Bouchard

. Childhood obesity: are genetic differences involved? Am J Clin Nutr, 2009; 89:1494S–1501S.

Charbonneau-Roberts

, Saudny-Unterberger

, Kuhnlein

, Egeland

. Body mass index may overestimate the prevalence of overweight and obesity among the Inuit. Int J Circumpolar Health, 2005; 64:107–109.

10.

Composición étnica de Argentina. http://es.wikipedia.org/wiki/Composición_étnica_de_Argentina. 2010 January 9.

11.

Kuczmarski

, Ogden

, Guo

, Grummer-Strawn

, Flegal

, Mei

, Wei

, Curtin

, Roche

, Johnson

. 2000 CDC growth charts for the United States: methods and development. Vital Health Stat, 2002; 11:1–190.

12.

McCarthy

, Ellis

, Cole

. Central overweight and obesity in British youth aged 11–16 years: cross sectional surveys of waist circumference. BMJ, 2003; 326:624–627.

13.

Tanner

. Growth at Adolescence: With a General Consideration of the Effects of Hereditary and Environmental Factors upon Growth and Maturation from Birth to Maturity, 2 ^nd. Oxford, UK: Blackwell Scientific, 1962.

14.

Sun

, Schubert

, Chumlea

, Roche

, Kulin

, Lee

, Himes

, Ryan

. National estimates of the timing of sexual maturation and racial differences among US children. Pediatrics, 2002; 110:911–919.

15.

National High Blood Pressure Education Working Group on High Blood Pressure in Children and Adolescents. The fourth report on the diagnosis, evaluation and treatment on high blood pressure in children and adolescents. Pediatrics, 2004; 114:555–576.

16.

World Health Organization. Report of a WHO Consultation on Obesity. Obesity: Preventing and Managing the Global Epidemic. Geneva: World Health Organization, 1998.

17.

Han

, van Leer

, Seidell

, Lean

MEJ

. Waist circumference action levels in the identification of cardiovascular risk factors: prevalence study in a random sample. BMJ, 1995; 311:1401–1405.

18.

Ashwell

, Lejeune

, McPherson

. Ratio of waist circumference to height may be better indicator of need for weight management. BMJ, 1996; 312:377.

19.

Hsieh

, Yoshinaga

. Abdominal fat distribution and coronary heart disease risk factors in men—waist/height ratio as a simple and useful predictor. Int J Obes, 1995; 19:585–589.

20.

Ashwell

, Hsieh

. Six reasons why the waist-to-height ratio is a rapid and effective global indicator for health risks of obesity and how its use could simplify the international public health message on obesity. Int J Food Sci Nutr, 2005; 56:303–307.

21.

Zimmet

, Alberti

, Kaufman

, Tajima

, Silink

, Arslanian

, Wong

, Bennett

, Shaw

, Caprio

. IDF Consensus Group. The metabolic syndrome in children and adolescents—an IDF consensus report. Pediatr Diabetes, 2007; 8:299–306.

22.

Cook

, Weitzman

, Auinger

, Nguyen

, Dietz

. Prevalence of a metabolic syndrome phenotype in adolescents. Arch Pediatr Adolesc Med, 2003; 157:821–827.

23.

The Expert Committee on Diagnosis and Classification of Diabetes Mellitus. Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care, 2003; 26:3160–3167.

24.

Young-Hyman

, Schlundt

, Herman

, De Luca

, Counts

. Evaluation of the insulin resistance syndrome in 5 to 10 year old overweight/obese African American children. Diabetes Care, 2001; 24:1359–1364.

25.

Matthews

, Hosker

, Rudenski

, Naylor

, Treacher

, Turner

. Homeostasis model assessment; insulin resistance and beta cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia, 1985; 28:412–419.

26.

McAuley

, Williams

, Mann

, Walker

, Lewis Barned

, Temple

, Duncan

. Diagnosing insulin resistance in the general population. Diabetes Care, 2001; 24:460–464.

27.

Burke

, Beilin

, Dunbar

. Family lifestyle and parental body mass index as predictors of body mass index in Australian children: a longitudinal study. Int J Obes Relat Metab Disord, 2001; 25:147–157.

28.

Francis

, Lee

, Birch

. Parental weight status and girls' television viewing, snacking, and body mass indexes. Obes Res, 2003; 11:143–151.

29.

Semmler

, Ashcroft

, van Jaarsveld

, Carnell

, Wardle

. Development of overweight in children in relation to parental weight and socioeconomic status. Obesity (Silver Spring), 2009; 17:814–820.

30.

, Law

, Lo Conte

, Power

. Intergenerational influences on childhood body mass index: the effect of parental body mass index trajectories. Am J Clin Nutr, 2009; 89:551–557.

31.

Norgan

, Jones

. The effect of standardising the body mass index for relative sitting height. Int J Obes Relat Metab Disord, 1995; 19:206–208.

32.

Valsamakis

, Chetty

, Anwar

, Banerjee

, Barnett

, Kumar

. Association of simple anthropometric measures of obesity with visceral fat and the metabolic syndrome in male Caucasian and Indo-Asian subjects. Diabet Med, 2004; 21:1339–1345.

33.

Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA, 2001; 285:2486–2497.

34.

Higgins

, Kannel

, Garrison

, Pinsky

, Stokes

III . Hazards of obesity—the Framingham experience. Acta Med Scand Suppl, 1988; 723:23–36.

35.

Schneider

, Klotsche

, Stalla

, Wittchen

. Obesity and risk of myocardial infarction: the INTERHEART study. Lancet, 2006; 367:1052.

36.

Hsieh

, Muto

. The superiority of waist-to-height ratio as an anthropometric index to evaluate clustering of coronary risk factors among non-obese men and women. Prev Med, 2005; 40:216–220.

37.

Bosy-Westphal

, Geisler

, Onur

, Korth

, Selberg

, Schrezenmeir

, Müller

. Value of body fat mass vs anthropometric obesity indices in the assessment of metabolic risk factors. Int J Obes (Lond), 2006; 30:475–483.

38.

Carey

, Walters

, Colditz

, Solomon

, Willett

, Rosner

, Speizer

, Manson

. Body fat distribution and risk of non-insulin-dependent diabetes mellitus in women. The Nurses' Health Study. Am J Epidemiol, 1997; 145:614–619.

39.

Folsom

, Kushi

, Anderson

, Mink

, Olson

, Hong

, Sellers

, Lazovich

, Prineas

. Associations of general and abdominal obesity with multiple health outcomes in older women: the Iowa Women's Health Study. Arch Intern Med, 2000; 160:2117–2128.

40.

Mo-suwan

, Tongkumchum

, Puetpaiboon

. Determinants of overweight tracking from childhood to adolescence: a 5 y follow-up study of Hat Yai schoolchildren. Int J Obes Relat Metab Disord, 2000; 24:1642–1647.

41.

Malik

, Schulze

, Hu

. Intake of sugar-sweetened beverages and weight gain: a systematic review. Am J Clin Nutr, 2006; 84:274–288.

42.

Nicklas

, O'Neil

, Myers

. The importance of breakfast consumption to nutrition of children, adolescents, and young adults. Nutr Today, 2004; 39:30–39.

43.

Hirschler

, Oestreicher

, Beccaria

, Hidalgo

, Maccallini

. Inverse association between insulin resistance and frequency of milk consumption in low-income Argentinean school children. J Pediatr, 2009; 154:101–105.

44.

Salmon

, Timperio

, Telford

, Carver

, Crawford

. Association of family environment with children's television viewing and with low level of physical activity. Obes Res, 2005; 13:1939–1951.

45.

Songul Yalcin

, Tugrul

, Nacar

, Tuncer

, Yurdakok

. Factors that affect television viewing time in preschool and primary schoolchildren. Pediatr Int, 2002; 44:622–627.

46.

Sallis

, Prochaska

, Taylor

. A review of correlates of physical activity of children and adolescents. Med Sci Sports Exerc, 2000; 32:963–975.