Anthropometric and clinical correlates of fat mass in healthy term infants at 6 months of age

This descriptive study derives from a prospective cohort of pregnant women conducted at the National Institute of Perinatology (Mexico City, 2009–2012).

We consecutively selected women in their first trimester of pregnancy (< 14 weeks), who did not have comorbidities (diabetes mellitus, renal, hepatic and/or autoimmune diseases), did not use medication that affected metabolism (metformin, prednisone, etc.) and did not have multiple pregnancies. All adult women that accepted to participate signed an informed consent; for those < 19 years old, the consent of parents was also required. Women were followed monthly until the end of pregnancy.

Newborns were selected by convenience from healthy mothers participating in the afore-mentioned cohort. We invited to participate only those mother-baby binomials that did not develop gestational diabetes or preeclampsia and whose babies were born at the institutional hospital. We included healthy term newborns (≥37 weeks). Those babies born prematurely (< 37 weeks) or with congenital/metabolic diseases were excluded. Babies that lacked body composition assessment at 6 months were eliminated from the analysis.

A nutritionist with prior experience and adequately trained in anthropometric measurement standardization performed all the babies’ anthropometric procedures at birth (first 72 h) and at 6 months of age, following the methods described by Lohman [23].

Infants were measured without clothes. The average of two measurements was recorded, except for weight and FM. At birth, weight was recorded in grams to the nearest 0.1 kg using a pediatric scale (1582 Baby/Mommy Scale, TANITA Corp., Tokyo, Japan). Low birth weight (< 2500 g) and macrosomia (> 4000 g) were recorded. At 6 months, weight was measured using the digital scale from the PEAPOD Infant Body Composition System (software version 3.1.0, Lifestyle Measurement Instruments, COSMED, California, USA). Recumbent length was measured to the nearest 0.1 cm using an infantometer (SECA 207, SECA, Hamburg, Germany). With the help of a nurse, the infant was laid down and her/his head was placed in the Frankfort plane (as the best anatomical indicator of a physiological position for a population without facial deformities) [24]. Length was measured from the crown of the head to the foot’s heel, while both were appropriately attached to the infantometer. BMI was computed (weight/length²) and sex-specific z-scores for BMI/A and W/L were calculated using the World Health Organization (WHO) Anthro software program (version 3.0.1, WHO, Geneva, Switzerland). Risk of overweight (> 1 standard deviation scores), overweight (> 2 standard deviation scores) and obesity (> 3 standard deviation scores) was defined based on z-scores.

Waist circumference (WC) was measured around the umbilicus to the nearest 0.1 cm with a fiberglass anthropometric measuring tape (Gulick, Sammons Preston, Chicago, USA). WC-to-length ratio (WLR) was obtained dividing WC by length.

Skinfolds (biceps –BSF–, triceps –TSF– and subscapular –SSF) were measured with a Lange caliper (Beta Technology, California, USA) from the child’s left arm holding the skin between the index finger and thumb (one centimeter above the measurement site) while the baby was laid down. BSF and TSF were measured taking as reference the mid-point of the arm grasping the skin parallel to the floor. SSF was taken at a 45° to the horizontal plane. A difference > 2 mm between both measurements required a third measurement. Skinfold sum (SFS) was obtained by adding the values from the three skinfolds aforementioned.

The assessment of FM (percentage -%- and kilograms -kg-) and fat-free mass (FFM, % and kg) was performed at 6 months using the PEAPOD Infant Body Composition System [25, 26]. Every day, before starting measurements, the PEAPOD was calibrated according to the manufacturer’s protocol. Before placing the infant in the PEAPOD scale, (s) he was undressed and a head cap was put on to minimize air trapped in the hair. Then, (s) he was placed in the test chamber tray in order to start with the measurement of body composition. After volume and weight were measured by de PEAPOD, the system’s software calculated the infant’s body composition based on Fomon’s body density values. FMI (FMkg/length²) was computed to correct for variations in FM related to body size [27].

At 1-, 3- and 6-months-old visits, mothers were asked a series of questions regarding their infant’s feeding practices. Using the information collected from the 6-months questionnaire, infants were classified within 3 breastfeeding categories according to the WHO definition [28]: a) exclusive and predominant breastfed for 6 months (E/P BF), b) breastfed for up to 6 months + formula feeding (BF + F), and c) breastfed for less than 6 months + formula feeding (BF < 6 m + F).

Mothers participating in the cohort study reported their age and pre-gestational weight in their first visit. Pre-gestational BMI was computed and classified according to WHO criteria (low, normal, overweight or obesity).

Descriptive statistics and frequencies were performed for all variables. Mean differences were analyzed using Student’s t-test for parametric variables or U Mann-Whitney for non-parametric ones. One-way ANCOVA was performed to evaluate differences in anthropometric indices and FM according to type of feeding. Pearson’s and Spearman’s correlations were used to evaluate bivariate associations between FM (% and index) and anthropometric measures.

A total of 222 healthy mother-baby binomials from the cohort study were recruited following birth and invited to participate. Nevertheless, only 155 agreed to participate. Fifty-three infants that lacked body composition measurement at 6 months were eliminated. This analysis included a total of 102 infants. Due to different circumstances not attributable to our staff, it was not possible to assess weight and length of eight infants at birth (n = 94) (Table 1).

Mothers had a mean age of 27.1 ± 8.7 years and their mean pre-gestational BMI was 24.7 ± 5.2 kg/m². The prevalence of pre-gestational overweight/obesity was 40.2%. At birth, infants’ mean gestational age, weight and length was 39.0 ± 1.1 weeks, 2.89 ± 0.35 kg and 46.9 ± 2.0 cm, respectively. Low birth weight was present in 10.6% (n = 10) of newborns and no infants were classified as macrosomic. Most of the newborns had normal BMI/A at birth (93.4%, n = 95), and there were no infants classified with obesity during the first 6 months of age.

Descriptive data of mothers, newborns and infants at 6 months is presented in Table 1. Boys had a higher weight and length at birth, as well as a higher weight and FFM at 6 months (p < 0.05). No other anthropometric differences were observed by sex (p > 0.05).

FM (% and kg) was higher and FFM (% and kg) was lower in infants classified as E/P BF when compared to BF + F or BF < 6 m + F. Except for WC, all anthropometric measurements were higher in the E/P BF group (adjusted by sex, BMI/A at birth and pre-gestational BMI) (Table 2).

Correlations of anthropometric measurements with FM% at 6 months were all statistically significant (p < 0.001). The strongest correlation coefficients were for BMI (r = 0.691), SFS (r = 0.603) and WC (r = 0.591), followed by WLR (r = 0.589), SSF (r = 0.588) and TSF (r = 0.581). The weakest correlations were for weight (r = 0.569) and BSF (r = 0.409).

All anthropometric measurements had significant correlations with FMI (p < 0.0001). In order of strength, correlation coefficients were as follows: BMI (r = 0.816), WLR (r = 0.693), WC (r = 0.691), weight (r = 0.668), SFS (r = 0.640), SSF (r = 0.631), TSF (r = 0.624) and BSF (r = 0.415).

The strongest models of FM% included SFS + WC + length (model 2) followed by TSF + WC + length (Model 4) (Table 3). Regarding FMI, the strongest associations were shown for SFS + WC (model 2) followed by TSF + WC (Model 4) (Table 4).