The association of intrauterine and postnatal growth patterns and nutritional status with toddler body composition

Data for this study was a secondary analysis of data drawn from the International Atomic Energy Agency Multicentre Body Composition Reference Study (IAEA-MBCRS). IAEA-MBCRS is a longitudinal study on singleton infants born in Soweto between 2014 and 2019. A total of 411 black mothers of infants born at Chris Hani Baragwanath Academic Hospital, South Africa, were recruited into the study. The commitment to the IAEA was to recruit 100 mother-infant pairs for inclusion into the global study. For the current study, a total of 115 infants with data at birth, 12 and 24 months were included. Sensitivity analyses were performed by comparing excluded participants to those who were included in the analytical sample. Mothers who were 18 years and above living in Soweto urban township, which is part of the Johannesburg Metropolitan Council Area in South Africa, were randomly approached and recruited at the maternity ward at Chris Hani Baragwanath Academic Hospital. Eligibility was based on the following criteria: (i) mothers had to be residents of Soweto, 18 years and above, must have delivered a full term (between 37- and 42-weeks) singleton healthy baby; (ii) A non-smoker and willing to breastfeed for up to 6 months and must have attained at least secondary school education level. Eligible mother-baby pairs were then followed up from birth until they reached 24 months. Infants with significant morbidity and congenital abnormalities were excluded from the study. All mothers provided written informed consent, and the ethical approval to conduct the study was granted by the University of the Witwatersrand Ethics Committee for Research on Human Subjects (M200997). All processes and methods were carried out in accordance with the IAEA-MBCRS guidelines and regulations.

Anthropometric measurements were taken at 3, 6, 9, 12, 15, 18, 24 months follow-up visits as per World Health Organization (WHO) Multicentre Growth Reference Study (MGRS) standardized methodology [24, 25]. A portable electronic scale (Seca 376) placed on a flat surface and covered with a soft paper towel to protect the baby from direct contact with the cold surface was used to measure weight. The scale was then tared before placing a naked baby and measurement was recorded when the baby was calm. For measurement reliability, all infant measurements were taken in duplicate and recorded blinded by two research assistants. The supine length was measured from the crown of the head to the heels using a calibrated Harpenden Infantometer (Range 300-1100 mm) with digital counter readings to the nearest 0.1 cm. For measuring of length, the Infantometer with a perpendicular fixed headboard and a movable footboard was placed on a flat surface with a soft paper towel on top to prevent the baby from direct contact with the cold surface. The measurement was taken with the head touching an immovable heard board, positioned in the Frankfort vertical plane, secured by the supporting research assistant standing at the head position. The lead measurer stood on the side to hold the baby’s legs together making sure that the knees were straightened, and the footboard moved gently towards the heels of the baby. Quality control of the instruments was performed daily before use. Weighing scale (Seca 376) was calibrated using different weights of 0.5 kg, 1 kg, 2 kg, and 5 kg. Harpenden Infantometer was calibrated using two aluminium metallic rods of 75 cm and 40 cm.

Birth size categories were classified as small for gestational age (SGA), appropriate for gestational age (AGA), and large for gestational age (LGA) using INTERGROWTH-21 standards [26]. Length-for-age (HAZ) was computed using 2006 WHO child growth standards for children from birth to five years. Stunting was defined as HAZ <-2 [27]. Conditional relative weight and conditional length) were obtained as standardised residuals by regressing previous length and weight on weight and length 12 and 24 months [28]. Conditional relative weight is the weight at current age accounting for current height and previous weight, while conditional length does not account for current weight [28]. For example, conditional relative weight at 24 months was obtained by regressing weight at birth and 12 months, and length at 12 and 24 months on weight at 24 months. Comparison by nutritional status (e.g., stunting vs. non-stunting) was assessed using the independent t-tests.

Total body water (TBW) was measured by stable isotope dilution method using 99.8 atom % D, Sigma-Aldrich deuterium oxide isotope as outlined in the IAEA standard operating procedures [29]. Infants’ saliva samples were collected at 3, 6, 9, 12, 15, 18, and 24 months follow-up visits. Sample collection was a 3-hour process that was based in three stages: Baseline sample collection, dose administration and post dose collection. Baseline saliva sample was collected 20 min after the baby’s last feed by placing a ball of clean cotton wool inside the baby’s mouth until it was soaked wet. The soaked cotton wool was carefully removed and transferred into 20ml syringe. Saliva sample was then squeezed out into a 2ml vial labelled with a unique infant identification code. The baby was slowly given 1ml dose of undiluted deuterium oxide making sure that no spillages occurred. Another saliva sample was then collected 3 h post dose administration, assuming the enrichment distribution reached equilibrium. Samples were stored in -20 degrees Celsius until analysis.

Saliva samples were analysed in duplicate using portable 4500 Tumbl-IR Fourier transform infrared (FTIR) spectrometer (Agilent Technologies, CA, USA) according to manufactures instructions. The instrument is fitted with optics suitable for absorbance measurements on liquid samples in the mid-infrared range, and only takes about 20–30 µl of sample. To prevent cross-contamination between samples, the sampling window was cleaned with 99.9% ethanol after each analysis. Within day coefficient of variations were run daily before and after saliva analysis on the FTIR.

FFM was calculated from the measurement of TBW by the hydration factors [23]. FM was then derived as the difference of body weight and FFM and presented as kg or percentage. Body composition indices (FMI and FFMI) were defined as fat mass or lean mass in relation to height and weight. Log-log index (FM/FFM^P) was calculated by performing a natural log transformation of FM and FFM then regressing log FM on log FFM. The regression coefficient was used as P in the ratio of absolute FM and FFM, as FM/FFM^P. To present readable results, the derived index values were multiplied by 1000 [30].

Continuous data were assessed for normality and where appropriate presented as means (95% CI) and as percentages for categorical variables. The independent t-test and Mann-Whitney U test were used to assess differences between groups for continuous normally and non-normally distributed variables from birth to 2 years, respectively. Differences between groups for categorical variables were compared using the Pearson chi-square test. The one-way analyses of variance (ANOVA), with Bonferroni multiple comparisons test, was used to assess differences in body composition between categories of birthweight (SGA, AGA, and LGA). FM, FMI, FFM, FFMI, and FM/FFM^P were each regressed on birthweight z-score, conditional relative weight, and conditional length at 12 months and 24 months using multiple linear regression analyses. All data were analysed using Stata 14.2 version (Stata-Corp LP, College Station, Texas, USA). Statistical differences were accepted at p < 0.05.