ObjectivesChildhood growth reflects nutritional status such as under- or overnutrition and is a determinant of life health. Thus, early detection of sub-optimal growth patterns can help identify modifiable risk factors. Current approaches rely on the trajectory intercept and struggle to capture differences in trajectory shape, which better reflects change over time. This study implemented a novel approach to identify shape-based growth patterns in children from birth to 11 years of age. MethodsWe used anthropometric measures from 3,945 children at postnatal age 3 months, 1, 2, 4, 6 and 11 years from the 2004 Pelotas Birth Cohort (Brazil). Weight-, height- and BMI-for-age Z-scores were calculated using INTERGROWTH21st and WHO growth standards. The impact of gestational age-correction (CA) in preterm-born children (n = 461) was tested, and trajectories were either baseline anchored at zero or not. This produced 4 analytical pipelines for pattern detection (k-means, Frechet’s distance): 1) no CA–no baseline anchoring; 2) no CA–with baseline anchoring; 3) with CA–no baseline anchoring; 4) with CA–with baseline anchoring. ResultsWe found 3 patterns for each Z-score, with “High” (increasing), “Middle” (stable) and “Low” (declining) trajectories. The use of both CA and baseline anchoring highlighted sub-groups of children with specific patterns related to early growth (first 3 months of life). In the non-anchored approach, CA influenced the mean trajectories detected by increasing the intercept and reducing the slope between birth and 3 months. Baseline anchoring identified mean trajectories with steeper slopes during this period, thus, capturing differential growth patterns in early life. ConclusionsDepending on the question, incorporating CA with or without baseline anchoring when analyzing growth patterns can better highlight aspects of early growth versus overall patterns across childhood and can improve growth analyses standardization in population-based studies. Funding SourcesJoannah and Brian Lawson Centre for Child Nutrition, Lawson Family Chair in Microbiome Nutrition Research, CONACyT scholarship, Wellcome Trust, World Health Organization, National Support Program for Centers of Excellence, Brazilian National Research Council, Brazilian Ministry of Health, and Children’s Pastorate.
Read full abstract