Metrics of early childhood growth in recent epidemiological research: A scoping review

Abstract

Metrics to quantify child growth vary across studies of the developmental origins of health and disease. We conducted a scoping review of child growth studies in which length/height, weight or body mass index (BMI) was measured at ≥ 2 time points. From a 10% random sample of eligible studies published between Jan 2010-Jun 2016, and all eligible studies from Oct 2015-June 2016, we classified growth metrics based on author-assigned labels (e.g., 'weight gain') and a 'content signature', a numeric code that summarized the metric's conceptual and statistical properties. Heterogeneity was assessed by the number of unique content signatures, and label-to-content concordance. In 122 studies, we found 40 unique metrics of childhood growth. The most common approach to quantifying growth in length, weight or BMI was the calculation of each child's change in z-score. Label-to-content discordance was common due to distinct content signatures carrying the same label, and because of instances in which the same content signature was assigned multiple different labels. In conclusion, the numerous distinct growth metrics and the lack of specificity in the application of metric labels challenge the integration of data and inferences from studies investigating the determinants or consequences of variations in childhood growth.

Fig 1. Components and ranges of possible values of the 8-digit content signature.

Each component of the signature is represented by a 1- or 2-digit code, and the component codes were concatenated to generate the 8-digit content signature for each metric.

Fig 2. Flow of study selection.

Fig 3. A Sankey diagram to illustrate the heterogeneity among published metrics for child growth in length, weight or body mass index (n = 235) and relative prevalences overall and within each component.

Moving from left to right, content signatures are deconstructed into their individual components (i.e., standardization, level of estimation, metric type, quantity of data, metric subtype, analytic approach), where the width of the band is proportional to the frequency of the approach. The most common approach was the calculation of each child’s incremental change in the standardized anthropometric parameter, which is represented by the band that flows through the following nodes: ‘standardized parameter’ (dark blue), ‘individual level of analysis’ (dark red), ‘continuous variable’ (dark green), ‘2 data points’ (light purple), ‘incremental change’ (dark orange), and ‘manual calculation’ (pink). The range of growth metrics presented is based on a random sample of published studies, and therefore is not exhaustive.

Fig 4. Decision tree for selection of metrics of growth in length (n = 87).

Percentages represent the relative prevalence of the approach at each branching point. For example, the most common approach for growth in length as an exposure with 2 data points is to first standardize the anthropometric parameter, then calculate the incremental change.

Fig 5. Decision tree for selection of metrics of growth in weight (n = 99).

Percentages represent the relative prevalence of the approach at each branching point. For example, the most common approach for estimating growth in weight as an outcome with >2 data points was to calculate the incremental rate of change of unstandardized weight using a linear mixed effects model.

Fig 6. Decision tree for selection of metrics of growth in BMI (n = 49).

Percentages represent the relative prevalence of the approach at each branching point. For example, the most common approach for expressing growth in BMI as an exposure with >2 data points was to first standardize BMI, then analyze it in relation to an outcome using latent class analysis.