Auxology of small samples: A method to describe child growth when restrictions prevent surveys

BACKGROUND: Child growth in populations is commonly characterised by cross-sectional surveys. These require data collection from large samples of individuals across age ranges spanning 1–20 years. Such surveys are expensive and impossible in restrictive situations, such as, e.g. the COVID pandemic or limited size of isolated communities. A method allowing description of child growth based on small samples is needed. METHODS: Small samples of data (N~50) for boys and girls 6–20 years old from different socio-economic situations in Africa and Europe were randomly extracted from surveys of thousands of children. Data included arm circumference, hip width, grip strength, height and weight. Polynomial regressions of these measurements on age were explored. FINDINGS: Polynomial curves based on small samples correlated well (r = 0.97 to 1.00) with results of surveys of thousands of children from same communities and correctly reflected sexual dimorphism and socio-economic differences. CONCLUSIONS: Fitting of curvilinear regressions to small data samples allows expeditious assessment of child growth in a number of characteristics when situations change rapidly, resources are limited and access to children is restricted.