Document Type



Inter-observer and intra-observer reliability in obtaining skinfold thickness measurements on newborn infants were examined. As part of ongoing growth studies,20 duplicate sets of eight skinfolds were measured by an index observer and one of three other experienced anthropometrists. Additionally, ten sets of duplicate measurements were obtained several hours apart by the index observer. No consistent biases among the four observers were noted, although occasional differences at the arm sites were observed. Although the technical errors of measurement were smaller, the coefficients of variation were of the same magnitude as those previously reported for skinfold measurements in older children. Intra-observer variation was markedly less than inter-observer variation for all sites. To minimize intra-observer bias, skinfold thickness measurements should be obtained at multiple rather than single sites. The data also suggest measurement error will be minimized by using a single observer for longitudinal, and multiple observers for cross-sectional studies.