Family resemblance for patterns of growth in early childhood

Parameters of the Jenss curve representing linear and non-linear components of growth in early childhood were fitted to longitudinal recumbent length and weight measurements from participants in the Fels Longitudinal Growth Study. Nuclear family correlations were estimated for each of four Jenss parameters (a linear intercept and slope and a non-linear intercept and degree of curve) for length and for weight. The TAU path model was fit to the correlations, in order to estimate path coefficients representing transmissibility from parent to child (t² ) and non-transmissible sibling resemblance (s). Likelihood ratio χ² tests suggest that significant family resemblance exists for each Jenss parameter for both variables. Transmissibility is the major component in family resemblance for the linear and non-linear intercepts for length, whereas non-transmissible sibling environmental effects are dominant for three of the four weight parameters. The source of the family resemblance for the remaining parameters could not be pinpointed. The linear slope and degree of curve for weight are more similar for same sex siblings than for opposite sex pairs, suggesting that the degree of shared sibling environment affecting the shape of the growth curve for weight may be greater for same sex pairs.