Parental height modifies the association between linear growth and neurodevelopment in infancy

Abstract

Aim: To estimate the extent to which maternal and paternal height modify the association between length-for-age Z-score (LAZ) and neurodevelopmental outcomes assessed by the Bayley Scales of Infant and Toddler Development, 3rd Edition (Bayley-III).

Methods: Baseline data from a clinical trial in 600 Nepalese infants aged six to 11 months with LAZ less than -1 were utilised. The primary exposure was the LAZ score, interaction variables were maternal and paternal height, and the outcomes were Bayley-III cognitive, language and motor scaled scores. Linear regression and generalised additive model (GAM) were used to identify potential interactions.

Results: Linear regression analysis stratified by parental height categories showed that association between unit increase in LAZ and cognitive scaled score differed across maternal (normal height: ß 1.16, 95% CI; 0.75, 1.57 and short height: ß 0.67, 95% CI; 0.28, 1.05) and paternal (normal height: ß 1.32, 95% CI; 0.91, 1.72 and short height: ß 0.61, 95% CI; 0.03, 1.18) height categories. Maternal height also modified the association between LAZ and fine motor scaled score.

Conclusion: The association between LAZ and neurodevelopmental outcomes was attenuated when maternal and paternal height was taken into account. Parental stature should be considered when using LAZ as a proxy for neurodevelopment among infants.

Keywords: Infancy; Interaction; Maternal height; Neurodevelopment; Paternal height.