The effects of family environments on structural covariance of cortical networks in late childhood

© 2017 Dr. Sally RichmondThe impact of adversity on brain development is well established, however relatively little attention has been directed toward the neurobiological effects of normative variation in both negative and positive parenting behaviours. Given the rapid brain reorganisation during late childhood, parenting behaviours are particularly likely to impact the structure of the brain during this time. The aim of this thesis was to investigate the association between normative variations in parenting and structural brain networks in late childhood. Observational data were collected from a cross-sectional sample of 163 mother-child dyads [76 male children, mean (M) age 8.4 years, standard deviation (SD) 0.3 years] recruited from across Melbourne, Australia. The dyads were specifically recruited from areas in Melbourne that scored within the lowest tertile on a measure of socioeconomic disadvantage. Parenting characteristics were determined from two lab-based interaction tasks, completed by dyads that were video recorded and subsequently coded. T1 weighted images were acquired from children in a 3T TIM Trio Siemens scanner and processed using FreeSurfer to extract cortical thicknesses. Structural brain networks were based on structural covariance (SC) of cortical thickness using sparse regularised methods. Exploratory Factor Analysis identified the following four parenting variables from the observational data: (a) negative affect displayed in a positive interaction task, (b) positive affect, (c) negative affect displayed in a negative interaction task, and (d) communication. Children were grouped based on the distributions of the parenting variables into low-, moderate-, and high- parenting groups. Structural brain networks were then determined for each group. Graph theory was then applied to analyse the properties of the structural brain networks between the low-, moderate- and high- parenting groups. Moderate levels of observed negative and positive maternal behaviours were associated with increased local efficiency relative to low- and high-levels, whereas moderate levels of maternal communication were associated with decreased local efficiency relative to low- and high-levels. The findings for small-worldness followed a similar pattern to the local efficiency associations. Minimal support was found for an association between global efficiency and observed maternal behaviours. Moderate levels of observed negative and positive maternal behaviours were associated with increased modularity; whereas high levels of maternal communication were associated with increased modularity. Alteration to brain structure has been proposed as one mechanism by which parenting influences the development of psychopathology. The current study offers support for this mechanism, as associations between normative variations in parenting behaviours and structural brain networks were found. As the results are cross-sectional however, further longitudinal research is required to explore whether parenting behaviour has a prospective relationship with the development of structural brain networks