Project outline: The longitudinal course of depressive symptoms from late childhood to early adolescence

The proposed 24-month project will investigate the developmental changes in depressive symptoms from late childhood (9 years) to adolescence (13 years) in the Avon Longitudinal Study of Parents and Children (ALSPAC). The research will exploit recent innovations in statistical modelling of longitudinal data that now make it possible to empirically test whether there are meaningful subgroups of individuals in the ALSPAC cohort that follow distinct developmental trajectories of depressive symptoms.

Using the statistical package Mplus (Muthen and Muthen 1998), the study will employ a range of approaches (both parametric, semi-parametric and non-parametric) to describe both the within-child and between child heterogeneity observed in a set of repeated measures of depressive symptoms. Techniques such as these will permit us to examine whether distinct developmental trajectories of depressive symptoms can be empirically identified in the ALSPAC cohort, and furthermore, allow us to test whether individuals cluster into distinct subgroups, each with a different trajectory profile of depressive symptoms. It is expected that there will be at least four distinct trajectory groupings relating to a) little or no depressive symptoms from childhood to adolescence (persistent low level); b) increasing levels of depressive symptoms from childhood into adolescence (increasing); c) onset of depressive symptoms in childhood with consistently high levels of depressed mood (chronic high); and d) moderate levels of depressed symptoms throughout the assessment period (moderate). Gender differences in the trajectory groupings will also be examined using two approaches: Firstly, by modelling the sample as a whole, it will be possible to examine the gender ratio observed within each trajectory group. Secondly, by fitting gender-specific models, we will be able to examine differences in trajectory shape. It is predicted that girls will show higher mean levels and sharper increases in depressive symptoms from childhood to adolescence than boys.

The second aim of the proposed study is to link distinctive trajectories with risk factors that might account for individual differences in the development of depressive symptoms.

By taking advantage of extensive longitudinal data from ALSPACon socio-demographic background, family adversity and parental psychopathology, the proposed study will compare the risk factors for depressive symptoms beginning in childhood (9-11 years) compared to adolescence (12-13 years). We will empirically test whether social disadvantage and family adversity are more strongly associated with childhood onset depressive symptoms (i.e. persistent depressive symptoms beginning in childhood) compared to those with depressive symptoms emerging in adolescence.

The study will use longitudinal data from the Mood and Feelings Questionnaire to measure depressive symptoms (MFQ, Angold et al. 1995; Sharp et al, 2006). The use of this instrument allows both a categorical view of likely depressive disorder, as well as a dimensional measure of total depressive symptoms with severity ranging from none to severe.