Developmental complexity, structural simplicity: a longitudinal multi-method investigation of internalising and externalising symptoms in young people

Praveetha Patalay

Supervisors: Peter Fonagy & Jessica Deighton

My PhD research involves a series of studies investigating various aspects of the development of mental disorder during late childhood and early adolescence (8-14 years) in a large nationwide longitudinal study in England. The main themes of this thesis, as implicated by the title, revolve around the development and underlying structure of internalising and externalising symptoms: the two key dimensions in child psychopathology. The four studies that fall within these themes are preceded by two studies investigating the measurement properties of the key outcome measure used in the subsequent studies. All through this work, I make an attempt to think about the limitations of the more commonly used approaches – in many cases the accepted ways of analysing these types of data, to allow myself to consider the alternatives.

The first two studies examine the clinical validity and survey format equivalence of a child self-report measure of mental health, the Me and My School questionnaire. Results of the first study support the measure’s reliability and validity. In study 2 using scale- and item-level analyses the findings suggest that overall children disclose more problems when online questionnaires are used, a finding with implications for combining and comparing results from studies where different survey methods have been used.

The next two studies investigate the complexity of individuals’ symptom development by examining the developmental trajectories of internalising and externalising symptoms using latent class growth analysis, a method that allows the estimation of different person-centred trajectories of symptom development. Subsequently, the studies identify the socio-demographic correlates (e.g. gender, socio-economic status) of individuals with different trajectories. This is followed by examining the impact of these different types of trajectories on another key domain of child functioning: academic attainment. This innovative approach to investigating these relationships led to uncovering nuanced details in these longitudinal relationships. For instance, stable moderate internalising symptoms also predict worse subsequent academic attainment in children, supporting the need for early intervention strategies to prevent problems from impacting on children’s learning.

The fifth study focusses on the co-development of internalising and externalising symptoms, aiming to uncover patterns in their association and development over time. Findings indicate moderate associations between these two domains – domains that are generally conceptualised as distinct dimensions of disorder, leading to questions regarding how the structure of psychopathology is currently conceptualised. The last study, aims to investigate the underlying structure of child psychopathology. Using hierarchical bi-factor analysis, the study explores the possibility of a general propensity for psychopathology that not only is a better predictor of future psychopathology but can lead to better models of specific disorders as well.

The thesis contributes to increasing the understanding of complexities in symptom development and proposes a simpler structural model underlying symptoms. The analytic approaches used highlight the value of modern data analytic techniques in increasing our understanding of the development of psychopathology.

References/ Publications

  • Patalay, P., Deighton, J., Fonagy, P. & Wolpert, M. (in press). The relationship between internalising symptom development and academic attainment in early adolescence. PLOS ONE.
  • Patalay, P., Fonagy, P., Deighton, J., Belsky, J., Vostanis, P., & Wolpert, M.(in press). A general psychopathology factor in early adolescence. The British Journal of Psychiatry.
  • Patalay, P., Deighton, J., Fonagy, P., Vostanic, P. & Wolpert, M. (2014).Clinical validity of the M&MS questionnaire: a child self-report mental health measure. Child and Adolescent Psychiatry and Mental Health. DOI: 10.1186/1753-2000-8-17
  • Patalay, P., Deighton, J., Fonagy, P., & Wolpert, M.(advance online). Equivalence of paper and computer survey formats of a child self-report mental health measure. European Journal of Psychological Assessment.doi: 10.1027/1015-5759/a000206
  • Patalay, P., Fink, E., Fonagy, P., & Deighton, J. (2015). Unpacking the associations between heterogeneous externalising symptom development and academic attainment in middle childhood. European child & adolescent psychiatry, advance online.