Repository logo
 

Understanding mental health treatment effectiveness in young people


Type

Thesis

Change log

Authors

Neufeld, Sharon Ann Scovil  ORCID logo  https://orcid.org/0000-0001-5470-3770

Abstract

Mental disorders are one of the leading causes of non-communicable disease burden worldwide, with distress emerging as a common factor among such disorders. Most lifetime mental disorders emerge during adolescence and young adulthood, yet no studies have robustly assessed changes in common distress over time during this broad developmental epoch. Common distress is particularly pertinent when assessing mental health treatment effectiveness, as comorbidity rates are higher amongst those who seek treatment. In this dissertation, I seek to redress the paucity of evidence regarding mental health treatment effectiveness in young people. Herein I longitudinally validate a common distress factor, enabling me to use specific and broad mental health outcomes to assess the effectiveness of treatment-as-usual mental health services in two separate samples. In chapter 2, I use a community sample of adolescents to test the relationship of age 14 service contact on depressive symptoms by age 17. I found that 14-year-old adolescents who had contact with mental health services in the past year had a greater decrease in depressive symptoms than those without contact. By age 17 years, the odds of reporting clinical depression were higher in individuals without contact than in service users who had been similarly depressed at baseline. In addition to propensity score weighting to adjust for participants’ initial likelihood to access services, I used a clinically relevant cutoff and adjusted for a wide range of time-varying confounding variables. These adjustments give greater confidence than previous studies to the notion that mental health service contact is related to meaningful improvements in subsequent mental health. Policy implications of these findings are addressed. In chapter 3, I develop a transdiagnostic measure of mental health to be used to assess treatment effectiveness with relevancy across multiple disorders. This transdiagnostic measure is the general factor (common distress or p) from a bifactor model of 118 self-report items from previously validated measures of depression, anxiety, behavioural problems, obsessions, psychotic symptoms, wellbeing, and self-esteem. Longitudinal and gender measurement invariance and reliability of this model was demonstrated in a sample of 14- to 24-year olds assessed annually three times. Predictive validity of the general and specific factors was demonstrated using an extensive set of external variables covering factors such as social environment, personality, and risk behaviour. Accelerated growth modelling revealed developmental changes in the factors from ages 14-27 largely consistent with epidemiological patterns of the associated disorders. I use this validated distress factor in chapter 4 to further test the effectiveness of mental health treatment-as-usual, in a broader age range with a broader outcome than the adolescent sample. Family functioning and friendship support were also explored as potential mechanisms of action. Treatment was related to decreased distress over two years, in unadjusted and adjusted models. In order to propensity weight, I focused analyses on those above the population mean in baseline distress, which contained 85% of young people reporting treatment for a mental disorder. Such individuals were more likely to have required treatment, so treated and untreated groups are more meaningful and comparable. In propensity score weighted models, treatment was related to decreased subsequent distress. Treatment was also related to improved family functioning over two years, but only in adolescents aged 14-18, more likely to be living with their family. In this younger group, the best fitting model revealed distress as a significant mediator: the majority of improvements in family functioning occurred through improvements in distress. However, treatment was not related to a change in friendship support over time. Thus, not only is mental health treatment-as-usual related to a significant decline in young people’s distress, but also improvements in adolescent family functioning. Finally, in chapter 5 I discuss the broader implications of the findings. Analyses in two separate community samples of young people strongly support the notion that treatment-as-usual is related to improvements in both specific and broad measures of mental health. Such treatment-related improvements in mental health appear to also yield benefits in family functioning in adolescents. Findings as a whole argue for increased access to mental health services.

Description

Date

2019-09-30

Advisors

Wilkinson, Paul Oliver
Jones, Peter Brian

Keywords

mental health services, adolescent depression, transdiagnostic mental health, mental health treatment effectiveness, bifactor modelling, propensity weighting, accelerated growth modelling, social support, mediation, emerging adults, measurement invariance

Qualification

Doctor of Philosophy (PhD)

Awarding Institution

University of Cambridge
Sponsorship
Funded by the National Institute for Health Research Collaboration for Leadership in Applied Health Research & Care for East of England.