Adverse life events and mental health in middle adolescence Eirini Flouri a, * , Constantinos Kallis b a Department of Psychology and Human Development, Institute of Education, University of London, 25 Woburn Square, London WC1H 0AA, UK b Medical Statistics Unit, London School of Hygiene and Tropical Medicine, University of London, Keppel Street, London WC1E 7HT, UK Keywords: Cumulative contextual risk Multivariate response models Psychopathology abstract This study’s aim was to search for the appropriate functional form of the effect of proximal cumulative contextual risk (PCCR), measured with number of adverse life events experi- enced in the last 6 months, on adolescent psychopathology and prosocial behavior, measured with the Strengths and Difficulties Questionnaire. The study sample was 171 year ten (aged 14–15) adolescents from predominantly socio-economically disadvantaged families in the UK. Adjustment was made for parental education, and for child’s age, gender, and academic achievement, which was measured with results in Standard Attainment Tests in English, mathematics and science taken in the previous year. PCCR predicted total difficulties, emotional symptoms, conduct problems and hyperactivity. The relationship between PCCR and total difficulties and emotional symptoms was non- quadratic; the PCCR/externalizing problems relationship was quadratic. The findings highlight the importance of considering both outcome specificity and non-linear patterns of associations when modelling cumulative contextual risk effects on adolescent psychopathology. Ó 2010 The Association for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights reserved. Introduction Contextual risk factors do not occur in isolation, and it is the combination of various contextual risk factors that portends negative psychological outcomes in children (Rutter, 1979). Despite this, the body of literature that examines the relationship between multiple risk exposure and children’s psychopathology using a cumulative risk approach is still relatively small (Ackerman, Izard, Schoff, Youngstrom, & Kogos,1999; Atzaba-Poria, Pike, & Deater-Deckard, 2004; Burchinal, Roberts, Hooper, & Zeisel, 2000; Deater-Deckard, Dodge, Bates, & Pettit, 1998; Evans, 2003; Flouri & Kallis, 2007; Morales & Guerra, 2006; Sameroff, Seifer, Baldwin, & Baldwin, 1993; Tiet et al., 1998; Tiet, Bird, Hoven, Wu, et al., 2001). Even when studies use a cumulative risk approach they rarely test whether the relation between cumulative risk and child psychopathology is non- linear (Burchinal, Roberts, Zeisel, & Rowley, 2008, for a review). Although there is evidence for a linear effect whereby increments in risk factors have a steady, additive impact on mental health problems in children (Deater-Deckard et al., 1998), few researchers (e.g., Flouri & Kallis, 2007; Gerard & Buehler, 2004; Simmons, Burgeson, Carlton-Ford, & Blyth, 1987) actually report whether their investigations included appropriate tests for non-linear patterns of cumulative risk. Therefore, the possibility of a non-linear relationship that might manifest itself as an acceleration of problems at a critical level of risk is ignored. For example, there is evidence that cumulative life transitions impact children’s mental health problems in a curvilinear, accelerated manner (Simmons et al., 1987), indicating that a high level of life change is especially difficult to manage. In contrast, complicating the issue even more, a recent study with at-risk elementary school children (Morales & * Corresponding author. Tel.: þ44 0 2076126289; fax: þ44 0 2076126304. E-mail address: e.flouri@ioe.ac.uk (E. Flouri). Contents lists available at ScienceDirect Journal of Adolescence journal homepage: www.elsevier.com/locate/jado 0140-1971/$ – see front matter Ó 2010 The Association for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.adolescence.2010.04.001 Journal of Adolescence 34 (2011) 371–377