Poster #63 - Predicting Depression among Adolescents in Post-conflict Northern Uganda

• In Event: 3-159 - Poster Session 12
  In Poster Session: PS 12 Section - Developmental Psychopathology

Integrative Statement

Previous research studying increases in depression through adolescence reports mixed results (e.g., Galambos et al., 2004). However, one consistent finding is that of gender: Girls show higher levels of depression than boys, with this gap emerging in early adolescence (e.g., Hankin et al., 1998). Part of the explanation for this gender gap could be due to stress exposure. Hankin et al. (2007) found that stress exposure was related to depression, particularly for girls. Further research suggests that positive peer status may act as a buffer, thereby reducing depressive symptoms (La Greca & Harrison, 2005).

The above research has been predominately conducted in developed countries, and it’s unclear whether the findings can generalize to other cultural contexts. Our work focuses on adolescents living in northern Uganda. Since 2006 northern Ugandans have been recovering from war. Yet despite this period of peace, these adolescents continue to experience high rates of depression (see Betancourt et al., 2009). The purpose of this research was to study depression among northern Ugandan adolescents over an 18-month period and to consider whether there are similar gender differences. We also explored whether stress factors increase depression, and whether having positive peer relationships could act as a buffer from the effects of this stressor.

One hundred and seventy adolescents (M = 13 years, SD =1.7 years; 46% girls) recruited from a primary school in northern Uganda completed measures assessing self-report depressive symptoms (adapted from the APAI, Betancourt et al., 2009) 4 times over an 18-month period (α’s > .7). In addition, other-report positive peer status (Parkhurst & Asher, 1992) was measured at time 2 (i.e., after 6 months). Finally, students listed 5 stressors in their lives at time 1. Since violence is considered a highly stressful event in this northern Ugandan context, we recorded the frequency of violence-related stressors for each student and used that as our measure of stress.

Exploratory latent growth curve model, using a cube root function, was found to fit our depression data, χ2 (5) = 5.02, p = .41; CFI < .99; RMSEA = .01, revealing an initial decline in depression over the first 6 months that then leveled off over the following 18 months. Next, gender, frequency of stress, and positive peer status were added as predictors of initial levels of depression (i.e., intercept) along with slope. While none of the variables predicted slope, all the variables negatively predicted initial intercept: Gender (males = 1, females = 0), b = -.21, p = .03); Positive Peer Status, b = -1.20, p = .04; Stress, b = -0.1, p = .04. Initial levels of depression were higher among girls than boys, those with less positive peer status, and those who reported less violent-related stressors. This pattern of results differs somewhat from previous work, and is discussed in the context of growing up in post-conflict northern Uganda.

Authors

• Connor Mitchell, Kwantlen Polytechnic University
  Presenting Author

• Kyle Matsuba, Kwantlen Polytechnic University
  Non-Presenting Author

• Andrea Yet Jang Mah, Kwantlen Polytechnic University
  Non-Presenting Author