Predictive model for depression symptom severity.

BackgroundStudents in sub-Saharan African countries experienced online classes for the first time during the COVID-19 pandemic. For some individuals, greater online engagement can lead to online dependency, which can be associated with depression. The present study explored the association between problematic use of the internet, social media, and smartphones with depression symptoms among Ugandan medical students.MethodsA pilot study was conducted among 269 medical students at a Ugandan public university. Using a survey, data were collected regarding socio-demographic factors, lifestyle, online use behaviors, smartphone addiction, social media addiction, and internet addiction. Hierarchical linear regression models were performed to explore the associations of different forms of online addiction with depression symptom severity.ResultsThe findings indicated that 16.73% of the medical students had moderate to severe depression symptoms. The prevalence of being at risk of (i) smartphone addiction was 45.72%, (ii) social media addiction was 74.34%, and (iii) internet addiction use was 8.55%. Online use behaviors (e.g., average hours spent online, types of social media platforms used, the purpose for internet use) and online-related addictions (to smartphones, social media, and the internet) predicted approximately 8% and 10% of the severity of depression symptoms, respectively. However, over the past two weeks, life stressors had the highest predictability for depression (35.9%). The final model predicted a total of 51.9% variance for depression symptoms. In the final model, romantic relationship problems (ß = 2.30, S.E = 0.58; p0.01) and academic performance problems (ß = 1.76, S.E = 0.60; p0.01) over the past two weeks; and increased internet addiction severity (ß = 0.05, S.E = 0.02; pTwitter use was associated with reduced depression symptom severity (ß = 1.88, S.E = 0.57; p0.05).ConclusionDespite life stressors being the largest predictor of depression symptom score severity, problematic online use also contributed significantly. Therefore, it is recommended that medical students’ mental health care services consider digital wellbeing and its relationship with problematic online use as part of a more holistic depression prevention and resilience program.背景 撒哈拉以南非洲国家的学生在2019冠状病毒病（COVID-19）大流行期间首次接触线上授课。对部分个体而言，更高程度的线上参与可能引发线上依赖，而这与抑郁症状存在关联。本研究旨在探讨乌干达医学生的互联网、社交媒体及智能手机问题性使用与抑郁症状之间的关联。 方法 本研究针对乌干达某公立大学的269名医学生开展了预试验研究。通过问卷调查收集了社会人口学因素、生活方式、线上使用行为、智能手机成瘾、社交媒体成瘾及互联网成瘾相关数据。采用分层线性回归模型，探究不同类型线上成瘾与抑郁症状严重程度的关联。 结果 研究结果显示，16.73%的医学生存在中度至重度抑郁症状。各类成瘾风险检出率分别为：(i) 智能手机成瘾风险45.72%，(ii) 社交媒体成瘾风险74.34%，(iii) 互联网成瘾风险8.55%。线上使用行为（如日均线上时长、使用的社交媒体平台类型、互联网使用目的）及各类线上相关成瘾（智能手机成瘾、社交媒体成瘾、互联网成瘾）分别可解释抑郁症状严重程度约8%和10%的变异量。然而，过去两周内，生活压力源对抑郁症状的预测力最高，可达35.9%。最终模型总共可解释抑郁症状51.9%的变异量。在最终模型中，过去两周内的恋爱关系问题（标准化回归系数β=2.30，标准误（standard error）=0.58；p<0.01）、学业表现问题（β=1.76，标准误=0.60；p<0.01）以及互联网成瘾严重程度升高（β=0.05，标准误=0.02；p<0.05）均为抑郁症状严重程度的显著预测因子。此外，使用推特（Twitter）与抑郁症状严重程度降低存在显著关联（β=1.88，标准误=0.57；p<0.05）。 结论 尽管生活压力源是抑郁症状评分严重程度的最强预测因子，但问题性线上使用同样具有显著贡献。因此，建议针对医学生的心理健康服务可将数字健康（digital wellbeing）及其与问题性线上使用的关系纳入更全面的抑郁预防与心理韧性提升项目中。