S1 Data -

Background
There is documented evidence of the increase of alcohol and substance use among college students globally. Increased morbidity and associated maladaptive socio-occupational outcomes of the habit with early dependence and mortality have also been reported. Majority of the substance use related studies conducted in low- and middle- income countries mainly look at health- related risk behaviour control mechanisms that focus on the social environment domain, with few or almost none focusing on those embedded within the person (self- control). This study focuses on the relationship between substance use and personality traits (in the self-control domain), among college students in a low- middle- income country.
Methods
Design. A cross- sectional descriptive study that used the self- administered WHO Model Core and the Big Five Inventory Questionnaires to collect information among students in Colleges and Universities in Eldoret town, Kenya.
Setting. Four (1- university campus; 3- non- university) tertiary learning institutions were randomly selected for inclusion.
Subjects. Four hundred students, 100 from each of the 4 institutions; selected through a stratified multi-stage random sampling, who gave consent to participate in the study. Associations between various variables, personality traits and substance use were tested using bivariate analysis, while the strength/ predictors of association with substance use was ascertained through multiple logistic regression analyses. A finding of p ≤ 0.05 was considered statistically significant.
Results
The median age was 21 years (Q1, Q3; 20, 23), approximately half 203 (50.8%) were male, with majority 335 (83.8%) from an urban residence and only 28 (7%) gainfully employed. The lifetime prevalence of substance use was 41.5%, while that of alcohol use was 36%. For both, a higher mean neuroticism score [substance use- (AOR 1.05, 95%CI; 1, 1.10: p = 0.013); alcohol use- (AOR 1.04, 95%CI; 0.99, 1.09: p = 0.032)] showed increased odds of lifetime use, while a higher mean agreeableness score [substance use- (AOR 0.99, 95%CI; 0.95, 1.02: p = 0.008); alcohol use- (AOR 0.99, 95%CI; 0.95, 1.02: p = 0.032)] showed decreased odds of lifetime use. A higher mean age (AOR 1.08, 95% CI; 0.99, 1,18: p = 0.02) of the students also showed an 8% increase in odds of lifetime alcohol use. The lifetime prevalence of cigarette use was 8.3%. Higher mean neuroticism (AOR 1.06, 95%CI; 0.98, 1.16: p = 0.041) and openness to experience (AOR 1.13, 95%CI; 1.04, 1.25: p = 0.004) scores showed increased odds of lifetime cigarette smoking, whereas being unemployed (AOR 0.23, 95%CI; 0.09, 0.64: p<0.001) had a decreased odd. Other substances reported included cannabis 28 (7%), sedatives 21 (5.2%), amphetamines 20 (Catha edulis) (5%), tranquilizers 19 (4.8%), inhalants 18 (4.5%), cocaine 14 (3.5%), with heroin and opium at 10 (2.5%) each. Among the 13 participants who reported injecting drugs, 10 were female and only 3 were male; this finding was statistically significant (p = 0.042).
Conclusions
The prevalence of substance use among college and university students in Eldoret is high and associated with high neuroticism and low agreeableness personality traits. We provide directions for future research that will examine and contribute to a deeper understanding of personality traits in terms of evidence- based approach to treatment.

背景：已有文献证实，全球范围内大学生的酒精与物质使用行为呈上升趋势。该类行为可引发更高发病率，同时伴随早期依赖、死亡风险以及适应不良的社会职业结局，相关报道屡见不鲜。目前在中低收入国家开展的物质使用相关研究，大多聚焦于社会环境维度的健康风险行为控制机制，极少或几乎未涉及个体内部（如自我控制）层面的相关因素。本研究旨在探讨某中低收入国家大学生的物质使用与自我控制维度人格特质之间的关联。

方法：
研究设计：本研究为横断面描述性研究，采用自填式世界卫生组织核心问卷（WHO Model Core）与大五人格量表（Big Five Inventory）收集数据，研究对象为肯尼亚埃尔多雷特镇各学院及高校的学生。
研究场景：随机选取4所高等教育学习机构纳入研究，其中1所为大学校园，3所为非校园类高等教育机构。
研究对象：通过分层多阶段随机抽样方法，从4所机构各抽取100名学生，共计400名签署知情同意书的参与者。采用双变量分析检验各类变量、人格特质与物质使用之间的关联；通过多重logistic回归分析明确物质使用相关关联的强度与预测因素，以p≤0.05作为统计学显著性阈值。

结果：
研究对象的年龄中位数为21岁（四分位间距Q1、Q3：20、23）；约半数（203名，50.8%）为男性，多数（335名，83.8%）居住于城市地区，仅有28名（7%）拥有有偿职业。物质使用的终生患病率为41.5%，酒精使用的终生患病率为36%。
针对两类使用行为，较高的神经质得分均与更高的终生使用风险相关：物质使用（调整后比值比[AOR]=1.05，95%置信区间[CI]：1、1.10；p=0.013）；酒精使用（AOR=1.04，95%CI：0.99、1.09；p=0.032）；而较高的宜人性得分则与更低的终生使用风险相关：物质使用（AOR=0.99，95%CI：0.95、1.02；p=0.008）；酒精使用（AOR=0.99，95%CI：0.95、1.02；p=0.032）。学生年龄每增加一岁，终生酒精使用的风险即升高8%（AOR=1.08，95%CI：0.99、1.18；p=0.02）。
卷烟使用的终生患病率为8.3%。较高的神经质得分（AOR=1.06，95%CI：0.98、1.16；p=0.041）与经验开放性得分（AOR=1.13，95%CI：1.04、1.25；p=0.004）均与更高的终生卷烟吸烟风险相关；而无业状态则与更低的吸烟风险相关（AOR=0.23，95%CI：0.09、0.64；p<0.001）。
其他被报告使用的物质包括：大麻28例（7%）、镇静剂21例（5.2%）、苯丙胺类（恰特草[Catha edulis]）20例（5%）、镇静催眠药19例（4.8%）、吸入剂18例（4.5%）、可卡因14例（3.5%），海洛因与鸦片各10例（2.5%）。在13名报告注射吸毒的参与者中，10名为女性，3名为男性，该差异具有统计学显著性（p=0.042）。

结论：埃尔多雷特地区高校大学生的物质使用患病率较高，且与高神经质、低宜人性的人格特质显著相关。本研究为未来研究指明了方向，未来可通过循证治疗路径深入探究人格特质的作用机制，为相关干预提供科学依据。