Table 3.

Adjusted associations between scores on social media addiction, perceived social support, loneliness, and social media violence with depressive outcomes among college students in Shaanxi, China.

Variables
Model 1: Depressive symptoms (A continuous dependent variable) a	Beta b	SE	p
Problematic social media use	0.18	0.01	<0.001
Perceived social support	−0.17	0.01	<0.001
Loneliness	0.36	0.01	<0.001
Social media violence	0.14	0.03	<0.001
Model 2: Probable depression (A binary dependent variable; CESD ≥ 16) a	OR	95% CI	p
Problematic social media use	1.083	1.075, 1.092	<0.001
Perceived social support	0.967	0.965, 0.970	<0.001
Loneliness	1.24	1.23, 1.25	<0.001
Social media violence	1.22	1.19, 1.25	<0.001

^a: Linear regression models were used, the covariates included age, sex, grade, time spent using social media in the past week, primary method used to access social media, number of people known on social media networks, smoking, perceived academic stress, parental relationship satisfaction, and physical exercise time per week. ^b: Logistic regression models were used, the covariates included age, sex, grade, time spent using social media in the past week, primary method used to access social media, number of people known on social media networks, smoking, perceived academic stress, parental relationship satisfaction, and physical exercise time per week.