[Subtypes of social media addiction in adolescents and their characteristics of emotional reinforcement and attentional bias].

Objectives

To analyze the characteristics of emotional reinforcement and attentional bias across different subtypes of social media addiction in adolescents.

Methods

Using stratified cluster sampling, 2,254 adolescents from four schools in Fujian Province, were recruited, and 2,028 valid samples were retained after screening. A self-designed Social Media Use Driving Situations Questionnaire was used to measure the types and intensity of emotion-driven motivations for social media use. The Bergen Social Media Addiction Scale (BSMAS) was used to assess the severity of social media addiction, and the Positive and Negative Affect Schedule (PANAS) was used to measure trait affect. With the six BSMAS items as observed variables, latent profile analysis (LPA) was performed in Mplus 8.3 to identify heterogeneous subtypes of social media addiction. One-way analysis of variance (ANOVA), Pearson correlation, and linear regression were used to compare social media use duration and trait affect across subtypes and to examine the association between the motivational intensity of emotion-driven situations and addiction severity. From each subtype, 20 participants were randomly selected by stratified sampling. A 3 (prime type: positive/neutral/negative) × 2 (target word type: social media word/neutral control word) × 3 (group: normal/at-risk/addicted) mixed design was used in an emotional Stroop color-word test. A visual analog scale (VAS) was used to measure craving and impulsivity before and after the test. Repeated-measures ANOVA, Pearson correlation, and Spearman rank correlation were used to analyze changes in social media craving and impulsivity before and after the test and the correlations of reaction times with immediate post-task social media craving and behavioral impulsivity.

Results

LPA identified three heterogeneous subtypes: the normal group (low addiction severity, short use duration, neutral emotional drive), the at-risk group (moderate addiction severity, moderate use duration, positive emotional dominance), and the addicted group (high addiction severity, long use duration, negative emotional avoidance). The addicted group showed significantly prolonged reaction times to social media words under negative emotional priming (P<0.01), which were significantly positively correlated with cue-induced craving (r=0.625, P<0.01) and impulsivity (rs=0.567, P<0.01). The at-risk group showed significantly prolonged reaction times to social media words under positive emotional priming (P<0.05), which were significantly positively correlated with cue-induced craving (r=0.547, P<0.05). In addition, craving and impulsivity increased significantly after the test in the at-risk group (both P<0.05)and addicted groups (both P<0.01), whereas no significant changes were observed in the normal group (both P>0.05).

Conclusions

Adolescent social media addiction can be classified into three heterogeneous subtypes, which differ significantly in emotional reinforcement characteristics and exhibit corresponding emotion-specific attentional bias patterns.