Abstract
This longitudinal cohort study investigates the psychological and physiological toll of algorithmic 'doomscrolling'—the compulsive consumption of negative online content—on Generation Z (Gen Z) users. Over a six-month period, we monitored a cohort of 450 Gen Z participants (aged 18–25) using passive smartphone tracking technology alongside weekly self-reported assessments of sleep quality via the Pittsburgh Sleep Quality Index (PSQI) and anxiety via the Generalized Anxiety Disorder 7-item scale (GAD-7). Linear mixed-effects modeling revealed that increased engagement with algorithmically curated feeds, particularly during pre-sleep hours, was significantly associated with a decline in sleep quality (β = 0.42, p < 0.001) and a concurrent rise in generalized anxiety levels (β = 0.38, p < 0.001). Lag-sequential analysis demonstrated that spikes in doomscrolling behavior consistently preceded elevated anxiety states by 24 to 48 hours. These findings highlight the critical socio-technical challenges of the attention economy, demonstrating how algorithmic optimization for engagement exploits cognitive biases to the detriment of user well-being. The study concludes with recommendations for ethical human-computer interaction (HCI) design, advocating for platform-level 'circuit breakers' and proactive digital literacy frameworks to mitigate these pervasive digital harms.
