Digital wellbeing in the social media era: dual pathway analysis of emotional and behavioural distress indicators and public health utility
- 1Ulster University
Abstract
Social media language and digital lifestyle behaviours are promoted as scalable signals of population mental health, but predictive models are often evaluated in ways that overstate accuracy. We analysed 20,000 tweets from 72 users, labelled by whether the account belonged to a user identified as depressed, and two synthetic datasets (100,000 and 30,000 records) linking screen time, sleep and social media use with self-reported stress. Under a conventional tweet-level split, a term-weighted logistic regression appeared accurate (area under the receiver operating characteristic curve 0.835), but when whole users were held out its performance fell to near chance (0.595; 95% confidence interval 0.505-0.686), because the model largely recognised users and their topics. Depression-associated tweets were more often negative (31.7% versus 24.2%), yet sentiment alone could not identify unseen users (0.539). The two behavioural datasets gave opposite answers: strong discrimination in one (0.926), driven by short sleep, and chance in the other (0.506), so synthetic results reflect how the data were generated. At a population prevalence of 10%, positive predictive values would be 0.10-0.30. Digital distress signals should be validated across individuals, on real linked data and at realistic prevalence before informing public health surveillance or prevention.
Related
- Psychometric Properties of a 20-Item Adapted Version of the Ontario Child Health Study Emotional Behavioural Scales for Population-Based Research
- Professionals' perspectives on barriers to improving psychosocial care for adolescents with emotional and behavioural problems - a concept mapping case study.
- Social relationships, emotion regulation, and emotional wellbeing in adolescence: associations in youth with and without a migratory background.
- “We teach the students that mainstream can’t teach”: a mixed-methods examination of staff wellbeing, resources, and digital support potential in Ireland’s alternative education sector
- Effects of cognitive-behavioural therapy for insomnia on cognitive and emotional functioning in adults with insomnia: behavioural and neuroimaging evidence
BibTeX
@misc{younas2026digital,
title = {{Digital wellbeing in the social media era: dual pathway analysis of emotional and behavioural distress indicators and public health utility}},
author = {S. Younas and T. J. Anande},
howpublished = {medRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.09.26365101},
url = {https://www.medrxiv.org/content/10.64898/2026.10.09.26365101v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Digital wellbeing in the social media era: dual pathway analysis of emotional and behavioural distress indicators and public health utility
AU - S. Younas
AU - T. J. Anande
PY - 2026
DA - 2026/10/09
PB - medRxiv
DO - 10.64898/2026.10.09.26365101
UR - https://www.medrxiv.org/content/10.64898/2026.10.09.26365101v1
AB - Social media language and digital lifestyle behaviours are promoted as scalable signals of population mental health, but predictive models are often evaluated in ways that overstate accuracy. We analysed 20,000 tweets from 72 users, labelled by whether the account belonged to a user identified as depressed, and two synthetic datasets (100,000 and 30,000 records) linking screen time, sleep and social media use with self-reported stress. Under a conventional tweet-level split, a term-weighted logistic regression appeared accurate (area under the receiver operating characteristic curve 0.835), but when whole users were held out its performance fell to near chance (0.595; 95% confidence interval 0.505-0.686), because the model largely recognised users and their topics. Depression-associated tweets were more often negative (31.7% versus 24.2%), yet sentiment alone could not identify unseen users (0.539). The two behavioural datasets gave opposite answers: strong discrimination in one (0.926), driven by short sleep, and chance in the other (0.506), so synthetic results reflect how the data were generated. At a population prevalence of 10%, positive predictive values would be 0.10-0.30. Digital distress signals should be validated across individuals, on real linked data and at realistic prevalence before informing public health surveillance or prevention.
ER -