Bidirectional Influences Between Adolescent Social Media Use and Mental Health - PROJECT SUMMARY
In the U.S., adolescents' loneliness increased 100%, from 18% to 37% from 2012 to 2018. Meanwhile,
social media use has grown sharply. In 2022, adolescents reported being constantly online twice as often
compared to 2014-15, with 95 % of U.S. teens using YouTube, 67% TikTok, 62% Instagram, and 59%
Snapchat. However, little research investigated the relationship between the two trends. Loneliness is a
risk factor for psychological and physical problems with long-term health effects related to biological
mechanisms, including stress response. Large-scale cohort studies reported that online time was
positively associated with adolescents' loneliness and that online time displaced time for in-person social
interaction with peers. In adults, in-person social interactions are associated with higher well-being than
online communication. However, we do not know to what extent this link is similar in the adolescent
population. Also, previous measures were mainly based on self-reports of ill-being and screen time.
Hence, it is now urgent to establish to what extent in-person and online social interactions predict
adolescents' loneliness as a risk factor for mental health problems, using objective indicators. Our long-
term goal is to determine which social media experiences promote adolescents' mental well-being,
defined as the presence of happiness, thriving, and flourishing. Using an intensive longitudinal mixed-
method design, biomarker-informed, we will address the following aims:
1. Quantify the short- and long- term bi-directional relationships between online/offline
interaction on loneliness and mental well-being. EMAs and trace data will be collected for 2 weeks
and individual and contextual factors will be assessed through a baseline survey. The same
participants will be followed up after 1 year to determine long-term effects on loneliness, social media
use, mental ill- and well- being, and resilience.
2. Assess the validity of the short- and long-term quantitative findings in Aim 1 by collecting
qualitative data24 on the bi-directional relationships between online/offline social interactions,
loneliness, and well-being, by running focus groups after each quantitative data analysis phase.
3. Test the feasibility and utility of a “digital biomarker” using hair cortisol concentration (HCC).
HCC has been increasingly used as an objective indicator of stress. Loneliness is experienced as a
stressor, and we will determine to what extent digital footprints and loneliness correlate with
adolescents' HCC.
1