Research from the University of Bristol, published in Scientific Reports, shows that posting on Twitter (now called X) during the night is tied to lower mental well‑being.
The study found that nighttime tweeting accounted for almost 2 % of the variation in participants’ mental health scores—a figure similar to the effects of binge drinking or marijuana use reported in prior research.
Authors suggest that using the platform after dark can both disturb and postpone sleep, which may reduce both the quality and quantity of rest and, in turn, worsen mental health. Although the link between after‑night posting and symptoms of depression or anxiety was weaker overall than its connection to general well‑being, the association grew stronger when data were split by age and sex.
According to a 2022 YouGov survey, 74 % of UK adults keep a phone in their bedroom at night, and 26 % admit they would check it if they wake up during the night.
Findings add to growing calls for clearer rules and guidance on nighttime social‑media use. For instance, TikTok introduced a “Wind Down” feature earlier this year that shows meditation videos in the evenings to help younger users reduce scrolling.
Researchers emphasize that top‑down changes to app design—such as TikTok’s wind‑down mode—combined with education campaigns aimed at vulnerable groups could make social‑media use safer.
The lead author, Dr. Daniel Joinson, noted, “While social media is often treated like a single entity, its impact on mental health depends on the specific behaviours users engage in. Our work highlights the potential harm of one particular act: posting at night.”
“Studies like ours could inform policies that discourage risky use while encouraging positive interactions,” he added, stressing the importance of accessing real social‑media data for a deeper understanding of the platform‑mental‑health link.
The analysis used longitudinal data from 310 adults aged 18–60+ who were part of the Children of the ‘90s cohort and had agreed to share their Twitter data. The dataset comprised 18 288 tweets. Participants’ mental health was assessed at multiple points using validated questionnaires such as the Short Mood and Feelings Questionnaire (SMFQ), measuring well‑being on a continuous scale rather than a simple depressed/not‑depressed dichotomy. Only tweets posted within two weeks of a questionnaire were included.
Because the researchers accessed actual Twitter data with participants’ permission, they could precisely log the time of day each tweet was made.
Limitations noted include the sample’s demographic skew toward adult, mostly white, female respondents, and the fact that data were collected during the COVID‑19 pandemic—a period of atypical social‑media use and mental‑health dynamics.
The team plans to further investigate how patterns of emotional expression and online interaction relate to mental well‑being.