The average adult in the United Kingdom spends approximately 3 hours and 40 minutes per day on their smartphone, with social media accounting for roughly two-thirds of that total. Over the course of a year, that amounts to more than 55 full days — nearly two months — spent scrolling, swiping, and tapping through algorithmically curated content. And yet, despite the staggering scale of this time investment, surprisingly little clinical research has examined what happens when it stops.
That changed in early 2025, when researchers at King's College London published the results of a 90-day social media abstinence trial — one of the longest and most rigorously designed studies of its kind. The findings have significant implications for anyone concerned about the relationship between digital behaviour and psychological wellbeing.
The Study Design
The trial enrolled 640 participants between the ages of 18 and 55, all of whom reported using social media for at least two hours per day. Participants were randomly assigned to one of three groups: complete social media abstinence for 90 days, a reduced-use group limited to 30 minutes per day, and a control group with no restrictions.
Crucially, the study went beyond self-reported wellbeing measures. Participants underwent neuropsychological testing at baseline, 30 days, 60 days, and 90 days, including standardised assessments of sustained attention, working memory, emotional regulation, and sleep quality. A subset of 120 participants also provided salivary cortisol samples and underwent functional brain imaging.
The First Two Weeks: Withdrawal Is Real
The most consistent finding across the abstinence group was a clearly defined withdrawal period during the first 10 to 14 days. Participants reported increased restlessness, difficulty concentrating, and a persistent urge to check their phones — even when they knew the relevant apps had been deleted. Some described the experience as comparable to quitting caffeine, with similar patterns of irritability and mental fogginess.
"The withdrawal response was remarkably consistent across participants and strongly suggests that habitual social media use produces genuine neurochemical dependency, not merely a psychological habit."
— Dr. Sarah Linden, Department of Psychology, King's College London
Neuroimaging data from the subset group confirmed what participants were reporting subjectively. During the first two weeks, the abstinence group showed increased activity in the anterior cingulate cortex — a brain region associated with conflict monitoring and craving — consistent with the neural signature of other withdrawal states.
Weeks Three Through Six: The Cognitive Rebound
After the initial withdrawal period subsided, a different pattern emerged. By the third week, the abstinence group began outperforming both the reduced-use and control groups on measures of sustained attention and working memory. By week six, the gap had widened considerably.
On the Continuous Performance Task — a standard measure of sustained attention — the abstinence group showed a 23% improvement from baseline, compared to 9% for the reduced-use group and no significant change in the control group. Working memory scores, measured by the N-back task, showed a 17% improvement in the abstinence group.
Key Finding: Attention Restoration
The researchers hypothesise that social media's constant stream of novel stimuli trains the brain's attention systems toward rapid context-switching at the expense of sustained focus. When this stimulus is removed, the attentional system gradually recalibrates toward deeper, more sustained engagement. This recalibration appears to require approximately three to four weeks of abstinence — significantly longer than most "digital detox" challenges, which typically last only 7 days.
The Emotional Shift
The emotional effects were equally striking, though they followed a different timeline. While attention improvements appeared relatively quickly, meaningful changes in mood stability and emotional regulation didn't emerge until weeks five through eight.
By the 60-day mark, the abstinence group reported a 34% reduction in anxiety symptoms (measured by the GAD-7 scale) and a 28% reduction in depressive symptoms (measured by the PHQ-9). Salivary cortisol levels — a biological marker of stress — showed a significant reduction in the abstinence group, with average morning cortisol dropping by approximately 18% from baseline.
The researchers attribute this delayed emotional improvement to the gradual restoration of what they call "self-referential stability" — the brain's capacity to maintain a consistent sense of identity and self-worth without constant external validation. Social media, by design, provides a continuous stream of social comparison cues that keep the brain's self-evaluation systems in a state of chronic activation. When these cues are removed, the system eventually settles into a more stable baseline.
Social Connection: The Paradox
Perhaps the most counterintuitive finding was that participants who completely abandoned social media reported feeling more socially connected, not less. At the 90-day mark, the abstinence group scored significantly higher on measures of relationship satisfaction, perceived social support, and loneliness (in the positive direction) than either comparison group.
Qualitative interviews revealed the mechanism. Without the passive social monitoring that social media enables — scrolling through others' updates without directly interacting — participants were prompted to initiate more direct, intentional social contact. Phone calls increased. In-person meetups became more frequent. The quality of social interaction shifted from broad-but-shallow to narrow-but-deep.
What Happened After the 90 Days
A follow-up assessment conducted three months after the trial's conclusion found that the majority of abstinence group participants (68%) had returned to some social media use — but at dramatically reduced levels. Average daily social media time in this group had dropped from a pre-trial baseline of 2.4 hours to just 38 minutes. Most participants reported implementing structured boundaries: specific times of day for checking social media, removal of apps from home screens, and the use of screen-time tracking tools.
The minority who maintained complete abstinence (32%) reported the highest sustained improvements across all measured outcomes.
Implications
The King's College study is not a call for universal social media abandonment. The researchers are careful to note that social media provides genuine value for many users, particularly those in geographically isolated communities or those managing specific health conditions where online support groups play a critical role.
What the study does demonstrate, with considerable rigour, is that the default level of social media consumption in most Western adults produces measurable cognitive and emotional costs — costs that are largely invisible because they accumulate gradually and because the brain adapts to the impaired state as its new normal. A sustained break of sufficient duration allows those costs to become visible by contrast, and provides the cognitive and emotional space to make more intentional choices about digital behaviour going forward.
This article has been reviewed by the Wellness Research editorial board for scientific accuracy. It is intended for informational purposes and does not constitute medical advice.