You open your phone for a quick check. Forty-five minutes later you're deep in videos you'll never remember.
Oxford made "brain rot" its 2024 Word of the Year, and a genuinely interesting study came out in late 2025 about what happens when young adults quit social media for a week. Here's what it found, along with the limitations that most coverage skipped.
What is brain rot?
Oxford University Press defines it as "the supposed deterioration of a person's mental or intellectual state, especially the result of overconsumption of trivial or unchallenging online content." The term isn't new. It appears in Thoreau's Walden in 1854.
One thing to be clear about: brain rot has no clinical or diagnostic status. The peer-reviewed review on the subject treats it explicitly as a cultural construct rather than a medical condition (Yousef et al., 2025, Brain Sciences). We previously described it as having a clinical definition. That was wrong.
Surveys put average US phone use at roughly five hours a day, though these figures come from consumer surveys with undisclosed methodology rather than research, and estimates vary between about 4.5 and 5.5 hours depending on who's asking.
The one-week detox study
In late 2025, researchers at Beth Israel Deaconess Medical Center, a Harvard Medical School teaching hospital, published results from 373 young adults aged 18 to 24 who took a one-week break from social media, tracked with objective phone data rather than self-report of usage (Calvert et al., 2025, JAMA Network Open).
Daily social media use dropped from 1.9 hours to 0.5 hours. Self-reported symptoms improved:
All figures from Calvert et al., 2025:
| Symptom | Change |
|---|---|
| Depression | 24.8% decrease |
| Anxiety | 16.1% decrease |
| Insomnia | 14.5% decrease |
For scale, Mitch Prinstein of the American Psychological Association told NPR: "It usually takes eight to 12 weeks of intensive psychotherapy to see those kinds of reductions in mental health symptoms. So if you can get those with just one week of change in behavior, wow!"
The caveats that matter
We previously wrote that this study "proved" a detox "reverses anxiety and depression." That overstated it considerably, and here's why:
- No control group and no randomization. Participants volunteered, and symptoms were self-reported. Uncontrolled pre-post designs are vulnerable to regression to the mean and expectancy effects. The senior author, John Torous, has said it "was not meant to be a treatment study but rather a methodological study."
- Loneliness didn't budge.
- Individual responses varied enormously. Benefits concentrated in participants with worse baseline depression, where the effect was much larger. Torous emphasized that "the heterogeneity of responses was tremendous." The averages aren't what everyone gets.
- Total screen time went up, not down. Overall daily screen time rose about 4.5%. We previously reported it "didn't decrease significantly," which soft-pedals the actual finding.
The substitution problem
That last result is the most useful thing in the study. People who quit social media didn't sit staring at a wall. They substituted: some toward exercise and time outdoors, and much of it toward other apps.
This is why "just use your phone less" so often fails. The underlying need, whether boredom, stimulation, or escape, doesn't disappear because one app does. It relocates.
Which suggests a different strategy: instead of trying to eliminate the reach for your phone, change what happens when you reach for it.
About "language learning rebuilds your brain"
This article used to claim that language learning delays Alzheimer's by five years, strengthens executive function, and increases gray matter in ways that improve overall cognition. Those claims need heavy qualification, and one needs removing entirely.
The dementia delay: retrospective clinic-based studies did find lifelong bilinguals showing dementia symptoms roughly 4 to 5 years later than monolinguals (Craik et al., 2010, Neurology). But prospective population studies have largely failed to replicate this. Across the literature, 7 of 8 retrospective studies found a delay while 4 of 5 prospective studies found none, and a meta-analysis found no protective effect on dementia incidence. The apparent effect may reflect education, immigration status, and socioeconomic differences rather than language itself.
Executive function: we've deleted this claim. A meta-analysis of 891 effect sizes from 152 studies found the bilingual executive-function advantage disappeared after correcting for publication bias (Lehtonen et al., 2018, Psychological Bulletin). It's now a textbook example in psychology's replication crisis.
Gray matter: a small study did find increased gray-matter density in the left inferior parietal cortex in bilinguals (Mechelli et al., 2004, Nature). We've dropped our added claim that this "improves overall cognition," which the paper doesn't show.
The most important caveat: every one of these studies is about lifelong bilingualism, not adult app-based vocabulary practice. Applying them to a phone app is a bait-and-switch. The honest source on adult second-language learning is a systematic review concluding the evidence is mixed and methodologically weak (Ware et al., 2021, Frontiers in Aging Neuroscience).
We've also removed a quote previously attributed to "Neuroscience research, Verywell Mind," which is not a valid citation.
What actually helps
The strongest evidence for improving attention specifically comes from a two-week trial that blocked mobile internet on participants' phones. Screen time fell from 314 to 161 minutes a day, objectively measured sustained attention improved, and participants spent more time socializing in person, exercising, and outdoors (Castelo et al., 2025, PNAS Nexus). Worth knowing: only about a quarter of participants maintained the block for the full two weeks.
The practical takeaways:
- Removing access beats resolving to do better. Both good studies here worked by making the apps unavailable, not by strengthening intention.
- Expect substitution, and plan for it. Decide in advance what fills the gap.
- One to two weeks is the realistic timescale, not 24 hours.
- Quality of screen use may matter more than quantity. A systematic review found that passive fast-paced consumption fatigued attention while interactive, demanding content improved it (Dewi et al., 2025).
Where ScrollScholar fits
Given the substitution finding, the design question isn't how to use your phone less. It's what your phone should ask of you when you pick it up.
ScrollScholar blocks the apps you choose and unlocks them only when you answer AI quiz questions generated from your own notes: PDFs, Word docs, PowerPoints, or photos of handwritten pages. Correct answers earn points, and points buy screen time. If you'd rather practice a language, language mode works the same way with vocabulary and translation exercises.
The honest pitch: no app reverses "brain rot," and we're not going to claim ours does. What it does is make the substitution deliberate. When the reach for Instagram happens anyway, and it will, the thing on the other side is your coursework.
It's iOS only, with a 3-day free trial.