What happens to your brain when you put your phone down for 90 days?

The first thing many people notice is the phantom buzz. A vibration in the pocket where the phone used to be, except the phone is sitting on the counter or on a desk across the room.

Most people who carry a phone have felt it at least once. The sensation is so common among people who give up their smartphones that researchers have a name for it. Roughly nine out of ten college students report feeling vibrations that never occurred.

It is the body expecting a signal that is not coming. And it is only the beginning.

A growing body of research now shows that putting down a smartphone after heavy use produces measurable physical withdrawal: rising heart rate, climbing blood pressure, spiking anxiety, and disrupted sleep. The responses look less like a bad habit being broken and more like a dependency being interrupted. 

Treatment centers have started borrowing the same 90-day recovery framework built for drug and alcohol addiction, and addiction specialists are increasingly pushing for smartphones to be treated with the same clinical seriousness as substances.

The medical establishment has not caught up yet. But the neuroscience is getting harder to ignore.

Augustine Recovery, an addiction treatment center specializing in alcohol and drug rehabilitation, reviewed the published neuroscience and behavioral research to examine the brain science behind smartphone use and how it compares to the substance addictions that clinicians treat every day.

The Itch that Shows Up on a Heart Monitor

In one experiment, researchers handed 40 iPhone users a series of word puzzles, then quietly removed their phones from the room. When the researchers called the phones so participants could hear them ringing but could not answer, heart rates rose. Blood pressure climbed. Anxiety scores jumped. Performance on the puzzles dropped.

The participants were not in danger. They were not being threatened. They were sitting in a lab, hearing their phone ring in the next room, and not being allowed to pick it up.

A separate experiment took devices away from 163 college students and measured anxiety three times over the course of an hour. Light users barely noticed. Heavy users showed anxiety that climbed with each measurement, getting worse the longer the phone was gone.

That pattern has a name in addiction research: dose-response. The more you use, the harder the withdrawal hits. It is the same relationship that shows up in alcohol, nicotine and opioids.

Anyone who has left their phone in the car and felt a low-grade unease for the next hour has experienced a mild version of what these studies measured. The difference is that in a lab, with monitors attached, you can see the body responding the way it would to any other threat: heart pumping faster, palms sweating, focus fragmenting. All because a phone rang and nobody could answer it.

Why the Brain Treats a Notification Like a Drug

The physical symptoms have an explanation, and it runs deeper than habit.

When researchers put 32 teenagers inside a brain scanner and showed them social media photos with lots of “likes,” the same part of the brain lit up that activates during drug use: a small region called the nucleus accumbens, which is the brain’s primary reward center. A follow-up found that giving likes and receiving them both triggered the same circuitry.

This is the same mechanism that makes a slot machine compelling. The phone delivers rewards on an unpredictable schedule. Sometimes the notification is nothing. Sometimes it is a message that makes your day. The brain cannot tell in advance which one it will be, so it stays alert, releasing a small hit of dopamine every time something new appears. Over time, the brain adjusts by dialing down its own sensitivity. It takes more stimulation to feel the same effect.

Addiction researchers call that tolerance. In everyday language, it is the reason an hour of scrolling that used to feel satisfying now feels like nothing.

The structural picture points in the same direction. Heavy smartphone users tend to show less grey matter in the parts of the brain that govern self-control and emotional regulation, a pattern that overlaps with what researchers see in substance addiction. The science comes with an important caveat: Most of these studies are snapshots, not long-term tracking, so they cannot prove the phone caused the changes. But the overlap is consistent enough across multiple research groups that scientists have stopped treating it as a coincidence.

Why 90 Days

The 90-day recovery timeline did not come from a hunch. It came from decades of federal research tracking tens of thousands of patients trying to get clean from drugs.

The pattern emerged in the 1970s and was confirmed again in the 1990s: Patients who stayed in treatment fewer than 90 days did about as well as those who never started. Only those who stayed 90 days or longer showed lasting improvement, and the results held up years later. That evidence became the backbone of federal treatment policy. The National Institute on Drug Abuse now states that treatment lasting less than 90 days is “of limited effectiveness.”

The biology lines up. After drug use stops, the brain’s dopamine system begins recovering within the first month. Mood improvements typically show up between one and three months. Full emotional stability takes three to six months. Researchers have also found that teenagers treated for compulsive phone habits showed their brain chemistry returning to normal after about nine weeks of therapy.

The reasoning is straightforward: if the brain needs roughly 90 days to start recovering from cocaine, and heavy phone use disrupts many of the same circuits, the same timeline is a reasonable starting point.

No study has proven that the 90-day model works specifically for phone addiction. But treatment programs are not waiting for the proof.

What Recovery Actually Looks Like

The first residential treatment center for digital addiction in the United States opened in 2009, on a five-acre property outside Seattle. It started with six beds and a model borrowed directly from substance abuse recovery: surrender the device on arrival, then work through therapy, life skills coaching, fitness and mindfulness training over weeks and months.

The program now serves teens and young adults aged 13 to 30, with stays ranging from the initial residential phase to extended care spanning six to twelve months. Other programs have followed. A hospital-based program opened in Pennsylvania in 2013. South Korea declared internet addiction a national crisis and built government-funded treatment camps for teenagers.

None of these programs have published independent, peer-reviewed outcome studies. What they have is a clinical framework that makes sense on paper, a growing number of patients who say it worked, and a waiting list that keeps getting longer.

The gap between the demand for treatment and the evidence supporting it is one of the defining tensions in this field. Clinicians see people whose lives have been disrupted by compulsive phone use. Researchers say the science is not yet strong enough to call it a formal addiction. Meanwhile, the patients keep showing up.

A Habit Almost Everyone Recognizes

Ask Americans whether they spend too much time on their phones and most of them will say yes. In a national survey of more than 34,000 adults, 58% said they use their phone “too much.” Among adults under 30, that figure was 81%. Half said they cannot imagine life without their smartphone. Eight in ten keep it within arm’s reach during every waking hour.

Those numbers describe a population that is aware of the problem and unable to stop. It sounds less like the way people talk about a casual preference and more like the way people talk about a habit they wish they could break.

The harder question is where to draw the line. Broad screening tools suggest roughly a quarter of people worldwide qualify as “problematic” smartphone users. But when the bar is raised to documented functional impairment, the kind that costs someone a job or a relationship, the number drops dramatically. Even by the most conservative clinical estimate, the affected population in the United States is larger than the number of Americans currently receiving treatment for opioid addiction. The debate is not about whether the problem exists. It is about how far it has already spread.

The Diagnosis That Does Not Exist Yet

The medical world is in an awkward spot. Doctors are treating it. Researchers are scanning brains and measuring withdrawal symptoms. Treatment centers are filling beds.

And the systems that formally decide what counts as a psychiatric disorder have not made room for it.

The handbook that American psychiatrists use to diagnose mental illness does not include smartphone addiction. The only behavioral addiction it recognizes is gambling. It lists internet gaming disorder as a “condition for further study,” which is the psychiatric equivalent of a maybe. The most recent update, in 2022, did not change that.

The WHO added gaming disorder to its own classification system in 2019, but drew the line there. No broader smartphone or internet addiction category exists.

Professional organizations are moving faster than the manuals. The American Society of Addiction Medicine expanded its definition of addiction over a decade ago to include behaviors, not just substances, calling it a disease “not limited to psychoactive substances.” Its next treatment manual will include a full volume on behavioral addictions. The president of the American Psychiatric Association described technological addictions in 2024 as “the new frontier in addiction medicine.”

The clinical world is treating it as real. The diagnostic books have not caught up. For the millions of Americans who feel the pull every time they try to put the phone down, the distinction may not matter much.

Methodology

This article draws on peer-reviewed neuroscience, behavioral psychology, and addiction medicine research alongside national survey data to examine the evidence for smartphone withdrawal and the clinical response to compulsive phone use.

Withdrawal symptom evidence relies primarily on two experimental studies. Clayton, Leshner, and Almond (2015), “The Extended iSelf: The Impact of iPhone Separation on Cognition, Emotion, and Physiology,” published in the Journal of Computer-Mediated Communication (DOI: 10.1111/jcc4.12109, N=40), measured heart rate, blood pressure, anxiety and cognitive performance during involuntary iPhone separation. Cheever, Rosen, Carrier, and Chavez (2014), “Out of sight is not out of mind: The impact of restricting wireless mobile device use on anxiety levels among low, moderate and high users,” published in Computers in Human Behavior (DOI: 10.1016/j.chb.2014.05.002, N=163), established the dose-response relationship between usage levels and separation anxiety. 

Additional withdrawal evidence is from Eide et al. (2018), published in Frontiers in Psychology (DOI: 10.3389/fpsyg.2018.01444, N=127, 72-hour restriction protocol), and Cheever et al. (2021), published in the International Journal of Humanities and Social Science, which measured galvanic skin response during smartphone separation.

Phantom vibration prevalence is documented in Drouin, Kaiser, and Miller (2012), published in Computers in Human Behavior (DOI: 10.1016/j.chb.2012.03.013, 89% prevalence among 290 undergraduates); Rothberg et al. (2010), published in BMJ (DOI: 10.1136/bmj.c6914, 68% prevalence among medical staff); and Lin et al. (2013), published in PLOS ONE (PMC3677878).

Brain imaging evidence draws on Sherman et al. (2016), “The Power of the Like in Adolescence,” published in Psychological Science (DOI: 10.1177/0956797616645673, N=32, fMRI); Sherman et al. (2018), published in Social Cognitive and Affective Neuroscience (DOI: 10.1093/scan/nsy051); Kim et al. (2011), published in NeuroReport (DOI: 10.1097/WNR.0b013e328346e16e, N=12, PET imaging, reduced D2 receptor availability); Horvath et al. (2020), published in Addictive Behaviors (DOI: 10.1016/j.addbeh.2020.106334, N=48, structural MRI); and Solly et al. (2022), published in Molecular Psychiatry (DOI: 10.1038/s41380-021-01315-7, meta-analysis of 15 VBM studies, 718 participants). The cross-sectional design of brain imaging studies means causation cannot be inferred from observed structural differences.

Neurotransmitter normalization after therapy is from Seo et al. (2020), published in the American Journal of Neuroradiology (PMID: 32616578, N=38, GABA-to-glutamate ratio normalization after 9 weeks of CBT in smartphone-addicted youth).

The 90-day treatment threshold traces through the Drug Abuse Reporting Program (DARP, approximately 44,000 admissions, 1969-1973), the Treatment Outcome Prospective Study (TOPS), and the Drug Abuse Treatment Outcome Studies (DATOS, 10,010 admissions, 1991-1993). The federal guideline is from NIDA’s Principles of Drug Addiction Treatment: A Research-Based Guide, Third Edition (2018). Dopamine receptor recovery timelines are from Lu, Grimm, Shaham, and Hope (2003), published in the Journal of Neurochemistry. The 90-day model has been validated for substance use disorders but not independently for digital or behavioral addiction.

National prevalence data is from Gallup (2022, probability-based panel, N=34,591). Additional context is from Pew Research Center (2025/2026 NPORS survey, N=5,022; 2024 teen survey, N=1,453). The global pooled prevalence of self-reported problematic smartphone use (approximately 27%) is from Meng et al. (2022), published in Clinical Psychology Review (498 articles, N=2,123,762). The functional impairment prevalence estimate (0.75-1.2%) is from Lu, An, and Chen (2024, PMID 39700606, N=1,989). The comparison to opioid treatment population is based on SAMHSA National Survey on Drug Use and Health data.

Treatment program information for reSTART Life (Fall City, Washington, opened 2009) and Bradford Regional Medical Center (opened 2013) is from published reporting. No independent peer-reviewed outcome evaluations were identified for either program. South Korea’s government response is from published government statistics.

DSM-5 classification is from the APA’s Diagnostic and Statistical Manual, Fifth Edition (2013) and Text Revision (2022). ICD-11 Gaming Disorder (6C51) was approved by the WHO May 25, 2019, effective January 1, 2022. ASAM’s expanded addiction definition is from its August 2011 public policy statement. The APA presidential address on technological addictions was delivered at the 2024 Annual Meeting by President Petros Levounis.