SSRIs for Skin Picking: What the Evidence Actually Shows

The two randomized trials

Fluoxetine, 1997. Twenty-one adults with chronic skin picking were randomized to fluoxetine or placebo for ten weeks, on a flexible schedule up to 80 mg a day. Seventeen finished, six on fluoxetine and eleven on placebo, at an average dose of 55 mg. The researchers used three separate measures of picking.

That second sentence matters more than the headline. Whatever fluoxetine did or didn't do to the picking in that trial, it wasn't doing it by lifting mood.

Citalopram, 2008. Forty-five people, 20 mg a day, four weeks. Against placebo, the citalopram group showed significant improvement in quality of life, general health status and obsessive-compulsive severity. Skin picking severity dropped more in the citalopram group too, but not by enough to reach statistical significance, and on the trial's own primary outcome measure, a visual analogue scale, there was no significant separation between the groups at all.

Read those two results together and a shape appears that recurs throughout this page. People felt better. The picking, measured directly, did not clearly move.

The open-label trials, which look far better

Now the other half of the literature, and the half most pages quote.

Fifteen people took open-label fluoxetine for six weeks in a 2001 study. Eight responded. Those eight were then randomized to keep taking fluoxetine or switch to placebo for another six weeks. The four who stayed on the drug held their improvement. The four switched to placebo went back to where they started.

Twenty-nine people took escitalopram for eighteen weeks in an open-label trial, at an average maximum tolerated dose of 25 mg. Nineteen completed. Roughly 45 percent met full response criteria and another quarter were partial responders.

Fourteen people took fluvoxamine for twelve weeks in an uncontrolled trial, and all fourteen improved. An open-label sertraline study reported a 68 percent response rate.

Why the two sets of studies disagree

People with skin picking get better during short trials whether or not the treatment does anything.

That is not a figure of speech. In the randomized trials Schumer's team analyzed, the inactive arms, meaning the people getting a placebo or sitting on a waitlist, improved over time with an effect size of 0.52. The authors then ran the arithmetic on what that implies for uncontrolled research: with an effect that size, an open-label trial of 25 people would come out positive about 80 percent of the time, a trial of 12 people about half the time, and a trial of 5 people a quarter of the time. Even for a treatment that does nothing at all.

Every study in the previous section was open-label except one small discontinuation phase involving eight people. Those response rates aren't fabricated. They're what this condition does in a research setting when someone is paying close attention to it.

Several things are probably happening at once. Picking waxes and wanes, and people tend to enrol in trials when it's at its worst, so the only direction left is down. Being assessed every fortnight by someone who understands the condition is itself a form of treatment. And being asked to count your own episodes changes them, an effect documented directly in this condition, where weeks of daily self-monitoring alone shifted picking measurably in a 2025 crossover trial. The same problem shapes how to read the NAC evidence for skin picking, and it runs through the whole medication picture for hair pulling too.

Why an SSRI still ends up on the prescription

Often the thing the drug is really being asked to treat isn't the picking.

In a study of 262 adults with skin picking disorder assessed in person rather than by online survey, comorbidity was the norm rather than the exception, with hair pulling, depression, generalized anxiety, ADHD and OCD all common. Those are conditions with large, well-replicated drug evidence behind them. Depression and anxiety are worth treating in their own right, entirely apart from what they do or don't do to your skin, and the relationship between anxiety and picking runs in both directions.

The Merck Manual's professional edition states the current position in two sentences: N-acetylcysteine and memantine, which act on glutamate rather than serotonin, "are increasingly considered the first-line medication treatment," while SSRIs "may be useful for coexisting depression or anxiety disorders, and limited evidence suggests that these medications can also reduce skin picking."

There's a second reason, less flattering and probably more common in practice. In that same study of 262 adults, 87 percent had never received any treatment for their picking at all. A GP with ten minutes, a distressed patient, no local therapist trained in body-focused repetitive behaviors and a prescription pad is going to reach for the thing they know. That isn't incompetence. It's the shape of what's available.

What to realistically expect

Two separate outcomes, worth tracking separately.

Mood and anxiety. This is where SSRIs have decades of large-trial evidence, and where you're most likely to notice something: sleep, appetite, the volume of the anxiety, the size that small problems seem to be.

The picking itself. Expect less. It may improve indirectly, since stress is one of the two most commonly reported triggers in this condition. It may not move at all. The 1997 fluoxetine trial found no relationship between how much someone's mood improved and how much their picking changed, which is a caution against assuming one will drag the other along.

The realistic best case is that an SSRI makes the emotional weather less extreme, which makes the behavioral work more possible. It isn't that the urge disappears. And if you pick largely out of boredom or the feel of your skin rather than distress, there is very little in this literature to suggest a serotonin drug will do much about it.

Timelines, and when to judge it

NHS guidance for SSRIs generally is that they usually need two to four weeks before any benefit is felt, and that four to six weeks with no change is the point to go back to your GP.

For the picking specifically, the trials ran anywhere from four weeks to eighteen. The citalopram trial that failed to move picking significantly was only four weeks long, short enough that it may have been testing the timeline as much as the drug. The escitalopram trial that reported the best response rate ran eighteen.

Four to six weeks tells you something about your mood and very little about your picking. Three months is a fairer test of both.

Side effects

What follows is the general SSRI picture rather than skin-picking-specific data. Trials this small can't detect anything but the most common effects, and every SSRI differs from the next.

When an SSRI makes the picking worse

A small signal in the literature deserves naming, because if it happens to you it helps to know it's been described before.

Two case reports published in 2003 documented people with OCD whose skin picking emerged or intensified after starting SSRI treatment. Case reports are the weakest form of evidence there is, and two of them prove nothing about how often this happens. The reason to know about them is practical: if your picking clearly stepped up within weeks of starting a medication or raising a dose, that's a specific, dateable observation worth reporting rather than dismissing.

The same applies in a milder form to the early weeks generally. SSRIs can increase restlessness at the start, and restlessness is fuel for a body-focused repetitive behavior.

Questions worth asking

Prescribing is your prescriber's decision. The quality of that decision depends heavily on what you bring into the room.

Worth raising specifically:

• Which goal is being treated. The picking, the depression or anxiety, or both, and how each will be judged. This often goes unstated and then gets confused later.
• What improvement would look like, and by which week. Agreed in advance, so neither of you is guessing at the review appointment.
• Everything else you take, including supplements. N-acetylcysteine is bought off a shelf, which is exactly why people forget to mention it, and it has interactions.
• Whether behavioral treatment is available alongside. Habit reversal training is a specific, learnable protocol, not an instruction to be more mindful, and it's the only approach that has beaten inactive controls in a randomized meta-analysis of skin picking treatments.
• Whether the glutamate drugs are worth discussing. NAC has the one positive randomized trial in skin picking, 47 percent much or very much improved against 19 percent on placebo. Memantine produced the largest single result in this field so far, in a trial that mixed hair pulling and skin picking.

Frequently asked questions

What to do next

The single most useful thing you can do before a medication review is arrive with a record instead of an impression. Episodes, sites, how long, and the urges you had without acting on them. That's what turns "maybe a bit better" into something a prescriber can measure a decision against, and it's the same record that would let you tell a real effect apart from the improvement everyone shows in the first few weeks of paying attention.

Logging that in SkinAware takes a few taps per entry, and it tracks resisted urges as well as episodes, which is often where the first change appears. It also walks through habit reversal training module by module, which is the part of this with the strongest evidence behind it, and there's a moderated community and accountability friends if doing it alone is the problem. If you're still working out whether what you have meets the criteria, start with the foundational overview of skin picking disorder.

Medication decisions belong to your prescriber. The data you bring to that conversation is yours.