Gabapentin is the drug American psychiatry is in a quiet relationship with that nobody really wants to talk about publicly. It's prescribed constantly.
Sections
Gabapentin is the drug American psychiatry is in a quiet relationship with that nobody really wants to talk about publicly. It’s prescribed constantly. It’s prescribed for things it wasn’t approved for. It mostly works for those things, kind of, sometimes. And the evidence base for most of what we use it for is, if we’re being honest, not great. It’s still everywhere.
It was approved decades ago as an antiepileptic, then later for postherpetic neuralgia (the nerve pain that hangs around after shingles), then it just kept showing up in other clinical situations and people kept prescribing it. Anxiety. Insomnia. Alcohol withdrawal. Restless legs. Hot flashes. Hot flashes. It’s the duct tape of psychiatry. Not particularly elegant. Holds things together.
What we actually know
For neuropathic pain (nerve pain, the burning shooting kind from diabetes or shingles or a pinched nerve), the evidence is decent. For seizures, it works for some kinds and not others, and there are better options for most. For everything else, the data is mostly mediocre studies, small samples, and a lot of clinical experience saying it helps some patients sometimes, which is the kind of evidence base you take with one eyebrow raised.
For anxiety, there’s a small but real signal in social anxiety. For generalized anxiety, the data is weaker than SSRIs (selective serotonin reuptake inhibitors, the modern first-line antidepressant class, things like Lexapro and Zoloft) but better than nothing. For alcohol withdrawal, there’s actually pretty good data and some addiction specialists use it as a primary tool. For sleep, the data is mediocre, but a lot of patients find it sedating and that solves their sleep problem at least for a while.
Mechanism: it doesn’t actually act on GABA receptors despite the name, which is one of pharmacology’s funnier jokes. It modulates calcium channels in a way that reduces neurotransmitter release. Nobody fully understands why this helps anxiety. The honest answer is we have a drug that works for some things, we sort of know why for some of those things, and for the rest we shrug and call the prescription in.
The doses are weird
Gabapentin’s bioavailability (how much of the pill actually makes it into your bloodstream) is dose-dependent and gets worse at higher doses. At 100mg, you absorb most of it. At 1200mg, you absorb maybe half. This is why the doses sound huge to people who know other psych meds. 300mg three times a day is a normal starting dose. 600mg three times a day is common. Some patients end up at 3600mg a day. It’s not that they’re tolerating it badly, it’s that you have to take more to get the same blood level, and the math gets steeper as the dose climbs.
Most guys start at 300mg at bedtime, see how the sleep goes, then build up from there if needed. For anxiety, 300mg three times a day. For alcohol withdrawal we use it scheduled and higher, and for short windows.
The dirty side
Gabapentin has an abuse profile the field is slowly admitting. At high doses, particularly in people with a substance use history, particularly in combination with opioids or benzos, it can produce a kind of dissociative high. The DEA hasn’t scheduled it federally but several states have it as a controlled substance now. In Oregon and Washington it’s still unscheduled, but pharmacies and prescribers are paying more attention to who’s filling what and how often.
The other dirty part is that a lot of people who get it for one thing stay on it forever. It’s not addictive in the way benzos are, but coming off it after long-term use isn’t fun. Headaches, anxiety, insomnia, sometimes for weeks. We taper, slowly, slower than most patients want. The taper isn’t dramatic per dose, but it’s real, and rushing it produces a couple of bad weeks that scare people back onto the medication.
It also makes you stupid at high doses. Cognitive slowing, word-finding problems, the kind of brain fog patients describe as feeling underwater. If a patient has a desk job that requires sharp thinking, we keep the dose low or pick something else. The pattern is sometimes the patient who took it for anxiety and got their anxiety dialed back at the cost of feeling like they’re two seconds slow at work, which is the trade-off worth naming up front instead of letting them figure it out three months in.
The pattern of who ends up on it
The kind of guy who comes in already on gabapentin from somebody else is often the long-time drinker who finally quit, white-knuckled his way through the first six months on willpower and naltrexone, and then ran into the anxiety that had been sitting under the alcohol the whole time. The drinking was self-medicating it for two decades. The drinking is gone, the anxiety is raw and all-the-time, the Lexapro helped some but not enough, and somewhere along the way a doctor added gabapentin 300mg three times a day and the evenings got manageable. He can sit through a movie with his wife without his chest pounding. He sleeps. He’s been on it three years. He’d like to come off it. He’s tried twice, and twice the anxiety came back hard enough that he picked the medication over the discomfort. He might try again. He might stay on it forever. Both are real options, and the calculus isn’t to chase him off a medication that’s letting him stay sober and married just because the data isn’t elegant.
It’s the duct tape of psychiatry, not particularly elegant, but it holds things together for a lot of patients who’d otherwise be on something worse.
Nerve pain, partial seizures, alcohol withdrawal
FDA-approved indications. Everything else (anxiety, sleep, hot flashes) is off-label and clinical-experience driven.
300mg-3600mg, divided
Absorption drops as the dose climbs. 100mg absorbs almost fully, 1200mg absorbs half. That’s why the totals look enormous.
Brain fog, abuse pattern, taper required
Real abuse potential with opioid or benzo on board. Cognitive slowing at higher doses. Don’t stop cold after long-term use.
What’s nice to hear about this one
For the right patient, gabapentin earns its keep. The white-knuckling-it-after-quitting-drinking patient who finally sleeps. The chronic nerve pain guy who got off opioids and needed something for the burning in his feet. The patient with the specific kind of anxiety that lives in the chest and won’t quit, who couldn’t get there on Lexapro alone, who adds 300mg twice a day and finds he can sit on the couch in the evening without his hands clenching. It’s not glamorous. The drug is honest about what it is. Cheap, generic, available at every pharmacy, easy on the liver, doesn’t tank the libido, doesn’t gum up the sleep architecture the way benzos do. There are worse drugs in everybody’s medicine cabinet, including a lot of the ones with bigger marketing budgets.
When this prescriber won’t write it
Active opioid use. The combination raises overdose risk significantly. Active benzo dependence, same reason… two CNS depressants stacked is asking for a respiratory problem. Patients with a clear pattern of escalating dose without a medical reason. Patients who are very explicit about wanting it for the high.
If you’ve been on 3600mg from a previous prescriber and you’re sharp and stable and have been for years, fine. If you’re asking to start at 3600mg, we’re going to have a different conversation. The starting point isn’t the maintenance dose, and a request to skip the titration is information about something other than the pharmacology.
On the autonomy piece
Plenty of patients want to keep taking gabapentin and don’t want a lecture about whether the evidence base for the specific thing they’re taking it for is elegant. That’s fair. If the drug is helping, you’re not escalating, you’re not stacking it with respiratory depressants, and the labs and the mental status are clean, the right move isn’t to pull it because it doesn’t look great on a slide deck. The work is in the room, not in the academic literature. I’m a provider, not a parent… if the drug is working and you want to keep taking it, keep taking it. Disapproving yes, sometimes, when I’d have voted against starting it, but the appointment isn’t mine.
The flip side of that autonomy is the conversation that does need to happen, which is the honest one about what the drug is. Patients who think they’re on a clean, low-risk, totally benign anxiety pill should know that the abuse profile is real, that the discontinuation is real, that the cognitive cost at higher doses is real. Not to scare anybody. Just so the decision is informed instead of vibes-based.
Bottom line
Gabapentin is a useful drug with a soft evidence base and a real but underestimated abuse pattern. It helps a lot of patients. It also gets handed out like candy in some clinics and that’s a problem. If you’re on it and it’s working and you’re stable, fine. If you’re starting it, know that the goal is usually short to medium term, the dosing is weird because absorption drops as the dose climbs, you’ll absorb less if you take more, and getting off it later is harder than getting on it. Like most things in this work, if a prescriber’s writing it for you, it’s because the next-best option was worse, not because the medication is some clean shining tool.
Sources
- Mason BJ, Quello S, Goodell V, et al. Gabapentin treatment for alcohol dependence: a randomized clinical trial. JAMA Intern Med. 2014;174(1):70-77. PMID 24190578.
- Pande AC, Davidson JR, Jefferson JW, et al. Treatment of social phobia with gabapentin: a placebo-controlled study. J Clin Psychopharmacol. 1999;19(4):341-348. PMID 10440462.
- Smith RV, Havens JR, Walsh SL. Gabapentin misuse, abuse and diversion: a systematic review. Addiction. 2016;111(7):1160-1174. PMID 27265421.
