Sunday, February 25, 2018

The Abuse Potential of Gabapentinoids

I first started prescribing gabapentin in the 1990s, as part of an early attempt to see if it worked for bipolar disorder.  It was an off-label approach and did not have that indication.  At the time anticonvulsant approaches to bipolar disorder (valproate, carbamazepine) were being heavily used.  I was following a number of people who could not take lithium and on anticonvulsants and they seemed to do surprisingly well.  Gabapentin seemed to have significant advantages in terms of toxicity, it was well  tolerated by most people.  Unfortunately, it was completely ineffective for bipolar disorder and I stopped trying it almost immediately.

The next off label application that surfaced was for chronic pain.  Any psychiatrist is exposed to a number of patients with chronic pain or chronic pain and addictions, and it became apparent that it was being used successfully for chronic back pain, chronic headaches, and post herpetic neuralgia.  Over the next decade, gabapentin and then pregabalin was prescribed for chronic pain indications and people seemed to do reasonably well with it - even at relatively high doses.

At some point, physicians working in detox and the addiction field started to use gabapentinoids for chronic pain, anxiety, and withdrawal.  It is not uncommon to see patient with all of these problems who is not able to tolerate antidepressants for those symptoms or who needs more immediate relief.  In fact, in residential addiction treatment it is common to see patients come in on high doses of gabapentin for chronic back pain.   They are there for treatment of an opioid use disorder, but during that time have not escalated the gabapentin dose.

In the literature reports of gabapentin misuse have been surfacing over the past 5 years (1-7).  A large review (4) suggests that 1.3% of the treated population is at risk for gabapentinoid misuse with the number being much higher is some populations such as opioid users.  There is a report (3) that patients with opioid use disorder will attempt to augment the eurphorigenic effect of methadone in a similar way that they use benzodiazepines.  Benzodiazepine use with methadone in methadone maintenance clinics is a chronic and at times lethal problem.  There is a report from Norway (5) that gabapentinoids may be useful is reducing benzodiazepine use.  The report generally suggests that the abuse potential is low and greater for pregabalin than gabapentin.  There is an insurance database report (6) that looks at an overuse metric comparing gabapentinoids to other abused drugs.  Goodman and Brett (7) comment on the epidemiology of gabapentinoid prescribing, specifically an increase in gabapentin prescriptions from 39 million in 2012 to 64 million in 2016 with an associated doubling in the sales of pregabalin during the same period.  They attribute the increase to attempts to treat chronic pain without opioids in primary care, suboptimal non-opioid medications (acetaminophen and NSAIDs). They cite mixed evidence in clinical trials, side effects, misuse or diversion, and an excessive focus on pharmacological measures for pain as being concerns.       

Are there biases in these report?  There certainly are.  I don't have access to the full text of the most comprehensive paper (2), but I would be interested in looking at the actual numerator and denominator for their numbers and how much was based on actual pharmacovigiliance/pharmacosurveillance as opposed to case reports, case series and reports of complications.  The other issue is that all of these papers seem to come from the same publisher.  I have not encountered that before.

The only study that I could find that looked at the direct question of concomitant use of opioids and gabapentin came from Canada (8).  It studies a large group of patients on a database that records the prescriptions and looked at all opioid users that died of opioid related causes between 1997 - 2013.  The big picture is that there were a total of 2,914,971 opioid users during the study time frame and 6,745 died of opioid related causes.  Then by selection criteria they identified 1,256 cases and matched them to 4,619 controls. They defined gabapentin exposure as concomitant gabapentin use in the 120 days preceding the index date.  They also looked for a dose response relationship of gabapentin doses considered as low (<900 mg daily), moderate (900 to 1,799 mg daily), or high (≥1,800 mg daily).  They also did a comparison with nonsteroidal anti-inflammatory drugs (NSAIDs) used as an adjunctive pain medication instead of gabapentin.  Their results are summarized in the following tables excerpted directly from the article (click to enlarge):

As noted from the data and analysis, 12.3% of the controls and 6.8% of the cases were prescribed gabapentin in the 120 days, representing a 50% increased risk of death in the gabapentin treated cases.  In the case/control comparison both groups have roughly the same levels of mental illness but the case group had higher utilization of antidepressants (all types), benzodiazepines, and other drugs/CNS depressants. They were also taking substantially more high dose opioid therapy (>200 MME).  Higher dose gabapentin nearly doubled the risk. There was no added effect from NSAID use.  The authors conclude that caution needs to be exercised in deciding to prescribe this combination (opioids + gabapentin) and that if that decision is made it needs to be carefully monitored.  From my perspective I had some concerns about the controlling for benzodiazepine use in the case/control comparison and did not see any risk attributable to benzodiazepines.  The authors do cite a reference that led to FDA warnings about the benzodiazepine-opioid combination. 

Given the concerns about gabapentin why use it at all?  The main reason is that it is effective for some of the most difficult problems in medicine.  It is very difficult to see people with extreme anxiety and insomnia go for weeks without sleep and experience continuous panic attacks all day long.  When a person stops taking benzodiazepines that they have been taking for years that is a frequent result.  The same is true for people who have decided to stop drinking and suddenly have very high levels of anxiety and insomnia now that their baseline anxiety is back.  More to the point, unless something can be done to provide them with timely relief, relapse to drug and alcohol use is certain.  Finally does high levels of abuse by some patients with addictions suggest that the medication is unsafe?  It is probably safer then other medications in this population and extremely safe outside of those populations.  In either case safety depends on whether there is a physician involved or the medication is acquired from nonmedical sources.   

Standard practice with gabapentin should be to tell all patients (in addition to the usual discussion and detailed information) the following information.  I point out here that I do not prescribe pregabalin:

1.  Take this medication exactly as it is prescribed.
2.  Do not accelerate the dose of the medication.
3.  Do not mix this drug with alcohol or any other intoxicants or street drugs.  If a relapse occurs call to discuss and set up a plan as soon as possible.
4.  Do not stop the medication abruptly it needs to be slowly tapered.  There is a seizure risk if it is not.
5.  This medication is potentially addictive to some people. If you notice any tendency to take more of this medication than prescribed contact me immediately.
6.  This medication is monitored on the state Prescription Monitoring Program and all prescriptions are recorded even though it is not technically a scheduled drug at this time.

At least that is the way that I think it should be handled.  If I was still seeing a lot of patients with chronic pain on moderate to high doses of opioids I would add in a line or two about the the Canadian study (9) and greater chances of death from the gabapentin + opioid combination.  In my current practice, psychiatric treatment is split off from buprenorphine detox and maintenance treatment - but I still see a lot of patients on buprenorphine + gabapentin and can attest to the fact that in a controlled environment we have not observed the complications suggested by the Canadian study over a period of months.  None of these patients receive benzodiazepines or sedative hypnotics beyond a period of detox.  In fact, doing that study might be a significant contribution to the research.  It also probably means that those patients when they are discharged should hear that the risk of taking that combination may increase substantially in the outpatient setting.

There is plenty of politics and confusion surrounding the gabapentinoids issue.  It should not be surprising that this medication is showing up in the toxicology of opioid overdose victims. It should not be surprising that some people try to get "high" on it, even though the people doing that do not have typical ideas about the utility of medications. It should not be surprising that people try to use gabapentin like benzodiazepines to augment the effect of what they are using to get high especially opioids.  It should not be surprising that when some people decide to stop buprenorphine or methadone that they will buy somebody's gabapentin to try to treat withdrawal effects.  It should not be surprising that in some areas it is currency on the street (What can I get for a month's worth of gabapentin?).  It should not be surprising that it has become a political football in the social media on pain ("See it's not opioids as the problem - it is gabapentin") or the social media on weed ("See these are Big Pharma solutions, marijuana is much safer").  It should not be surprising that you can read about it on drug culture web sites where everyone is an expert pharmacologist and provides you with anonymous advice on how to get high. It should not be surprising that you can buy it online and have it delivered to your door, although you can never really be certain that it is the same stuff you get at Walgreens.  I am always amazed at how easy it is to sell some Americans drugs, if they think there is the slightest possible chance they can get high on it. That is also why it should not be surprising that children and teens will take it out of medicine cabinets - use it to get high and brag about it even though there were probably not high at all.

The features about the gabapentinoids that make sense to me is that they are medically useful  and have low toxicity, for people with nearly impossible problems in desperate situations.  It is a less toxic drug on the street than those mentioned in the above paragraph. Even then these drugs need to be carefully prescribed and closely monitored.  And even then some people will escalate the dose. There are no perfect solutions in medicine and in this area in particular - nothing seems to be coming down the pike.  The probability statement is always - does the use of gabapentin result in more people with improved symptoms, better quality of life and less addiction?  At this time unless presented with compelling evidence I would say that it does with the qualifier that its application needs to be carefully done by a physician who knows what they are doing and is aware of the potential for misuse. In  the current era, that can all be subject to the next social media fad.

There is not a big push by the pharmaceutical industry at this time to discover a drug that has limited toxicity that can be used for severe chronic pain, insomnia, and anxiety associated with addictive disorders.  There is also the question of what medications are being used for these problems if not gabapentin.  The answer is atypical antipsychotics (mostly quetiapine), hydroxyzine (a first generation antihistamine), and clonidine (primary use is hypertension and opioid withdrawal).  If the comprehensive toxicology of overdoses is available I would expect to see these compounds listed.  In any search of drug interactions both quetiapine and hydroxyzine are flagged as potentially affecting cardiac conduction. Clonidine can cause hypotension if used excessively and rebound sympathetic symptoms (tachycardia, hypertension, diaphoresis).  Looking at that group of medications gabapentin would appear to have the preferred side effect profile.

There also appears to be a big push to make gabapentin a controlled substance according to the Controlled Substances Act (CSA).  Pregabalin is currently a Schedule V drug (see page 14) or considered to have the lowest abuse liability.  Getting on that list depends on how the DEA currently sees the addictive behavior towards gabapentin versus pregabalin.  Putting a drug on Schedule V will probably have no impact on how it is used in medical practice or out on the street.  The fact that pregabalin is ranked so lowly is a sign of regulatory opinion on abuse liability.

That's my current opinion on the topic.  I may add more to this post in the future or to a post I am working on about the basic science of gabapentinoids.

George Dawson, MD, DFAPA


1:  Schifano F. Misuse and abuse of pregabalin and gabapentin: cause for concern? CNS Drugs. 2014 Jun;28(6):491-6. doi: 10.1007/s40263-014-0164-4. Review. PubMed PMID: 24760436.

2:  Chiappini S, Schifano F. A Decade of Gabapentinoid Misuse: An Analysis of the European Medicines Agency's 'Suspected Adverse Drug Reactions' Database. CNS Drugs. 2016 Jul;30(7):647-54. doi: 10.1007/s40263-016-0359-y. PubMed PMID:27312320.

3:  Baird CR, Fox P, Colvin LA. Gabapentinoid abuse in order to potentiate the effect of methadone: a survey among substance misusers. Eur Addict Res. 2014;20(3):115-8. doi: 10.1159/000355268. Epub 2013 Oct 31. PubMed PMID: 24192603.

4:  Evoy KE, Morrison MD, Saklad SR. Abuse and Misuse of Pregabalin and Gabapentin. Drugs. 2017 Mar;77(4):403-426. doi: 10.1007/s40265-017-0700-x. Review. PubMed PMID: 28144823.

5: Smith, R. V., Havens, J. R., and Walsh, S. L. (2016) Gabapentin misuse, abuse and diversion: a systematic review. Addiction, 111: 1160–1174. doi: 10.1111/add.13324.

6: Bramness JG, Sandvik P, Engeland A, Skurtveit S. Does Pregabalin (Lyrica(®) ) help patients reduce their use of benzodiazepines? A comparison with gabapentin using the Norwegian Prescription Database. Basic Clin Pharmacol Toxicol. 2010 Nov;107(5):883-6. doi: 10.1111/j.1742-7843.2010.00590.x. PubMed PMID: 22545971.

7: Peckham AM, Fairman KA, Sclar DA. Prevalence of Gabapentin Abuse: Comparison with Agents with Known Abuse Potential in a Commercially Insured US Population. Clin Drug Investig. 2017 Aug;37(8):763-773. doi: 10.1007/s40261-017-0530-3. PubMed PMID: 28451875.

8: Goodman CW, Brett AS. Gabapentin and Pregabalin for Pain - Is Increased Prescribing a Cause for Concern? N Engl J Med. 2017 Aug 3;377(5):411-414. doi: 10.1056/NEJMp1704633. PubMed PMID: 28767350.

9: Gomes T, Juurlink DN, Antoniou T, Mamdani MM, Paterson JM, van den Brink W.  Gabapentin, opioids, and the risk of opioid-related death: A population-based nested case-control study. PLoS Med. 2017 Oct 3;14(10):e1002396. doi: 10.1371/journal.pmed.1002396. eCollection 2017 Oct. PubMed PMID: 28972983; PubMed Central PMCID: PMC5626029.

Graphics Credit:

Figure 2 about is excerpted directly from the work in reference 8 above per the Creative Commons Attribution License.  The authors are listed as the copyright holders.


Publication from the above content?  If you are a psychiatrist or pharmacologist and think you can rework the above article into a publication.  Contact me and let's write that paper!


  1. Patients I have spoken to working recovery tell me the average dose for abuse is at least 1200mg at a time. I met one pt who took 2400mg and apart from severe sedation and most likely nystagmus-induced gait dysfxn for a day, survived it.

    Your experiences with dosages, if you feel comfortable noting it in print?

    Joel H

    1. You are correct - I am told about very high doses usually in polysubstance users or people trying to mitigate coming off one drug and going on to another. Various compounds are used for this including hydroxyzine, diphenhydyramine, quetiapine, benzodiazepines, and various muscle relaxants.

      You are also correct that I can't post the amounts here. There are far too many people out there who see themselves as their own irrational psychopharmacologist and would consider a large dose listed somewhere as a challenge. I would rather they see it in the chat room power stories on the drug user web sites.

      I did have a patient on gabapentin 600 mg TID become acutely delirious (confusion, sedation, ataxia, nystagmus, incoherent speech). After a stat evaluation I discovered that he had gone onto acute renal failure from starting a statin and stopped excreting gabapentin. I sent him to medicine where he recovered uneventfully.

      Publicly available data from Toxnet documents the low toxicity of the compound:

  2. Thanks for the feedback, I am sorry for what I wrote earlier, and appreciate you taking the comment as forwarded. Just for any readers interest here, I once worked as a consultant at a rehab over 10 years ago, and had a patient ask me how much Lexapro he could take to achieve some form of high. My response was to the point, basically saying "one, a dose what would cause much more side effects and risk for consequences well beyond the brief time achieving what you frame as a high, and two, why are you here under what seems a pretense for recovery when you are seeking abuse of substances intended solely for treatment?!" He left the next day, as he actually went back to the next group session in the program and admitted what he asked to me, what I responded with, and then told the group that he has the sole right to seek recovery when he wants it, not when others demand it.

    And thus why recovery is about voluntary choice. Am I one of the few and far between psychiatrists treating patients who are at least equally seeking out mental health for primary addiction issues who are tired of these damn court referrals to community mental health clinics for people who just aren't interested in honest recovery??

    Again, sorry to take up Dr Dawson's site for my rantings, but, when idiots in the legal system present treatment in a punitive manner, what the hell is to be accomplished here?!

    Again, completely venting, but, there are times when I have frankly antisocial drug abusing cretins trying to manipulate, if not intimidate me for controlled substance Rxs, and ask what can I give them to make them feel good, I just wish I could be a complete jerk and reply "well, I could advise cyanide, but, that would really kill the moment!"

    But, Won't be a complete jerk, risk losing my license, so just take the time here to note what is the real moment we might want to have in the office. Kinda reminds me of the beginning of "Analyze This" with Billy Crystal who imagines what he wanted to say to the rambling woman who doesn't have the capacity to hear the truth...

    Joel H

    1. Treating people with substance use disorders is tough work because it involves limit setting around medications that other physicians are unlikely to do. That is tough and it is also an opportunity to convince the person to try it a different way.

      Responses do vary based on population and you can end up in a tough position if the staff you work with are ineffective in maintaining a recovery environment. That does take the right mix of people and I agree, court ordering a large group of very young people with substance use problems who spend the entire time talking about the positives of substance use is a waste of time.

      Settings like that also present the risk of a much greater likelihood or active use of contraband on site and all the associated risk. It is very difficult in those circumstances to recommend high quality psychiatric care since those attitudes end up in consultation sessions and evidence themselves as requests/demand for controlled substances especially stimulants, not showing up, and in the extreme fabricating statements about what happened during the office visit.

      I have had that conversation about taking a drug to feel good many times. They generally recognize that as transient and associated with very negative phenomena with repeat episodes. The old George Koob - positive-negative reinforcement cycle + dysphoria phenomenon.

  3. Heard this one recently?

    As I wrote at my own blog earlier today, not lunacy, just disgusting.

    Joel H

  4. "Treating people with substance use disorders is tough work because it involves limit setting around medications that other physicians are unlikely to do."

    That's putting it mildly. It's limit setting around medications that other physicians continue to prescribe without the input of any social history. And which family members encourage.

    I keep running into this issue but for me it's more about Soma and the muscle relaxants. Of course, they're also taking Xanax and an opioid so it's a mess. But that's just the beginning. I saw a young woman with depression on Xanax, Soma, opioids, Prozac and Ritalin (all from a hometown GP) in combination to help with performance at college. That's basically driving down the freeway with the brake and gas pedals fully floored at the same time, while Mom and Dad cheer away from the back seat.

    1. Have definitely seen the muscle relaxant problem - when they don't work for purported spasm - benzodiazepines are added. Nothing is ever stopped and then the stimulants are added to counter polypharmacy induced fatigue.

      Some associated patterns:

      1. No pain relief from anything - Is there anything else you can give me for spasm?
      2. Insomnia - Can I get Ambien with the Soma and Xanax?
      3. Anxiety and depression of course.
      4. ADHD - the patient did well in school, college and in some cases professional school but was now "diagnosed" with ADHD by some extensive neuropsych battery or neurometric testing.

      So limit setting in these cases, including the necessary medication revisions and discontinuations is yeoman's work. Parental involvement is another variable. They may have been sold on polypharmacy and believe all of the medication is helpful. I have attended seminars where the heritability of ADHD was emphasized and know that some clinic approaches involve treating the entire family if an index case is found. True disability from a disorder is never really considered.

  5. The popularity of Soma is the stealth return of Miltown, which is its metabolite. So now there's the return of carbamates with all the mental and physical depression in addition to the CYP450 enzymatic effects on the other drugs, making for a real mess.

    For readers unfamiliar with the history of downers, Miltown was the "Mother's Little Helper" of the sixties, replaced by the benzo Valium which was safer.

    1. I'm familiar with Miltown, but I thought valium was the "Mother's Little Helper" in the Stones song, since it mentions "a little yellow pill." Don't recall what color miltown was, but valium was out then if I recall.

  6. According to wiki, we may both be right:

    My sense was that Miltown was more sixties and Valium was more seventies in terms of general popularity.

    The song came out in 1965 and Valium released in 1963. Valium peaked a few years later.

  7. Unfortunately, Miltown seemed to go the way of the Dodo bird, it is a shame that Valium hasn't been forced into extinction as of 2018? Really, what psychiatrists are writing for diazepam in this decade if not providing coexisting neurological care, I get the spasm impact it has when legitimately indicated. That is not my expertise in my office, and doubtful for what, 80% or more of colleagues who are just board certified in psychiatry??

    Joel H

    1. I would not leave Quaaludes (methaqualone) out of that ugly mix. Wiki describes it more of a "club drug" of the 60s and 70s. I met a young man travelling through Africa who said that taking it was the best way to travel in adverse circumstances.

      All three point out the ongoing history that people will take something that is essentially unknown or dangerous in order to alter their conscious state.

      I am sure that at times people ask for Qualuudes and got Miltown, diazepam, diphenhydramine or whatever. That should be an expected result even today.