Skip to main content
NCBI home page

This is a preprint.

It has not yet been peer reviewed by a journal.

The National Library of Medicine is running a pilot to include preprints that result from research funded by NIH in PMC and PubMed.

medRxiv logoLink to medRxiv
[Preprint]. 2026 Jan 15:2026.01.13.26344031. [Version 1] doi: 10.64898/2026.01.13.26344031

Comparative effectiveness of gabapentin and pregabalin on reduction in alcohol use: A nationwide observational cohort study

Tommy Gunawan 1,2, Joshua C Gray 1, Mingjian Shi 1,2, Thomas Wingo 3, Aliza P Wingo 4, David A Fiellin 5, John Tazare 6, Mehdi Farokhnia 7,8, Lorenzo Leggio 7,9,10,11, Henry R Kranzler 12, Christopher T Rentsch 5,6,13
PMCID: PMC12870570  PMID: 41646699

Abstract

Background:

Gabapentin and pregabalin have potential utility for treating alcohol use disorder (AUD), but their comparative effectiveness in reducing alcohol consumption in real-world settings is unknown. We compared changes in alcohol consumption associated with gabapentin or pregabalin treatment with those of matched unexposed comparators.

Methods:

We identified patients who were prescribed gabapentin or pregabalin for ≥60 days for any indication between 1 January 2009 and 30 June 2022 using electronic health records from the Veterans Aging Cohort Study (VACS-National). Alcohol consumption was measured using routinely-collected Alcohol Use Disorders Identification Test-Consumption (AUDIT-C) questionnaires. We used propensity score matching to balance baseline characteristics between groups. Three comparisons were conducted: gabapentin vs. unexposed, pregabalin vs. unexposed, and pregabalin vs. gabapentin. Changes in AUDIT-C scores were estimated using multivariable difference-in-difference linear regression models. Analyses were stratified by baseline AUD diagnosis and AUDIT-C categories (lower-risk, at-risk, hazardous/binge).

Results:

We identified 592,957 gabapentin initiators, 14,923 pregabalin initiators, and 2,959,006 eligible unexposed comparators who were eligible for matching. Compared to unexposed individuals, patients who received gabapentin (DiD: 0.09, 95% CI 0.06, 0.11, p<0.0001) or pregabalin (DiD: 0.14, 95% CI 0.02, 0.26, p=0.0279) reported greater reductions in AUDIT-C scores. When compared head-to-head, pregabalin initiators reported greater reductions in AUDIT-C scores than gabapentin initiators with the largest difference observed among those with AUD (DiD: 0.86, 95% CI 1.22, 0.50, p<0.0001) and those who report baseline hazardous/binge drinking (DiD: 1.74, 95% CI 2.21, 1.27, p<0.0001).

Discussion:

In this large, nationwide cohort, treatment with gabapentin and pregabalin were associated with reductions in reported alcohol consumption, compared to matched unexposed comparators. Initiation of pregabalin was associated with greater reductions than with gabapentin, particularly among those with AUD and those with highest severity of alcohol use. Known safety concerns and risk of misuse should be considered when prescribing these medications. Randomized clinical trials directly comparing these medications are necessary to validate these findings.

Introduction

Globally, alcohol use disorder (AUD) affects 400 million individuals, with 2.6 million deaths attributable to alcohol consumption 1. Over 28 million U.S. adults had past-year AUD in 2023, and excessive alcohol consumption costs approximately $250 billion in 2010 2,3. In addition to behavioral treatments 4, three medications – disulfiram, acamprosate, and naltrexone - have been approved by the U.S. Food and Drug Administration for the treatment of AUD 5, though, their number and effectiveness are suboptimal 6–8. Drug repurposing is a potential strategy to address the need for novel AUD pharmacotherapies 9–13.

Recently, our group developed a genetically informed drug-repurposing pipeline that integrates genome-wide association study (GWAS) data with network-based methods and pharmacologic resources to identify drug repurposing candidates for AUD 14. Using the pipeline, we identified calcium channel subunit genes CACNA1C, CACNA1D, CACNB2 and CACNB3 as alcohol-related molecular targets. A recent study integrated brain proteogenomic data with AUD GWAS data identified CACNA2D1 as a candidate causal protein for AUD 15. Together, these findings converge on calcium channel signaling as an underlying mechanism driving alcohol use risk and suggests that drugs targeting these channels may hold therapeutic potential for AUD. Gabapentin and pregabalin represent two drugs targeting these channels and for which preexisting literature suggests a potential pharmacotherapeutic role for AUD.

Gabapentin, a gabapentinoid approved for postherpetic neuralgia, partial seizures, and restless leg syndrome 16–18, has been investigated for AUD and used off-label for the treatment for alcohol withdrawal syndrome 19 and to reduce drinking in individuals with AUD 20. Gabapentin binds to the α2δ subunit of voltage-gated calcium channels 21. Some randomized clinical trials (RCTs) of gabapentin for treating AUD showed evidence that the drug reduced alcohol craving, number of drinking days and heavy drinking days, and increasing abstinence rates relative to placebo 22–28. However, other studies, including a large multi-site RCT that tested an extended-release formulation of gabapentin 29, failed to show an effect 30. A meta-analysis of seven RCTs comparing gabapentin to placebo showed a medium effect size of gabapentin reducing the percentage of heavy drinking days (Hedge’s g = −0.64), with smaller, non-significant effects on abstinence, relapse to heavy drinking and drinks per day 20. The American Psychiatric Association also recommends gabapentin (as well as topiramate) as second-line medication for AUD 5.

Pregabalin, also a gabapentinoid active at the α2δ subunit of voltage-gated calcium channels 21, is approved for the treatment of neuropathic pain, partial onset seizures, and fibromyalgia 31–34. Three open-label trials evaluating the efficacy of pregabalin for alcohol dependence or AUD showed that patients treated with pregabalin experienced a reduction in alcohol craving, withdrawal symptoms, and heavy drinking days 35–37. A single-blind RCT also showed that individuals prescribed pregabalin had higher rates of abstinence compared to those prescribed tiapride or lorazepam 38. However, evidence from double-blind RCTs is mixed. Pregabalin showed no effect on alcohol withdrawal symptoms relative to placebo 39, but a longer period of abstinence than naltrexone-treated individuals 40. The small study samples and inconsistencies in study design preclude drawing conclusions regarding the efficacy of pregabalin for AUD.

Despite similarities in the pharmacodynamic effects of gabapentin and pregabalin, there are notable differences in their pharmacokinetics and clinical use. Pregabalin has higher bioavailability and more predictable linear absorption than gabapentin, whose bioavailability decreases at higher doses 21,41. Thus, pregabalin achieves more consistent plasma concentrations and may require a lower dosage to achieve therapeutic efficacy. Pregabalin also has a faster onset of action, which may be advantageous for managing acute symptoms of alcohol withdrawal or craving 21,41. However, pregabalin carries a higher potential for abuse liability, misuse, and dependence, and is classified as a Schedule V controlled substance in the United States, while gabapentin is unscheduled 42, though this differs across states and countries 43,44. Regulatory differences impact prescribing practices and access, and gabapentin is less expensive and more widely available than pregabalin, whose controlled status may limit its availability despite a more favorable pharmacokinetic profile. Whether differences between gabapentin and pregabalin translate to differential effectiveness to treat AUD is unknown, as no study has directly compared their association with a reduction in alcohol use.

Thus, we investigated associations between gabapentin or pregabalin prescriptions for any indication and changes in alcohol use using data from the US Department of Veterans Affairs’ (VA) electronic health record (EHR). We evaluated whether associations differed by AUD diagnosis, AUDIT-C risk categories at baseline, and medication dosage. We hypothesized that receipt of gabapentin and pregabalin would each be associated with a reduction in alcohol use, and that this effect would be greater among individuals with AUD, those in higher AUDIT-C risk categories, and those who were prescribed higher doses of the medications. Given pregabalin’s greater bioavailability, faster onset of action, and more predictable pharmacokinetics than gabapentin’s, we also hypothesized that receipt of pregabalin would be associated with greater reductions in alcohol use than gabapentin in a head-to-head comparison.

Methods

Study Sample.

We conducted an observational cohort study using data from the Veterans Aging Cohort Study (VACS-National), which includes all >14.1 million individuals who have received any VA care since 1999. The VA is the largest integrated U.S. healthcare system serving approximately 6 million patients annually in over 1,300 inpatient and outpatient medical centers. All care is recorded in the EHR with daily uploads to the VA Corporate Data Warehouse. Available data include demographics, International Classification of Diseases Ninth and Tenth Revisions (ICD-9/-10) diagnostic codes, pharmacy dispensing records, laboratory results, procedures, vital signs, smoking status, and routinely collected information on alcohol consumption. The VA study was approved by the institutional review boards of Yale University (ref #1506016006) and VA Connecticut Healthcare System (ref #AJ0013) with a waiver of informed consent.

Exposure groups.

Our study comprised three groups: gabapentin initiators, pregabalin initiators, and matched comparators unexposed to either medication. Medication receipt was based on VA outpatient pharmacy dispensing data. We excluded individuals with no outpatient care in the year prior to index date due to the inability to capture baseline data, individuals with no measurement of alcohol consumption in the 2 years prior to index date, and those who reported no alcohol consumption based on the closest measurement to index date. For the gabapentin- and pregabalin-exposed groups, we included all patients who received the medications for at least 60 continuous days for any indication between 1 January 2009 and 30 June 2022, requiring a 180-day washout period from either medication to ensure new exposure episodes. For constructing the unexposed comparator group, we first identified outpatient clinics that were the largest sources of gabapentin and pregabalin prescriptions. We then selected all individuals who attended at least one of these clinics, but did not receive a gabapentin or pregabalin prescription to ensure that unexposed individuals came from the same source population. We randomly selected one visit per unexposed individual to be included in the analyses. Index date was defined as the first dispensed date for gabapentin and pregabalin recipients and the randomly selected outpatient visit date for unexposed individuals.

Covariates.

Data on multiple potential confounders were extracted, including age, race, ethnicity, sex, urban/rural residence, Area Deprivation Index 45, geographic region, year of index date, history of clinical conditions and procedures in the two years prior to index date (i.e., AUD, asthma, bariatric surgery, cancer, cerebrovascular disease, chronic obstructive pulmonary disease, congestive heart failure, dementia, diabetes, epilepsy, hemiplegia or paraplegia, human immunodeficiency virus (HIV) infection, liver disease, internalizing disorders (mood disorders, anxiety disorders, and stress-related disorders, including post-traumatic stress disorder), multiple sclerosis, myocardial infarction, opioid use disorder, peptic ulcer disease, peripheral vascular disease, renal disease, rheumatic disease, and traumatic brain injury), Charlson Comorbidity Index, and VACS Index. The Charlson Comorbidity Index is a measure of overall comorbidity based on 17 clinical domains 46,47. The VACS Index is a summary score that captures physiologic frailty computed using a validated algorithm that incorporates routine laboratory measures 48,49. We also extracted data on medication use at index date (i.e., AUD pharmacotherapies, neurocognitively-active medications, and anticholinergic medications), total medication count, and site-level gabapentin and pregabalin prescribing frequency. Neurocognitively-active medications included prescription opioids, antipsychotics, lithium, anticonvulsants (excluding gabapentin and pregabalin), anti-Parkinson’s, antidepressants, sedatives/hypnotics, muscle relaxants, amphetamine derivatives, antihistamines, and antivertigo agents. We also derived variables capturing substance use at index date (i.e., level of alcohol consumption, smoking status), any substance use treatment program visit, vital signs (body mass index, systolic and diastolic blood pressure, numeric rating scale pain score), laboratory findings (i.e., albumin, total cholesterol, HDL cholesterol, glycated hemoglobin, hemoglobin, total bilirubin, triglycerides, white blood cell count, fibrosis-4 (FIB-4) score, and estimated glomerular filtration rate [eGFR]). We created variables denoting whether the index prescription/visit was in primary care, total number of visits to a prescribing clinic, total number of visits to any clinic, and any hospitalization in the two years prior to index date.

Outcome Measure and Follow-up.

Primary outcome was change in alcohol consumption captured by the AUDIT-C. The VA has performed annual AUDIT-C screening on all patients in primary care since 2008 50. The AUDIT-C is a 3-item questionnaire that assesses frequency and quantity of alcohol use and heavy episodic drinking in the past year 51. Scores range from 0 to 12, with the likelihood of alcohol-related morbidity and mortality increasing as scores increase 52–54. An AUDIT-C score of 0 indicates no current alcohol use, 1–3 suggests lower-risk drinking, 4–7 suggests at-risk drinking, and ≥8 suggests hazardous or binge alcohol use.

Patients were followed from index date until the earliest of the following: 2 years post-index, their last VA visit, death, or 30 June 2024. Additionally, gabapentin and pregabalin recipients were censored at medication discontinuation. To ensure equal follow-up time within matched pairs, unexposed comparators were censored at the total follow-up time of their matched exposed individual. Although evidence of alcohol consumption at baseline (i.e., AUDIT-C score > 0) was an inclusion criterion, availability of a follow-up AUDIT-C was not required for matching eligibility as such a restriction would not translate to an analogous RCT.

Statistical Analyses.

We conducted propensity score matching to balance the distribution of potential confounders across exposure groups. Propensity scores were estimated using separate multivariable logistic regression models for each exposure contrast: gabapentin versus unexposed (contrast A), pregabalin versus unexposed (contrast B), and pregabalin versus gabapentin (contrast C). A missing category was included for 18 of the 60 covariates with missing data used for matching. Under the additional assumption that associations between fully observed covariates and exposure do not differ across missingness patterns, this approach produces unbiased estimates 55,56. For each contrast, individuals from each group were matched 1:1 on the logit of the propensity score using a caliper width 0.20 times the standard deviation of the logit of the propensity score in the region of common support and applying a greedy matching algorithm 57. Exact matching was performed on AUD diagnosis and baseline AUDIT-C categories. Given the minimization of potential confounding by indication comparing pregabalin initiators to gabapentin initiators (contrast C), we believed a priori that this comparison would provide the most robust evidence whether either medication reduced alcohol use.

We calculated absolute standardized mean difference (SMD) to examine balance between each exposure contrast in the unmatched, matched, and final analytic cohorts, with SMDs ≤0.1 indicating balance 57, and average pre- and post-index AUDIT-C scores of individuals in the final analytic cohort. Pre-index AUDIT-C scores were those on or closest to but before the index date, within a maximum of 2 years prior. Post-index AUDIT-C scores were those during and closest to the end of follow-up. Multivariable difference-in-difference DiD linear regression models 58,59 were used to estimate the differential change in AUDIT-C scores for each exposure contrast. We performed subgroup analyses by AUD diagnosis, baseline AUDIT-C category, and average daily dose category. Average daily doses were categorized based on previous trial and observational evidence as low (<1,200mg for gabapentin, <300mg for pregabalin), moderate (1,200–1,799mg for gabapentin, 300–599mg for pregabalin), and high (≥1,800mg for gabapentin, ≥600mg for pregabalin). Microsoft SQL Server Management Studio v20.2 and SAS Enterprise Guide v8.3 (SAS Institute) were used for data management and analysis, respectively.

Results

Sample.

The study flow diagram is presented in Figure 1. We identified 592,957 gabapentin initiators, 14,923 pregabalin initiators, and 2,959,006 eligible unexposed comparators who reported any alcohol consumption in the 2 years prior to the index date. After propensity score matching, contrast A (gabapentin vs. unexposed) included 494,611 matched patients per group, contrast B (pregabalin vs. unexposed) comprised 14,915 matched patients per group, and contrast C (pregabalin vs. gabapentin) included 14,923 patients per group. After excluding individuals with no eligible follow-up AUDIT-C, the final analytic cohorts included 177,189 gabapentin initiators and 140,025 unexposed comparators for contrast A (Supplemental Table 1), 5,270 pregabalin initiators and 4,009 unexposed for contrast B (Supplemental Table 2), and 3,611 pregabalin initiators and 3,384 gabapentin initiators for contrast C (Table 1).

Figure 1.

Figure 1.

Open in a new tab

Study flow diagram

Table 1.

Characteristics between initiators of gabapentin and pregabalin, before and after propensity score matching and recorded follow-up AUDIT-C

Unmatched Matched Matched with follow-up AUDIT-C
Gabapentin Pregabalin SMD Gabapentin Pregabalin SMD Gabapentin Pregabalin SMD
Sample size, n 592957 14,923 14923 14923 3384 3611
Demographics
Age, years
 20–49 162603 (27.4) 4458 (29.9) 0.05 4417 (29.6) 4458 (29.9) 0.006 924 (27.3) 1032 (28.6) 0.028
 50–59 131783 (22.2) 3206 (21.5) 0.02 3172 (21.3) 3206 (21.5) 0.006 716 (21.2) 798 (22.1) 0.023
 60–69 179146 (30.2) 3924 (26.3) 0.09 4097 (27.5) 3924 (26.3) 0.026 990 (29.3) 1003 (27.8) 0.033
 70–74 58381 (9.8) 1703 (11.4) 0.05 1667 (11.2) 1703 (11.4) 0.008 395 (11.7) 404 (11.2) 0.015
 75–79 29470 (5.0) 843 (5.6) 0.03 822 (5.5) 843 (5.6) 0.006 198 (5.9) 199 (5.5) 0.015
 ≥80 31574 (5.3) 789 (5.3) 0.00 748 (5.0) 789 (5.3) 0.012 161 (4.8) 175 (4.8) 0.004
Race/ethnicity
 White 403078 (68.0) 11348 (76.0) 0.18 11023 (73.9) 11348 (76.0) 0.050 2491 (73.6) 2757 (76.4) 0.063
 Black 102799 (17.3) 1771 (11.9) 0.16 1935 (13.0) 1771 (11.9) 0.033 458 (13.5) 434 (12.0) 0.045
 Hispanic 41687 (7.0) 708 (4.7) 0.10 841 (5.6) 708 (4.7) 0.040 185 (5.5) 161 (4.5) 0.046
 Asian 4900 (0.8) 87 (0.6) 0.03 86 (0.6) 87 (0.6) 0.001 17 (0.5) 24 (0.7) 0.021
 AI/AN 4080 (0.7) 127 (0.9) 0.02 128 (0.9) 127 (0.9) 0.001 30 (0.9) 31 (0.9) 0.003
 PI/NH 3965 (0.7) 97 (0.7) 0.00 104 (0.7) 97 (0.7) 0.006 24 (0.7) 18 (0.5) 0.027
 Mixed race 4969 (0.8) 120 (0.8) 0.00 125 (0.8) 120 (0.8) 0.004 28 (0.8) 31 (0.9) 0.003
 Other/unknown 27479 (4.6) 665 (4.5) 0.01 681 (4.6) 665 (4.5) 0.005 151 (4.5) 155 (4.3) 0.008
Sex
 Female 52089 (8.8) 2126 (14.2) 0.17 1842 (12.3) 2126 (14.2) 0.056 435 (12.9) 504 (14.0) 0.032
 Male 540868 (91.2) 12797 (85.8) 0.17 13081 (87.7) 12797 (85.8) 0.056 2949 (87.1) 3107 (86.0) 0.032
Urban residence 376730 (63.5) 9392 (62.9) 0.01 9449 (63.3) 9392 (62.9) 0.008 2188 (64.7) 2316 (64.1) 0.011
Area Deprivation Index
 1st quintile (lowest) 108094 (18.2) 2994 (20.1) 0.05 2944 (19.7) 2994 (20.1) 0.008 663 (19.6) 743 (20.6) 0.025
 2nd quintile 115006 (19.4) 3226 (21.6) 0.06 3210 (21.5) 3226 (21.6) 0.003 756 (22.3) 768 (21.3) 0.026
 3rd quintile 118047 (19.9) 2998 (20.1) 0.00 2992 (20.0) 2998 (20.1) 0.001 682 (20.2) 728 (20.2) 0.000
 4th quintile 125492 (21.2) 3131 (21.0) 0.00 3101 (20.8) 3131 (21.0) 0.005 686 (20.3) 770 (21.3) 0.026
 5th quintile (highest) 122206 (20.6) 2492 (16.7) 0.10 2595 (17.4) 2492 (16.7) 0.018 579 (17.1) 583 (16.1) 0.026
 Missing 4112 (0.7) 82 (0.5) 0.02 81 (0.5) 82 (0.5) 0.001 18 (0.5) 19 (0.5) 0.001
Census region
 Midwest 123302 (20.8) 3906 (26.2) 0.13 3642 (24.4) 3906 (26.2) 0.041 831 (24.6) 956 (26.5) 0.044
 Northeast 61942 (10.4) 1883 (12.6) 0.07 1866 (12.5) 1883 (12.6) 0.003 437 (12.9) 509 (14.1) 0.035
 South 276028 (46.6) 5909 (39.6) 0.14 6096 (40.8) 5909 (39.6) 0.026 1361 (40.2) 1357 (37.6) 0.054
 West 131685 (22.2) 3225 (21.6) 0.01 3319 (22.2) 3225 (21.6) 0.015 755 (22.3) 789 (21.8) 0.011
Year of index date
 2009–2012 159232 (26.9) 2681 (18.0) 0.21 2974 (19.9) 2681 (18.0) 0.050 705 (20.8) 704 (19.5) 0.033
 2013–2015 146577 (24.7) 2751 (18.4) 0.15 3034 (20.3) 2751 (18.4) 0.048 651 (19.2) 689 (19.1) 0.004
 2016–2018 149730 (25.3) 4266 (28.6) 0.08 4257 (28.5) 4266 (28.6) 0.001 989 (29.2) 1040 (28.8) 0.009
 2019–2021 137418 (23.2) 5225 (35.0) 0.26 4658 (31.2) 5225 (35.0) 0.081 1039 (30.7) 1178 (32.6) 0.041
Clinical characteristics
Alcohol use disorder 112423 (19.0) 1848 (12.4) 0.18 1848 (12.4) 1848 (12.4) 0.000 502 (14.8) 544 (15.1) 0.006
Asthma 19482 (3.3) 652 (4.4) 0.06 646 (4.3) 652 (4.4) 0.002 170 (5.0) 169 (4.7) 0.016
Bariatric surgery 551 (0.1) 13 (0.1) 0.00 9 (0.1) 13 (0.1) 0.010 6 (0.2) 5 (0.1) 0.010
Cancer, localized 49636 (8.4) 1279 (8.6) 0.01 1261 (8.5) 1279 (8.6) 0.004 330 (9.8) 345 (9.6) 0.007
Cancer, metastatic 5026 (0.8) 148 (1.0) 0.02 148 (1.0) 148 (1.0) 0.000 42 (1.2) 55 (1.5) 0.024
Cerebrovascular disease 34680 (5.8) 922 (6.2) 0.01 856 (5.7) 922 (6.2) 0.019 244 (7.2) 277 (7.7) 0.018
COPD 106982 (18.0) 2887 (19.3) 0.03 2844 (19.1) 2887 (19.3) 0.007 765 (22.6) 804 (22.3) 0.008
Congestive heart failure 29786 (5.0) 796 (5.3) 0.01 730 (4.9) 796 (5.3) 0.020 194 (5.7) 246 (6.8) 0.045
Dementia 4480 (0.8) 139 (0.9) 0.02 133 (0.9) 139 (0.9) 0.004 47 (1.4) 37 (1.0) 0.033
Diabetes, complications 61548 (10.4) 2339 (15.7) 0.16 2092 (14.0) 2339 (15.7) 0.047 555 (16.4) 640 (17.7) 0.035
Diabetes, w/o complications 164737 (27.8) 4238 (28.4) 0.01 4147 (27.8) 4238 (28.4) 0.014 1057 (31.2) 1089 (30.2) 0.023
Epilepsy 4849 (0.8) 253 (1.7) 0.08 189 (1.3) 253 (1.7) 0.036 53 (1.6) 76 (2.1) 0.040
Hemiplegia or paraplegia 6147 (1.0) 439 (2.9) 0.14 293 (2.0) 439 (2.9) 0.064 103 (3.0) 169 (4.7) 0.085
HIV 3105 (0.5) 122 (0.8) 0.04 109 (0.7) 122 (0.8) 0.010 34 (1.0) 38 (1.1) 0.005
Liver disease, mild 36680 (6.2) 929 (6.2) 0.00 875 (5.9) 929 (6.2) 0.015 252 (7.4) 266 (7.4) 0.003
Liver disease, moderate to severe 3662 (0.6) 81 (0.5) 0.01 75 (0.5) 81 (0.5) 0.006 30 (0.9) 29 (0.8) 0.009
Internalizing disorders1 213977 (36.1) 6191 (41.5) 0.11 5869 (39.3) 6191 (41.5) 0.044 1603 (47.4) 1719 (47.6) 0.005
Multiple sclerosis 2360 (0.4) 147 (1.0) 0.07 94 (0.6) 147 (1.0) 0.040 29 (0.9) 45 (1.2) 0.038
Myocardial infarction 13783 (2.3) 345 (2.3) 0.00 318 (2.1) 345 (2.3) 0.012 93 (2.7) 111 (3.1) 0.019
Opioid use disorder 14311 (2.4) 550 (3.7) 0.07 441 (3.0) 550 (3.7) 0.041 129 (3.8) 149 (4.1) 0.016
Peptic ulcer disease 7033 (1.2) 196 (1.3) 0.01 198 (1.3) 196 (1.3) 0.001 53 (1.6) 63 (1.7) 0.014
Peripheral vascular disease 45962 (7.8) 1276 (8.6) 0.03 1204 (8.1) 1276 (8.6) 0.017 346 (10.2) 385 (10.7) 0.014
Renal disease 35595 (6.0) 1079 (7.2) 0.05 956 (6.4) 1079 (7.2) 0.033 253 (7.5) 289 (8.0) 0.020
Rheumatic disease 9421 (1.6) 392 (2.6) 0.07 347 (2.3) 392 (2.6) 0.019 106 (3.1) 126 (3.5) 0.020
Traumatic brain injury 6317 (1.1) 229 (1.5) 0.04 205 (1.4) 229 (1.5) 0.013 60 (1.8) 76 (2.1) 0.024
Charlson Comorbidity Index
 0 6317 (1.1) 229 (1.5) 0.04 205 (1.4) 229 (1.5) 0.013 60 (1.8) 76 (2.1) 0.024
 1 272467 (46.0) 6397 (42.9) 0.06 6678 (44.7) 6397 (42.9) 0.038 1281 (37.9) 1348 (37.3) 0.011
 2 129639 (21.9) 2936 (19.7) 0.05 3020 (20.2) 2936 (19.7) 0.014 690 (20.4) 699 (19.4) 0.026
 3 86711 (14.6) 2418 (16.2) 0.04 2350 (15.7) 2418 (16.2) 0.012 592 (17.5) 636 (17.6) 0.003
 4 45705 (7.7) 1303 (8.7) 0.04 1245 (8.3) 1303 (8.7) 0.014 336 (9.9) 359 (9.9) 0.000
 ≥5 25309 (4.3) 762 (5.1) 0.04 637 (4.3) 762 (5.1) 0.040 171 (5.1) 235 (6.5) 0.062
VACS Index
 1st quartile (lowest frailty) 124624 (21.0) 3286 (22.0) 0.02 3287 (22.0) 3286 (22.0) 0.000 691 (20.4) 752 (20.8) 0.010
 2nd quartile 122704 (20.7) 3018 (20.2) 0.01 2955 (19.8) 3018 (20.2) 0.011 650 (19.2) 757 (21.0) 0.044
 3rd quartile 124370 (21.0) 3040 (20.4) 0.01 3119 (20.9) 3040 (20.4) 0.013 730 (21.6) 752 (20.8) 0.018
 4th quartile (highest frailty) 124341 (21.0) 3111 (20.8) 0.00 3048 (20.4) 3111 (20.8) 0.010 782 (23.1) 852 (23.6) 0.011
 Missing 96918 (16.3) 2468 (16.5) 0.01 2514 (16.8) 2468 (16.5) 0.008 531 (15.7) 498 (13.8) 0.054
Other medications at index date
Any alcohol pharmacotherapy 28205 (4.8) 944 (6.3) 0.07 840 (5.6) 944 (6.3) 0.029 224 (6.6) 265 (7.3) 0.028
Any neurocognitively-active2 365459 (61.6) 10249 (68.7) 0.15 10025 (67.2) 10249 (68.7) 0.032 2419 (71.5) 2708 (75.0) 0.079
Any anticholinergic 134401 (22.7) 4197 (28.1) 0.13 3936 (26.4) 4197 (28.1) 0.039 1020 (30.1) 1163 (32.2) 0.045
Medication count
 0–1 88261 (14.9) 2432 (16.3) 0.04 2407 (16.1) 2432 (16.3) 0.005 445 (13.2) 462 (12.8) 0.011
 2–4 232440 (39.2) 5562 (37.3) 0.04 5672 (38.0) 5562 (37.3) 0.015 1244 (36.8) 1272 (35.2) 0.032
 5–9 217112 (36.6) 5143 (34.5) 0.04 5232 (35.1) 5143 (34.5) 0.013 1270 (37.5) 1343 (37.2) 0.007
 ≥10 55144 (9.3) 1786 (12.0) 0.09 1612 (10.8) 1786 (12.0) 0.037 425 (12.6) 534 (14.8) 0.065
Site-level prescribing intensity
 1st quartile (lowest) 94519 (15.9) 3682 (24.7) 0.22 3311 (22.2) 3682 (24.7) 0.059 779 (23.0) 908 (25.1) 0.050
 2nd quartile 133801 (22.6) 3967 (26.6) 0.09 3811 (25.5) 3967 (26.6) 0.024 830 (24.5) 1012 (28.0) 0.080
 3rd quartile 180415 (30.4) 3894 (26.1) 0.10 4187 (28.1) 3894 (26.1) 0.044 965 (28.5) 953 (26.4) 0.048
 4th quartile (highest) 184222 (31.1) 3380 (22.6) 0.19 3614 (24.2) 3380 (22.6) 0.037 810 (23.9) 738 (20.4) 0.084
Substance use
Alcohol consumption
 Lower-risk 418465 (70.6) 11829 (79.3) 0.20 11829 (79.3) 11829 (79.3) 0.000 2615 (77.3) 2819 (78.1) 0.019
 At-risk 123638 (20.9) 2477 (16.6) 0.11 2477 (16.6) 2477 (16.6) 0.000 589 (17.4) 615 (17.0) 0.010
 Hazardous/binge 50854 (8.6) 617 (4.1) 0.18 617 (4.1) 617 (4.1) 0.000 180 (5.3) 177 (4.9) 0.019
Substance use treatment program visit 103723 (17.5) 1664 (11.2) 0.18 1675 (11.2) 1664 (11.2) 0.002 464 (13.7) 497 (13.8) 0.002
Smoking status
 Never 155287 (26.2) 4095 (27.4) 0.03 4122 (27.6) 4095 (27.4) 0.004 901 (26.6) 960 (26.6) 0.001
 Former 185616 (31.3) 5159 (34.6) 0.07 5074 (34.0) 5159 (34.6) 0.012 1168 (34.5) 1242 (34.4) 0.003
 Current 250354 (42.2) 5626 (37.7) 0.09 5689 (38.1) 5626 (37.7) 0.009 1308 (38.7) 1399 (38.7) 0.002
 Missing 1700 (0.3) 43 (0.3) 0.00 38 (0.3) 43 (0.3) 0.006 7 (0.2) 10 (0.3) 0.014
Vital signs
Body mass index, kg/m2
 ≤26 135901 (22.9) 3118 (20.9) 0.05 3166 (21.2) 3118 (20.9) 0.008 713 (21.1) 787 (21.8) 0.018
 >26–32 236167 (39.8) 5798 (38.9) 0.02 5773 (38.7) 5798 (38.9) 0.003 1302 (38.5) 1412 (39.1) 0.013
 >32 198691 (33.5) 5387 (36.1) 0.05 5337 (35.8) 5387 (36.1) 0.007 1255 (37.1) 1283 (35.5) 0.032
 Missing 22198 (3.7) 620 (4.2) 0.02 647 (4.3) 620 (4.2) 0.009 114 (3.4) 129 (3.6) 0.011
Systolic blood pressure, mm Hg
 ≤127 207992 (35.1) 5681 (38.1) 0.06 5577 (37.4) 5681 (38.1) 0.014 1297 (38.3) 1417 (39.2) 0.019
 >127–141 217871 (36.7) 5442 (36.5) 0.01 5515 (37.0) 5442 (36.5) 0.010 1260 (37.2) 1339 (37.1) 0.003
 >141 131571 (22.2) 3048 (20.4) 0.04 3041 (20.4) 3048 (20.4) 0.001 739 (21.8) 748 (20.7) 0.027
 Missing 35523 (6.0) 752 (5.0) 0.04 790 (5.3) 752 (5.0) 0.012 88 (2.6) 107 (3.0) 0.022
Diastolic blood pressure, mm Hg
 ≤82 366232 (61.8) 9581 (64.2) 0.05 9477 (63.5) 9581 (64.2) 0.015 2243 (66.3) 2335 (64.7) 0.034
 >82 191202 (32.2) 4590 (30.8) 0.03 4656 (31.2) 4590 (30.8) 0.010 1053 (31.1) 1169 (32.4) 0.027
 Missing 35523 (6.0) 752 (5.0) 0.04 790 (5.3) 752 (5.0) 0.012 88 (2.6) 107 (3.0) 0.022
Pain score category
 None (0) 202342 (34.1) 3489 (23.4) 0.24 3914 (26.2) 3489 (23.4) 0.066 953 (28.2) 854 (23.6) 0.103
 Mild to moderate (1–6) 215215 (36.3) 6255 (41.9) 0.12 6101 (40.9) 6255 (41.9) 0.021 1406 (41.5) 1543 (42.7) 0.024
 Severe (7–10) 129237 (21.8) 4200 (28.1) 0.15 3874 (26.0) 4200 (28.1) 0.049 913 (27.0) 1083 (30.0) 0.067
 Missing 46163 (7.8) 979 (6.6) 0.05 1034 (6.9) 979 (6.6) 0.015 112 (3.3) 131 (3.6) 0.017
Laboratory findings
Albumin, g/dL
 ≤4.0 217830 (36.7) 5566 (37.3) 0.01 5586 (37.4) 5566 (37.3) 0.003 1399 (41.3) 1472 (40.8) 0.012
 >4.0–4.4 196000 (33.1) 4805 (32.2) 0.02 4789 (32.1) 4805 (32.2) 0.002 1030 (30.4) 1156 (32.0) 0.034
 >4.4 103116 (17.4) 2567 (17.2) 0.00 2499 (16.7) 2567 (17.2) 0.012 550 (16.3) 573 (15.9) 0.010
 Missing 76011 (12.8) 1985 (13.3) 0.01 2049 (13.7) 1985 (13.3) 0.013 405 (12.0) 410 (11.4) 0.019
Total cholesterol, mg/dL
 ≤137 86464 (14.6) 2316 (15.5) 0.03 2266 (15.2) 2316 (15.5) 0.009 541 (16.0) 598 (16.6) 0.016
 >137–220 364569 (61.5) 8816 (59.1) 0.05 8933 (59.9) 8816 (59.1) 0.016 2056 (60.8) 2158 (59.8) 0.020
 >220 84782 (14.3) 2171 (14.5) 0.01 2180 (14.6) 2171 (14.5) 0.002 484 (14.3) 543 (15.0) 0.021
 Missing 57142 (9.6) 1620 (10.9) 0.04 1544 (10.3) 1620 (10.9) 0.017 303 (9.0) 312 (8.6) 0.011
HDL cholesterol, mg/dL
 <35 116708 (19.7) 2966 (19.9) 0.00 3039 (20.4) 2966 (19.9) 0.012 743 (22.0) 735 (20.4) 0.039
 >35–45 173129 (29.2) 4424 (29.6) 0.01 4417 (29.6) 4424 (29.6) 0.001 978 (28.9) 1086 (30.1) 0.026
 >45 244667 (41.3) 5869 (39.3) 0.04 5897 (39.5) 5869 (39.3) 0.004 1357 (40.1) 1467 (40.6) 0.011
 Missing 58453 (9.9) 1664 (11.2) 0.04 1570 (10.5) 1664 (11.2) 0.020 306 (9.0) 323 (8.9) 0.003
HbA1c, %
 <6.5 277432 (46.8) 7272 (48.7) 0.04 7187 (48.2) 7272 (48.7) 0.011 1675 (49.5) 1830 (50.7) 0.024
 6.5–<8.0 68617 (11.6) 1706 (11.4) 0.00 1712 (11.5) 1706 (11.4) 0.001 413 (12.2) 440 (12.2) 0.001
 ≥8.0 43743 (7.4) 1000 (6.7) 0.03 1057 (7.1) 1000 (6.7) 0.015 242 (7.2) 234 (6.5) 0.027
 Missing 203165 (34.3) 4945 (33.1) 0.02 4967 (33.3) 4945 (33.1) 0.003 1054 (31.1) 1107 (30.7) 0.011
Hemoglobin, g/dL
 ≤13.5 146084 (24.6) 3997 (26.8) 0.05 3841 (25.7) 3997 (26.8) 0.024 997 (29.5) 1094 (30.3) 0.018
 >13.5–14.3 101992 (17.2) 2487 (16.7) 0.01 2484 (16.6) 2487 (16.7) 0.001 570 (16.8) 650 (18.0) 0.031
 >14.3 296344 (50.0) 7105 (47.6) 0.05 7247 (48.6) 7105 (47.6) 0.019 1570 (46.4) 1650 (45.7) 0.014
 Missing 48537 (8.2) 1334 (8.9) 0.03 1351 (9.1) 1334 (8.9) 0.004 247 (7.3) 217 (6.0) 0.052
Total bilirubin, mg/dL
 ≤0.6 303784 (51.2) 8097 (54.3) 0.06 7883 (52.8) 8097 (54.3) 0.029 1892 (55.9) 2040 (56.5) 0.012
 >0.6 221089 (37.3) 4955 (33.2) 0.09 5099 (34.2) 4955 (33.2) 0.020 1098 (32.4) 1194 (33.1) 0.013
 Missing 68084 (11.5) 1871 (12.5) 0.03 1941 (13.0) 1871 (12.5) 0.014 394 (11.6) 377 (10.4) 0.038
Triglycerides, mg/dL
 ≤118 234282 (39.5) 5390 (36.1) 0.07 5587 (37.4) 5390 (36.1) 0.027 1256 (37.1) 1342 (37.2) 0.001
 >118–184 149721 (25.2) 3832 (25.7) 0.01 3818 (25.6) 3832 (25.7) 0.002 900 (26.6) 940 (26.0) 0.013
 >184 147756 (24.9) 3991 (26.7) 0.04 3886 (26.0) 3991 (26.7) 0.016 909 (26.9) 995 (27.6) 0.016
 Missing 61198 (10.3) 1710 (11.5) 0.04 1632 (10.9) 1710 (11.5) 0.017 319 (9.4) 334 (9.2) 0.006
White blood cell count, K/μL
 ≤5.7 141565 (23.9) 3450 (23.1) 0.02 3508 (23.5) 3450 (23.1) 0.009 836 (24.7) 867 (24.0) 0.016
 >5.7–7.2 163690 (27.6) 4089 (27.4) 0.00 4036 (27.0) 4089 (27.4) 0.008 901 (26.6) 1030 (28.5) 0.042
 >7.2 238754 (40.3) 6031 (40.4) 0.00 6007 (40.3) 6031 (40.4) 0.003 1397 (41.3) 1493 (41.3) 0.001
 Missing 48948 (8.3) 1353 (9.1) 0.03 1372 (9.2) 1353 (9.1) 0.004 250 (7.4) 221 (6.1) 0.051
Fibrosis-4 score
 <1.45 314019 (53.0) 8289 (55.5) 0.05 8188 (54.9) 8289 (55.5) 0.014 1841 (54.4) 2045 (56.6) 0.045
 1.45–3.25 158600 (26.7) 3716 (24.9) 0.04 3754 (25.2) 3716 (24.9) 0.006 883 (26.1) 951 (26.3) 0.006
 >3.25 22753 (3.8) 452 (3.0) 0.04 460 (3.1) 452 (3.0) 0.003 126 (3.7) 115 (3.2) 0.030
 Missing 97585 (16.5) 2466 (16.5) 0.00 2521 (16.9) 2466 (16.5) 0.010 534 (15.8) 500 (13.8) 0.054
eGFR, mL/min
 ≥60 477844 (80.6) 11697 (78.4) 0.05 11775 (78.9) 11697 (78.4) 0.013 2696 (79.7) 2877 (79.7) 0.000
 30–59 62035 (10.5) 1679 (11.3) 0.03 1637 (11.0) 1679 (11.3) 0.009 399 (11.8) 430 (11.9) 0.004
 <30 5418 (0.9) 165 (1.1) 0.02 170 (1.1) 165 (1.1) 0.003 48 (1.4) 47 (1.3) 0.010
 Missing 47660 (8.0) 1382 (9.3) 0.04 1341 (9.0) 1382 (9.3) 0.010 241 (7.1) 257 (7.1) 0.000
Health Service Utilization
Index visit in primary care 425156 (71.7) 9470 (63.5) 0.18 9771 (65.5) 9470 (63.5) 0.042 1910 (56.4) 1995 (55.2) 0.024
No. visits to prescribing clinic in previous 2 years
 0 35149 (5.9) 652 (4.4) 0.07 709 (4.8) 652 (4.4) 0.018 91 (2.7) 92 (2.5) 0.009
 1–2 69862 (11.8) 1752 (11.7) 0.00 1783 (11.9) 1752 (11.7) 0.006 247 (7.3) 296 (8.2) 0.034
 3–4 75286 (12.7) 1639 (11.0) 0.05 1709 (11.5) 1639 (11.0) 0.015 303 (9.0) 345 (9.6) 0.021
 5–6 74070 (12.5) 1647 (11.0) 0.05 1686 (11.3) 1647 (11.0) 0.008 356 (10.5) 351 (9.7) 0.027
 7–12 162326 (27.4) 3781 (25.3) 0.05 3888 (26.1) 3781 (25.3) 0.016 947 (28.0) 942 (26.1) 0.043
 ≥12 176264 (29.7) 5452 (36.5) 0.14 5148 (34.5) 5452 (36.5) 0.043 1440 (42.6) 1585 (43.9) 0.027
No. visits to any clinic in previous 2 years
 0 26946 (4.5) 507 (3.4) 0.06 554 (3.7) 507 (3.4) 0.017 51 (1.5) 65 (1.8) 0.023
 1–3 69890 (11.8) 1882 (12.6) 0.03 1889 (12.7) 1882 (12.6) 0.001 233 (6.9) 271 (7.5) 0.024
 4–5 47299 (8.0) 1176 (7.9) 0.00 1208 (8.1) 1176 (7.9) 0.008 217 (6.4) 197 (5.5) 0.041
 6–10 103852 (17.5) 2220 (14.9) 0.07 2303 (15.4) 2220 (14.9) 0.016 443 (13.1) 476 (13.2) 0.003
 11–20 141820 (23.9) 3067 (20.6) 0.08 3186 (21.3) 3067 (20.6) 0.020 756 (22.3) 776 (21.5) 0.021
 ≥20 203150 (34.3) 6071 (40.7) 0.13 5783 (38.8) 6071 (40.7) 0.039 1684 (49.8) 1826 (50.6) 0.016
Any hospitalization 98095 (16.5) 2387 (16.0) 0.01 2312 (15.5) 2387 (16.0) 0.014 730 (21.6) 783 (21.7) 0.003
Open in a new tab
1

Internalizing disorders consisted of mood disorders, anxiety disorders, and stress-related disorders.

2

Neurocognitively-active medications consisted of prescription opioids, antipsychotics, lithium, anticonvulsants, anti-Parkinson’s, antidepressants, sedatives/hypnotics, muscle relaxants, amphetamine derivatives, antihistamines, and antivertigo agents.

Abbreviations: SMD, absolute value of the standardized mean difference; AUDIT-C, Alcohol Use Disorder Identification Test - Consumption; AI/AN, American Indian/Alaska Native; PI/NH, Pacific Islander/Native Hawaiian; COPD, chronic obstructive pulmonary disease; HIV, human immunodeficiency virus; VACS, Veterans Aging Cohort Study; HDL, high-density lipoprotein; HbA1c, glycated hemoglobin; eGFR, estimated glomerular filtration rate.

Note: Reported as n (%)

Distribution of baseline characteristics differed between groups before propensity score matching. After propensity score matching and restricting patients to those with a follow-up AUDIT-C, the distribution of baseline characteristics was balanced between groups (nearly all SMDs≤0.1 except for six variables with SMDs≤0.16 for contrast A [Supplemental Table S1], and three variables with SMD≤0.11 for contrast B [Supplemental Table S2]). The distribution of propensity scores for each contrast before and after matching are shown in Supplemental Figure S1.

Changes in alcohol consumption – Contrast A (Gabapentin vs. Unexposed).

Pre-index and post-index AUDIT-C scores among gabapentin initiators was 2.93 (standard error [SE]=0.01) and 2.20 (SE=0.01), respectively, resulting in a difference of −0.72 (SE=0.01). Pre-index and post-index AUDIT-C scores among unexposed comparators was 2.83 (SE=0.01) and 2.19 (SE=0.01), respectively, resulting in a difference of −0.63 (SE=0.01). Gabapentin initiators had a significantly greater reduction in AUDIT-C scores than unexposed comparators (DiD: 0.09 points, 95% CI: 0.06, 0.11, p<0.0001; Table 2, Figure 2A). This effect was consistent in subgroup analyses based on diagnosed AUD and by baseline AUDIT-C category (Table 2, Figure 2A). A dose-response relationship was observed, with a small effect at low average daily doses (DiD: 0.04 points, 95% CI: 0.01, 0.07, p=0.0027), an effect consistent to the overall estimate at moderate doses (DiD: 0.11 points, 95% CI: 0.06, 0.16, p<0.0001), and the largest effect at high doses (DiD: 0.22 points, 95% CI: 0.18, 0.26, p<0.0001).

Table 2.

Estimated average pre- and post-index date AUDIT-C scores and difference-in-differences (DiD), overall, by current alcohol use diagnosis, and level of alcohol consumption

Gabapentin Unexposed Pregabalin Unexposed Pregabalin Gabapentin
n=177,189 n=140,025 n=5,270 n=4,009 n=3,611 n=3,384
All patients Pre 2.93 (0.01) 2.83 (0.01) 2.44 (0.03) 2.47 (0.03) 2.51 (0.04) 2.63 (0.04)
Post 2.20 (0.01) 2.19 (0.01) 1.76 (0.03) 1.92 (0.03) 1.83 (0.04) 2.05 (0.04)
Dn −0.72 (0.01) −0.63 (0.01) −0.69 (0.04) −0.55 (0.05) −0.68 (0.05) −0.58 (0.06)
DiD (95% CI) 0.09 (0.06, 0.11), p<0.0001 0.14 (0.02, 0.26), p=0.0279 0.10 (−0.05, 0.26), p=0.1834
By baseline AUD
 No AUD n=147,365 n=117,980 n=4,537 n=3,452 n=3,067 n=2,882
Pre 2.41 (0.01) 2.38 (0.01) 2.06 (0.03) 2.10 (0.03) 2.08 (0.04) 2.18 (0.04)
Post 1.85 (0.01) 1.89 (0.01) 1.54 (0.03) 1.64 (0.03) 1.58 (0.04) 1.64 (0.04)
Dn −0.56 (0.01) −0.49 (0.01) −0.52 (0.04) −0.46 (0.05) −0.50 (0.05) −0.53 (0.06)
DiD (95% CI) 0.07 (0.05, 0.10), p<0.0001 0.06 (−0.06, 0.19), p=0.3253 −0.03 (−0.18, 0.12), p=0.6750
 AUD n=29,824 n=22,045 n=733 n=557 n=544 n=502
Pre 5.45 (0.01) 5.21 (0.02) 4.78 (0.07) 4.74 (0.08) 4.93 (0.09) 5.22 (0.09)
Post 3.93 (0.01) 3.79 (0.02) 3.09 (0.07) 3.67 (0.08) 3.24 (0.09) 4.39 (0.09)
Dn −1.52 (0.02) −1.42 (0.02) −1.69 (0.10) −1.08 (0.12) −1.70 (0.13) −0.83 (0.13)
DiD (95% CI) 0.10 (0.04, 0.16), p=0.0005 0.61 (0.31,0.92), p=0.0001 0.86 (0.50, 1.22), p<0.0001
By baseline AUDIT-C
 Lower-risk n=126,159 n=101,690 n=4,169 n=3,180 n=2,819 n=2,615
Pre 1.64 (0.00) 1.66 (0.01) 1.56 (0.02) 1.59 (0.03) 1.57 (0.03) 1.63 (0.03)
Post 1.53 (0.00) 1.58 (0.01) 1.35 (0.02) 1.44 (0.03) 1.40 (0.03) 1.49 (0.03)
Dn −0.12 (0.01) −0.08 (0.01) −0.21 (0.03) −0.16 (0.04) −0.17 (0.04) −0.14 (0.04)
DiD (95% CI) 0.03 (0.01,0.05), p=0.0023 0.05 (−0.05, 0.15), p=0.3041 0.03 (−0.09, 0.15), p=0.5990
 At-risk n=37,886 n=29,200 n=857 n=654 n=615 n=589
Pre 4.75 (0.01) 4.72 (0.01) 4.64 (0.05) 4.78 (0.06) 4.68 (0.06) 4.80 (0.07)
Post 3.37 (0.01) 3.39 (0.01) 3.05 (0.05) 3.44 (0.06) 3.10 (0.06) 3.30 (0.07)
Dn −1.38 (0.01) −1.33 (0.01) −1.59 (0.07) −1.33 (0.08) −1.58 (0.09) −1.50 (0.09)
DiD (95% CI) 0.05 (0.02, 0.09), p=0.0028 0.26 (0.04, 0.47), p=0.0188 0.08 (−0.18, 0.34), p=0.5419
 Hazardous/binge n=13,144 n=9,135 n=244 n=175 n=177 n=180
Pre 9.95 (0.01) 9.71 (0.02) 9.82 (0.10) 9.73 (0.11) 9.96 (0.12) 10.00 (0.12)
Post 5.31 (0.01) 5.19 (0.02) 4.17 (0.10) 5.07 (0.11) 4.31 (0.12) 6.09 (0.12)
Dn −4.64 (0.02) −4.53 (0.02) −5.65 (0.14) −4.66 (0.16) −5.65 (0.17) −3.91 (0.17)
DiD (95% CI) 0.11 (0.05, 0.18), p=0.0004 0.98 (0.57, 1.40), p<0.0001 1.74 (1.27, 2.21), p<0.0001
By average daily dose
 Low n=117,509 n=140,025 n=4,117 n=4,009 n=2,214 n=2,322
Pre 2.85 (0.01) 2.83 (0.01) 2.40 (0.03) 2.47 (0.03) 2.40 (0.05) 2.56 (0.05)
Post 2.17 (0.01) 2.19 (0.01) 1.78 (0.03) 1.92 (0.03) 1.83 (0.05) 1.99 (0.05)
Dn −0.68 (0.01) −0.63 (0.01) −0.62 (0.05) −0.55 (0.05) −0.58 (0.07) −0.56 (0.07)
DiD (95% CI) 0.04 (0.01,0.07), p=0.0027 0.08 (−0.06, 0.21), p=0.2617 0.02 (−0.17, 0.20), p=0.8669
 Moderate n=21,539 n=140,025 n=1,064 n=4,009 n=1,219 n=365
Pre 3.02 (0.02) 2.83 (0.01) 2.56 (0.07) 2.47 (0.03) 2.63 (0.07) 2.85 (0.12)
Post 2.28 (0.02) 2.19 (0.01) 1.72 (0.07) 1.92 (0.03) 1.85 (0.07) 2.24 (0.12)
Dn −0.74 (0.02) −0.63 (0.01) −0.83 (0.09) −0.55 (0.05) −0.78 (0.09) −0.61 (0.17)
DiD (95% CI) 0.11 (0.06, 0.16), p<0.0001 0.29 (0.08, 0.49), p=0.0058 0.17 (−0.21, 0.55), p=0.3806
 High n=38,141 n=140,025 n=89 n=4009 n=178 n=697
Pre 3.10 (0.01) 2.83 (0.01) 3.12 (0.23) 2.47 (0.03) 3.08 (0.17) 2.75 (0.09)
Post 2.25 (0.01) 2.19 (0.01) 1.25 (0.23) 1.92 (0.03) 1.76 (0.17) 2.13 (0.09)
Dn −0.86 (0.02) −0.63 (0.01) −1.88 (0.32) −0.55 (0.05) −1.31 (0.24) −0.62 (0.12)
DiD (95% CI) 0.22 (0.18, 0.26), p<0.0001 1.33 (0.70, 1.97), p<0.0001 0.70 (0.16, 1.23), p=0.0105
Open in a new tab

Abbreviations: AUDIT-C - Alcohol Use Disorders Identification Test - Consumption; AUD, alcohol use disorder; Pre - pre-index AUDIT-C score; Post - post-index AUDIT-C score; Dn - change in AUDIT-C score; DiD - difference-in-difference estimate; CI - confidence interval

Notes: Average pre- and post-index date AUDIT-C scores reported as mean (standard error); average daily doses were categorized as low (<1,200 mg for gabapentin, <300 mg for pregabalin), moderate (1,200–1,799 mg for gabapentin, 300–599 mg for pregabalin), and high (≥1,800 mg for gabapentin, ≥600 mg for pregabalin).

Figure 2. Association between receipt of gabapentin and pregabalin and alcohol use.

Figure 2.

Open in a new tab

Difference-in-difference estimates and 95% confidence intervals of changes in AUDIT-C scores stratified by baseline AUD diagnosis and by baseline AUDIT-C risk categories. (A) Gabapentin recipients vs. unexposed individuals, (B) pregabalin recipients vs. unexposed individuals, (C) pregabalin recipients vs. gabapentin recipients.

Changes in alcohol consumption – Contrast B (Pregabalin vs. Unexposed).

Patients from both groups reported a reduction in AUDIT-C scores (pregabalin: Dn = −0.69 [SE=0.04]; unexposed: Dn = −0.55 [SE=0.05]). Thus, patients who initiated pregabalin had a significantly greater reduction in AUDIT-C scores than unexposed comparators (DiD: 0.14 points, 95% CI: 0.02, 0.26, p = 0.0279; Table 2, Figure 2B). This effect was greater for patients with a diagnosis of AUD at baseline (DiD: 0.61 points, 95% CI: 0.31, 0.92, p=0.0001) and those who reported baseline hazardous/binge consumption (DiD: 0.98 points, 95% CI: 0.57, 1.40, p<0.0001; Table 2, Figure 2B). A dose-response relationship was observed, with no effect at low average daily doses, an effect at moderate doses (DiD: 0.29 points, 95% CI: 0.08, 0.49, p=0.0058), and the largest effect at high doses (DiD: 1.33 points, 95% CI: 0.70, 1.97, p<0.0001).

Changes in alcohol consumption – Contrast C (Pregabalin vs. Gabapentin).

Patients from both groups reported a reduction in AUDIT-C scores (pregabalin: Dn = −0.68 [SE=0.05]; gabapentin: Dn = −0.58 [SE=0.06]). Overall, there was no statistical difference in changes in AUDIT-C scores between pregabalin initiators and gabapentin initiators (DiD: 0.10 points, 95% CI: −0.05, 0.26, p=0.1834; Table 2, Figure 2C). However, there was strong evidence that pregabalin initiators had a greater reduction in AUDIT-C scores than gabapentin initiators among those with a diagnosis of AUD at baseline (DiD: 0.86 points, 95% CI: 0.50, 1.22, p<0.0001) and those who reported baseline hazardous/binge consumption (DiD: 1.74 points, 95% CI: 1.27, 2.21, p<0.0001; Table 2, Figure 2C). Comparative effectiveness between gabapentin and pregabalin was evident at high average daily doses (DiD: 0.70 points, 95% CI: 0.16, 1.23, p=0.0105), with no observed differences at low or moderate doses.

Discussion

We compared changes in alcohol use among patients who initiated gabapentin or pregabalin for any indication. Using real-world EHR data from the largest integrated U.S. healthcare system, we applied propensity-score matching to balance baseline characteristics across three exposure contrasts and used DiD analyses to compare changes in drinking captured by routinely-collected AUDIT-C scores. Individuals who initiated gabapentin reported, on average, lower alcohol use than unexposed comparators; this difference did not differ as a function of AUD diagnosis or AUDIT-C risk category. Individuals who initiated pregabalin also reported a greater reduction in alcohol use than unexposed comparators, with a greater effect observed among those with AUD and those reporting the highest alcohol use at baseline. In a direct comparison, pregabalin initiators reported significantly greater reductions in alcohol use than gabapentin initiators, and those with AUD and those reporting the highest alcohol use at baseline experienced the greatest decrease. Additionally, we observed a dose-dependent relationship for both medications, with greater reductions in alcohol use among patients prescribed higher doses of either gabapentin or pregabalin. These results provide evidence of the effectiveness of either gabapentin or pregabalin to reduce alcohol use, with pregabalin as more effective than gabapentin, especially among patients who may benefit the most, i.e., those with AUD or those who report hazardous or heavy episodic drinking.

Our findings build upon and extend prior work on the use of gabapentinoids for treating AUD. Gabapentin has previously been shown to reduce craving and heavy drinking among individuals with AUD, though effects were not always consistent across all clinical trials 20,22–30. We previously estimated the real-world effectiveness of gabapentin using similar methods in a different VA cohort enriched with people living with HIV 60. In the present study, we replicated our previous findings in a broader, more generalizable cohort including all patients who have accessed VA care and expanded our analysis to compare its effectiveness against pregabalin.

Similarly, prior clinical trials of pregabalin have shown inconsistent but promising results. Pregabalin was previously shown to reduce heavy drinking frequency 36, and to have mixed effects on alcohol withdrawal and craving 35,37–40. Our study extends prior work by leveraging a large, real-world EHR dataset and applying statistical approaches that minimize confounds and approximate causal effects. We found evidence for the use of pregabalin to reduce alcohol use especially in patients with AUD. Additionally, this was the first study to directly compare the effects of gabapentin to pregabalin on reducing alcohol consumption, which addresses an important gap in the literature regarding the relative effectiveness of these two medications.

Machine learning-based secondary analyses of the Falk et al.’s multisite RCT 29 have suggested potential baseline factors that could help identifying treatment responders among those people with AUD receiving gabapentin 61,62. Additional studies are needed to shed light on potential subgroup(s) of patients who may respond to gabapentin and/or pregabalin, such as those based on genetic liability for alcohol consumption. For example, it has been shown that patients with history of alcohol withdrawal symptoms respond best to gabapentin 23 and preliminary works suggests that gabapentin may help people with other substance use disorders such as cannabis user disorder 63. Additionally, gabapentin may provide additional benefits on protracted withdrawal, alcohol-related sleep disturbances and negative affect 64,65. Furthermore, these medications may offer an alternative for patients with AUD and comorbid neuropathic pain, or seizure conditions, provided that use is closely monitored and controlled.

The use of both gabapentin and pregabalin use carries some risks. Pregabalin has a higher recognized abuse liability consistent with it being a Schedule V controlled drug in the United States, while gabapentin, though federally unscheduled, has been associated with misuse in certain populations (e.g., in those with AUD and comorbid opioid use disorder; 66) and is regulated in some states 42,67. In addition, our previous work leveraging the same data source as the present study found that gabapentin use was associated with increased risk of falls, fractures, and altered mental status 68. These established safety profiles underscore the need for careful monitoring and judicious prescribing for both medications. Importantly, our findings suggest modulation of the α2δ subunit may be a promising therapeutic pathway. The side-effect and misuse risks of current gabapentinoids highlight the potential value of developing or identifying alternative agents targeting similar mechanisms with improved safety profiles.

Although the magnitude of effect for gabapentin was comparable to that observed in other repurposing studies for AUD, including recent work that evaluated glucagon-like peptide-1 receptor agonists 69 and spironolactone 70, the effects observed for pregabalin, particularly among individuals with AUD and hazardous drinking, were larger than those typically reported in the literature. These findings suggest that while gabapentin’s effects may be modest at the population level, targeting the α2δ calcium channels may hold greater promise, either through more potent agents or optimized treatment strategies. Future studies could investigate whether the effect of gabapentinoids may become more robust if these medications are combined with others; for example, an initial clinical study suggested so when combining gabapentin with naltrexone 71. Similar approaches could be taken with other emerging new targets under investigation in AUD 72,73.

This study’s strengths include very large sample sizes and that its results are generalizable to real-world patients with different comorbid medical and mental health conditions and taking concurrent medications. However, there are limitations to the study. First, despite our use of propensity-score matching to control for baseline differences between groups, we cannot rule out unmeasured variables that could confound the results, such as patient motivation to reduce alcohol use and participation in non-pharmacological treatments. Second, this study was conducted on Veterans receiving VA care, who are, on average, older and have a higher prevalence of chronic health conditions and risk behaviors than the general US population 74–76. However, previous research has established that after adjusting for age, sex, race, ethnicity, region, and residence type, all of which were accounted for in this study, total disease burden between Veterans and non-Veterans does not differ 76. Thus, effects reported in this study may be generalizable to non-Veteran populations. Third, although individuals in VA care represent a diversity of backgrounds, women represented a small proportion of individuals in the cohort, which prevented us from examining sex differences. Finally, gabapentin and pregabalin were examined without restriction to indication. The inability to capture the clinical rationale for prescribing limited adjustment for indication-related factors, which may have introduced residual confounding by indication in contrasts with unexposed comparators.

In conclusion, we found that both gabapentin and pregabalin are associated with reductions in drinking in a real-world setting, with stronger evidence for the effectiveness of pregabalin particularly for patients with diagnosed AUD or those who report hazardous or heavy episodic drinking. Randomized controlled trials directly comparing gabapentin to pregabalin are necessary to confirm these findings.

Supplementary Material

Supplement 1
media-1.xlsx (49.3KB, xlsx)
Supplement 2
media-2.pdf (152KB, pdf)

Acknowledgements:

This work uses data provided by patients and collected by the Veterans Affairs (VA) as part of their care and support.

Funding:

Dr. Kranzler is supported by the Veterans Integrated Service Network’s Mental Illness Research, Education and Clinical Center; U.S. Department of Veterans Affairs grant I01 BX004820 and NIAAA grant R01 AA030056. Dr. Gray is supported by NIAAA grant R01 AA030041 and Department of Defense grant HU0001–22-2–0066. Dr. Aliza Wingo is supported by grant I01BX005686. Dr. Thomas Wingo is supported by R01 AG075827. This research was in part supported by the Intramural Research Program (IRP) of the National Institutes of Health (NIH) as Drs. Farokhnia and Leggio are supported by the NIH IRP (NIDA/NIAAA). This research uses data from the Veteran Aging Cohort Study (VACS) and is supported by grant P01 AA029545 and U24 AA020794. The contributions of the NIH authors were made as part of their official duties as NIH federal employees, are in compliance with agency policy requirements, and are considered works of the United States Government. However, the findings and conclusions presented in this paper are those of the authors and do not necessarily reflect the views of the NIH or the U.S. Department of Health and Human Services.

Footnotes

Conflict of Interest Disclosures: Dr. Kranzler has been a member of advisory boards for Altimmune and Clearmind Medicine; a consultant to Sobrera Pharmaceuticals, Altimmune, Lilly and Ribocure; and the recipient of research funding and medication supplies for an investigator-initiated study from Alkermes and company-initiated studies by Altimmune and Lilly. Dr. Leggio reports, outside his federal employment, honoraria from the UK Medical Council on Alcohol (Editor-in-Chief for Alcohol and Alcoholism) and book royalties from Routledge (as editor of a textbook).

Disclaimer: The contents of this publication are the sole responsibility of the authors and do not necessarily reflect the official policy or position of the Uniformed Services University of the Health Sciences, the Department of Defense, the U.S. Department of Veterans Affairs, or Henry M. Jackson Foundation for the Advancement of Military Medicine.

References

  • 1.World Health Organization. Global status report on alcohol and health and treatment of substance use disorders. June 25, 2024. Accessed September 24, 2025. https://www.who.int/publications/i/item/9789240096745 [Google Scholar]
  • 2.Sacks JJ, Gonzales KR, Bouchery EE, Tomedi LE, Brewer RD. 2010 National and State Costs of Excessive Alcohol Consumption. Am J Prev Med. 2015;49(5):e73–e79. doi: 10.1016/j.amepre.2015.05.031 [DOI] [PubMed] [Google Scholar]
  • 3.Substance Abuse and Mental Health Services Administration (SAMHSA) C for BHS and Q (CBHSQ). 2024 National Survey on Drug Use and Health. August 6, 2025. Accessed December 11, 2025. https://www.samhsa.gov/data/report/2024-nsduh-detailed-tables [Google Scholar]
  • 4.Witkiewitz K, Litten RZ, Leggio L. Advances in the science and treatment of alcohol use disorder. Sci Adv. 2019;5(9). doi: 10.1126/SCIADV.AAX4043 [DOI] [Google Scholar]
  • 5.Reus VI, Fochtmann LJ, Bukstein O, et al. The American Psychiatric Association Practice Guideline for the Pharmacological Treatment of Patients With Alcohol Use Disorder. Am J Psychiatry. 2018;175(1):86–90. doi: 10.1176/APPI.AJP.2017.1750101 [DOI] [PubMed] [Google Scholar]
  • 6.Kranzler HR, Soyka M. Diagnosis and Pharmacotherapy of Alcohol Use Disorder: A Review. JAMA. 2018;320(8):815–824. doi: 10.1001/JAMA.2018.11406 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.McPheeters M, O’Connor EA, Riley S, et al. Pharmacotherapy for Alcohol Use Disorder: A Systematic Review and Meta-Analysis. JAMA. 2023;330(17):1653–1665. doi: 10.1001/JAMA.2023.19761 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Winslow BT, Onysko M, Hebert M. Medications for Alcohol Use Disorder. Am Fam Physician. 2016;93(6):457–465. Accessed October 22, 2025. https://www.aafp.org/pubs/afp/issues/2016/0315/p457.html [PubMed] [Google Scholar]
  • 9.Kranzler HR, Hartwell EE. Medications for treating alcohol use disorder: A narrative review. Alcohol, clinical & experimental research. 2023;47(7):1224–1237. doi: 10.1111/ACER.15118 [DOI] [Google Scholar]
  • 10.Litten RZ, Wilford BB, Falk DE, Ryan ML, Fertig JB. Potential medications for the treatment of alcohol use disorder: An evaluation of clinical efficacy and safety. Subst Abus. 2016;37(2):286–298. doi: 10.1080/08897077.2015.1133472 [DOI] [PubMed] [Google Scholar]
  • 11.Mason BJ. Emerging Pharmacotherapies for Alcohol Use Disorder. Neuropharmacology. 2017;122:244–253. doi: 10.1016/J.NEUROPHARM.2017.04.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Soyka M, Müller CA. Pharmacotherapy of alcoholism - an update on approved and off-label medications. Expert Opin Pharmacother. 2017;18(12):1187–1199. doi: 10.1080/14656566.2017.1349098 [DOI] [PubMed] [Google Scholar]
  • 13.Swift RM, Aston ER. Pharmacotherapy for alcohol use disorder: current and emerging therapies. Harv Rev Psychiatry. 2015;23(2):122–133. doi: 10.1097/HRP.0000000000000079 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Rentsch CT, Malone SG, Shi M, et al. Bridging Genomics and Pharmacoepidemiology to Expand Treatment Options for Alcohol Use Disorder. medRxiv. Published online January 11, 2026:2026.01.09.26343615. doi: 10.64898/2026.01.09.26343615 [DOI] [Google Scholar]
  • 15.Wingo AP, Liu Y, Vattathil SM, et al. Multiancestry brain pQTL fine-mapping and integration with genome-wide association studies of 21 neurologic and psychiatric conditions. Nat Genet. 2025;57(9):2156–2165. doi: 10.1038/S41588-025-02291-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Cruccu G, Truini A. A review of Neuropathic Pain: From Guidelines to Clinical Practice. Pain Ther. 2017;6(Suppl 1):35–42. doi: 10.1007/S40122-017-0087-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Garcia-Borreguero D, Silber MH, Winkelman JW, et al. Guidelines for the first-line treatment of restless legs syndrome/Willis–Ekbom disease, prevention and treatment of dopaminergic augmentation: a combined task force of the IRLSSG, EURLSSG, and the RLS-foundation. Sleep Med. 2016;21:1–11. doi: 10.1016/J.SLEEP.2016.01.017 [DOI] [PubMed] [Google Scholar]
  • 18.Kanner AM, Ashman E, Gloss D, et al. Practice guideline update summary: Efficacy and tolerability of the new antiepileptic drugs I: Treatment of new-onset epilepsy: Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology and the American Epilepsy Society. Neurology. 2018;91(2):74–81. doi: 10.1212/WNL.0000000000005755 [DOI] [PubMed] [Google Scholar]
  • 19.Fischler PV, Soyka M, Seifritz E, Mutschler J. Off-label and investigational drugs in the treatment of alcohol use disorder: A critical review. Front Pharmacol. 2022;13:927703. doi: 10.3389/FPHAR.2022.927703/FULL [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Kranzler HR, Feinn R, Morris P, Hartwell EE. A meta-analysis of the efficacy of gabapentin for treating alcohol use disorder. Addiction (Abingdon, England). 2019;114(9):1547–1555. doi: 10.1111/ADD.14655 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Calandre EP, Rico-Villademoros F, Slim M. Alpha2delta ligands, gabapentin, pregabalin and mirogabalin: a review of their clinical pharmacology and therapeutic use. Expert Rev Neurother. 2016;16(11):1263–1277. doi: 10.1080/14737175.2016.1202764 [DOI] [PubMed] [Google Scholar]
  • 22.Mason BJ, Light JM, Williams LD, Drobes DJ. Proof-of-concept human laboratory study for protracted abstinence in alcohol dependence: effects of gabapentin. Addiction biology. 2009;14(1):73–83. doi: 10.1111/J.1369-1600.2008.00133.X [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Anton RF, Latham P, Voronin K, et al. Efficacy of Gabapentin for the Treatment of Alcohol Use Disorder in Patients With Alcohol Withdrawal Symptoms: A Randomized Clinical Trial. JAMA Intern Med. 2020;180(5):728–736. doi: 10.1001/JAMAINTERNMED.2020.0249 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Brower KJ, Myra Kim H, Strobbe S, Karam-Hage MA, Consens F, Zucker RA. A randomized double-blind pilot trial of gabapentin versus placebo to treat alcohol dependence and comorbid insomnia. Alcohol Clin Exp Res. 2008;32(8):1429–1438. doi: 10.1111/J.1530-0277.2008.00706.X [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Chompookham P, Rukngan W, Nilaban S, Suwanmajo S, Yoosom P, Kalayasiri R. A randomized trial of low-dose gabapentin for post hospitalization relapse prevention in a Thai clinical sample of alcohol dependence. Psychiatry Res. 2018;270:34–40. doi: 10.1016/j.psychres.2018.09.002 [DOI] [PubMed] [Google Scholar]
  • 26.Furieri FA, Nakamura-Palacios EM. Gabapentin reduces alcohol consumption and craving: a randomized, double-blind, placebo-controlled trial. J Clin Psychiatry. 2007;68(11):1691–1700. doi: 10.4088/JCP.V68N1108 [DOI] [PubMed] [Google Scholar]
  • 27.Mariani JJ, Pavlicova M, Basaraba C, et al. Pilot randomized placebo-controlled clinical trial of high-dose gabapentin for alcohol use disorder. Alcohol Clin Exp Res. 2021;45(8):1639–1652. doi: 10.1111/ACER.14648 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Mason BJ, Quello S, Goodell V, Shadan F, Kyle M, Begovic A. Gabapentin treatment for alcohol dependence: a randomized clinical trial. JAMA Intern Med. 2014;174(1):70–77. doi: 10.1001/JAMAINTERNMED.2013.11950 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Falk DE, Ryan ML, Fertig JB, et al. Gabapentin Enacarbil Extended-Release for Alcohol Use Disorder: A Randomized, Double-Blind, Placebo-Controlled, Multisite Trial Assessing Efficacy and Safety. Alcohol Clin Exp Res. 2019;43(1):158–169. doi: 10.1111/ACER.13917 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Trevisan LA, Ralevski E, Keegan K, et al. Alcohol detoxification and relapse prevention using valproic acid versus gabapentin in alcohol-dependent patients. Addict Disord Their Treat. 2008;7(3):119–128. doi: 10.1097/ADT.0B013E31812E6A3C [DOI] [Google Scholar]
  • 31.Bragg S, Marrison ST, Haley S. Diabetic Peripheral Neuropathy: Prevention and Treatment. Am Fam Physician. 2024;109(3):226–232. Accessed October 22, 2025. https://www.aafp.org/pubs/afp/issues/2024/0300/diabetic-peripheral-neuropathy.html [PubMed] [Google Scholar]
  • 32.Derry S, Cording M, Wiffen PJ, Law S, Phillips T, Moore RA. Pregabalin for pain in fibromyalgia in adults. Cochrane Database Syst Rev. 2016;9(9). doi: 10.1002/14651858.CD011790.PUB2 [DOI] [Google Scholar]
  • 33.French J, Kwan P, Fakhoury T, et al. Pregabalin monotherapy in patients with partial-onset seizures: A historical-controlled trial. Neurology. 2014;82(7):590. doi: 10.1212/WNL.0000000000000119 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Tong C, Zhengyao Z, Mei L, Dongpo S, Qian H, Fengqun M. Pregabalin and Gabapentin in Patients with Spinal Cord Injury-Related Neuropathic Pain: A Network Meta-Analysis. Pain Ther. 2021;10(2):1497–1509. doi: 10.1007/S40122-021-00302-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Di Nicola M, Martinotti G, Tedeschi D, et al. Pregabalin in outpatient detoxification of subjects with mild-to-moderate alcohol withdrawal syndrome. Hum Psychopharmacol. 2010;25(3):268–275. doi: 10.1002/HUP.1098, [DOI] [PubMed] [Google Scholar]
  • 36.Mariani JJ, Pavlicova M, Choi CJ, et al. An open-label pilot study of pregabalin pharmacotherapy for alcohol use disorder. American Journal of Drug and Alcohol Abuse. 2021;47(4):467–475. doi: 10.1080/00952990.2021.1901105, [DOI] [PubMed] [Google Scholar]
  • 37.Martinotti G, Nicola M, Tedeschi D, Mazza M, Janiri L, Bria P. Efficacy and safety of pregabalin in alcohol dependence. Adv Ther. 2008;25(6):608–618. doi: 10.1007/S12325-008-0066-2 [DOI] [PubMed] [Google Scholar]
  • 38.Martinotti G, Di Nicola M, Frustaci A, et al. Pregabalin, tiapride and lorazepam in alcohol withdrawal syndrome: A multi-centre, randomized, single-blind comparison trial. Addiction. 2010;105(2):288–299. doi: 10.1111/J.1360-0443.2009.02792.X, [DOI] [PubMed] [Google Scholar]
  • 39.Förg A, Hein J, Volkmar K, et al. Efficacy and safety of pregabalin in the treatment of alcohol withdrawal syndrome: A randomized placebo-controlled trial. Alcohol and Alcoholism. 2012;47(2):149–155. doi: 10.1093/ALCALC/AGR153, [DOI] [PubMed] [Google Scholar]
  • 40.Martinotti G, Di Nicola M, Tedeschi D, et al. Pregabalin versus naltrexone in alcohol dependence: A randomised, double-blind, comparison trial. Journal of Psychopharmacology. 2010;24(9):1367–1374. doi: 10.1177/0269881109102623, [DOI] [PubMed] [Google Scholar]
  • 41.Bockbrader HN, Wesche D, Miller R, Chapel S, Janiczek N, Burger P. A comparison of the pharmacokinetics and pharmacodynamics of pregabalin and gabapentin. Clin Pharmacokinet. 2010;49(10):661–669. doi: 10.2165/11536200-000000000-00000 [DOI] [PubMed] [Google Scholar]
  • 42.Evoy KE, Sadrameli S, Contreras J, Covvey JR, Peckham AM, Morrison MD. Abuse and Misuse of Pregabalin and Gabapentin: A Systematic Review Update. Drugs. 2021;81(1):125–156. doi: 10.1007/S40265-020-01432-7 [DOI] [PubMed] [Google Scholar]
  • 43.Grauer JS, Cramer JD. Association of State-Imposed Restrictions on Gabapentin with Changes in Prescribing in Medicare. J Gen Intern Med. 2022;37(14):3630–3637. doi: 10.1007/S11606-021-07314-2/TABLES/3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.NHS England. Handling of gabapentin and pregabalin as schedule 3 controlled drugs in health and justice commissioned services. January 1, 2019. Accessed October 20, 2025. https://www.england.nhs.uk/wp-content/uploads/2019/02/handling-pregabalin-and-gabapentin.pdf [Google Scholar]
  • 45.Singh GK. Area Deprivation and Widening Inequalities in US Mortality, 1969–1998. https://doi.org/102105/AJPH9371137. 2011;93(7):1137–1143. doi: 10.2105/AJPH.93.7.1137 [DOI] [Google Scholar]
  • 46.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–383. doi: 10.1016/0021-9681(87)90171-8 [DOI] [PubMed] [Google Scholar]
  • 47.Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43(11):1130–1139. doi: 10.1097/01.MLR.0000182534.19832.83 [DOI] [PubMed] [Google Scholar]
  • 48.Justice AC, Tate JP. Strengths and Limitations of the Veterans Aging Cohort Study Index as a Measure of Physiologic Frailty. AIDS Res Hum Retroviruses. 2019;35(11–12):1023–1033. doi: 10.1089/AID.2019.0136 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Tate JP, Justice AC, Hughes MD, et al. An internationally generalizable risk index for mortality after one year of antiretroviral therapy. AIDS. 2013;27(4):563. doi: 10.1097/QAD.0B013E32835B8C7F [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Bradley KA, Debenedetti AF, Volk RJ, Williams EC, Frank D, Kivlahan DR. AUDIT-C as a brief screen for alcohol misuse in primary care. Alcohol Clin Exp Res. 2007;31(7):1208–1217. doi: 10.1111/J.1530-0277.2007.00403.X [DOI] [PubMed] [Google Scholar]
  • 51.Bush K, Kivlahan DR, McDonell MB, Fihn SD, Bradley KA. The AUDIT alcohol consumption questions (AUDIT-C): an effective brief screening test for problem drinking. Ambulatory Care Quality Improvement Project (ACQUIP). Alcohol Use Disorders Identification Test. Arch Intern Med. 1998;158(16):1789–1795. doi: 10.1001/ARCHINTE.158.16.1789 [DOI] [PubMed] [Google Scholar]
  • 52.Bradley KA, Rubinsky AD, Lapham GT, et al. Predictive validity of clinical AUDIT-C alcohol screening scores and changes in scores for three objective alcohol-related outcomes in a Veterans Affairs population. Addiction (Abingdon, England). 2016;111(11):1975–1984. doi: 10.1111/ADD.13505 [DOI] [PubMed] [Google Scholar]
  • 53.Jack HE, Oliver MM, Berger DB, Bobb JF, Bradley KA, Hallgren KA. Association between clinical measures of unhealthy alcohol use and subsequent year hospital admissions in a primary care population. Drug Alcohol Depend. 2023;245. doi: 10.1016/j.drugalcdep.2023.109821 [DOI] [Google Scholar]
  • 54.Justice AC, McGinnis KA, Tate JP, et al. Risk of Mortality and Physiologic Injury Evident with Lower Alcohol Exposure Among HIV Infected Compared with Uninfected Men. Drug Alcohol Depend. 2016;161:95. doi: 10.1016/J.DRUGALCDEP.2016.01.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Blake HA, Leyrat C, Mansfield KE, et al. Propensity scores using missingness pattern information: a practical guide. Stat Med. 2020;39(11):1641. doi: 10.1002/SIM.8503 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.D’Agostino RB, Rubin DB. Estimating and Using Propensity Scores with Partially Missing Data. J Am Stat Assoc. 2000;95(451):749. doi: 10.2307/2669455 [DOI] [Google Scholar]
  • 57.Austin PC . Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med. 2009;28(25):3083–3107. doi: 10.1002/SIM.3697 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Donald SG, Lang K. Inference with Difference-in-Differences and Other Panel Data. Rev Econ Stat. 2007;89(2):221–233. doi: 10.1162/REST.89.2.221 [DOI] [Google Scholar]
  • 59.Lechner M, Lechner, Michael. The Estimation of Causal Effects by Difference-in-Difference Methods. Foundations and Trends(R) in Econometrics. 2011;4(3):165–224. doi: 10.1561/0800000014 [DOI] [Google Scholar]
  • 60.Rentsch CT, Fiellin DA, Bryant KJ, Justice AC, Tate JP. Association Between Gabapentin Receipt for Any Indication and Alcohol Use Disorders Identification Test-Consumption Scores Among Clinical Subpopulations With and Without Alcohol Use Disorder. Alcohol Clin Exp Res. 2019;43(3):522–530. doi: 10.1111/ACER.13953 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Ray LA, Grodin EN, Baskerville WA, Donato S, Cruz A, Montoya AK. Identifying responders to gabapentin for the treatment of alcohol use disorder: an exploratory machine learning approach. Alcohol Alcohol. 2025;60(3):agaf010. doi: 10.1093/ALCALC/AGAF010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Laska EM, Siegel CE, Lin Z, Bogenschutz M, Marmar CR. Gabapentin Enacarbil Extended-Release Versus Placebo: A Likely Responder Reanalysis of a Randomized Clinical Trial. Alcohol Clin Exp Res. 2020;44(9):1875–1884. doi: 10.1111/ACER.14414 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.Mason BJ, Crean R, Goodell V, et al. A proof-of-concept randomized controlled study of gabapentin: effects on cannabis use, withdrawal and executive function deficits in cannabis-dependent adults. Neuropsychopharmacology. 2012;37(7):1689–1698. doi: 10.1038/NPP.2012.14 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Hoffman M, Voronin K, Book SW, Prisciandaro J, Bristol EJ, Anton RF. Sleep as an Important Target or Modifier in Alcohol Use Disorder Clinical Treatment: Example From a Recent Gabapentin Randomized Clinical Trial. J Addict Med. 2024;18(5):520–525. doi: 10.1097/ADM.0000000000001316 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Mason BJ, Quello S, Shadan F. Gabapentin for the treatment of alcohol use disorder. Expert Opin Investig Drugs. 2018;27(1):113–124. doi: 10.1080/13543784.2018.1417383 [DOI] [Google Scholar]
  • 66.Castillo F, Jones JD, Luba RR, et al. Gabapentin increases the abuse liability of alcohol alone and in combination with oxycodone in participants with co-occurring opioid and alcohol use disorder. Pharmacol Biochem Behav. 2022;221. doi: 10.1016/J.PBB.2022.173482 [DOI] [Google Scholar]
  • 67.Peckham AM, Evoy KE, Ochs L, Covvey JR. Gabapentin for Off-Label Use: Evidence-Based or Cause for Concern? Subst Abuse. 2018;12. doi: 10.1177/1178221818801311 [DOI] [Google Scholar]
  • 68.Rentsch CT, Morford KL, Fiellin DA, Bryant KJ, Justice AC, Tate JP. Safety of Gabapentin Prescribed for Any Indication in a Large Clinical Cohort of 571,718 US Veterans with and without Alcohol Use Disorder. Alcohol Clin Exp Res. 2020;44(9):1807–1815. doi: 10.1111/ACER.14408 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Farokhnia M, Tazare J, Pince CL, et al. Glucagon-like peptide-1 receptor agonists, but not dipeptidyl peptidase-4 inhibitors, reduce alcohol intake. J Clin Invest. 2025;135(9). doi: 10.1172/JCI188314 [DOI] [Google Scholar]
  • 70.Farokhnia M, Rentsch CT, Chuong V, et al. Spironolactone as a potential new pharmacotherapy for alcohol use disorder: convergent evidence from rodent and human studies. Mol Psychiatry. 2022;27(11):4642–4652. doi: 10.1038/S41380-022-01736-Y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 71.Anton RF, Myrick H, Wright TM, et al. Gabapentin combined with naltrexone for the treatment of alcohol dependence. Am J Psychiatry. 2011;168(7):709–717. doi: 10.1176/APPI.AJP.2011.10101436 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 72.Heilig M, Witkiewitz K, Ray LA, Leggio L. Novel medications for problematic alcohol use. J Clin Invest. 2024;134(11). doi: 10.1172/JCI172889 [DOI] [Google Scholar]
  • 73.Loften A, Farokhnia M, Vendruscolo LF, Leggio L. Neuroendocrinology meets addiction: Emerging pharmacotherapies on the horizon. J Intern Med. 2025;298(5):392–423. doi: 10.1111/JOIM.70021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 74.Hoerster KD, Lehavot K, Simpson T, McFall M, Reiber G, Nelson KM. Health and Health Behavior Differences: U.S. Military, Veteran, and Civilian Men. Am J Prev Med. 2012;43(5):483–489. doi: 10.1016/J.AMEPRE.2012.07.029 [DOI] [PubMed] [Google Scholar]
  • 75.Lehavot K, Hoerster KD, Nelson KM, Jakupcak M, Simpson TL. Health Indicators for Military, Veteran, and Civilian Women. Am J Prev Med. 2012;42(5):473–480. doi: 10.1016/J.AMEPRE.2012.01.006 [DOI] [PubMed] [Google Scholar]
  • 76.Eibner C, Krull H, Brown KM, et al. Current and Projected Characteristics and Unique Health Care Needs of the Patient Population Served by the Department of Veterans Affairs. Rand Health Q. 2016;5(4):13. doi: 10.7249/rr1165.1 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1
media-1.xlsx (49.3KB, xlsx)
Supplement 2
media-2.pdf (152KB, pdf)

Articles from medRxiv are provided here courtesy of Cold Spring Harbor Laboratory Preprints

RESOURCES