By Ellen Feldman, MD
Synopsis: This retrospective study examining fall risk among older adults taking gabapentin compared to duloxetine finds gabapentin users exhibit a markedly lower risk of fall-related visits at six months (hazard ratio = 0.52) yet finds no difference in the incidence of severe falls.
Source: Chaitoff A, Desai RJ, Choudhry NK, et al. Assessing the risk for falls in older adults after initiating gabapentin versus duloxetine. Ann Intern Med. 2025;178(2):187-198.
Falls in older adults are a critical concern for primary care professionals (PCPs), given their strong association with morbidity, mortality, and healthcare costs.1 Among the many risk factors, medications are among the most modifiable contributors to fall risk, particularly central nervous system-active drugs.1,2 Gabapentin, initially developed as an anticonvulsant, has seen widespread off-label use for neuropathic pain. However, many clinicians worry about side effects of this agent, such as dizziness and sedation that may add to the risk for falls in this vulnerable population.3
Previous studies comparing fall rates in patients taking gabapentin compared to unmedicated individuals confirmed the increased risk, but these studies may have been limited or skewed by confounding factors.4
Chaitoff et al sought to address some of these limitations by using an active comparator design, selecting duloxetine, a serotonin-norepinephrine reuptake inhibitor often used for treatment of neuropathic pain, as the comparator.5 This approach attempted to provide a more accurate assessment of gabapentin’s fall risk relative to another treatment option, rather than to untreated individuals.
The study used the IBM commercial claims database (2014-2021), which contains de-identified inpatient, outpatient, and pharmacy claims for insured patients. Included were older adults (ages 65 years and older) diagnosed with diabetic neuropathy, postherpetic neuralgia, or fibromyalgia who were newly prescribed either gabapentin or duloxetine. Notably, the focus on new users eliminated mixing in long-term users of these medications, thus avoiding another source of bias.
Exclusion criteria included conditions that could predispose a patient to fall risks or could influence medication choice. Thus, patients with diagnoses such as depression, anxiety, seizures, or cancer were not eligible for inclusion in the study. Patients with a prescription filled for gabapentin or duloxetine in the last year also were excluded. After applying these criteria, an analysis was conducted on 52,152 new gabapentin users and 4,934 new duloxetine users.
Since a randomized controlled trial (RCT) was not feasible with this dataset, Chaitoff et al employed a target trial emulation — a method designed to approximate the conditions of an RCT. This approach involved carefully adjusting for confounding variables (66 total in this study), establishing clear eligibility criteria and accounting for additional factors that could predispose patients to falls.
Key findings included:
- Primary outcome: The primary analysis assessed the hazard of experiencing any fall-related healthcare visit within six months of medication initiation, using diagnosis codes to identify falls. Gabapentin users had a significantly lower hazard ratio (HR) for falls compared to duloxetine users (HR, 0.52; 95% confidence interval [CI], 0.43-0.64).
- Severe fall-related events: No significant differences were found between the two groups in terms of hip fractures or fall-related hospitalizations.
- Subgroup analysis: Among people with a history of falls, there was no significant difference in fall risk between the two drugs, suggesting that prior falls may be a more powerful predictor than medication choice.
- Sensitivity analysis: Findings remained consistent across multiple adjustments, including alternate definitions of falls and subgroup analysis stratified by frailty level.
Commentary
This well-designed study has direct clinical relevance for the PCP managing neuropathic pain in older adults. The findings challenge previous assumptions that gabapentin is a primary contributor to fall risk, at least within the first six months of treatment. Instead, the results suggest that duloxetine may have a comparable or even higher fall risk in certain populations, warranting further investigation. The study also highlights that fall risk in the early months of therapy may be independent of the specific agent prescribed, emphasizing the importance of comprehensive fall risk assessment beyond medication choice alone. These findings could have significant implications for prescribing practices in older adults with neuropathic pain.
Notable strengths of this study include the robust methodology with an active comparator approach (comparing gabapentin to another neuropathic agent rather than non-users), which mitigates confounding by indication. The overall use of a target trial emulation framework strengthens causal interference by reducing biases associated with observational studies.
Additionally, the use of claims data from more than 57,000 patients allows for strong, real-world analysis and statistical power, leaning toward the ability to generalize results to older adults in routine clinical practice.
However, there are some notable limitations as well. Despite rigorous statistical adjustment, unmeasured factors (such as pain severity, mobility status, and physician prescribing biases) may have influenced fall risk. For example, clinicians concerned about fall risk may preferentially prescribe gabapentin to less frail patients, potentially skewing results.
The relatively short follow-up time of six months also limits conclusions about long-term fall risk. Gabapentin often is titrated gradually while duloxetine titration involves fewer steps. Future studies comparing these agents over time are necessary.
Finally, the reliance on claims data likely underestimates the true incidence of falls, particularly if medical attention was not required. Additionally, these databases may not fully capture the most vulnerable older adults, such as those in long-term care facilities, where gabapentin’s fall risk could be higher.
A key takeaway for the PCP from this study is that the decision to prescribe gabapentin or duloxetine should be individualized, considering patient facility, comorbidities, and potential for dose titration. Further research is needed to explore fall risk beyond the first six months of treatment and in a more diverse patient population, including those in skilled nursing facilities.
While this study suggests gabapentin’s relative safety in respect to falls, clinicians should remain cautious about polypharmacy, central nervous system depression, and dose-dependent effects that may not have been fully covered in this relatively short-term follow-up period. Ultimately, these findings reinforce the need for nuanced decision-making in the pharmacologic management of neuropathic pain. While gabapentin has been subject to increasing scrutiny, its risks should be weighed against those of alternative treatments rather than assumed to be universally harmful. This study provides an important contribution to the ongoing evaluation of gabapentin use in older adults, helping clinicians weigh efficacy and safety in a population highly vulnerable to adverse drug effects.
Ellen Feldman, MD, works for Altru Health System, Grand Forks, ND.
References
- Howland J, Peterson EW. The critical role of primary care health care professionals in referring older adults to community-based fall prevention programs. Front Public Health. 2024;12:1377972.
- Rodrigues D, Silvestre S, Monteiro C, Duarte AP. Medication and the risk of falls: An analysis of adverse drug reactions reported to the Portuguese pharmacovigilance system. J Clin Med. 2023;12(23):7268.
- Peckham AM, Evoy KE, Ochs L, Covvey JR. Gabapentin for off-label use: Evidence-based or cause for concern? Subst Abuse. 2018;12:1178221818801311.
- Nguyen VT, Engleton M, Davison M, et al. Risk of bias in observational studies using routinely collected data of comparative effectiveness research: A meta-research study. BMC Med. 2021;19(1):279.
- Wu C-S, Huang Y-J, Ko Y-C, Lee C-H. Efficacy and safety of duloxetine in painful diabetic peripheral neuropathy: A systematic review and meta-analysis of randomized controlled trials. Syst Rev. 2023;12(1):53.
This retrospective study examining fall risk among older adults taking gabapentin compared to duloxetine finds gabapentin users exhibit a markedly lower risk of fall-related visits at six months (hazard ratio = 0.52) yet finds no difference in the incidence of severe falls.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content