Efficacy and safety of duloxetine for osteoarthritis: a comparator-stratified systematic review and meta-analysis

Background Duloxetine is recommended for osteoarthritis pain, yet its role relative to placebo, active pharmacological comparators, and add-on therapy remains uncertain. We assessed the efficacy and safety of duloxetine for osteoarthritis using a comparator-stratified meta-analysis. Methods We searched major English and Chinese databases from inception to August 30, 2026, for randomised controlled trials of duloxetine in adults with osteoarthritis. Eligible trials compared duloxetine with placebo, active pharmacological treatments, or add-on/background therapy controls. Pain and physical function were the primary outcomes. We used random-effects meta-analysis, comparator-stratified and duration-based subgroup analyses, meta-regression, sensitivity analyses, and Grading of Recommendations Assessment, Development and Evaluation (GRADE) certainty assessment. Findings 20 studies, contributing 21 eligible randomised comparisons and 3455 participants, were included. Duloxetine was associated with reduced Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain (standardised mean difference −0.36, 95% CI −0.58 to −0.14) and improved physical function (−0.41, −0.58 to −0.25), but with more adverse events overall (risk ratio 1.36, 95% CI 1.03–1.79). Comparator-stratified analyses were conducted separately for placebo-controlled, active-comparator, and add-on/background-therapy settings, which addressed distinct clinical questions. Placebo-controlled studies provided the clearest evidence of benefit, whereas duloxetine did not show clear superiority over active pharmacological comparators. Add-on or combination-therapy estimates for pain and function were inconclusive: the confidence intervals crossed the null and the certainty of evidence was very low. No statistically significant associations were detected between maintenance dose, treatment duration, or study quality and WOMAC treatment effects; however, these exploratory analyses were based on a limited number of studies. Certainty of evidence ranged from low to high; evidence was low for WOMAC pain and total scores, and overall adverse events, and moderate for WOMAC physical function, stiffness. Interpretation Duloxetine was associated with improvements in pain and physical function in osteoarthritis, with the clearest benefit observed versus placebo. Evidence from active-comparator and add-on/background-therapy studies was limited, heterogeneous, and inconclusive; these analyses did not establish superiority over active treatment or a consistent additive treatment effect. Duloxetine may be considered a conditional pharmacological option for osteoarthritis pain management, particularly when non-steroidal anti-inflammatory drugs are unsuitable; however, its comparative benefit over active treatments remains uncertain.