Opioid initiation did not reduce cardiovascular risk in rheumatoid arthritis
In patients with rheumatoid arthritis, starting an opioid did not deliver a cardiovascular advantage over starting a non-steroidal anti-inflammatory drug. Across a large new-user, active-comparator cohort, the risk of major adverse cardiovascular events was essentially the same between treatment groups, while all-cause mortality was higher after opioid initiation. The finding cuts against a common bedside assumption: when pain is severe and inflammation is chronic, opioids are often viewed as the safer option for the heart. These data do not support that belief.
The cohort came from FORWARD, The National Databank for Rheumatic Diseases, and included patients with at least one year of participation between 1998 and 2021. Investigators used propensity score matching to compare 6866 opioid initiators with 13,689 NSAID initiators. Follow-up continued until a composite cardiovascular endpoint or death. The composite endpoint bundled myocardial infarction, stroke, heart failure, cardiovascular disease death, and venous thromboembolism, creating a broad view of vascular and cardiac harm rather than a narrow signal from one event type alone.
The event rates were close. MACE occurred in 212 opioid initiators and 253 NSAID initiators, corresponding to 20.6 and 18.9 events per 1000 person-years, respectively. The adjusted hazard ratio for MACE was 1.02, with a 95% confidence interval from 0.85 to 1.22. In practical terms, opioid initiation was neither protective nor clearly worse for the composite cardiovascular endpoint. The biology is plausible in both directions: opioids can worsen falls, sedation, respiratory status, and immobility, while NSAIDs can raise blood pressure, promote fluid retention, and aggravate thrombosis risk. In this cohort, those competing hazards converged into a neutral estimate for MACE.
The mortality signal was harder to dismiss.
All-cause death occurred in 144 opioid initiators and 150 NSAID initiators, with rates of 13.5 versus 10.8 per 1000 person-years. After weighting and adjustment, opioid initiation was associated with a 33% higher risk of death than NSAID initiation, with a hazard ratio of 1.33 and a 95% confidence interval from 1.06 to 1.67. In a chronic inflammatory disease where pain intensity, comorbidity burden, frailty, and functional decline often travel together, excess mortality after opioid use may reflect more than analgesic choice alone. Severe pain can mark advanced disease, but opioids also carry recognized risks of sedation, constipation, falls, confusion, respiratory depression, and downstream deconditioning, any of which can magnify vulnerability in a medically complex rheumatoid arthritis population.
Venous thromboembolism may be part of the excess harm signal
Among individual components of the composite endpoint, venous thromboembolism trended higher in opioid initiators than NSAID initiators. The hazard ratio was 1.41, with a 95% confidence interval from 0.84 to 2.35, so the estimate did not reach statistical certainty, but the direction of effect is clinically uncomfortable. Rheumatoid arthritis already carries a prothrombotic background driven by systemic inflammation, endothelial dysfunction, and immobility during flares. Adding opioids may worsen immobility and respiratory suppression, and severe pain itself can reduce activity further. Even without a definitive signal, the pattern supports caution when prescribing opioids to patients already at elevated thrombotic risk.
Dose intensity appeared to matter as well. Strong opioids carried a higher risk of all-cause mortality and venous thromboembolism than weak opioids when each was compared with NSAIDs, pointing toward a dose-dependent association. That gradient is difficult to ignore in clinical practice because it fits the expected pharmacology: stronger opioid exposure generally brings greater sedation, impairment, and physiologic stress. The signal does not prove causality, but it does argue against any reflexive assumption that escalating opioid potency is merely a stronger version of the same therapeutic tradeoff.
Why the cardiovascular comparison matters at the bedside
The central clinical question in rheumatoid arthritis is not whether pain should be treated; it is how to relieve pain without trading one hazard for another. NSAIDs have long been recognized as potentially problematic because of their effects on blood pressure, renal perfusion, and thrombotic balance. Opioids have often been cast as the fallback option when NSAIDs are undesirable. This study challenges that hierarchy. In a real-world RA cohort, opioids were not safer for cardiovascular outcomes and were associated with higher mortality. Clinicians looking for a clear cardiovascular escape route by switching from NSAIDs to opioids will not find one here.
The study design strengthens the relevance of the findings. A new-user active comparator approach reduces some confounding by indication because it compares patients starting one analgesic strategy with those starting another rather than with non-users who may differ dramatically in disease severity and baseline risk. Propensity score weighting further balanced measured covariates. Even so, residual confounding remains possible in observational research, especially when pain severity, frailty, functional limitation, and prescribing behavior are only partly captured in routine data. Patients selected for opioids may have had more severe or refractory pain, more disability, or more complex comorbidity than those starting NSAIDs.
Pain control in rheumatoid arthritis needs a safer hierarchy
For clinicians caring for patients with rheumatoid arthritis, the practical message is straightforward: opioid initiation should not be viewed as a cardiovascularly safer alternative to NSAID initiation. The absence of MACE excess does not make opioids benign, because mortality was higher and venous thromboembolism leaned in the same direction. Pain management in RA still needs to prioritize disease control, careful NSAID selection when appropriate, and selective opioid use only when other strategies have failed or are unsuitable. The study does not ban opioids, but it does strip away the assumption that they offer a cardiovascular free pass.
The open question is not whether opioids can reduce pain—they can—but which patients, at what dose, and for how long the benefit outweighs the combined burden of vascular, thrombotic, and systemic harm. This cohort suggests that in rheumatoid arthritis, the answer is rarely as reassuring as prescribing habits have implied.