Abstract Body: Background: Diabetes mellitus (DM) is a risk factor for heart failure (HF). People with DM frequently have chronic pain, often treated with non-steroidal anti-inflammatory drugs (NSAID). Past research suggests that NSAID use among people with DM may be associated with cardiomyopathy. Therefore, we examined the association of NSAID use with incident HF hospitalization overall and by HF subtype among adults with DM.
Methods: Incident HF hospitalizations and HF deaths were expert-adjudicated in the REasons for Geographic and Racial Differences in Stroke (REGARDS) biracial cohort (age ≥ 45 years), recruited in 2003-07 from the continental US. Baseline DM was defined by fasting or random serum glucose, anti-diabetic medication use, and/or self-report. Baseline NSAIDs use came from medication bottle review at an in-home visit and baseline pain interference was self-reported using the SF-12 scale. Propensity for NSAIDs use was computed using 54 baseline participant characteristics. Cox proportional hazards regression, adjusted for propensity score for NSAID use, examined risk for HF overall, by pain interference, and by opioid use.
Results: We studied 5483 HF-free REGARDS participants with DM, aged 64 years; 53% female, 56% Black, 15% on NSAIDs, and 27% reporting pain interference in the past week. Over median 15-year follow up, there were 584 incident HF events. We observed no statistically significant associations between NSAID use and incident HF overall (TABLE). However, among those with high pain interference, NSAID use was associated with 1.5-fold increased risk of incident HF compared to similar NSAID nonusers, without differences by HF subtype. NSAID use was associated with 2-fold higher risk of incident HF among prescription opioid users compared to nonusers.
Conclusions: Among adults with DM, NSAID use was associated with increased risk of incident HF among those with higher pain interference or concurrent opioid intake. NSAIDs should be used in people with DM with caution.