|
 Rheumatoid Arthritis - Other AgentsAbstract 2000: Hydroxychloroquine Reduces Diabetes Risk in RAHydroxychloroquine (Plaquenil; HCQ) has been shown to have beneficial effects outside of its immuno-modulatory properties, including lipid and glucose lowering. Clinical trials conducted in difficult to control diabetics without inflammatory arthritis have even shown a benefit of HCQ therapy. Here, Wasko et al examine the effect of HCQ on incident diabetes in RA. Methods Data were derived from the ARAMIS cohort, a multicenter cohort of RA subjects in which periodic health outcomes assessments are obtained by semi-annual mailed questionnaire. For the study presented here, subjects with RA and no history of diabetes were recruited from 1983 to 2003 and followed to July 2004 for the outcome of incident self-reported diabetes (or use of diabetes medications). The effect of HCQ on incident diabetes risk was determined using Cox proportional hazards modeling, accounting for pertinent covariates. Results 4905 subjects without diabetes were present at study entry. Of these, 1808 (37%) reported HCQ use for some period over the 29,677 person-years of follow-up observed. Subjects receiving HCQ, compared to those never receiving the drug, were significantly more likely to be female (80 vs. 75%, respectively), were younger (54 vs. 58 years of age, respectively), with more education (13.3 vs. 12.6 years, respectively), shorter RA duration (8.4 vs. 13 years, respectively), a higher proportion of subjects with prednisone exposure (72% vs. 58%, respectively), and a higher proportion with methotrexate co-exposure (64% vs. 49%, respectively). The incidence of diabetes was 5.2 per 100 patient-years in the HCQ group compared to 8.9 per 100 patient-years in the group that was never exposed to HCQ (p=0.001). In adjusted multivariable models, the risk of diabetes was reduced by 33% in HCQ users compared to non-users (HR 0.67 (95%CI 0.48 – 0.93)). Other significant predictors of diabetes in multivariable models were increasing age, HAQ, and maximum BMI. Interestingly, methotrexate use was also protective against incident diabetes, independent of HCQ, (HR 0.61 (95% CI 0.41 – 0.90)). Conclusions In a large cohort of RA subjects, any exposure to HCQ was protective against incident self-reported diabetes, independent of confounding factors. Editorial Comment These are interesting results that help to confirm with a hard outcome the observations of glucose lowering observed with HCQ. There are, however, criticisms inherent to the study design that mean the findings are not definitive. The primary issue is confounding by indication. In clinical practice, HCQ use is often reserved for those with milder disease, or in combination with other DMARDs. Thus, the effect observed might not be due to the effect of the drug but attributable to other unmeasured effects associated with the drug, such as milder disease. In support of this, the West of Scotland Coronary Prevention Study (ref Sattar) elegantly demonstrated that incident diabetes was increased in subjects with elevated CRP, an effect that could explain the relationships observed here. That study also demonstrated that both hypertension and hyperlipidemia were potent risk factors for incident diabetes, confounders that were not available for adjustment in the ARAMIS cohort that could account for the effect of HCQ observed here. In this case, more definitive proof of the ability of HCQ to reduce the risk of diabetes would require a randomized clinical trial. It is interesting to note that the effect may not necessarily by attributable to HCQ, but could be generic for any agent capable of reducing inflammation. This is suggested by the finding that methotrexate was also protective of incident diabetes with the same magnitude as HCQ.
| |
