Background Studies indicate that different RA-treatments are associated with a decreased risk of type 2 diabetes mellitus (DM), a strong risk factor for cardiovascular disease in the general population.ion.

Objectives To estimate the incidence rate and standardized morbidity ratio (SMR) of DM in RA, and to assess the influence of RA-treatment on DM risk in RA.

Methods We identified RA-patients participating in a longitudinal prospective observational study of rheumatic diseases in the US from 2000 through 2010. All were diagnosed by a rheumatologist and completed at least two semi-annual questionnaires. DM was determined by patient self-report or use of any DM-specific medication. Patients with DM at enrollment were excluded from analyses. Participants provided all treatments taken. SMRs were estimated using the Centers for Disease Control and Prevention reported incidence rate of DM for the US stratified by calendar year, sex and age. Cox proportional hazard models were used to estimate hazard ratios (HR) to compare the 6-month lagged association of treatment with DM risk. Adjustments were made for baseline age, sex, RA-duration, ethnicity, employment status, total income, smoking status, comorbidity index, HAQ and body mass index (BMI).

Results In total, 10,853 RA-patients were included with 695 (6.4%) having DM during followup.At baseline, mean (SD) RA duration was 13.6 (11.3) years and age was 59.5 (13.2) years. Baseline mutually exclusive treatment groups included 24.6% on at least hydroxychloroquine (HCQ), 30.0% on methotrexate (MTX) alone or with TNF-inhibitor, and 42.4% on other or no DMARDs. The incidence rate (95% CI) for DM was 1.34 (1.2-1.5) per 100 person-years, and the corresponding SMR was 1.2 (1.1-1.3). After adjustments, the only individual treatments significantly associated with DM were HCQ, HR 0.59 (0.48-0.73); MTX, HR 0.8 (0.7-0.9); golimumab, HR 12.3 (3.6-41.5); and prednisone, HR 1.3 (1.1-1.5). Mutually exclusive treatment groups showed a differential effect of MTX and TNF based on being on prednisone or not; protective effects of MTX with or without TNF (HR 0.8, 0.7-1.0) in individuals not on prednisone, but not in individuals on prednisone (HR 1.1, 0.8-1.3). HCQ with prednisone was still protective, HR 0.6 (0.4-0.9). Of all risk factors tested, BMI had the greatest association with DM followed by HCQ usage.

Conclusions HCQ usage in RA patients was associated with a decrease in DM incidence independent of other treatment and covariate adjustments while MTX alone and MTX with anti-TNF therapy was similarly protective when prednisone was not concomitant.

Disclosure of Interest None Declared