Background Chronic inflammation, the basic feature of rheumatoid arthritis (RA), plays a major role in accelerated atherosclerosis through its influence on insulin resistance (IR), lipid status, and other risk factors.
Objectives To investigate the prevalence of IR in RA pts with normal glycoregulation and to determine its association with inflammatory disease burden, or therapy exposure.
Methods The study population included 90 RA pts (mean age 52.4±9.9 yrs), 86.7% females, with median disease duration of 9 yrs (range 4-13). Clinical work-up included determination of the body mass index, waist cucumference (WC), and blood pressure (TA). RA specific evaluation comprised assessment of disease activity (mDAS28-SE, mDAS28-CRP, mHAQ) as well as antirheumatic treatment. All pts were on disease modifying antirheumatic drugs, 59/90 (65.6%), steroids (median 5 mg/day, range 5-10), and biologic therapy 25/90 (27.8%). Pts on steroids >10 mg/day were excluded. Laboratory analyses included inflammation markers (SE, hsCRP) and lipids. Serum specific insulin was measured by a sensitive ELISA, which employs a monoclonal antibody to insulin. IR was calculated using the updated-computer Homeostasis Model Assessment (HOMA2-IR), based on fasting plasma glucose and serum specific insulin concentrations. The output of the HOMA2 model was calibrated to give IR of 1 as normal. Therefore, values were considered abnormal when HOMA2-IR was >1. Because of high coefficient of variation (53%) values of HOMA2-IR were logarithmically transformed.
Results Increased logHOMA2-IR was noted in 71/90 (74.4%) pts with median of 1.4 (range 1.0-2.3). Pts with high disease activity (DAS28-SE≥5.1) had statisticaly higher concentration of specific insulin 79 (58-120) pmol/l and higher logHOMA2-IR=1.7 (1.2-2.5) than those with DAS28-SE<5.1: insulin 57 (39-91), logHOMA2-IR=1.3 (0.9-1.9) (p<0.01 for both). Importantly, both groups were comparable regarding all other factors that might influence IR (age, BMI, blood pressure, WC, and triglicerides), duration of RA and anti-inflammatory treatment inlcuding glycocorticoids. Univariant regression revealed association of logHOMA2-IR with all IR risk factors, as well as with disease activity. After adjustment for age, BMI, TA, and triglycerides in multivariant regression analysis, association of logHOMA2-IR persisted for the number of tender joints (β 0.008 (95% CI, 0.001-0.016), p=0.029), values for visual analogue scale (β 0.003 (95% CI, 0.001-0.005) p=0.010), HAQ (β 0.086 (95% CI, 0.007-0.165) p=0.034, mDAS28-SE (β 0.034 (95% CI, 0.006-0.061) p=0.017, and mDAS28-CRP (β 0.034 (95% CI, 0.006-0.063) p=0.019. On the other hand we did not find significant association with the counts of swollen joints, SE rate and hsCRP-a. Of note, we obtained a negative association of logHOMA2-IR with duration of anti-inflammatory therapy, including steroids.
Conclusions We demonstrated a significant association of disease activity and impaired insulin sensitivity in RA patients, which remains valid after correction for factors with well established influence on insulin resistance. The lack of association with swollen joint counts and acute phase reactants may implicate that the influence of inflammatory disease burden is important but not the only risk factor for IR in RA.
Disclosure of Interest None declared