Background/Purpose: Adipose tissue deposited around the mesentery is highly associated with insulin resistance and cardiovascular disease (CVD). Compared to BMI, waist circumference (WC) has been shown to be a better marker of adiposity around the abdomen in the general population. However, it is not known if it is similarly associated with CVD and type 2 diabetes (T2D) in RA patients.  We sought to determine whether WC is a better T2D and CVD predictor than BMI.

Methods: RA patients with ≥1 year participation in FORWARD, National Databank for Rheumatic Diseases without baseline T2D from 1998 through 2017 were assessed for T2D (self-report or initiating of antidiabetic medication) and CV events (myocardial infarction, stroke and heart failure validated from hospital/ death records). WC was measured by the patients according to World Health Organization (WHO) 2008 measurement guideline. WC ≥102cm in men and ≥88cm in women was considered as abdominal obesity. BMI was categorized according to WHO classification. Cox proportional hazard models with adjustment for sociodemographics, comorbidities, RA severity measures and treatment (DMARDs and glucocorticoids) constructed to estimate T2D and CVD risk. WC and BMI were evaluated in different models. WC and BMI interaction in the same model was also assessed.

Results: The study included 2,177 RA patients (mean [SD] age 66 [12] years) of which 28% were obese (BMI≥30kg/m²) and 52% had abdominal obesity. During a median (IQR) 5.9 (2.6-11.1) years of follow-up, 229 incident T2D cases and 94 CV events were observed. The incidence rate (95%CI) of T2D was slightly higher in obese patients than patients with abdominal obesity (33.4 [27.8-40.1] vs. 24.1 [20.6-28.2]); but, CVD incidence (8.6 [6.5-11.1] vs.8.7 [6.1-12.4]) was similar.  In adjusted models both obesity (HR 1.86 [1.33-2.59]) and abdominal obesity (HR 1.59 [1.19-2.13]) were significantly associated with incident T2D, however the risk with obesity was higher than the abdominal obesity. For the CVD, neither BMI based obesity nor WC based abdominal obesity were significantly associated with increased risk (Table). When women and men analyzed separately, abdominal obesity was a more prominent predictor of T2D in women (HR 1.75 [1.25-2.44] than men (1.05 [0.54-2.06]). In analysis of interaction of BMI and WC in normal weight and overweight patients, abdominal obesity in the presence of normal weight tended to increase the T2D risk.

Conclusion: Both BMI based obesity and WC based abdominal obesity were strong risk factors for T2D in RA patients but BMI predicted T2D better. Neither of them was associated with CVD. Obesity in RA can have a paradoxical effect in RA in terms of mortality which might be a reason why obesity and abdominal obesity were not associated with CVD, the leading cause of death in RA. Regardless, WC measurement in RA might be helpful in particularly nonobese female patients to estimate T2D risk.