Background/Purpose: Data on body composition (BC) in Ankylosing Spondylitis (AS) are sparse and controversial. Female AS patients have less response to TNF-α blockers and a shorter drug survival. This might be explained by differences in BC, as women have higher body fat (BF) levels and fat produces TNF-α. Our objective was to assess the BC in a cohort of AS patients naïve to TNF-α blockers, and to evaluate the differences between genders and with the reference population.

Methods: Patients included fulfilled the Modified New York criteria and were naïve to TNF-α blockers. Clinical assessments included age, disease duration, extra articular manifestations, therapies, hypertension, diabetes, dyslipidemia, lifestyle (tobacco, alcohol and marihuana use, physical activity), body mass index (BMI), ASDAS CRP and ESR, BASDAI, BASFI, BASMI. BC was assessed by whole body Dual-Energy X-ray Absorptiometry. BF%, Fat Mass Index (FMI), Fat Free Mass Index (FFMI), and android/gynoid (A/G) fat ratio were reported and compared with the reference population (percentiles, stratified by ethnicity, age and gender). Association between variables and differences between groups were assessed using t-test, Chi-Square, Mann-Whitney U test or linear regression depending on the variable.

Results: Seventy consecutive patients were included, 42 (62%) were men. The mean age was 43.9 (11.7 SD). By BMI 35.7% were overweight and 15.7% obese and there were no underweight patients. 62.9% had a BASDAI >4, and the median for ASDAS CRP was 3.4, for BASFI 4 and for BASMI 2. Disease duration, severity, extra articular manifestations, therapies, comorbidities and lifestyle were similar between men and women, except for dyslipidemia, present in 57.1% of men and in 14.3% of women (p<0.001). Women had a higher proportion of obese patients by BMI, higher BF % and FMI, and lower FFMI, but the percentile levels for BF% and FMI were similar to men, and better for FFMI. Men had a higher proportion of overweight patients and a higher A/G fat ratio (Table 1). After multivariate analysis, a higher A/G fat ratio was related to dyslipidemia in all patients. In men there was a significant association between BF% and categorical ASDAS CRP, and for FMI and FFMI with HTA. In women there was a significant association between FMI and categorical ASDAS ESR, and between FFMI and psoriasis. In the univariate analysis BASMI was significantly associated with BF% and FMI in men, and BASFI with BF % in women. 

Conclusion: Women had higher BF% and FMI than men, but when BF%, FMI and FFMI were compared to the reference population they had a better than average BC. For men it was the opposite, with a trend for higher BF% and FMI percentile levels than women. There was a trend between higher fat percentages and increased disease activity measurements. The association of A/G fat ratio and dyslipidemia was significant independently of sex, which might contribute to the increased cardiovascular risk in AS.