Sex-specific disparities in disease activity scores among patients with axial spondyloarthritis and their implications for evaluating the response to tumor necrosis factor alpha inhibitor therapy

Table 3 Multivariate linear regression analysis and inverse probability of treatment weighting (IPTW) of change in each outcome from baseline to the first follow-up by sex. All analyses corrected for variables^a that were significantly different between sexes in baseline characteristics, baseline value of each outcome, and the time interval between baseline to the first follow-up

Outcome	Multivariate linear regression analysis			IPTW
Outcome	β^b (95% CI)	SE	p-value	β^b	SE	p-value
BASDAI	0.343 (0.079–0.607)	0.135	0.011	0.262 (0.028–0.496)	0.119	0.028
ASDAS	0.079 (-0.501–0.209)	0.066	0.235	-0.011 (-0.121–0.099)	0.056	0.846
PGA	0.491 (0.208–0.775)	0.145	0.001	0.028 (-0.218–0.273)	0.125	0.823

IPTW inverse probability of treatment weighting, CI confidence interval, SE standard error, BASDAI Bath Ankylosing Spondylitis Disease Activity Score, ASDAS Ankylosing Spondylitis Disease Activity Score, CRP C-reactive protein, PGA Patient Global Assessment of Disease.
^aIncluded variables were age, disease duration, HLA-B27 positivity, radiographic changes, cigarette smoking, peripheral arthritis, uveitis, family history of spondyloarthritis, enthesitis, and type of tumor necrosis factor alpha inhibitors.
^bThe regression coefficient was calculated for the change in the disease activity measure in males compared to females. A positive coefficient indicated a greater change in the disease activity measure in males, while a negative coefficient indicated a smaller change in the disease activity measure in males.

ISSN: 1478-6362