Men Showed Slightly Higher ASDAS-CRP, While Other Measures Overlapped Substantially
In a real-world outpatient cohort of Chinese patients with ankylosing spondylitis, sex differences were present but narrow. The clearest signal appeared in ASDAS-CRP, which was modestly higher in men than in women in the primary cohort of 530 patients. By contrast, BASDAI, BASFI, and ASDAS-ESR showed no multiplicity-supported between-sex differences, with substantial overlap across groups.
The finding matters because axSpA clinics often rely on both symptom-based and inflammation-linked indices to judge disease burden. Here, the men and women were not separated by a broad clinical gulf. Instead, the data point to a narrower pattern: one inflammation-sensitive score shifted modestly, while the broader measures of symptoms and function stayed largely aligned.
Physical Function Tracked Disease Activity in Both Sexes
The study directly tested whether the cross-sectional relationship between BASFI and ASDAS differed by sex. It did not. Sex-specific Spearman correlations were similar, and the men-minus-women differences in correlation were small for both ASDAS-CRP and ASDAS-ESR, with confidence intervals spanning no difference. Sex-by-ASDAS interaction models reached the same conclusion.
This is clinically important at the bedside. Functional limitation in ankylosing spondylitis often reflects a blend of active inflammation, stiffness, pain behavior, and accumulated structural change. If sex had materially altered the BASFI-ASDAS relationship, one might expect the same level of disease activity to map differently onto disability in men and women. The present data do not support that interpretation. Disease activity and physical function appear to move together in a comparable way across sexes, even if the absolute level of one inflammatory index is slightly different.
A Younger Outpatient Cohort With Long Diagnostic Lag
The primary analytical cohort included 387 men and 143 women, with a median age of 33 years. Median age at symptom onset was 24 years, and median disease duration was 8 years. The interval from symptom onset to diagnosis was 2 years in the patients with available data, reflecting a persistent delay between first symptoms and formal recognition of disease.
Several baseline features were consistent with the known epidemiology of ankylosing spondylitis. HLA-B27 positivity was common overall, but it was numerically lower in women than in men. Smoking and alcohol use were also far more frequent in men, while peripheral arthritis and family history were somewhat more common in women, although the study was not designed to make causal claims from those descriptive differences. Acute anterior uveitis, psoriasis, and inflammatory bowel disease were uncommon across the cohort.
The sample came from an outpatient rheumatology setting and was assembled from routine electronic medical records, which gives the findings practical relevance. These are not trial-selected patients filtered by narrow inflammatory thresholds. They are the kind of patients clinicians actually see: younger adults with heterogeneous symptom profiles, uneven biomarker expression, and different patterns of extra-articular disease.
The ASDAS-CRP Signal Was Real, But Small
In the primary cohort, the men-minus-women Cliff’s delta for ASDAS-CRP was 0.142, with a 95% confidence interval from 0.033 to 0.248 and a Holm-adjusted P value of 0.0476. The other primary between-sex comparisons did not remain significant after multiplicity adjustment. BASDAI and BASFI were especially similar between groups, and ASDAS-ESR showed no supported separation.
A small difference in ASDAS-CRP can still be biologically plausible. CRP-linked disease activity scores are more sensitive to systemic inflammation than purely patient-reported indices, and men with axSpA often show stronger inflammatory phenotypes in some cohorts. Even so, the effect size here was modest, not dramatic. It suggests a difference in inflammatory burden rather than a different disease entity.
One sentence captures the pattern cleanly: men had slightly higher CRP-linked disease activity, but the overall clinical picture was far more overlapping than divergent.
Sensitivity Analysis Reinforced the Same Pattern
The investigators repeated the analysis in a sensitivity cohort of 567 patients, including 424 men and 143 women. Results were concordant. The between-sex pattern remained modest and outcome-specific, and the BASFI-ASDAS association still did not show evidence of sex modification.
That replication matters because cross-sectional studies can be sensitive to missingness and cohort construction. Here, the cohort workflow was explicit: 2,135 outpatient axSpA records narrowed to 1,755 rheumatologist-documented AS cases, then to 576 patients with complete core measures, and finally to the 530-patient primary analytic sample after exclusion of records with unavailable smoking, alcohol, or inflammatory bowel disease history. The sensitivity cohort, which retained more patients, did not change the message.
What the Pattern Means for Routine Care
The results support joint interpretation of disease activity and physical function rather than sex-specific reinterpretation of their relationship. In practice, this means clinicians should not assume that a woman with ankylosing spondylitis will have a fundamentally different BASFI-to-ASDAS linkage than a man. The scales remain complementary across sexes.
At the same time, the study also argues against over-reading a single inflammatory score. ASDAS-CRP was the only measure that showed a modest male excess after adjustment for multiple comparisons. BASDAI and BASFI, which are often more sensitive to symptom burden and function, did not separate cleanly by sex. That combination suggests that sex may influence some inflammatory readouts without fundamentally changing how inflammation and function relate at the individual level.
Real-World Evidence Filled a Gap RCTs Often Miss
The cohort adds value because sex-related axSpA data often come from randomized trials with restrictive eligibility criteria. Such trials can underrepresent women who have substantial symptoms but lower inflammatory biomarkers or less obvious radiographic damage. The real-world outpatient setting used here captures a broader spectrum of disease expression, including patients who might be missed by tighter trial entry rules.
The study also fits a broader clinical concern: women with ankylosing spondylitis may experience diagnostic delay and a less classic biomarker profile, even when symptoms are clinically meaningful. Yet in this cohort, those differences did not translate into a different BASFI-ASDAS relationship. The clinical signal is subtler than stereotypes about male and female disease phenotypes would suggest.
Limits Built Into the Design
This was a cross-sectional analysis, so it cannot determine how sex influences progression over time or whether function changes differently as inflammation evolves. Medication exposure was not included in the regression models, and the study drew from a single outpatient center in China, which limits generalizability. Some baseline variables also had missingness, and the work relied on charted clinical data rather than prospectively standardized collection.
Those constraints leave specific questions open: whether the modest ASDAS-CRP difference persists longitudinally, whether treatment history explains part of the pattern, and whether similar findings appear in other Asian or non-Asian cohorts. The present data answer a narrower question well. In this outpatient population, sex differences in ankylosing spondylitis were limited, and sex did not alter the way disease activity and physical function tracked together.