Background/Purpose: Smoking and systemic inflammation have been shown to independently associate with radiographic spinal progression in patients with axSpA. Evidence suggests that certain socioeconomic (SE) factors (e.g. physically demanding jobs) may modify these associations.The study investigates the relationship between smoking and imaging outcomes (SI joints, spine, MRI and radiographs) over time in axSpA, to assess if SE factors first modify and if not, confound, such a relationship.

Methods: Patients with axSpA from the DESIR cohort who fulfil the ASAS axSpA classification criteria were included. Four imaging continuous outcomes (X-ray spine [mSASSS, range 0-72]; X-ray SIJs modified New York grading [mNY, 0-8]; MRI-Spine [SPARCC, range 0-414] and MRI-SIJ [SPARCC, range 0-72]) have been scored by 3 central readers independently (average score used) in one session, blinded for time order (baseline, 2 and 5 years). Smoking was tested as a binary variable: smoking status since last visit. SE variables were: age, gender, ethnicity (Caucasian vs other), job type based on ‘collar’ (blue [manual labour work] vs white [office-based work]); educational status (low vs high [university]); marital status (married vs not) and parental (number of children) status. Potential interactions between smoking and SE factors were first investigated and, if statistically (p< 0.15) and clinically relevant, models were stratified. The effect of smoking on imaging outcomes was assessed in multivariable time-varying models using generalized estimating equations adjusted for other possible confounders.

Results: In total, 425 axSpA patients were included: 225 [53%] male, 167 [40%] smokers and 287 [68%] blue collar. The mean baseline (SD) MRI-SIJ SPARCC was 4.67 (7.85), MRI-spine SPARCC: 2.73 (7.95)); mSASSS: 0.39(1.66); mNY: 1.70(1.84)). A significant interaction was found between smoking and job type with MRI-SIJ inflammation as the outcome (p=0.031) as well as with mNY grading (p=0.096). Similarly, educational status also proved to modify the association between smoking and MRI-SIJ inflammation (p=0.026). In the multivariable models, smoking was significantly associated with more MRI-SIJ inflammation over 5 years of follow-up but only in patients with a blue-collar job (β [95% CI]: 3.35 [0.54,6.17]) (table). Results were similar in the low education stratum (β [95% CI]: 2.69 [0.48, 4.91]. Smoking was not significantly associated with any of the other imaging outcomes over time. Male gender was positively associated with MRI-SIJ inflammation regardless of job type, and also with MRI-Spine inflammation and structural damage although only for blue-collar patients in the case of SIJ damage.

Conclusion: There is a strong association between smoking and MRI-SIJ inflammation over time in axSpA patients with blue collar job type or low education, irrespective of other socio-economic factors, systemic inflammation and treatment. These findings suggest a possible role for mechanical stress (seen with manual jobs) amplifying the effect of smoking on axial inflammation in axSpA. No significant relationship was found between smoking and spinal inflammation or axial damage, possibly also due to the limited imaging changes in this cohort.

« Back to 2019 ACR/ARP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/do-smoking-and-socio-economic-factors-independently-influence-imaging-outcomes-in-axial-spondyloarthritis-five-year-data-from-the-desir-cohort/