Introduction

Spondyloarthritis is a family of chronic arthritis diseases, characterized by inflammatory back pain, peripheral arthritis, and enthesitis [1]. Ankylosing spondylitis (AS) is the typical disease among the family of spondyloarthritis, and it predominantly involves the axial joints and bilateral sacroiliac joints. It is a potentially debilitating disease, which may lead to progressive limitation of the spinal mobility, loss of the functional ability, and reduced quality of life. Inflammatory rheumatic diseases are considered to be due to a complex interaction between environment and genetic factors, which may lead to immune reactions and cause different rheumatic disorders [2]. The environmental factor might be quite important in the development of chronic rheumatic and immune disease.

Cigarette smoking, one of the most serious health problems, has been identified as one of the major environmental risk factor of rheumatic diseases, including rheumatoid arthritis (RA) [3–6] and systemic lupus erythematosus (SLE) [7, 8]. Previous studies also showed that smoking was associated with rapid disease progression with poor functional outcome of AS [9–12]. The present data about the association of cigarette smoking with AS disease entity in Chinese population is absent. The association of smoking intensity with the AS smokers has rarely been reported [13]. The aim of this study was to assess the differences between the smoking and non-smoking Chinese AS patients, including the systemic inflammation, disease activity, functional ability, and physical mobility, and also evaluate its association with smoking intensity.

Patients and methods

We consecutively collected 75 male AS patients who fulfilled the 1984 modified New York criteria [14] and visited the Outpatient Department of Division of Allergy–Immunology–Rheumatology, Taipei Veterans General Hospital. These patients were asked to fill out a questionnaire which contains the demographic data, disease activity, functional ability, and patient’s global assessment. Disease activity, functional ability, and patient’s global assessment in the AS patients were assessed by using the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) [15], Bath Ankylosing Spondylitis Functional Index (BASFI) [16], and Bath Ankylosing Spondylitis Patient Global Score (BAS-G) [17], with visual analogue scale. Meanwhile, physical examinations were performed to determine the patient’s physical mobility, including tragus-to-wall distance, lumbar flexion (modified Schober’s index), intermalleolar distance, cervical rotation (CROT), lateral lumbar flexion, fingertip-to-floor distance (FFD), chest expansion, and occiput-to-wall distance (OWD). The BASDAI, BASFI, or BAS-G scores had a final range from 0 to 10. Acute-phase reactants, including erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) levels were also measured in these AS patients. History of peripheral joint involvement was obtained through the questionnaire inquiry.

Smoking habit was evaluated by face-to-face interview: 1. □ Smoker (A. □ Current smoker: Pack per day, duration years (age started smoking); B. □ Past smoker: Pack per day, duration years (age started smoking, number of years since quitting); 2. □ Non-smoker. We calculated the smoking duration and smoking intensity (pack–years of smoking, the product of years of smoking and packs of cigarette per day).

Statistical analysis

Statistical analyses were carried out using the SPSS statistical package. The Mann–Whitney U test or Fisher’s exact test was used to analyze group differences. Correlations between variables were determined by the Spearman’s rank correlation test. Multivariate logistic regression analysis was performed to calculate the odds ratios (OR) of different clinical parameters, and the positive cutoff values of each clinical parameter were obtained by receiver operating characteristic (ROC) curve analysis. P values were regarded as being significant if they were less than 0.05.

Results

Table 1 shows the demographics and clinical characteristics of the 75 male AS patients. These AS patients were divided into two subgroups with smoking (including current smoker and past smoker) and non-smoking. Clinical parameters were compared between the two subgroups (Table 2). There was no statistically significant difference between smoking AS patients and non-smoking in age (p = 0.068), onset age (p = 0.56), and disease duration (p = 0.261). There was also no significant difference of peripheral joint involvement rate between smoking AS patients and non-smoking (p = 0.251).

Table 1 Demographic and disease characteristics of the 75 male AS patients
Full size table
Table 2 Comparison of the clinical parameters between smoking AS patients (current and past smokers) and non-smoking
Full size table

Among the physical mobility parameters (Table 2), it is of interest that modified Schober’s index (p < 0.001), cervical rotation (p = 0.034), later lumbar flexion (p = 0.002), and chest expansion (p = 0.016) were significantly reduced in smoking AS patients as compared to those with non-smoking. In addition, occiput-to-wall distances were significantly increased in smoking patients than those with non-smoking (p = 0.003). Taken together, smoking AS patients showed relatively poor physical mobility than those with non-smoking. ESR was significantly higher in smoking AS patients than those with non-smoking [mean (SD), 29.63 (17.97) vs. 20.63 (17.05) mm/h, p = 0.03]. The values of BASDAI (p = 0.283), BASFI (p = 0.240), BAS-G (p = 0.305), and CRP (p = 0.171) levels were higher in smoking AS patients than those with non-smoking, but these difference did not show statistical significance.

Furthermore, among the 35 smoking AS patients, the smoking intensity correlated significantly with BASFI (r = 0.481, p = 0.005), cervical rotation (r = −0.401, p = 0.031), fingertip-to-floor distance (r = 0.485, p = 0.004), and occiput-to-wall distance (r = 0.473, p = 0.005) (Table 3). The smoking duration also correlated significantly with BASFI (r = 0.409, p = 0.018), fingertip-to-floor distance (r = 0.482, p = 0.004), and occiput-to-wall distance (r = 0.402, p = 0.021).

Table 3 Correlation between smoking status and clinical parameters in the 35 smoking AS patients
Full size table

Using multivariate logistic regression analysis, the odds ratio for the relationship between smoking status and different clinical parameters has also been calculated (Table 4). The smoking AS patients showed significantly elevated odds ratio (OR, 95 % confidence interval) than those with non-smoking in advanced ESR (3.34, 1.08 to 10.34), CRP (6.06, 1.23 to 29.78), modified Schober’s index (12.17, 2.60 to 56.98), lateral lumbar flexion (4.61, 1.51 to 14.08), fingertip-to-floor distance (4.06, 1.14 to 14.43), chest expansion (4.44, 1.05 to 18.78), and occiput-to-wall (4.61, 1.36 to 15.6).

Table 4 Variables associated with smoking in the logistic regression analysis
Full size table

Discussion

Recent studies demonstrated that cigarette smoking could be a trigger of autoimmune disease, such as RA and SLE and related to the presence of rheumatoid factor, anti-citrulline antibody, and anti-dsDNA antibodies [3, 18]. Smoking could be an important environmental factor in the disease process of inflammatory rheumatic disease, including AS. Our data showed that smoking AS patients suffered from poor physical mobility compared with those with non-smoking. The length of modified Schober’s index, cervical rotation, later lumbar flexion, and chest expansion were significantly decreased in smoking AS patients as compared to those with non-smoking. In addition, distances of occiput-to-wall significantly increased in smoking AS patients. Moreover, multivariate logistic regression analysis revealed that the odds ratio of advanced modified Schober’s index, lateral lumbar flexion, fingertip-to-floor distance, chest expansion, and occiput-to-wall were significantly elevated in smoking AS patients compared to those with non-smoking. In a UK cohort of 53 AS patients, Averns HL et al. demonstrated that smoking patients have poor outcome of finger–floor distance, Schober’s test, total spinal movement, occiput–wall distance, functional index, stiffness, and spine X-ray scores [12]. One study of 48 AS patients in Turkey showed that there were significantly decreased lumbar modified Schober’s test and chest expansion, and increased hand–ground distance in smokers than non-smokers [11]. Ward MM et al. in the USA showed that smoking was associated with more rapid progression of functional ability and a poor prognostic factor for radiographic severity in the AS patients [10, 19]. Previous studies also showed that cigarette smoking is associated with poor functional status outcome in AS [20] and radiographic spinal progression in early axial spondyloarthritis [21]. Bodur H et al. also showed that smoking negatively affected the quality of life in AS [22].

In our study, ESR was significantly higher in smoking AS patients than those with non-smoking. In addition, BASDAI, BASFI, BAS-G, and CRP levels had trend to be higher in smoking AS patients than those with non-smoking, but these differences did not show statistical significance. Multivariate logistic regression analysis showed that the odds ratio of higher ESR, and CRP were significantly elevated in smoking AS patients compared to those with non-smoking. Previous reports ever showed that smoking AS patients had significantly higher BASDAI and BASFI scores [9, 11]. However, significantly higher BASDAI and BASFI scores were not observed in our results, which may be due to our small patient number or possible confounders. In the previous cohort of 48 AS patients, ESR (p = 0.094) and CRP levels (p = 0.082) were higher in the smokers than non-smokers but did not reach statistical difference [9]. In another recent cohort of 647 early axial spondyloarthritis patients, smoking was independently associated with higher disease activity, increased axial inflammation and structural damage on magnetic resonance imaging (MRI), poorer functional status, and poorer quality of life [23]. Smoking in AS patients could be related with chronic inflammation, which lead to progressive joint damage and new bone formation, resulting in ankylosing of adjacent vertebral body and poor spinal mobility. Furthermore, our study demonstrated that the AS smokers with higher smoking intensity have poorer disease outcome, such as functional ability (BASFI), cervical movement (CROT, OWD), and spinal mobility (FFD). Our results were compatible with a recent study which showed that smoking has a dose-dependent relationship with measures of disease severity in AS, including decreased function and poor quality of life [13].

In conclusion, cigarette smoking, one of the most concerned public health problems, is associated with higher systemic inflammation and poor physical mobility in the Chinese AS patients. Also, smoking intensity has a relationship with disease severity in AS patients, including poor functional ability and physical mobility. Smoking in AS patients may be associated with increase systemic inflammation, poor functional, and physical status; thus, quitting smoking is an important recommendation for the AS patients to improve their long-term disease outcome. However, the effect of smoking on AS disease process cannot be established on such a cross-sectional study only. To assess the association between the smoking and AS disease outcome, a larger-scale study with longitudinal follow-up in disease activity, functional ability, and physical mobility is necessary.