Rheumatoid arthritis (RA) is a systemic autoimmune disease characterised by chronic inflammation of the joints. Its origin is multifactorial involving the intervention of genetic and environmental factors whose interaction is determinant in the development of the disease.1 The most important genetic factor for it to develop is a sequence of HLA alleles called “shared epitope” (SE).2 In addition, a wide variety of environmental factors influence the genetic base for the disease's autoimmune process to initiate. Smoking is the most important environmental factor associated with increased risk of RA, particularly in individuals who are SE-carriers.3

The discovery of antibodies against proteins and/or cyclic citrullinated peptide (anti-CCP) has amended the hypothesis of the role of SE as a susceptibility factor. It has been proposed that SE alleles predispose exclusively to the development of anti-CCP positive RA. Furthermore, rheumatoid factor (RF) and anti-CCP antibodies frequently occur simultaneously but the association between SE and RF is secondary to the association between SE and anti-CCP antibodies. In recent years a new model has been formulated for the pathogenesis of RA, where SE and anti-CCP antibodies play an essential role, and where genetic and environmental factors intervene, called the “citrullinated SE model”.4

The aim of our study was to evaluate the association of SE, smoking and the interaction between the two on the presence of specific autoantibodies associated with RA.

This study evaluated the association of SE and smoking on the presence of autoantibodies associated with RA in our southern Spanish population. Most studies on RA have been performed within the European continent, especially in the northern countries, showing consistency in their results.3,7,8 The results obtained have validated previous results in our population and indicate that the presence of SE implies a greater risk of developing RA with positive anti-CCP antibodies (OR=2.68). In addition, the increase in gene dose had an associated increase in risk, coinciding with previous studies. The patients with heterozygous SE had a risk of 2.50, whereas this risk rose to 3.48 for the patients with homozygous SE (Table 2). By contrast, no significant association was observed between SE and the presence of RF. These data show that SE is principally associated with the presence of anti-CCP antibodies, and not the presence of RF in our geographical area.

Without doubt, smoking is the most prominent and contended environmental risk factor in the development of RA. From the first observation at the end of the nineteen eighties of the association of smoking with RA in a group of women,9 consecutive case, control and cohort studies have identified smoking as a major risk factor in RA.1,10,11 Heliövaara et al.12 found that exposure to tobacco smoke can trigger production of RF and contribute to the development of RA, whereas Uhlig et al.13 identified SH as an independent risk factor for developing RA in men, particularly RF-seropositive RA. Likewise, in a study performed in women they found that the duration, but not the intensity, of SH was associated with an increased risk of RA, both RF-seropositive and RF-seronegative.1 However, a study performed by Costenbader et al.11 found that the duration as well as the intensity of the SH were directly related with the development of RA, seropositive RA in particular. In this study the association occurred in both active and ex-smokers.

Therefore, in our patients we also studied the association of smoking with the presence of antibodies (Tables 2 and 3). From the results obtained it was extracted that SH was associated with the presence of anti-CCP antibodies as well as with RF. By contrast, when the smokers were stratified as ex- and active smokers, the results did not become significant. This would support the finding by Costenbader et al.11 who highlight that the association occurs regardless of current tobacco exposure, and affects both active smokers and ex-smokers. SD is a parameter routinely used in patients who are smokers and measures the amount of cigarettes smoked in packs/year and takes into account the number of years that the patient has been smoking. Therefore, in patients with RA we studied the relationship between SD and the presence of the different antibodies. It was found that a high SD (>20packs/year) was associated with the presence of anti-CCP and RF (Tables 2 and 3). However, for a low SD, a consumption of between 1 and 20packs/year, no significant association was obtained with the presence of the antibodies.

Although the aetiology of RA is not entirely known, the findings of recent years indicate that SH acts as an environmental trigger for RA inducing immunity to citrulline residue in SE allele carriers. As described by Klareskog et al.8; SE and smoking do not separately influence the production of autoantibodies but rather interact with each other, and this interaction confers a risk of developing RA but restricted only to what is known as seropositive disease, i.e., to the subgroup of patients with RF-positive RA and/or anti-CCP antibodies but not for RA seronegative to these autoantibodies. The multivariate model obtained in this study by ordinal logistic regression (Table 4) emphasises the importance of the increase in gene dose and intensity of tobacco consumption over the presence of anti-CCP in the patients, supporting previous hypotheses. It is interesting to note, together with these factors, that the patients’ ethnicity and the presence of RF positivity also influence. By contrast, in the patients with seronegative RA no significant association has been described between smoking and SE on the presence of anti-CCP antibodies.3,8,14–16

An interesting finding from the study was that the intensity of smoking quantified by SD was more strongly associated than SH with an increased risk of positive anti-CCP antibodies (Table 2). In fact, the multivariate regression model obtained confirmed this finding including SD but not SH within the independent factors associated with the presence of anti-CCP positivity (Table 4). In the patients with positive SE and high SD, we observed a higher percentage of patients with anti-CCP positivity (95%) compared with the rest of the groups (Fig. 1C). Nevertheless, the high percentage of patients with positive anti-CCP observed in the group of patients with negative SE and high SD (Fig. 1D) is due to the individual effect of accumulated cigarette smoking. Although many studies only assess the interaction between SE and SH and find a significant interaction between the two factors, there are fewer studies that take both SH and SD into account. In one of these studies, Karlson et al.1 found a strong genetic-environmental interaction between SE and smoking when stratifying according to SD rather than SH for the development of seropositive RA. Recently, a study performed in various cohorts of patients with RA found that smoking is not associated with the presence of anti-CCP per se but rather with the concurrent presence of several autoantibodies in RA.17

Anti-CCP antibodies play a fundamental role in the pathogenesis of RA.8,18,19 The first stage of this process involves the induction of anti-CCP antibodies through the generation of citrullinated neoepitopes either due to a local inflammatory event or the action of environmental agents. These proteins and citrullinated peptides would constitute the substrate for activating a local immune response with the production of anti-CCP. This primary stage can take place several years before RA develops. Subsequently, the second stage involving the presence of a joint inflammatory process citrullinisation of peptides and proteins would be triggered in the inflamed joints that would be recognised by the previously generated anti-CCP giving rise to an immune response with release of inflammation mediators such as interleukins and TNF-alpha. However, the joint immune response would enter a vicious circle where the inflammation would produce more antigens to be processed resulting in the possible perpetuation of the response.18,19 Anti-CCP antibodies have excellent specificity (90%–100%) with very good sensitivity (41%–88%) for diagnosing RA.20 In addition, anti-CCP antibodies can precede the onset of RA for many years, before the first symptoms of the disease start.21 Although several diseases can be confused with RA in its initial stages, especially in cases where RF is not discriminating, it is useful to test for anti-CCP antibodies in these patients. However there can be anti-CCP positivity in a small proportion of patients with psoriatic arthritis and erosive arthritis.22,23

A limitation of this study was the small amount of patients studied as it was undertaken in only one hospital. A consequence of this would be the lack of significance observed between the presence of anti-CCP positivity and gene dose (Table 2, marginal significance for homozygous SE). In addition, one of the environmental factors that has become relevant in recent years and appears to play an important role in the aetiology of RA is periodontitis caused by P. gingivalis bacterial infection.24 We could consider a limitation the fact that this data was not compiled in all the patients with RA included in our study, it being impossible to assess this variable. Another limitation in our results is that we did not include in the study other genes relating to the disease, which we do not test for in our hospital in daily clinical practice.

In conclusion, Latin American mestizo ethnicity, positive SE, positive RF, and SD of more than 20 packs/year, which characterises heavy smokers, are independent risk factors for developing RA with high titres of positive anti-CCP antibodies (>200U/mL). In patients who are SE carriers, smoking intensity is associated more strongly than SD with an increased risk of anti-CCP positivity, and an interaction was observed between the two factors.

No financing was received to undertake this study.

The authors have no conflict of interests to declare.