Statistics from Altmetric.com
Hutchinson et al concluded that prolonged heavy cigarette smoking, but not smoking itself, is strongly associated with rheumatoid arthritis (RA), particularly in patients without a positive family history.1 The authors proposed that increased rheumatoid factor (RF) production resulting from heavy smoking exposures explains, in part, the relation of increasing cumulative pack years smoked and the greater association with RA.1
No data were presented in that study on the extent of smoking and RF positivity or its titres.1 The proposal1would be strengthened if heavy smoking were associated with RF, either when clinical disease began or when patients were studied at hospital rheumatology clinics. Others have proposed that tobacco smoke exposure triggers RF production, thereby contributing to the onset of RA.2 3 However, no significant association was seen between current smoking and IgM RF positivity in the earlier multicase family study,2 either among 41 patients with RA or their non-rheumatoid relatives—168 blood and 36 non-blood relatives.4
Although heavy cigarette smoking may be associated with RF during clinical disease, it is still relevant to determine whether it is associated with RA, either in the presence or absence of RF positivity. A further question remains as to the sequence of occurrences. Does heavy smoking first induce RF production, which later contributes to RA?1-3 Alternatively, might RA be induced first and RF produced later? Prospective, rather than cross sectional, studies are needed to answer these questions. Prospective data suggest that reported smoking of 30 or more cigarettes daily (CS 30+/day) predisposes to RA risk independently from RF positivity or positive family history.5 6
These complex relationships were investigated in a case-control study nested within a community based cohort (n=21 061 adults) enrolled in 1974. For each of the 18 male and 36 female unrelated incident patients who satisfied American College of Rheumatology criteria for RA, identified in 1994, four controls from the entry cohort were matched for age, sex, and race (all white subjects).5 6 Table 1shows the number of patients before they developed RA and their respective controls who reported heavy cigarette smoking (CS 30+/day) at baseline. Heavy smoking was not associated with pre-RA RF+ status, but was associated significantly (p=0.001) with patients who were RF− at baseline. The highest observed odds ratio (OR) was in 15 sets in which the patient was RF− at baseline and continued to be RF− after active disease developed (OR 21.5, 95% CI 1.9 to 122.5, p=0.005). The ORs were similar for sets in which the patients had positive or negative FDR status, but was significant (p=0.012) only in the larger FDR− subset (table 1).
The hypothesis that cigarette smoking contributes to RA partly by RF production1-3 is attractive. However, critical substantiation in prospective and cross sectional studies is currently lacking. Available prospective data (table 1)5 6 suggest that alternative mechanisms may be more likely. For example, long term cigarette smoking causes general vascular endothelial damage,7 and smoking is significantly associated with vasculitis in active RA.8 9 Heavy smoking was proposed4 to contribute to RA risk through its endothelial and microvascular effects, perhaps through nitric oxide pathways,10 rather than by RF production primarily.1-3
Whether or not heavy smoking differentially associates with RA depending upon family history of disease1 is as complex as the dilemmas of RF contributions to onset (table 1). Our FDR+ female patients had a significantly (p<0.001) younger mean age at clinical onset (45.6 years) than their counterparts (57.1 years). Might such earlier onset of RA among patients with a positive family history, as also noted by Hutchinson et al, 1 have influenced their behaviour to lower cumulative exposures to cigarette smoking compared with their counterparts?1
We read the letter of Masi et alwith interest and are pleased to have an opportunity to discuss the questions they have raised. Our study group was derived from an area of northwest England made up principally of people in a lower socioeconomic class, in contrast with other UK studies.1-1Although we did not record the presence of rheumatoid factor (RF) in our patients for the purpose of this study, seropositivity in our RA patient group was high, approximately 80–90%. This is comparable with Glasgow, an area in Scotland with a similarly high level of social deprivation, where 96% of randomly selected patients with RA were found to be seropositive.1-2 We therefore decided to compare the smoking history of familial and sporadic patients with RA rather than compare seropositive and seronegative patients.
Published reports almost uniformly suggest that cigarette smoking is associated with seropositive rather than seronegative RA. Cigarette smoking is associated with the development of seropositivity in healthy subjects1-3 1-4 and, furthermore, there is a dose related phenomenon for the development of seropositive RA.1-5 It has also been established that the development of seropositive RA is greatly increased in healthy subjects who are persistently seropositive.1-6 Wolfe noted a significant trend in patients with RA of increasing RF titre with pack years smoked.1-7Yet although the development of rheumatoid joint erosions, nodules, and disability was significantly increased by cigarette smoking, he found that this was independent of RF production.
We suspect that cigarette smoking and RF are strongly interlinked, but other mechanisms, as suggested by Masi, may also be at work. For example, cigarette smoke contains numerous oxidising agents that can inactivate α1-proteinase inhibitor (α1-PI),1-8 the natural inhibitor of neutrophil elastase (NE), a serine proteinase that can degrade articular cartilage.1-9 Cigarette smoke can also prime neutrophils to degranulate and discharge NE,1-10 activate macrophages to produce matrix metalloproteinases,1-11 up regulate production of interleukin 1β and interleukin 81-12 and down regulate interleukin 1 receptor antagonist,1-13 and interleukin 10.1-14Furthermore, cigarette smoking induces disease processes in a specific dose dependent fashion (independent of current smoking status), such as pulmonary emphysema, in which there is increased neutrophil priming, increased oxidised α1-PI and α1-PI-NE complexes (indicative of increased NE activity).1-8Therefore a heavy smoker may have an otherwise benign short lived inflammatory arthritis modified by the mechanisms outlined above and develop RA.
Whether RA increases or decreases cigarette consumption remains uncertain. Our controls had a pack year total estimated at entry to the study and not at the time of their disease onset. We are, however, unaware of any data to suggest that RA increases cigarette consumption. Indeed, a study by Harrison et al observed that 18% of all smokers with polyarthritis stopped smoking within three years of disease onset as opposed to <1% of non-smoking patients who started smoking during this period.1-15
Other important questions remain unanswered. For example, does increased cumulative cigarette consumption increase RA susceptibility independently of RF production? (Data presented here by Masiet al only consider cigarette consumption at one time point.) If so, do these subjects have an increased prevalence of circulating levels of α1-PI-NE complexes, high levels of oxidised and inactivated α1-PI complexes, and therefore pulmonary emphysema?
We welcome the heightened interest in the relationship between smoking and RA and look forward to the establishment of new studies designed to answer some of the interesting questions raised by recent studies.
Heavy cigarette smoking and RA
Masi AT, Aldag JC, Malamet RL. Ann Rheum Dis 2001;60:1154.
The authors of this letter, in a further analysis of their data, found that four heavy smokers in the control group were incorrectly included in the 168 subjects matched to the 42 pre-RA cases who had baseline negative rheumatoid factor (RF-) status. They should be correctly reassigned to the 48 matched controls for the 12 pre-RA cases who had baseline positive rheumatoid factor (RF+) status.
The correct assignments place 11 (23%) heavy smokers in the 48 controls for the 12 pre-RA RF+ cases. Those 12 cases include two (17%) heavy smokers. The 168 controls for the 42 pre-RA cases who had baseline negative rheumatoid factor (RF-) status should correctly include eight (5%) heavy smokers. Those 42 cases include 11 (26%) heavy smokers. The new correct figures are shown in bold in the table:
Table 1 Numbers of pre-RA cases and matched controls reporting heavy cigarette smoking (CS 30+/day) at baseline by relevant categories and odds ratios (ORs) with 95% confidence intervals (95% CIs) for developing ACR+ rheumatoid arthritis
Respective matched controls Number
0.1 to 3.5
2.6 to 19.1
Entry and post-RA RF-
2.2 to 210.6
Conversion of pre-RA RF- to RF+�
1.6 to 16.0
*No association of CS 30+/day with pre-RA RF+ (p=0.99).
�Conversion of RF- at baseline to RF+ after clinical onset of RA.
The correct assignments strengthen the findings in this prospective, community based study that baseline heavy cigarette smoking was an independent risk factor from baseline positive rheumatoid factor status.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.