Smoking is implicated in over a third of cases of the most severe and common form of rheumatoid arthritis, researchers in Sweden have found, and in one in five cases of RA overall.
Results from a population-based study published online Dec. 14 in the Annals of the Rheumatic Diseases strengthened the growing body of evidence that links smoking with development of anti–citrullinated protein/peptide antibody (ACPA)–positive rheumatoid arthritis. In a dose-response manner, the link became stronger with heavier smoking, regardless of allele status.
The investigators, led by Henrik Källberg, Ph.D., of the Karolinska Institute in Stockholm, determined the excess fraction of RA cases attributable to smoking to be 20%, regardless of the presence of known genetic risk factors, which comprise single or dual copies of the HLA-DRB1 shared epitope.
Smoking was estimated to be responsible for 35% of ACPA-positive cases (31% for women and 42% for men), and for each copy of the HLA-DRB1 shared epitope (SE) that was found, smoking was dose-dependently associated with an increased risk of ACPA-positive RA. In people with two copies of the HLA-DRB1 SE, 55% of ACPA-positive RA was attributable to smoking, Dr. Källberg and colleagues determined.
Dr. Källberg and colleagues also found an increased risk of developing RA (OR, 1.9; 95% confidence interval, 1.1-3.5) among heavy smokers without any genetic risk factors. "That was one really interesting finding," Dr. Källberg said in an interview. "As a heavy smoker, you are almost two times more likely to develop RA" even without the HLA-DRB1 SE alleles, he said.
For their research, Dr. Källberg and colleagues collected blood samples and questionnaire information from 1,205 people who were diagnosed with RA according to the American College of Rheumatology’s 1987 criteria, as well as 872 healthy controls matched for age, sex, and geographic location. The cases were part of the Swedish EIRA (Epidemiological Investigation of Rheumatoid Arthritis) cohort study.
The questionnaires solicited information on past and current smoking, thereby allowing investigators to classify each subject by smoking history (current and former smokers of 0-9, 10-19, and 20 pack-years, with 1 pack-year defined as equaling 20 cigarettes per day for 1 year). The investigators tested blood samples for ACPA status and the presence of genotyped SE alleles.
The investigators calculated the odds ratios of developing RA associated with different smoking levels and SE alleles, together with 95% confidence intervals, by using logistic regression models. The interaction between smoking and the presence of SE alleles was evaluated as a departure from additive effects, and was estimated by calculating the attributable proportion due to interaction.
For former light and moderate smokers, the risk of developing RA declined and approached never-smoker levels the longer the person had not smoked since quitting. Former heavy smokers, however, continued to see elevated risk, even decades after quitting.
The dose-dependent association with smoking was "not a total surprise. We knew from earlier studies that there was some sort of relationship with the amount," Dr. Källberg said. "We just didn’t expect it to be so clear cut."
The fact that ex–heavy smokers continued to see elevated risk does not mean that smokers shouldn’t quit, Dr. Källberg said. "To some degree, the damage may be done, but we actually find that you gain something by quitting smoking. Quitting smoking can affect how well you respond to treatment."
Although smoking’s presence in RA is smaller than in lung cancer, it is "similar to that seen for ischemic heart disease," the investigators wrote in their analysis. Furthermore, cardiovascular disease is associated with RA and is the major cause of premature death in people with RA, according to the investigators (Ann. Rheum. Dis. 2010 Dec. 14 [doi:10.1136/ard.2009.120899]).
Dr. Källberg and colleagues noted that some other factors, such as air pollution, alcohol consumption, and hormonal differences could affect the smoking-RA interaction among populations. However, they wrote, they were confident – based on the age, sex, and residential matching of controls and cases within the same population – that their methodology was strong.
Dr. Källberg and colleagues’ study was funded by grants from the Swedish government; the insurance company AFA; the European Union; the Flight Attendant Medical Research Institute; National Institutes of Health; and the COMBINE (Controlling Chronic Inflammatory Diseases With Combined Efforts) project. Neither Dr. Källberg nor any of his colleagues declared conflicts of interest.