How does rheumatoid arthritis risk change after quitting smoking?

01

This study counted new diagnoses, not improvement of existing arthritis

The 2019 analysis included 230,732 female nurses without RA or another connective-tissue disease at baseline. During more than six million person-years, 1,528 new RA cases were confirmed through medical-record review. Smoking information was updated every two years rather than treating the first questionnaire as a lifelong smoking status.

Seropositive RA meant a recorded positive rheumatoid factor (RF) or anti-CCP antibody; seronegative cases lacked those positive findings in the available records. This is a research classification of confirmed cases, not an online blood-test rule. Some older cases had no anti-CCP test, which could misclassify the phenotype. A positive or negative test alone does not settle a clinical diagnosis; the NHS describes diagnosis as a broader assessment.

[1][2]

02

Keep the reference group attached to the number

A hazard ratio compares the rate of new diagnoses over follow-up among people still without RA; 1 means no difference in that rate. In the original Table 4, women who had quit at least 30 years earlier had a seropositive-RA hazard ratio of 0.63 (95% confidence interval 0.44–0.90) compared with women who had quit 0 to less than 5 years earlier. That is where the often-quoted ‘37% lower’ comes from. It is not a comparison with never smokers, and it is not a 37-percentage-point reduction in a person's chance of arthritis.

Table 3 asks the different question. Against never smokers, the same at-least-30-years category had a seropositive-RA hazard ratio of 1.30 (1.01–1.68), and an all-RA hazard ratio of 1.25 (1.02–1.53). A lower association relative to recent quitters therefore does not mean the previous exposure has been erased. Conversely, residual excess relative to never smokers does not make stopping smoking pointless.

The outcome also matters: for all RA combined, the at-least-30-years versus recent-quitter estimate was 0.78 (0.58–1.05). Its interval includes 1, unlike the seropositive estimate. The study did not find a clear cessation-duration association for seronegative RA. Do not turn a phenotype-specific result into an equally certain result for every kind of RA.

[1]

03

Why recent quitters can look worse than current smokers

In Table 3, recent quitters had a higher seropositive point estimate against never smokers than current smokers did. This was not evidence that quitting caused RA. The authors noted that recent quitters included heavy past smokers, some could resume smoking, and early symptoms or another illness could have prompted stopping. People were not randomly assigned to quit or continue.

Models accounted for several factors, including age, cohort, reproductive factors, body mass index, activity, income and alcohol. Adjustment cannot remove every difference between groups, and two-year questionnaires can miss intervening changes. Participants were predominantly White, well-educated US women in nursing; the exact estimates cannot be assigned unchanged to men or every population. The broad time categories are not anniversary switches.

[1]

04

A useful question for an appointment is about the endpoint

If you already have RA, this incidence study does not tell you whether joint inflammation, pain, antibody levels or treatment response will change after stopping. Ask the treating clinician which evidence addresses your actual question. New or persistent joint concerns also need assessment rather than waiting for a smoke-free anniversary or assuming they are withdrawal.

In England, start a joint-health discussion with a GP; the NHS describes specialist referral when appropriate. Separately, its stop-smoking page links to local support services. A practical note is: ‘Is this finding about developing RA, or about managing established RA?’ No smoking history, symptoms or test results need to be entered on this page.

[1][2][3]

05

What to keep in mind

• The strongest cessation-duration signal was for newly diagnosed seropositive RA.
• Always name recent quitters or never smokers as the reference.
• Thirty years is a study category, not a risk-reset deadline.
• Incidence research does not forecast existing arthritis activity.