Smoking and alcohol use are associated with structural and inflammatory hand osteoarthritis features.

OBJECTIVES

To explore whether smoking and alcohol use are associated with hand osteoarthritis (OA) features in two different OA cohorts.

METHOD

We studied 530 people with radiographic hand OA from the Musculoskeletal pain in Ullensaker STudy (MUST) and 187 people from the Oslo hand OA cohort [mean (sd) age 65 (8.0) and 62 (5.7) years, 71% and 91% women, respectively]. Smoking, alcohol use and hand pain were self-reported. Participants underwent conventional hand radiographs and ultrasound examination of 30 hand joints. The Kellgren-Lawrence sum score for radiographic OA severity (0-120 scale) and the proportion of participants having at least one joint with grey-scale synovitis (grade ≥1) were calculated. We studied whether smoking and alcohol use were cross-sectionally associated with radiographic OA, synovitis, and pain using adjusted linear and logistic regression analyses.

RESULTS

Smoking was associated with less radiographic OA in both cohorts [β = -4.71, 95% confidence interval (CI) -8.36 to -1.06 for current smoking in MUST and β = -0.15, 95% CI -0.29 to -0.02 for smoking pack-years in the Oslo hand OA cohort]. Stratified analyses indicated that the association was present in men only. Being a monthly drinker (examined in MUST only) was significantly associated with present synovitis compared to never drinkers (odds ratio = 2.35, 95% CI 1.27 to 4.34) (no gender differences). Neither smoking nor alcohol was associated with hand pain.

CONCLUSIONS

Smoking was associated with less radiographic hand OA whereas alcohol consumption was associated with present joint inflammation in hand OA. Future longitudinal studies are needed to explore the causal associations and explanatory mechanisms behind gender differences.