Risk factors for progression to new sites of radiographically defined osteoarthritis in women

James R. Cerhan, Robert B. Wallace, Georges Y. El-Khoury, Timothy E. Moore

Research output: Contribution to journalArticle

23 Scopus citations


Objective. To describe the association between hormonally related risk factors and the progression to new sites of radiographically defined full body (generalized) osteoarthritis (OA) in a cohort of older women. Methods. A retrospective cohort design was used to study former radium dial painters over the age of 40 years who had minimal radium exposure. At study entry and at varying followup times, clinical examinations were conducted and full body radiographs were taken. Two followup groups were defined: women with a followup radiograph 1-9 years after baseline (n = 75) and 10-19 years after baseline (n = 53). Fifty-five joints (10 joint groups) were independently graded at baseline and followup for OA by the method of Kellgren and Lawrence, and provided the basis for summary full body OA progression scores. Progression was defined as an increase in the number of sites with OA and in separate analyses as an increase in the number of joint groups with OA. Results. Increasing length of followup and lower baseline OA score were associated with greater OA progression, while age at baseline examination showed no clear relation to progression. Beyond these variables, increasing height and having ever smoked were inversely associated with OA progression, while body mass index (BMI) showed a weak positive association. In multivariable modeling for followup 1-9 years, only lower baseline OA score predicted greater OA progression to new sites (partial r 2 = 0.13, p = 0.0009). In followup 10-19 years, baseline OA score (partial r 2 = 0.12, p = 0.0011), height (partial r 2 = 0.057, p = 0.033), and smoking status (partial r 2 = 0.09, p = 0.035) were independent predictors of OA progression to new sites, while greater BMI was a positive, weak, and nonsignificant predictor (partial r 2 = 0.031, p = 0.29). History of prior cholecystectomy, hysterectomy, dilation and curetage, number of pregnancies, and change in BMI were not significantly related to progression of OA to new sites. Similar results were found for predictors of OA progression to new joint groups. Conclusion. Lower baseline level of OA is associated with greater OA progression to new sites or joint groups independent of age, suggesting a 'burnout' phenomenon. In addition, shorter height and having never smoked appear to be independent risk factors that predict the progression of radiographic OA to new sites or joint groups.

Original languageEnglish (US)
Pages (from-to)1565-1578
Number of pages14
JournalJournal of Rheumatology
Issue number9
StatePublished - Sep 14 1996



  • Aging
  • Longitudinal studies
  • Osteoarthritis
  • Risk factors

ASJC Scopus subject areas

  • Rheumatology
  • Immunology and Allergy
  • Immunology

Cite this