Arthritis & Rheumatism, Volume 60,
October 2009 Abstract Supplement

The 2009 ACR/ARHP Annual Scientific Meeting
Philadelphia October 16-21, 2009.

Associations of Education, Occupation and Community Poverty with Knee Osteoarthritis (OA)

Callahan,  Leigh F., Shreffler,  Jack, Schwartz,  Todd, Schoster,  Britta, Renner,  Jordan B., Jordan,  Joanne M.


Education and occupation, individual level measures of socioeconomic status (SES), have been demonstrated to be associated with OA outcomes in numerous studies, but to date community SES measures have not been evaluated in OA. In other chronic diseases and in self-report arthritis of any type, the poverty rate of one's community was found to be associated with disease prevalence and health status outcomes independent of a person's education level. This study examined associations between education, occupation, and community poverty with radiographic knee OA (rOA), knee symptoms, and symptomatic knee OA (sympOA) in the Johnston County OA Project.


A cross-sectional analysis was conducted on 4098 individuals (65% White and 35% African American [AA]) who resided in 67 of the 68 census block groups of Johnston County NC. Education (<high school (HS) degree or 3HS) and occupation (physically demanding or not) were used as individual measures of SES. Census block group household poverty rate (<12%, 12–25%, >25%) was used as a measure of community SES. Covariates included age, gender, race, current smoking, and BMI. Three outcomes were assigned as a finding in one or both knee joints: rOA defined as Kellgren-Lawrence grade 32, knee symptoms (pain, aching or stiffness on most days), and sympOA (symptoms and rOA in the same joint). Race was evaluated for effect modification and not found so multiple logistic regression models of the whole sample were used to determine associations of outcomes with each of the main SES effects, adjusting for covariates. Multivariate analyses were also conducted with all 3 SES variables in the model simultaneously, adjusting for the covariates and allowing for random intercepts based on block groups.


In bivariate analyses with education and the covariates, <HS was significantly associated with rOA (OR=1.4, CI 1.2, 1.6), symptoms (OR=1.7, CI 1.5,2.0), and sympOA (OR=1.6, CI 1.3,2.0). In bivariate analyses, occupation was significantly associated with symptoms (OR=1.5, CI, 1.3,1.8) and sympOA (OR=1.3, CI 1.03,1.6). Residing in a block group with >25% poverty (with referent <12% poverty) was significantly associated with rOA (OR=1.8, CI 1.3,2.5) and sympOA (OR=1.5 CI 1.1, 2.1) in bivariate analyses. In the multivariate regression models including all 3 SES variables simultaneously, adjusting for the covariates, <HS (OR=1.4, CI 1.1,1.6) and poverty >25% (OR=1.8 CI 1.3,2.4) were significantly associated with rOA. <HS (OR=1.6, CI 1.3, 1.9) and a physically demanding occupation (OR=1.3, CI 1.1, 1.5) were associated with symptoms, but poverty rate was not. And, <HS (OR=1.6, CI 1.3, 1.9) and poverty >25% (OR=1.5, CI 1.1, 2.0) were associated with sympOA.


Both individual and community SES measures were independently associated with knee OA in a population-based study of individuals from a rural community, after adjusting for some of the primary risk factors for knee OA including age, BMI, gender and smoking.

To cite this abstract, please use the following information:
Callahan, Leigh F., Shreffler, Jack, Schwartz, Todd, Schoster, Britta, Renner, Jordan B., Jordan, Joanne M.; Associations of Education, Occupation and Community Poverty with Knee Osteoarthritis (OA) [abstract]. Arthritis Rheum 2009;60 Suppl 10 :1885
DOI: 10.1002/art.26959

Abstract Supplement

Meeting Menu