Osteoarthritis (OA) is the most common form of arthritis in the elderly, and is a major cause of activity limitation, physical disability and health services utilization in the elderly. As part of the ongoing studies of OA in the Baltimore Longitudinal Study of Aging (BLSA), in order to test the hypothesis that genetic markers may be implicated in the expression of OA, especially generalized (polyarticular) OA, we have examined the association between alleles of the polymorphic human aggrecan gene (at least 11 alleles identified) and bilateral hand and knee OA in 74 Caucasian men aged 60 and over. DNA grown from fibroblast samples obtained from these subjects, who had bilateral hand and knee x-rays, has been analyzed by PCR and/or Southern analysis for the presence of alleles of the human aggrecan gene. After adjustment for age and body mass index (BMI), it appears that the presence of allele#5 is associated with increased odds of bilateral hand OA (Odds ratio [OR] 5.3, p=0.007), but not with bilateral knee OA, which would be the first association between a human aggrecan gene polymorphism and OA. We are currently analyzing several more samples in women as well. We also examined the history of knee surgery with knee OA in 545 Caucasian men and 350 Caucasian women. 36 subjects had knee surgery prior to the x-ray. Multivariate models, adjusting for age, BMI, gender and smoking status at the time of x-ray, demonstrated increased odds ratios (OR[95%CI]) of knee OA as measured on the Kellgren-Lawrence (KL) scale (108.4 [12.8, 916.8]), the Osteophyte scale (15.4 [5.9, 39.8]), and Joint space narrowing (JSN) (12.5 [5.5, 28.3]). Knee surgery is significantly associated with knee OA, although the indication for the surgery may be a contributing or confounding factor. We also examined factors predictive of total knee replacement surgery (TKR) in 177 Caucasians (126 men and 51 women), who had definite radiographic knee OA, and completed symptoms questionnaires at the time of x-ray. 11 of the subjects had a TKR 1-8 years following the knee x-ray. Multivariate models, adjusting for age, BMI, gender and smoking status at the time of the x-ray, demonstrated increased odds of TKR among subjects with the following: Pain (14.0 [2.8, 68.9]), KL grades 2-4 (23.4 [4,4, 114.2]), and JSN (19.5 [2.3, 162.0]), and Sclerosis (19.2 [4.4, 82.8). Moderate-to-severe knee OA, joint space narrowing, and subchondral sclerosis, with concurrent knee pain, are all predictive of total knee replacement. In addition to the above completed work, since January 1996, we have been collecting knee radiographs on all BLSA subjects aged 40 and over, in order to add to our limited collection of longitudinal knee x-ray data, and to provide more cross-sectional and baseline data on African American participants in the BLSA.
