Markers of atherosclerosis in relation to presence and progression of knee osteoarthritis: the Rotterdam study

Abstract

Purpose: Several observational studies have found an association between subclinical measures of atherosclerosis and osteoarthritis (OA) of the hands and knees, predominantly among women. Different mechanisms have been suggested to explain the potential relation between atherosclerosis and OA; systemic low-grade inflammation caused by visceral adipose tissue is particularly mentioned and may consequently highlight a route to improve prevention and treatment of atherosclerosis and OA. However, the reported associations between subclinical measures of atherosclerosis and OA were modest in effect size, derived mainly from cross-sectional studies and generally attenuated after adjustment for cardiovascular risk factors. Furthermore, previous results were inconsistent for the different imaging markers of atherosclerosis. In previous work, we found no relation between peripheral measurements of atherosclerosis (including carotid intima-media-thickness or carotid plaque) and progression of knee, hand, or hip OA in a large sample of a prospective cohort study. Hence, it remains unclear whether atherosclerosis and OA are related or whether they are simply independent disorders with shared risk factors. Coronary artery calcification (CAC) is considered a late stage pathognomonic feature of coronary atherosclerosis. CD40L, VCAM-1, and VEG-f are serum biomarkers that reflect initial inflammatory stages of vascular wall damage in the ischemic cascade. In the vascular endothelium, VEG-f is a protein that stimulates angiogenesis, CD-40L and VCAM-1 are molecules that initiate coagulation and immune responses. We therefore investigated the association between these markers of early and late atherosclerosis and presence and progression of knee OA, a possible cardiometabolic phenotype of OA, in a large population-based cohort study. Methods: The analyses on prevalence of knee OA were performed in 3,465 participants from the prospective population-based Rotterdam Study (mean age 73.1 years, 61% women). Data on coronary artery calcification (CAC) were available for 1,669 participants and plasma levels of CD40L, VCAM-1, and VEG-f in 975. For the analyses on progression of knee OA, data on CAC was available for 979 participants (17% progressors), and 246 participants (20% progressors) had plasma levels of CD40L, VCAM-1, and VEG-f. We scored radiographs of the knee with the Kellgren-Lawrence (K&L) score for osteoarthritis (knee OA present with a K&L graded score greater or equal to 2) at baseline and follow-up (average follow-up time 4.5 years ( 0.5)). Overall progression of knee OA was defined as the combination of the incidence and the progression of existing OA at baseline and was considered present if the K&L score increased 1 grade between baseline and follow-up visit. After stratification by gender, multivariate logistic regression models with generalized estimated equations on knee level were used to calculate odds ratios (95% confidence intervals) for prevalence and progression of knee OA per each SD increase in marker levels. Results: Within the study population, 18% had radiographic knee OA, 11% of the men, 23% of the women. CAC and VEG-f were not associated with prevalent knee OA. Among women, CD40L (adjusted odds ratio (aOR) 1.31 (1.12 to 1.56)) and VCAM-1 (aOR 1.31 (1.08 to 1.59)) were associated with prevalent knee OA (table). No associations with progression were found in women. In men, too few progressors were available to assess associations.