Metabolic Syndrome, Its Components, and Knee Osteoarthritis: The Framingham Osteoarthritis Study

Abstract

Objective: Previous studies have suggested that metabolic syndrome is associated with osteoarthritis (OA). However, analyses have often not included adjustment for body mass index (BMI) and have not addressed whether levels of individual metabolic syndrome components are related to OA. This study was undertaken to examine the relationship of metabolic syndrome and its components with radiographic and symptomatic knee OA.

Methods: Framingham Study subjects were assessed for OA in 1992-1995 and again in 2002-2005. Near the baseline visit, subjects had components of metabolic syndrome assessed. We defined incident radiographic OA as present when a knee without radiographic OA at baseline had a Kellgren/Lawrence grade of ≥2 at follow-up, and defined incident symptomatic OA as present when a knee developed the new combination of radiographic OA and knee pain. After excluding knees with prevalent OA at baseline, we tested the relationship of metabolic syndrome according to the National Cholesterol Education Program Adult Treatment Panel III criteria and its components with the risk of incident radiographic OA and symptomatic OA before and after adjusting for BMI using the risk ratio from a binary regression with generalized estimating equations.

Results: A total of 991 subjects (55.1% women) with a mean age of 54.2 years were studied, and 26.7% of men and 22.9% of women had metabolic syndrome. Metabolic syndrome and many of its components were associated with both incident radiographic OA and symptomatic OA, but after adjustment for BMI, almost all of these associations became weak and nonsignificant. An association of high blood pressure, especially diastolic pressure, with OA outcomes persisted in both men and women.

Conclusion: After adjustment for BMI, neither metabolic syndrome nor its components were associated with incident OA. There may be an association between OA and high blood pressure that needs further study.

Figure 1. Severity of metabolic syndrome components and incident radiographic knee OA by gender

Adjusted RRs of incident radiographic knee OA for the count (0, 1, 2-, 3, 4–5) of metabolic syndrome components based on ATP III criteria (■), count of metabolic syndrome components based on modified ATP III criteria (■) and gender-specific quartiles (lowest, 2^nd 3^rd, highest) of waist circumference (●), residual of waist circumference from regression on body weight (○), triglyceride (x), HDL (+), SBP (◆), DBP (◆), fasting blood glucose (▲). Figures 1a and 1c: RRs from models adjusting for age, education, smoking status, alcohol consumption, physical activity level. Figures 1c and 1d: RRs from models adjusting for age, education, smoking status, alcohol consumption, physical activity level and body weight. HDL: high density, SBP: systolic blood pressure, DBP: diastolic blood pressure.

Figure 2. Severity of metabolic syndrome components and incident symptomatic knee OA by gender

Adjusted RRs of incident radiographic knee OA for the count (0, 1, 2-, 3, 4–5) of metabolic syndrome components based on ATP III criteria (■), count of metabolic syndrome components based on modified ATP III criteria (■) and gender-specific quartiles (lowest, 2^nd 3^rd, highest) of waist circumference (●), residual of waist circumference from regression on body weight (○), triglyceride (x), HDL (+), SBP (◆), DBP (◆), fasting blood glucose (▲). Figures 1a and 1c: RRs from models adjusting for age, education, smoking status, alcohol consumption, physical activity level. Figures 1c and 1d: RRs from models adjusting for age, education, smoking status, alcohol consumption, physical activity level and body weight. HDL: high density, SBP: systolic blood pressure, DBP: diastolic blood pressure.