Background There has been very little lifecourse research looking at the risk of osteoarthritis (OA). Several studies have demonstrated the association between adult risk factors such as obesity and higher bone mineral density with subsequent knee OA.
Objectives We performed a lifecourse analysis of risk factors for knee OA (defined by osteophytes on ultrasound) acting at different stages of life, including early life factors, among members of the Newcastle Thousand Families birth cohort.
Methods Potential risk factors for knee osteoarthritis (including birth weight, breast feeding data and socioeconomic status) have been collected prospectively in this birth cohort of subjects aged 63 (born in May-June 1947) and an a priori conceptual framework was developed. Subjects had both knees scanned by a trained musculoskeletal sonographer. Ultrasound protocols were derived from EULAR guidelines. The presence of knee osteophytes was assessed at the tibial and femoral sites, medially and laterally. These data were analysed in relation to a range of factors from across the lifecourse using logistic regression models.
Results Among the 311 participants, the prevalence of knee osteophytes was 22%, 25% and 30% for right, left and “any” knee, respectively. While birth weight, exclusive breast feeding and social class at birth showed significant univariate associations with knee osteophytes, only exclusive breast feeding (among factors acting in early life) showed a significant association in the adjusted model (OR 0.81 per month; CI 0.68, 0.97; p=0.02). BMI (OR 1.11; CI 1.02, 1.20; p=0.01) and total hip bone mineral density at age 50 (OR 1.37 per 0.1 g/cm2; CI 1.06, 1.78; p=0.02) were the factors acting in adulthood that increased the risk of knee osteophytes at age 63. Serum fibrinogen levels at age 50 (a surrogate marker of inflammation) showed a borderline significant association in the multivariate model (OR 1.68 per g/L; CI 0.99, 2.85). The univariate effect of social class at birth on knee osteophytes was found to be mediated by its subsequent effect on breast feeding and total hip bone mineral density.
Conclusions This is the first study to perform a lifecourse analysis of knee OA risk using prospectively collected data. While exclusive breast feeding is known to decrease risk of adult obesity and therefore of knee OA, this study suggests that exclusive breast feeding is an independent predictor of knee OA. The mechanism might be reduced burden of infection and inflammation through the lifecourse, a testable hypothesis. Systemic inflammation itself (as measured by serum fibrinogen at age 50) might be an independent predictor of knee osteophytes.
Disclosure of Interest None Declared