Key determinants for lowering the risk of joint replacement in weight-bearing joints: a population-based cohort study

Hip and knee replacement (arthroplasty) is used to treat advanced joint disease by reducing pain and restoring joint function. This study aimed to identify determinants associated with a lower risk of joint replacement (JR).

Longitudinal data from the Geelong Osteoporosis Study (GOS) were used. JR was identified via linkage with the Barwon Joint Registry, medical records, and self-report. Anthropometry, body composition, and blood biomarkers were obtained. Demographics, lifestyle, and comorbidities were self-reported. Fractures were identified from radiological reports. Participants with JR before baseline were excluded, leaving 2,882 eligible participants (1,436 men, 1,446 women; ages 20-96 years). A time-dependent Cox regression model with age as the primary time scale and sex as a stratification variable was fitted. Forward stepwise and the least absolute shrinkage and selection operator (LASSO) regression identified relevant predictors.

Over a median follow-up of 16.7 years (IQR: 9.7-23.2), 223 participants (7.7%) underwent JR. Factors with the reduced risk of JR included lower body mass index (HR 0.96, 95% CI, 0.94-0.99), lower spine bone mineral density (BMD) (0.84, 0.77-0.92), lower procollagen type 1 N-terminal propeptide (0.69, 0.50-0.96), non-fallers (0.74, 0.55-0.99) and no history of cancer (0.66, 0.48-0.90). Lower dietary calcium intake (0.74, 0.52-1.04) was associated with the reduced risk of JR in the main models, but this association was attenuated after accounting for supplement use. Compared with low socioeconomic status (SES), participants with medium and high SES had higher JR risk (1.91, 1.04-3.50; 1.88, 1.13-3.15), although the association was weaker in women (interaction: 0.48, 0.25-0.91).

Age-related spine degeneration may explain the link between higher spine BMD and JR. The association between lower dietary calcium intake and reduced JR risk is likely driven by confounding by indication or reverse causality. Lower BMI and lower P1NP levels emerged as potentially modifiable pathways. The absence of falls and no history of cancer were associated with reduced JR risk; however, the cancer association was attenuated in the Fine-Gray analysis when competing mortality was accounted for, and the fall association should be interpreted as observational rather than as evidence of a modifiable pathway. Socioeconomic disparities also highlight potential avenues for reducing JR risk. Overall, these findings suggest that biological, lifestyle, and social determinants may play important roles in shaping long-term JR outcomes.