The interaction between knee osteoarthritis and multimorbidity with fall risk: A longitudinal study using data from the osteoarthritis initiative

Objective To examine the interaction between baseline Multiple Long-term Conditions (multimorbidity) and baseline knee osteoarthritis with longitudinal fall risk and fall counts. Design A longitudinal study using publicly available data from the osteoarthritis initiative. Setting Multisite centres. Participants Participants (45–79 years) with or without knee osteoarthritis at baseline were included ( n = 4449). Main outcome measures Multimorbidity was categorised as having ≥ two chronic conditions using the Charlson Comorbidity Index at baseline and having knee osteoarthritis (Kellgren and Lawrence grade ≥ two) or without knee osteoarthritis (Kellgren and Lawrence grades < two) were the main predictor variables. The occurrence of falls was the main dependent outcome variables as evaluated at the initial visit, 12, 24, 36, 48, 72, and 96 months. Results After adjustments for false discovery rate using Benjamini-Hochberg correction for p-values, having one chronic disease with knee osteoarthritis (odds ratio 1.21, p = . 016) and without knee osteoarthritis (odds ratio 1.29, p = . 004) were significantly associated with over time increased risk of fall. Having multimorbidity (≥ two chronic conditions) with knee osteoarthritis (odds ratio 1.21, p = . 010) was associated with increased over time number of falls. Having one chronic disease with knee osteoarthritis (Incidence Rate Ratio: 1.28, p < . 001) and without knee osteoarthritis (Incidence Rate Ratio: 1.33, p < . 001) were associated with increased over time number of falls. Conclusions The interaction between baseline one chronic disease and the presence or absence of baseline knee osteoarthritis was associated with a longitudinal increased risk of fall. Multimorbidity with knee osteoarthritis was linked with number of falls.