Development of a Model for Predicting the 4-year Risk of Symptomatic Knee Osteoarthritis in China

A Longitudinal Cohort Study

Limin Wang; Han Lu; Hongbo Chen; Shida Jin; Mengqi Wang; Shaomei Shang

Disclosures

Arthritis Res Ther. 2021;23(65) 

In This Article

Results

In CHARLS2011, physical activity measures were available for 3684 participants, while blood samples were available for 11,847 participants. Complete KOA data were available in CHARLS2011 and CHARLS2015 for 9204 of these participants. Seven participants were excluded because they declined to undergo body measurements assessments, rendering over 50% of participant's variables (including measurements of weight, height, waist circumference, assessments of depressive symptoms, physical activity, ADLs/IADLs, or the blood biomarkers) inaccessible. Among them, one patient was diagnosed with KOA in 2011, and one developed KOA in 2015. Among the remaining 9197 participants, an additional 1004 were excluded because they were diagnosed with KOA at baseline (CHARLS2011). Thus, data from a total of 8193 patients were included when developing the model. Among the 8193 included participants, 815 developed symptomatic KOA in the following 4 years. The overall 4-year cumulative incidence of symptomatic KOA was 9.95%, with 7.62% and 13.77% in males and females respectively.

The mean age was 58.82 years (standard deviation (SD) ± 9.01 years), and 4251 patients were female (51.89%). At baseline, 23.31% participants had difficulty with ADLs/IADLs, while 17.08% were diagnosed with metabolic syndrome, and 44.66% reported one or two chronic comorbidities. A history of hip fracture was reported by 252 (3.08%) participants. Other baseline characteristics are summarized in Table 1, along with the number of missing values for each variable.

Univariable and Multivariable Analysis

Table 2 shows the results of the univariable and multivariable analysis based on the imputed datasets. In the univariable analysis, age was identified as a risk factor for KOA, with the biggest difference occurring between the 60–64 and 65–69 age groups. Female sex, rural residence, history of hip fracture, ADL/IADL difficulty, severe depressive symptoms, more chronic comorbidities, poor health status, and higher levels of moderate physical activity (MPA) were significantly associated with an increased risk of developing KOA (all P values ≤ 0.01), while smoking was significantly associated with a decreased risk of developing KOA (P ≤ 0.01). Although high BMI/waist circumference and metabolic syndrome were also positively associated with the incidence of KOA, these associations were not significant (all P values > 0.05). Considering the important effects of metabolism and vigorous physical activity on the incident of KOA, we included metabolic syndrome and level of vigorous physical activity (VPA) in the multivariable logistic model, although the significance was not significant. As the clinical values of BMI and waist circumference are comparable, we selected waist circumference into the multivariable logistic regression given the relatively smaller P values.

The final prediction model included ten variables: age, sex, waist circumference, residential area, ADLs/IADLs difficulty, history of hip fracture, depressive symptoms, number of chronic comorbidities, health status, and level of MPA.

Model Performance

The discrimination and calibration curves for the model are shown as Figure 1a and b, respectively. The final prediction model achieved acceptable discrimination, AUC = 0.719 (95% CI, 0.700–0.737), with optimism = 0.007 and bias-corrected AUC = 0.712 after bootstrap validation. The apparent observed line was quite close to the ideal line, while the bias-corrected line was slightly further from the ideal line than the observed line after the bootstrap procedure.

Figure 1.

The discrimination and calibration curves of final risk model. a ROC curve analysis for predicting symptomatic KOA when using age, sex, waist circumference, residential area, ADL/IADL difficulty, history of hip fracture, depressive symptoms, number of chronic comorbidities, health status, and level of MPA. The AUC was 0.719 (95% CI 0.700–0.737), and optimism-corrected AUC was 0.712 after bootstrap validation. b The calibration curve. Area under the receiver characteristic curve, AUC; receiver operating characteristic curve, ROC

Clinical Score Model

We developed a simple clinical score model based on the ten variables included in the final multivariable model (Table 3). Total scores in this model range from 0 (lowest risk) to 51 (greatest risk). This clinical score model may aid in identifying patients at the greatest risk for developing KOA within the next 4 years. The AUC of the risk score model was 0.713 (95% CI, 0.695–0.731), and the optimal cut-off, where patients with a score ≥ 20.5 were most likely to develop KOA in 4 years, was obtained from the maximal Youden index. At the optimal cut-off, the sensitivity and specificity were 63.3% and 66.0%, respectively. Referring to the previous score model,[22] the incident probability of KOA within 4 years was calculated by dividing the total risk score by 51 and multiplying by 100%.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....