Skip to main content
. 2024 Nov 11;7(11):e2443899. doi: 10.1001/jamanetworkopen.2024.43899

Table 2. Logistic Regression Analysis Resultsa.

Possible risk factor (cutoff value) Univariate analysis Multivariate analysis
OR (95% CI) P value OR (95% CI) P value
Age (72), y 0.44 (0.12-1.57) .21 NA NA
Height (153.0), cm 1.17 (0.33-4.13) .81 NA NA
Body weight (56.6), kg 6.00 (1.20-29.97) .03 4.44 (1.04-18.96) .04
BMI (18.5) 1.63 (0.29-9.30) .58 NA NA
VCF
Prevalent 2.15 (0.25-18.82) .49 NA NA
Prior multiple 2.23 (0.60-8.23) .23 NA NA
Denosumab treatment duration, y
≥2.5 vs <2.5 3.23 (0.89-11.75) .08 NA NA
≥3.0 vs <3.0 8.17 (1.99-33.58) .004 6.91 (1.75-27.28) .006
Bone turnover marker serum level, ng/mL
CTX (0.30) 0.69 (0.20-2.44) .56 NA NA
P1NP (23.97) 0.37 (0.09-1.52) .17 NA NA
Osteoporosis before trial enrollment 0.45 (0.12-1.66) .23 NA NA

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CTX, C-terminal telopeptide; NA, not applicable; OR, odds ratio; P1NP, N-terminal propeptide of type 1 collagen; VCF, vertebral compression fracture.

a

Results are presented for the univariate analysis and the multivariable logistic regression using stepwise selection, which was conducted only for risk factors with P < .10 in the univariate analysis. With regard to denosumab treatment duration, there were 2 competing factors; only the more statistically significant factor was incorporated by the model.