Table 3.

Logistic regression analysis

	β	p value	Odds ratio	95% CI
DECT BMD	− 0.13817	< .001	0.8710	[0.8091;0.9375]
Age	0.075474	< .001	1.0784	[1.0246;1.1350]
Female sex	− 0.039467	0.9627	0.9613	[0.1837;5.0299]
Constant	9.10164	0.0094

Logistic regression shows a strong protective effect of increased BMD values against the occurrence fractures during a 2-year follow-up interval. Numbers in square brackets are confidence intervals. The regression model is statistically significant with high goodness of fit [χ² = 80.832, p < .001], Nagelkerke R² = 0.7899, AUC = 0.962. AUC, area under the curve; DECT, dual-energy CT; BMD, bone mineral density