Skip to main content
. 2021 Jul 4;31:102754. doi: 10.1016/j.nicl.2021.102754

Table 2.

Logistic regression modeling of imaging predictor variables.

OR (95% CI)a AUROC P-value
Single biomarker predictors
CIS ratio 4.06 (1.74, 9.45) 0.793 0.001
Medial Temporal SUVr 6.37 (2.50, 20.98) 0.844 <0.001
Substantia Nigra SUVr 0.09 (0.02, 0.41) 0.791 0.002
Two biomarker predictors
CIS ratio + Medial Temporal SUVr
CIS ratio 2.67 (1.12, 6.37) 0.884 0.026
Medial Temporal SUVr 4.33 (1.44, 13.00) 0.009
CIS ratio + Substantia Nigra SUVr
CIS ratio 4.14 (1.60, 10.67) 0.875 0.003
Substantia Nigra SUVr 0.10 (0.02, 0.52) 0.006
Medial Temporal + Substantia Nigra SUVrs
Medial Temporal SUVr 6.60 (1.95, 22.31) 0.902 0.002
Substantia Nigra SUVr 0.10 (0.02, 0.53) 0.007
All three biomarker predictors
CIS ratio + Medial Temporal + Substantia Nigra SUVrs 0.921
CIS ratio 2.56 (0.84, 7.78) 0.097
Medial Temporal 3.86 (1.10, 13.49) 0.035
Substantia Nigra 0.13 (0.02, 0.70) 0.018
a

In the logistic regression analysis, the odds ratios are based on 0.10 unit change. MCI-AD is the reference group. When OR is greater than 1 the values are higher in the MCI-DLB and conversely when the OR is<1 then the values are higher in MCI-AD individuals.