Skip to main content
. 2022 Mar 11;12(3):691. doi: 10.3390/diagnostics12030691

Table 1.

Comparison between training/validation and test datasets.

Variable All
(N = 196)
Training/
Validation Dataset (N = 96)
Test Dataset
(N = 100)
p-Value
Age 64.9 (8.7) 64.5 (8.8) 65.2 (8.7) 0.593
Race/ethnicity 0.248
American Indian/Alaskan native 1 (0.5%) 1 (1.0%) 0 (0.0%)
Asian 2 (1.0%) 0 (0.0%) 2 (2.0%)
Black/African-American 1 (0.5%) 0 (0.0%) 1 (1.0%)
Hispanic 2 (1.0%) 0 (0.0%) 2 (2.0%)
Other 1 (0.5%) 0 (0.0%) 1 (1.0%)
White 189 (96.4%) 95 (99.0%) 94 (94.0%)
Gender, male 23 (11.7%) 12 (12.5%) 11 (11.0%) 0.826
Height (m) 1.65 (0.08) 1.65 (0.08) 1.64 (0.08) 0.47
Weight (kg) 76.2 (19.2) 76.5 (19.1) 75.8 (19.4) 0.8
BMI (kg/m2) 27.9 (6.4) 27.9 (6.4) 27.9 (6.4) 0.991
Diagnosis 1
Osteoporosis 54 (27.6%) 26 (27.1%) 28 (28.0%)
Osteopenia 116 (59.2%) 57 (59.4%) 59 (59.0%)
Normal 26 (13.3%) 13 (13.5%) 13 (13.0%)

BMI—body mass index; osteoporosis—minimum BMD T-score ≤ −2.5; osteopenia—−2.5 < minimum BMD T-score < −1; normal—minimum BMD T-score ≥ −1; ***—p-value < 1 × 10−3; **—1 × 10−3 < p-value < 1 × 10−2; *—1 × 10−2 < p-value < 5 × 10−2.