Skip to main content
. 2008 Jun 5;8:26. doi: 10.1186/1472-6882-8-26

Table 7.

Test of discriminant validity between controls and patient groups and gender with estimation of the means of aR with a linear model with the discrete factors diagnosis and gender and the steady covariate age

Conditions of participants Estimation of the means of aR Total score 95%-Confidence interval p-value
No chronic condition 44.8 43.7 – 45.8
Breast cancer 41.6 40.3 – 42.9 0.0004
Colorectal cancer 42.5 40.8 – 44.1 0.1284
Diabetes mellitus 41.1 39.7 – 42.4 0.0002
Rheumatological condition 41.5 39.5 – 43.6 0.0304
Coronary heart disease 42.5 40.5 – 44.4 0.2645
Hashimoto disease 43.0 41.0 – 44.9 0.4827
Multiple conditions 38.3 36.0 – 40.5 <.0001
Gender: Women Men 40.5 43.4 39.7 – 41.2 42.3 – 44.4 <.0001
1) Orthostatic-circulatory aR
No chronic condition 17.4 16.8 – 18.0
Breast cancer 16.7 15.9 – 17.4 0.4443
Colorectal cancer 16.5 15.5 – 17.4 0.4036
Diabetes mellitus 16.4 15.6 – 17.1 0.1840
Rheumatological condition 15.7 14.6 – 16.8 0.0463
Coronary heart disease 16.3 15.3 – 17.4 0.4640
Hashimoto disease 17.3 16.2 – 18.4 1.0000
Multiple conditions 15.4 14.2 – 16.7 0.0306
Gender: Women Men 17.4 15.1 – 15.9 16.8 – 18.0 <.0001
2) Rest/activity aR
No chronic condition 19.8 19.1 – 20.4
Breast cancer 17.5 16.7 – 18.2 <.0001
Colorectal cancer 18.5 17.5 – 19.5 0.1695
Diabetes mellitus 17.6 16.8 – 18.4 0.0003
Rheumatological condition 18.4 17.2 – 19.6 0.2256
Coronary heart disease 18.7 17.5 – 19.9 0.5671
Hashimoto disease 18.9 17.8 – 20.1 0.7362
Multiple conditions 16.5 15.2 – 17.9 0.0001
Gender: Women Men 17.9 18.5 17.9 – 18.4 17.9 – 19.2 0.1406
3) Digestive aR
No chronic condition 7.6 7.3 – 7.9
Breast cancer 7.5 7.1 – 7.8 0.9917
Colorectal cancer 7.6 7.1 – 8.0 1.0000
Diabetes mellitus 7.1 6.7 – 7.5 0.2094
Rheumatological condition 7.5 6.9 – 8.0 0.9990
Coronary heart disease 7.4 6.9 – 8.0 0.9969
Hashimoto disease 6.8 6.2 – 7.3 0.0446
Multiple conditions 6.3 5.7 – 6.9 0.0020
Gender: Women Men 7.0 7.4 6.8 – 7.2 7.1 – 7.7 <.0001

A Dunnett adaption has been conducted on p-values for the error of first type including lower and upper limits of confidence intervall.