Table 4. Multinomial logistic regression analyses of significant T-cell marker of Prefrailty and Frailty from stepwise forward selection.
|
Prefrailty
|
Frailty
|
|||||||
|---|---|---|---|---|---|---|---|---|
| OR | 95% | CI | P-value | OR | 95% | CI | P-value | |
| Male sex | 1.01 | 0.67 | −1.55 | 0.94 | 0.56 | 0.24 | −1.32 | 0.184 |
| Age, years | 1.01 | 0.98 | −1.04 | 0.52 | 1.16 | 1.10 | −1.22 | 0.0001 |
| Number of comorbidity | 1.21 | 1.06 | −1.38 | 0.005 | 1.21 | 1.06 | −1.38 | 0.005 |
| CD8+CD28−CD27+ (tertile scores: 1,2,3) | 1.42 | 1.11 | −1.81 | 0.005 | 1.16 | 0.73 | −1.86 | 0.53 |
| CD8+CD28−CD27+ (low tertile) | 1 | — | — | — | 1 | — | — | — |
| CD8+CD28−CD27+ (mid tertile) | 0.85 | 0.52 | −1.39 | 0.512 | 0.81 | 0.27 | −2.44 | 0.714 |
| CD8+CD28−CD27+ (high tertile) | 1.72 | 1.03 | −2.87 | 0.037 | 2.56 | 0.96 | −6.81 | 0.060 |
Abbreviations: CI, confidence interval; OR, odds ratio.
Multinomial logistic regression models were employed with sex, age and number of comorbidities as fixed variables in the base model, The results are shown for forward conditional stepwise selections of immune markers (tertiles of CD4/CD8 ratio, CD4+CD28−CD27+, CD8+CD28−CD27+ and CD8+CD28−) with P<0.05 for entry and 0.10 for retention), and identical results were obtained with backward selection procedures, to produce final prediction model of frailty.