Multivariate analysis
| . | Estimate (95% CI) . | P . |
|---|---|---|
| CD34-selected | ||
| CD4 | 0.079 (0.016 to 0.143) | .015 |
| CD8 | 0.104 (0.049 to 0.159) | <.001 |
| BM and PBSC | ||
| CD4 | 0.016 (−0.004 to 0.036) | .111 |
| CD8 | 0.005 (−0.031 to 0.041) | .781 |
| . | Estimate (95% CI) . | P . |
|---|---|---|
| CD34-selected | ||
| CD4 | 0.079 (0.016 to 0.143) | .015 |
| CD8 | 0.104 (0.049 to 0.159) | <.001 |
| BM and PBSC | ||
| CD4 | 0.016 (−0.004 to 0.036) | .111 |
| CD8 | 0.005 (−0.031 to 0.041) | .781 |
GEE analysis of the association of periengraftment α-diversity with the separate outcomes of CD4 and CD8 counts over the first 200 days after allo-HCT. The multivariate model was adjusted for age, total body irradiation, graft (for the unmodified grafts), ATG exposure (posttransplant), exposure to high-dose steroids, anti-CD20 antibodies, GVHD prophylaxis (calcineurin inhibitor and mycophenolate mofetil, calcineurin inhibitor, methotrexate, post-HCT cyclophosphamide–based, others, or none), cytomegalovirus risk, and incidence of GVHD. Time was included in each model as a polynomial spline.