Table 3.
Results of multivariable regression models for serum hepcidin levels stratified by site
| Variable | Gambian children* | Kenyan children† | ||
|---|---|---|---|---|
| β (95% CI) | P | β (95% CI) | P | |
| Age in years | −0.05 (–0.02, −0.08) | .004 | −0.06 (–0.03, −0.09) | <.0005 |
| Log-ferritin, µg/L | 0.94 (0.84, 1.03) | <.0005 | 0.42 (0.21, 0.63) | <.0005 |
| Interaction ferritin × season | No interaction | 0.92 (0.65, 1.19) | <.0005 | |
| Log-sTfR, mg/L | −1.19 (–0.88, −1.51) | <.0005 | −2.14 (–1.76, −2.52) | <.0005 |
| Hb, g/dL | 0.10 (0.07, 0.13) | <.0005 | n/a | |
| MCV, fL | –0.006 (–0.01, 0.002) | 0.15 | n/a | |
| Log-ACT, g/L or log-CRP, mg/L | 0.16 (–0.21, 0.53) | 0.40 | 0.18 (0.09, 0.28) | <.0005 |
| Season, end of malaria season | −0.15 (–0.09, −0.21) | <.0005 | −1.48 (–1.83, −1.14) | <.0005 |
| P. falciparum parasitemia | −0.09 (–0.001, −0.19) | .05 | 0.05 (–0.12, 0.22) | .53 |
Terms in italics were not significant and were not included in the final model. Ninety-five percent CIs and P values were adjusted to take account of potential within-subject clustering of events using the sandwich estimator.34
*
R2 is 0.49 indicating that the model explains 49% of total plasma hepcidin variance.
†
R2 is 0.62 indicating that the model explains 62% of total plasma hepcidin variance.