Table 5. Relation between placental malaria infection or maternal peripheral parasitaemia at delivery and infant haemoglobin level (g/L) during the first year of life in district of Allada, Benin 2010–2012, N = 337 (multilevel linear regression).
| Factors | Adjusted † mean Hb difference (g/L) | 95% CI | P value | ||
|---|---|---|---|---|---|
| Fixed effects* | |||||
| Maternal peripheral parasitaemia at delivery | Reference | No | |||
| Yes | - 4.6 | [-7.9, -1.3] | 0.007 | ||
| Placental malaria infection | Reference | No | |||
| Yes | - 2.8 | [-5.3, -0.3] | 0.03 | ||
| Random effects** | |||||
| Child-to-child variation (π00) | 6.4 | [5.3, 7.7] | - | ||
| Residual variation (σ²) | 11.8 | [11.1, 12.5] | - |
Hb: Haemoglobin; 95% CI: Confidence Interval to 95%
† Adjusted for estimation of pre-pregnancy body max index, infant malaria infection, fever episode and inflammatory syndrome, acid folic concentration at birth and infant age
* Estimated by maximum likelihood method
** Estimated by restricted maximum likelihood method.
The intraclass coefficient of Hb variations was estimated at 0.35. Thus, 65% of the total variance could be explained by the model.