TABLE 3.
Correlates of log10 breast milk vitamin B-12 concentration (picomoles per liter) in lactating mothers 1–6 mo postpartum participating in a substudy of the WASH Benefits trial in rural western Kenya1
| β (95% CI) |
||||
| Exposure characteristics | Unadjusted model | Adjusted model 1 | Adjusted model 2 | Adjusted model 3 |
| Household hunger | ||||
| Little/none | Ref | |||
| Moderate/severe | 0.00 (−0.13, 0.13) | −0.01 (−0.14, 0.12) | ||
| Animal-source food score, item-d/wk | ||||
| Q1 (0–7) | Ref | Ref | ||
| Q2 (8–9) | −0.01 (−0.21, 0.18) | 0.02 (−0.16, 0.19) | ||
| Q3 (10–13) | 0.01 (−0.13, 0.14) | 0.01 (−0.14, 0.15) | ||
| Q4 (14 – 30) | 0.08 (−0.04, 0.20) | 0.09 (−0.05, 0.22) | ||
| Vitamin B-12 intake, μg/d | −0.01 (−0.03, 0.01) | −0.01 (−0.03, 0.02) | ||
1
β coefficients and Wald 95% CIs were computed with linear regression with the use of generalized estimating equations and robust SEs to account for correlation at the cluster level. Adjusted model 1 includes household hunger, maternal age, and months postpartum. Adjusted model 2 includes animal-source food score, maternal age, and months postpartum. Adjusted model 3 includes vitamin B-12 intake, maternal age, and months postpartum. Q, quartile; Ref, reference.