Table 2.
LNS-Zn0 | LNS-Zn5 | LNS-Zn10 | LNS-TabZn5 | P-value among 4 intervention groupsa,b | IC | NIC | P-value between cohortsa,b | ICCc | DEd | |
---|---|---|---|---|---|---|---|---|---|---|
Plasma zinc concentration (μg/dL)e | ||||||||||
N (9 and 18 mo) | 84 | 74 | 79 | 73 | 310 | 93 | ||||
Unadjusted pZC at 9 months | 66.2 (63.4, 69.0) | 66.0 (63.5, 68.6) | 68.9 (66.5, 71.3) | 68.8 (65.8, 72.0) | 0.255 | 67.4 (66.1, 68.8) | 65.1 (62.5, 67.7) | 0.078 | 0.00 | 1.0 |
Adjusted pZC at 9 months | 68.0 (65.2, 70.9) | 68.1 (65.6, 70.8) | 71.0 (68.6, 73.5) | 71.2 (68.1, 74.5) | 0.212 | 69.5 (68.2, 70.9) | 67.1 (64.5, 69.9) | 0.090 | 0.00 | 1.0 |
Unadjusted pZC at 18 months | 61.7 (59.2, 64.4) | 62.5 (60.1, 65.0) | 64.4 (61.9, 67.0) | 64.5 (62.4, 66.6) | 0.774 | 63.2 (62.0, 64.4) | 61.6 (59.4, 63.8) | 0.968 | 0.02 | 1.3 |
Adjusted pZC at 18 months | 63.3 (60.7, 65.9) | 64.2 (61.8, 66.7) | 66.0 (63.5, 68.7) | 65.9 (63.8, 68.1) | 0.742 | 64.8 (63.6, 66.0) | 63.6 (61.4, 65.8) | 0.831 | 0.03 | 1.4 |
Change in adjusted pZCf | −4.9 (−6.93, −2.78) | −4.0 (−6.58, −1.41) | −4.8 (−7.46, −2.13) | −6.0 (−8.89, −3.05) | 0.843 | −4.9 (−6.16, −3.64) | −3.9 (−6.23, −1.51) | 0.391 | – | – |
% Low plasma zinc concentration (pZC <65 μg/dL)e | ||||||||||
Low unadjusted pZC at 9 months, N (%) | 38 (45.3) | 34 (45.9) | 28 (35.4) | 33 (45.2) | 0.498 | 133 (42.9) | 44 (47.3) | 0.399 | – | – |
Low adjusted pZC at 9 months, N (%) | 29 (34.5) | 28 (37.8) | 22 (27.8) | 27 (37.0) | 0.587 | 106 (34.2) | 36 (38.7) | 0.370 | – | – |
Low unadjusted pZC at 18 months, N (%) | 53 (63.1) | 46 (62.2) | 46 (58.2) | 39 (53.4) | 0.909 | 184 (59.3) | 56 (60.2) | 0.379 | – | – |
Low adjusted pZC at 18 months, N (%) | 50 (59.5) | 42 (56.8) | 43 (54.4) | 34 (46.6) | 0.723 | 169 (54.5) | 51 (54.8) | 0.425 | – | – |
pZC plasma zinc concentration
aGeometric mean (95% confidence interval), n (%). Values in the same row with different superscript letters are significantly different (P < 0.05)
bAdjusting for the random effect of the village and baseline value
cInter-cluster coefficient calculated based on the cluster (village) variance and the residual variance from proc MIXED
dDesign effect calculated based on the standard error of the intervention with or without the random effect of the cluster
epZC adjusted for time of blood draw, time since last breastfeed and CRP and AGP concentration; and additionally for baseline value at 18 months
fMeans (95% CI) were calculated based on non-transformed adjusted values, but covariance analysis was done with the logarithmic-transformed variable