Skip to main content

Table 3 Mortality outcomes

From: Maternal antenatal multiple micronutrient supplementation for long-term health benefits in children: a systematic review and meta-analysis

Location Follow-up study Deaths in control groupa Livebirths in control group Deaths in intervention groupa Livebirths in MMN group Mortality differencea
MMN – control, per 1000 livebirths (95 % CI)
Bangladesh, JiVitA West et al. [26] 764
625 neonatal
139 at 29–180 days
14,142 741
626 neonatal
115 at 29–180 days
14,374 –2.5 (–8.1 to 3.1)b
Bangladesh, MINIMat Persson et al. [27] 33
23 neonatal
10 post-neonatal to 5 years
612 34
25 neonatal
9 post-neonatal to 5 years
595 3.2 (–22.6 to 29.0)
Burkina Faso Roberfroid et al. [35] 45
7 neonatal
38 post-neonatal to 30 months
644 55
11 neonatal
44 post-neonatal to 30 months
650 14.7 (–14.3 to 43.8)
Chinac Wang et al. [41] 16
16 neonatal
0 in children born in 2004
1499 20
18 neonatal
2 in children born in 2004
1469 6.10 (–3.9 to 17.3)b
Guinea Bissau Andersen et al. [36] 57
25 neonatal
9 at 29–91 days
10 at 92–365 days
13 at 366–730 days
519 65
29 neonatal
6 at 29–91 days
21 at 92–365 days
9 at 366–730 days
528 13.3 (–25.8 to 52.4)
Indonesiac Prado et al. [37] 580 infant
17 in subsample
14,053 490 infant
12 in subsample
14,373 –13.2 (–33.4 to 7.1)b
Mexicoc Ramakrishnan et al. [38] 7
6 early deaths
1 in control group of childhood randomisation
460 4
4 early deaths
456 –10.30 (–29.30 to 6.60)
Nepal Janakpur Devakumar et al. [39] 22
12 neonatal
6 post-neonatal
4 at 1–8 years
564 23
17 neonatal
4 post-neonatal
2 at 1–8 years
567 1.6 (–21.2 to 24.3)
Nepal Sarlahi Christian et al. [40] 54
34 before 91 days
20 at 91 days to 7 years
773 81
52 before 91 days
29 at 91 days to 7 years
872 23.0 (–5.2 to 51.3)b
  1. aMortality figures given for the most recent follow-up. In the case of the Bangladesh MINIMat study, this was the trial paper
  2. bThe analysis accounts for clustering in relevant trials. Design effects of 1.15, 1.20 and 1.20 were predicted in the Bangladesh JiVitA, Indonesia and Nepal Sarlahi trials, respectively [26, 29, 34]; the China trial did not report a figure and we used a design effect of 1.20
  3. cThe follow-up report described a subgroup of children; confidence intervals were created using a simulation approach, as described in more detail in the methods section