The largest zinc dataset there is found no drop in child deaths
Sixteen trials graded high certainty found deaths essentially unchanged, the range running from 16% fewer to 3% more. The children were mostly under five.
What the study found
In 143,474 children, preventive zinc did not reduce deaths
This is the strongest evidence anywhere about zinc, and it is a null. Sixteen trials, graded high certainty, found deaths essentially unchanged, the range running from 16% fewer to 3% more. The claim that zinc saves children's lives has the field's largest dataset pointing away from it.
The caveat · Mostly under-fives, poorer countries
The children were mostly under five, in low- and middle-income countries, on 10 to 15 mg of zinc a day for a median of 26 weeks, so this is not evidence about supplementing a well-fed child anywhere else. The review's own summary describes zinc as probably reducing deaths from chest infection and from malaria, but both of those ranges cross into no effect and the authors add that they could not rule out an increased risk of death. A null on mortality is not a null on everything the same trials measured.