The Evidence on Teaching

Effects of multiple micronutrient supplementation on cognitive function in children: a systematic review and meta-analysis of randomized controlled trials

Asgari Avini N, Shahinfar H, Yazdian Z, Ranjbar M, Shab-Bidar S · 2026

grade Cmeta-analysisindependentunreplicated
Sample
35 RCTs, 19,343 participants; searched to 11 May 2025
Population
Children and adolescents aged 18 and under, international, mixed baseline micronutrient status.
Design
Updated and much larger successor to Eilander 2010, restricted to RCTs comparing three or more micronutrients against placebo or control. Cochrane risk-of-bias tool applied by two independent reviewers - and ONLY FOUR OF THIRTY-FIVE TRIALS were rated low risk of bias. Outcomes split into nine cognitive domains, random-effects pooling, GRADE applied per domain. Graded C here because the pooled trials are overwhelmingly high or unclear risk of bias, which is what the GRADE ratings reflect.
Key findings
With 19,343 children and 35 trials, only two of nine cognitive domains move at all. Memory SMD 0.22 (95% CI 0.08 to 0.35, P = 0.002) but with I2 = 73.8% and GRADE rated VERY LOW. Language SMD 0.09 (95% CI 0.03 to 0.16, P = 0.006), I2 = 39%, GRADE high - a 0.09 SD effect on language is the single high-certainty result in the modern micronutrient literature. No significant effects in any other domain, including any intelligence measure. The authors conclude the data are insufficient to support regular MMN intake for cognitive improvement in children.
Genetic confound
None internally - randomized placebo-controlled trials. The residual risk is publication and reporting bias given that only 4 of 35 trials were low risk of bias.

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Memory (near transfer)pooled SMD0.22 (95% CI 0.08 to 0.35), P = 0.002, I2 = 73.8%, GRADE very lowmixedend of supplementationbusiness-as-usualend-of-treatmentnear-transfer
Language (domain skills)pooled SMD0.09 (95% CI 0.03 to 0.16), P = 0.006, I2 = 39%, GRADE highmixedend of supplementationbusiness-as-usualend-of-treatmentdomain-skill
All other cognitive domains including intelligencepooled SMDno significant effectmixedend of supplementationbusiness-as-usualend-of-treatmentg
Risk of bias of the evidence baseCochrane RoB rating4 of 35 trials low risk of biasstandardizedn/anonenot-applicabledomain-skill

Cited by