Bibliography	RoB Tier	Study design	Study name	Endpoint assessed	Additional comments
Dewey et al. ( 1998 )	2	RCT	n/a	SF < 12 μg/L Hb < 110 g/L	Same RCT as Cohen et al. (1995a), reported in previous sections
Dewey et al. ( 2004 )	2	RCT	n/a	SF < 12 μg/L	Same RCT as Dewey et al. (1999), reported in previous sections
Kattelmann et al. ( 2001 )	2	RCT	n/a	SF < 12 μg/L Hb < 110 g/L	Same RCT as Mehta et al. (1998), reported in previous sections
Jonsdottir et al. ( 2012 )	1	RCT	n/a	SF < 12 μg/L
Libuda et al. ( 2016 )	2	PC	Polyunsaturated fatty acids in child nutrition—A German multimodal optimisation study (PINGU)	SF < 12 μg/LSF < 12 +Hb < 105 g/L
Meinzen‐Derr et al. ( 2006 )	2	PC	n/a	Hb < 100 g/L	CV
Hong et al. ( 2017 )	3	RETRO: CS	n/a	SF < 12 ng/mL, MCV < 70 fl and TfS < 10%

CS: cross‐sectional study; CV: timing of introduction of CF used as a continuous variable in the analysis; Hb: haemoglobin; MCV: mean corpuscular volume; n/a: not applicable; PC: prospective cohort study; RCT: randomised controlled trial; RETRO: retrospective study; RoB: risk of bias; SF: serum ferritin; TfS: transferrin saturation.