Table 6.

Epidemiological studies examining associations between Se status during pregnancy, postpartum and in cord blood and birth/growth outcomes

Authors	Study design and sample size	Measures of Se status	Measures of birth/growth outcomes	Key findings
Bogden et al. 2006	CA: (< 37 weeks gestation, n 107; sub-set < 32 weeks, n 25) CO: n 126, ≥ 37 gestation) nested within cohort	Maternal erythrocyte GPx activity (microunits of GPx activity/mg of HgB): 20·9 ± 0·76; 22·4 ± 0·7 Dietary Se intake (24-h recall): 115 ± 9; 93 ± 5 Serum Se (umol/l): 1·37 ± 0·02; 1·34 ± 0·02 Collected in first and second trimesters of pregnancy	Serum Se and birth weight (g): −134 ± 182 (P = 0·46) (lowest decile v. others, CA, n 126) Serum Se and birth weight (g): −270 ± 117 (P = 0·02) (lowest decile v. others, CO, n 107)	Low-normal serum Se measured in first and second trimesters of pregnancy predicts lower birth weight for full-term infants. The risk of LBW was highest in the lowest decile of serum Se, no increase in risk above lowest decile Reduction in birth weight was considerable (–259 g for women in lowest decile of serum Se)
Klapec et al. 2008	Cross-sectional mother newborn pairs IUGR – n 57; normal – n 44	Placental Se levels (µg/g)	Placental level Se (µg/g) and IUGR (median, range): 0·14 (0·10, 0·20), n 49; AGA (n 36): 0·15 (0·10, 0·24) Se (µg/g) v. weight change (g/unit): 6954·92 (508·25, 11 125·61), P < 0·05, R 20·59	Gestational age, Se content in placenta, maternal BMI and newborn sex were associated with significant contribution to birth weight in AGA infants Maternal serum or placental Se – predict birth weight for full-term infants
Tsuzuki et al. 2013	Japan, observational 44 infants 14 premature 2 IUGR	Maternal serum Se at delivery: (pre-term: 79·3 ± 19·3 µg/l v. term: 94·1 ± 18·1 µg/l, P = 0·032) Cord blood Se at delivery: (pre-mature: 52·8 ± 9·7 µg/l v. term: 59·4 ± 10·0 µg/l, P = 0·071) Neonatal serum Se at day: (pre-term: 43·3 ± 7·0 µg/l v. term: 52·0 ± 8·9 µg/l, P = 0·001)	Birth weight (dependent): NeoSe: 17·783 (–5·590, 41·155); P = 0·128 MatSe: 10·985 (2·391, 19·578); P = 0·015 (maternal (serum Se) predictor of neonatal birth weight independent of maternal nutritional status) UmbSe: −14·524 (–33·134, 4·085); P = 0·119	Se status in neonates and mothers is positively associated with birth weight
Mistry et al. 2014	126 pregnant adolescent women Prospective observational study – About Teenage Eating Study	(Se µg/l) AGA (n 107) à 65·1 (62·7, 67·5) SGA (n 19) à 49·4 (45·9, 52·9)	SGA: individualised birth weight ratio < 10th percentile; plasma Se: 60·3 µg/l (White-European); 65·9 µg/l (Afro-Caribbean) OR for SGA: 1·15 (1·07, 1·23); 1·16 (1·08, 1·24), P < 0·0001	Odds of SGA are significantly higher among pregnant women with lower plasma Se
Sun et al. 2014	209 women, cross-sectional study (19–41 years)	Arithmetic mean (95 % CI) Maternal Se: 156·58 (145·82, 167·34) µg/l Cord Se: 8·71 (7·29, 10·14) µg/g Urine Se: 129·04 (124·30, 133·76) µg/l	Log cord Se and birth weight: 653·80. P = 0·01	Cord blood Se is positively associated with birth weight
Nazemi et al. 2015	Case: 91 mothers (LBW < 2500 g) Control: 86 mothers (normal birth weight neonates)	Maternal blood Se (µg/ml): CA: 80·69 ± 28; CO: 78·48 ± 25·54 Umbilical blood Se (µg/ml): CA: 77·32 ± 26·12; CO: 73·89 ± 24·37	Maternal blood Se, P-value: 0·961 Umbilical cord Se, P-value: 0·660	No significant associations found between maternal or umbilical cord Se and birth weight
Bermudez et al. 2015	Cross-sectional, n 54 paired samples SGA, n 11; AGA, n 30; Large for Gestational Age, n 13	Cord plasma Se (nM/l): 438·4 ± 78·6 Maternal plasma Se (nM/l): 704 ± 199·7 Cord/maternal ratio (mean ± sd): 1·08 ± 3·02	Umbilical cord blood Se and Small for Gestational Age (SGA) (n 11): 442·0 ± 79·81 nM/l Umbilical cord blood Se and Average for Gestational Age (AGA) (n 30): 439·0 ± 84·83 nM/l Umbilical cord blood Se and Large for Gestationnal Age (LGA) (n 13): 433·8 ± 67·37 nM/l Maternal peripheral blood Se and SGA (n 8): 680·9 ± 298·6 nM/l Maternal peripheral blood Se and AGA (n 25): 712·37 ± 160·2 nM/l Maternal peripheral blood Se and LGA (n 13): 701·92 ± 213·91 nM/l	Umbilical cord Se was higher in SGA neonates; however, maternal plasma Se was lower in SGA infants, suggesting preferential transfer to the fetus to sustain growth and development
Hu et al. 2016	n 81 mother–newborn pairs from 4 cities in China enrolled cross-sectionally	Maternal blood Se (ng/g): 140·8 Newborn cord blood Se (ng/g): 22·0	Birth weight and maternal blood: β: 1·0 (–3·6 to 1·6), P = 0·459 β: 0·5 (–3·2 to 2·2), P = 0·707	Positive non-significant associations observed between maternal Se levels and infant birth weight
Kobayashi et al. 2019	Japan Environmental and Children’s Study – prospective birth cohort, n 15 444 pregnant women	Blood Se: 170·0 (158·0–183·0) ng/g	Compared with infants of mothers with highest blood Se, mothers with lowest blood Se had neither significant increase in birth weight (9 g, 95 % CI −6·25) nor significant odds ratio of SGA (0·903, 95 % CI 0·748, 1·089)	No observed differences between birth weight and SGA and levels of maternal blood Se
Lewandowska et al. 2019	Nested prospective cohort of n 750 women recruited at 10–14th week of singleton healthy pregnancy in Poland	Mean Serum Se concentration of cohort: 61·95 µg/l (range: 41·14–89·17 µg/l)	SGA: (weight < 10th percentile for GA, sex of newborn in population): n 48 AGA: (weight between 10–90th percentile): n 192 Q1: Se: (41·14–56·60 µg/l); OR for SGA: 2·63 (1·08–6·42); P = 0·034 Q2: Se: (56·60–61·86 µg/l); OR for SGA: 0·87 (0·36–2·14); P = 0·794 Q3: Se: (61·86–66·62 µg/l); OR for SGA: 1·42 (0·55–3·66); P = 0·472 Q4: Se: (66·62–89·17 µg/l); OR for SGA: 1	Lowest concentrations of early pregnancy maternal serum Se associated with higher risk of SGA Threshold found at 56·1 µg/l below which steep increase in SGA risk observed
Sole-Navais et al. 2020	MoBa – prospective population-based pregnancy cohort, Norway	WBSe: (n 2572 women at 17–18 weeks gestation), median: 102 (89–117) µg/l	Maternal dietary Se intake in first 4–5 months of pregnancy: β: 0·027 (95 % CI 0·007, 0·014) Birth weight (g): β: −0·47 (–1·29, 0·34), 41·6; P = 0·254 Z-scores: β: −0·001 (–0·003, 0·001), 0·09, P = 0·195 SGA OR: 1·24 (0·53, 2·88), P = 0·623	Maternal dietary Se intake in the first 4–5 months of pregnancy was associated with higher birth weight and population-based birth weight z-scores (sd: 14·3 µg/d) of Se intake Negative associations observed between birth weight and maternal whole blood Se, additionally non-significant but higher OR for SGA
Moody et al. 2020	Community-based cohort of n 100 children 6–59 months	Venous blood Se (median, IQR) (µg/dl): 12·20 (10·69–15·02) Stunted (median, IQR) (µg/dl): 11·53 (9·65–12·13) Not stunted (median, IQR) (µg/dl): 13·05 (10·9–15·49)	β: 1·924 (P = 0·005)	Significant positive association between Se and HAZ score Se showed a U-shaped toxicity dose–response curve, with negative effects at deficient and toxic levels
Guo et al. 2021	Prospective birth cohort study n 1931 pregnant women (28–36 gestational weeks)	Maternal serum Se: 136·9 ± 47·9 µg/l	Birth weight (g) (3·4 ± 0·4 kg) β: 0·003 (–0·001, 0·006); −0·0002 (–0·0008, 0·00003) Recumbent length (0·1 cm) (49·9 ± 1·9 cm) β: 0·004 (–0·003, 0·012); 0·001 (–0·0002, 0·002) Maternal Se levels: ≥ 103·7 µg/l (highest 3 quartiles)	Effect sizes were small and non-significant for the associations between maternal serum Se and birth weight, recumbent length, although associations were positive for Se concentrations in the highest 3 quartiles compared with the lowest Se concentration quartile
Zhang et al. 2021	China-Anhui Birth Cohort Study – n 3133 mother–infant pairs	Maternal serum Se: 123 µg/l (2·0–440·6)	Birth weight (g): β: 0·58 (0·34, 0·82), P < 0·0001 Birth length (cm): β: 0·001 (0·000, 0·003), P = 0·03 Head circumference (cm): β: 0·0002 (0·000, 0·001), P = 0·61 Chest circumference (cm): β: 0·001 (0·000, 0·002), P = 0·13	Odds of low birth weight were 3·92 (2·03, 7·57) times higher OR for SGA were 2·77 (1·92, 4·02) times higher among infants with mothers who were Se deficient (< 45 µg/l) OR for LBW were 187 (1·02, 3·45) times higher and OR for SGA were 1·47 (1·07, 2·02) times higher among infants with mother who were Se insufficient (45–94·9 µg/l), all results were statistically significant (P < 0·001)

SGA, small-for-gestational age; LBW, low birth weight; IUGR, intra-uterine growth restriction; IQR; inter-quartile range; GPx, glutathione peroxidase.