Maternal selenium status in pregnancy, offspring glutathione peroxidase 4 genotype, and childhood asthma.

To the Editor:

Two prospective birth cohort studies have suggested that prenatal selenium status may play a role in the inception of childhood asthma. In the Avon Longitudinal Study of Parents and Children (ALSPAC), umbilical cord tissue concentration of selenium was negatively associated with risk of wheezing in early childhood; in another United Kingdom cohort, maternal and cord plasma selenium concentrations were negatively associated with wheezing in the second year of life, but not at 5 years. Associations between biomarkers of prenatal selenium exposure and asthma and wheezing later in childhood have not been reported.

Glutathione peroxidase (GPX) 4 is a unique membrane-associated selenium-dependent antioxidant enzyme that can directly reduce phospholipid hydroperoxides and protect against oxidative stress in mammalian cells. Aside from its antioxidant role, GPX4 has been implicated in lipoxygenase metabolism, which has major relevance to leukotriene synthesis and inflammatory signaling in asthma. Thus, demonstration of an interaction between selenium status and GPX4 genotype on asthma risk could strengthen the evidence for a causal role of selenium.

In the ALSPAC birth cohort, we investigated whether higher maternal blood concentration of selenium in pregnancy is associated with a lower risk of asthma and wheezing at age 7 years and whether associations are modified by the GPX4 genotype (rs713041). Details of the ALSPAC protocol can be found at http://www.alspac.bris.ac.uk. Methods for maternal selenium analysis have been described in detail previously; whole blood samples were obtained from 4484 women as early as possible in pregnancy, and selenium analysis was performed by the Centers for Disease Control and Prevention, Atlanta, Georgia, in 2009-2010. After excluding assay failures, selenium measurements were complete for 4287 women. The mean (interquartile range) of the gestational timing of blood samples was 11.7 (9-13) weeks (median, 11 weeks; range, 2-42 weeks). Maternal blood selenium concentrations were log transformed for analysis. Children were defined as having current doctor-diagnosed asthma at 7.5 years if mothers responded positively to the question “Has a doctor ever actually said that your study child has asthma?” and positively to questions about wheezing and/or asthma in the past 12 months at 7.5 years. A single nucleotide polymorphism (SNP) in GPX4 (rs713041, at position 718) was typed by LGC Genomics Ltd (Hoddesdon, Herts, United Kingdom; www.lgcgenomics.com) in mothers and children using a competitive allele-specific PCR system (KASPar). Maternal and child GPX4 genotype frequencies did not deviate from Hardy-Weinberg equilibrium. All analyses were restricted to white mothers and their offspring. The median blood selenium concentration was 110.37 μg/L (range, 27.4-324.07 μg/L); the arithmetic mean was 113.82 ± 24.53 μg/L. The 20% centile concentration was 96 μg/L, which approximates to the whole blood concentration of 100 μg/L needed to saturate glutathione peroxidase activity in adults in vivo. After controlling for confounders, including gestational timing of maternal blood samples (see this article's Online Repository at www.jacionline.org for the full list), it was found that maternal blood selenium concentration was not associated with childhood asthma (N = 2298) or wheezing (N = 2326) overall (see Table E1 in this article's Online Repository at www.jacionline.org), nor was maternal or child GPX4 genotype associated with these outcomes (see Table E2 in this article's Online Repository at www.jacionline.org). Table I presents the adjusted associations between maternal blood selenium concentrations and asthma and wheezing, stratified by maternal and child GPX4 genotypes. Maternal genotype did not modify the associations. However, there was evidence of interaction between maternal selenium status and child GPX4 genotype on asthma (P interaction .068) and wheezing (P interaction .011), with an odds reduction of 83% to 84% per doubling increase in blood selenium concentration in children who were homozygous for the minor T allele of rs713041; maternal blood selenium concentration was not associated with these outcomes in children carrying the C allele. Similarly, children of mothers below the 20th centile for blood selenium concentration were more likely to have asthma than were children of mothers with higher blood selenium concentration if they were homozygous for the T allele of rs713041 (odds ratio, 2.21; 95% CI, 0.98-4.99; P = .057), but not if they were carrying the C allele (P interaction .18) (Fig 1). The main findings were unchanged when we carried out 2 sensitivity analyses: first, excluding 3 mother-child pairs with outlying maternal blood selenium values (>300 μg/L); second, after controlling for 10 variables derived by principal-components analysis from the ALSPAC genomewide association study data to address possible residual confounding by population substructure.

Table E1

Associations between maternal blood selenium concentrations during pregnancy and childhood asthma and wheezing

Disease	Selenium concentration (μg/L)	Unadjusted OR (95% CI)	Adjusted OR (95% CI)
Asthma (N = 2298)	Q1 (≤96)	1.0	1.0
	Q2 (96-105.74)	0.77 (0.52-1.14)	0.82 (0.54-1.26)
	Q3 (105.8-115.47)	0.61 (0.40-0.92)	0.76 (0.49-1.18)
	Q4 (115.5-128.7)	0.73 (0.49-1.08)	0.85 (0.55-1.31)
	Q5 (128.8-324.07)	0.72 (0.48-1.07)	0.83 (0.53-1.30)
	Per doubling	0.66 (0.42-1.04)	0.74 (0.45-1.21)
		P value for trend = .07	P value for trend = .24
Wheezing (N = 2326)	Q1 (≤96)	1.0	1.0
	Q2 (96-105.74)	0.92 (0.60-1.40)	0.97 (0.61-1.54)
	Q3 (105.8-115.47)	0.77 (0.50-1.20)	0.96 (0.59-1.54)
	Q4 (115.5-128.7)	1.05 (0.70-1.59)	1.16 (0.73-1.84)
	Q5 (128.8-324.07)	0.95 (0.62-1.45)	1.07 (0.66-1.72)
	Per doubling	0.84 (0.52-1.33)	0.87 (0.52-1.45)
		P value for trend = .45	P value for trend = .59

OR, Odds ratio.

Table E2

Association between maternal and child GPX4 rs713041 genotype and childhood asthma and wheezing

Disease	Maternal GPX4 rs713041 genotype		OR (95% CI)
Asthma (N = 1738)	N
	560	C:C	1.0
	852	T:C	0.90 (0.64-1.25)
	326	T:T	1.29 (0.87-1.93)
		Per allele	1.11 (0.91-1.37)
Wheezing (N = 1761)	N
	567	C:C	1.0
	865	T:C	0.83 (0.59-1.19)
	329	T:T	1.28 (0.85-1.94)
		Per allele	1.10 (0.89-1.37)

OR, Odds ratio.

Table I

Associations between maternal blood selenium concentrations during pregnancy and childhood asthma and wheezing, stratified by maternal and child GPX4 rs713041 genotype

GPX4 rs713041 genotype	N	Adjusted OR (95% CI) per doubling selenium concentration	P value	N	Adjusted OR (95% CI) per doubling selenium concentration	P value
Asthma	Stratified by maternal genotype			Stratified by child genotype
C:C	560	0.51 (0.17-1.54)	.23	573	0.92 (0.31-2.74)	.88
T:C	852	0.95 (0.41-2.22)	.91	877	1.18 (0.50-2.78)	.70
T:T	326	0.92 (0.29-2.92)	.88	298	0.17 (0.04-0.72)	.02
P for interaction∗	.64			.068
Wheezing	Stratified by maternal genotype			Stratified by child genotype
C:C	567	0.74 (0.24-2.27)	.60	577	0.69 (0.26-1.86)	.46
T:C	865	1.41 (0.57-3.45)	.46	893	1.99 (0.81-4.93)	.13
T:T	329	1.14 (0.35-3.66)	.83	302	0.16 (0.04-0.66)	.01
P for interaction∗	.68			.011

OR, Odds ratio.

Testing difference in selenium associations according to genotype.

Fig 1

Odds ratio for asthma, comparing children of mothers in the bottom quintile for blood selenium concentration with children of mothers in the top 4 quintiles, and stratifying by child GPX4 rs713041 genotype.

In this population-based birth cohort study, we have found a novel, and plausible, interaction between maternal blood selenium concentration in pregnancy and child GPX4 genotype (rs713041) on the risk of asthma and wheezing at age 7 years. Our data indicate that low maternal blood selenium concentration increases the risk and higher blood selenium concentration reduces the risk but only in genetically susceptible children, namely, those who are homozygous for the minor T allele of GPX4. Villette et al proposed that the rs713041 SNP in the 3′ untranslated region of GPX4 may influence how efficiently selenocysteine is incorporated into GPX4, leading to altered GPX4 synthesis in response to variations in selenium supply. There is other evidence confirming that T/C variation (rs713041) has functional consequences, including a study showing that in individuals with the GPX4 (rs713041) TT genotype, oxidative stress decreased as in vivo selenium concentrations Furthermore, in vitro, selenium is necessary for the maximum expression of GPX4 in human lung epithelial cells. We therefore propose that low maternal blood selenium concentration leads to suboptimal GPX4 activity and impaired antioxidant defenses against oxidative stress in fetal airway epithelium, leading to epithelial damage, which, in turn, contributes to the pathogenesis of asthma. Alternatively, prenatal selenium status might influence the development of asthma through the involvement of GPX4 in lipoxygenase metabolism. However, because we cannot exclude the possibility that the main findings have arisen through type 1 error, replication of the interaction we observed is needed in another birth cohort of sufficient size.

Our findings suggest that in a susceptible subgroup of mother-child pairs in the population, namely, pregnant women with suboptimal selenium status and offspring with the homozygous minor variant of GPX4 rs71304, there may be potential to reduce the incidence of childhood asthma through selenium supplementation in pregnancy. Whether primary prevention of asthma can be achieved in this way can be determined only through a clinical trial; the efficacy of such a trial is likely to be maximized by adopting a stratified approach.