Maternal Selenium Intake And Selenium Status During Pregnancy In .

E. Holmquist, A.L. Brantsæter, H.M. Meltzer et al.Science of the Total Environment 798 (2021) 149271Table 1Studies on associations between selenium supplementation and pregnancy-induced hypertensive disorders.Ref (no)Author, title, journal and year ofpublicationStudy designMean Se status instudy group atbaselinea(gestational ageat measurement)OutcomeCountry(Han and Zhou, 1994)Han et al. Selenium supplement in theprevention of pregnancy- inducedhypertension. Chin Med J (Engl), 1994No informationavailableIncidence of PIH was 7.7% in theSe-treated group and 22.7% in theplacebo group, p 0.05.China(Tara et al., 2010)Tara et al. Selenium supplementation andthe incidence of preeclampsia in pregnantIranian women: a randomized, doubleblind, placebo-controlled pilot trial. TaiwanJ Obstet Gynecol, 2010Rayman et al. Effect of selenium on markersof risk of pre-eclampsia in UK pregnantwomen: a randomised, controlled pilot trial.Br J Nutr, 2014Randomisation of women at highrisk of PIH: 52 were given Se 100μg/d (natural dietetic liquid) and48 placebo for 6-8 weeks in latepregnancyRandomisation of 166 healthywomen to 100 μg Se/day (yeast)or placebo, from first trimester todeliverySe in serum122.5 23.2 μg/L(up to 12 weeks)No significant difference (no incidenceof preeclampsia in the Se group,compared with 4.7% (n 3) in thecontrol group, p 0.05)Iran230 primiparas randomised to Se60 μg/d (Se-enriched yeast) orplacebo, from 12 to 14 weeks ofgestation until deliverySe in whole blood1.32 μmol/L(mean 12 weeks 3 days)Same as above (Rayman et al.,2014)Same as above(Rayman et al.,2014)UKSe treatment significantly reduced theodds of either preeclampsia or PIH (OR0.350, 95% CI 0.126, 0.974; p 0.044).Analyses were adjusted for baseline Seconcentration and haematocrit.After exclusion of non-compliers withUKSe treatment, Se supplementation wasfound to significantly reduce the OR forpreeclampsia/PIH (OR 0.30, 95% CI 0.09,1.00, p 0.049)(Rayman et al., 2014)(Rayman et al., 2015)aRayman et al. Selenium status in U.K.pregnant women and its relationship withhypertensive conditions of pregnancy. Br JNutr, 2015.Se selenium; 1 μmol/L selenium 79 1 μg/L; 80 μg/L plasma selenium equals 100 μg/L whole blood selenium.Table 2Studies on associations between selenium status and pregnancy-induced hypertensive disorders.Ref noAuthor, title, journal and year of publicationStudy design(Rayman et al.,2003)Rayman et al. Low selenium status is associatedwith the occurrence of the pregnancy diseasepreeclampsia in women from the UnitedKingdom. Am J Obstet Gynecol, 2003(Atamer et al.,2005)Atamer et al. Lipid peroxidation, antioxidantdefense, status of trace metals and leptin levelsin preeclampsia. Eur J Obstet Gynecol ReprodBiol, 2005Maleki et al. The relationship between plasmalevel of Se and preeclampsia. HypertensPregnancy, 2011Significantly lower median toenail Seconcentrations in preeclamptic subjects thanin controls (p 0.001). Within thepreeclamptic group, lower Se levels weresignificantly associated (p 0.029) withseverity of disease.Serum levels of Se were significantly lower (pIn a cross-sectional, prospective study serum 0.001) in preeclamptic women (60.68 Se levels were measured in 32 preeclamptic,6.42 μg/L), compared to healthy women25 non-pregnant women and 28 pregnant(87.50 10.96 μg/L), p 0.001.women without PIHD, all previously healthy.Plasma Se levels measured in 40 preeclamptic Compared to the highest plasma Se tertile,and 40 healthy pregnant women at gestational women in the lowest tertile ( 47 μg/L) and inthe middle tertile ( 58 μg/L) had significantlyweek 34-39.increased risk of preeclampsia (OR 4.96; 95%CI 1.56–15.6 and OR 3.94; 95% CI 1.26-12.33respectively).Plasma Se concentrations were first measured Se plasma concentrations were significantlyat gestational week 24-28 in a group of 650lower in the case group (70.63 21.41 versushealthy primiparous women. Measured again 82.03 15.54 μg/L, p 0.05). The bottom3 months later in 38 women presenting withtertile of Se concentration (less than 62.2 μg/L)preeclampsia and in nested matched controls was associated with higher risk ofwithout preeclampsia.preeclampsia.Plasma Se levels measured at gestational week Mean Se in plasma [median, IQR] was 79.0[71.8, 87.4] μg/L in cases and 79.6 [73.1, 86.8]15 in 244 women who subsequentlyμg/L in controls; p 0.05.developed preeclampsia, and compared with472 matched normotensive controls.(Maleki et al.,2011)(Ghaemi et al.,2013)Ghaemi et al. A prospective study of seleniumconcentration and risk of preeclampsia inpregnant Iranian women: a nested case-controlstudy. Biol Trace Elem Res, 2013(Mistry et al.,2015)Mistry et al. Association between maternalmicronutrient status, oxidative stress, andcommon genetic variants in antioxidantenzymes at 15 weeks′ gestation in nulliparouswomen who subsequently developpreeclampsia. Free Radic Biol Med, 2015Haque et al. Low serum seleniumconcentration is associated with preeclampsiain pregnant women from Bangladesh. J TraceElem Med Biol, 2016(Haque et al.,2016)(da Silva et al.,2017)da Silva et al. Comparison of serum seleniumlevels among hypertensive and normotensivepregnant women. Hypertens Pregnancy, 2017(Lewandowskaet al., 2019)Lewandowska et al. Serum Selenium Level inearly Healthy pregnancy as a risk marker ofpregnancy induced hypertension. Nutrients,2019OutcomeMedian toenail Se concentrations measured in53 preeclamptic patients and 53 matchedpregnant controls (toenails laid down 3-12months previously).Serum Se levels measured in 74 preeclampsiapatients at gestational week / 20 (52 mildand 22 severe preeclapmisa) and 118normotensive pregnant women as controlswith the same gestational age.Serum Se levels measured at inclusion in 20hypertensive (chronic and gestationalhypertension) women, 38 preeclampticwomen and 32 controls.Serum Se levels measured at gestational week11-14. Comparison between 121 womensubsequently diagnosed with PIH and 363matched controls.Se selenium; 1 μmol/L selenium 79 1 μg/L; 80 μg/L plasma selenium equals 100 μg/L whole blood selenium.3Mean serum concentration of selenium inpreeclampsia patients was significantly lowerthan that of healthy pregnant women.CountryUKTurkeyIranIranUKBangladeshBrazilNo significant difference in mean Se levelsbetween controls (56.4 15.3μg/L),hypertensive women (53.2 15.2 μg/L) andpreeclamptic women (53.3 16.8 μg/L); p 0.67Mean Se level significantly lower in case group Poland(57.5 μg/L vs. 62,9 μg/L), p 2.6 10-10

E. Holmquist, A.L. Brantsæter, H.M. Meltzer et al.Science of the Total Environment 798 (2021) 149271Biobank (Caspersen et al., 2019), resulting in a study sample of n 2572 for analysis of selenium status (Fig. 1).The establishment of MoBa and initial data collection was based on alicense from the Norwegian Data Protection Agency and approval fromThe Regional Committees for Medical and Health Research Ethics. TheMoBa cohort is now based on regulations related to the NorwegianHealth Registry Act. The current study was approved by The RegionalCommittee for Medical and Health Research Ethics (2015/2425/Reksør-øst A).This study includes women with singleton gestations who gave birthto a live-born baby at gestational week 22–42 and who had filled in bothQ1 and the FFQ (n 82,828). Only a woman's first pregnancy enrolledin MoBa was included. Women with hypertension before pregnancy,chronic renal disease, systemic lupus erythematosus or any kind of diabetes were excluded. We also excluded women with reported energyintake less than 4500 kJ per day or more than 20,000 kJ per day, sincethese values were assumed to represent invalid reporting (Meltzeret al., 2008). This resulted in a study population of n 69,972 (85% ofeligible) for analysis of selenium intake. A severe form of preeclampsiamight cause a woman from refraining from another pregnancy ormight impact her behavior, e.g. to consume (more) supplements inorder to prevent recurrence of disease, introducing a possibility of reverse causality. Therefore, analyses were performed for all women andfor primiparas separately. Data on blood selenium status was availablefor n 2999 MoBa participants in the Norwegian Environmental2.2. Selenium intake from diet and supplementsEstimated dietary selenium intake during the first half of pregnancywas based on the MoBa FFQ, designed to record dietary habits and intake of supplements during the first half of pregnancy. Women wereasked to report how often per day, week, or month they consumed different food items, dishes and beverages. The PDF of the FFQ is availableat the MoBa homepage (Norwegian Institute of Public Health, 2021).Food frequencies were converted into amounts (μg/day) using standardportion sizes, and energy and nutrient intakes were calculated using theNorwegian Food Composition Table and FoodCalc (Lauritsen, 1998).Registered selenium supplements contained one or more forms of selenium, including inorganic selenite or selenate, selenomethionine, Semethylselenocysteine or selenised yeast, which differ in their impacton blood selenium concentration (Niedzielski et al., 2016). Therefore,intake was estimated and analysed separately for selenium originatingfrom inorganic and organic supplements (Sigrist et al., 2012).A validation study of the FFQ, using a four-day weighed food diaryand several biological markers as reference methods, has demonstratedthat the MoBa FFQ is a valid tool for assessing dietary intake of energy,Fig. 1. Flowchart of the study population.* The current FFQ was introduced in 2002, explaining the large drop in numbers from box 2 to 3.4

E. Holmquist, A.L. Brantsæter, H.M. Meltzer et al.Science of the Total Environment 798 (2021) 149271nutrients and food in the first half of pregnancy (Brantsaeter et al.,2008).Iodine intake, divided into quintiles, was therefore added as a confounder. Missing covariate values was treated as a category of its own,except for iodine intake, for which only 76 women had missing data.These women were excluded from the analyses.2.3. Whole-blood selenium concentrations2.6. Statistical analysesA MoBa subgroup was included into the Norwegian EnvironmentalBiobank, established by the Norwegian Institute of Public Health(Caspersen et al., 2019). Inclusion criteria for the Norwegian Environmental Biobank were available whole blood, urine and plasma samples,available genetic data and available data from the first six MoBa questionnaires and the fathers' questionnaire. A total of 2999 women metthese criteria (Caspersen et al., 2019). Whole blood was collected inheparin tubes and shipped in a vacutainer, unrefrigerated, by ordinarymail for long-term freezing at 20 C at a central biorepository(Paltiel et al., 2014; Ronningen et al., 2006). Selenium analyses wereconducted at Lund University, Sweden, by inductive coupled plasmamass spectrometry (ICP-MS; iCAP Q, Thermo Fisher Scientific, Bremen,GmbH) equipped with a collision cell with kinetic energy discriminationand helium that served as the collision gas. The detection limit was 3.2μg/L and the coefficient of variation was 1.5%. The analytical accuracywas verified using certified reference materials, i.e., Seronorm Trace elements Whole Blood L-1 and L-2 (SERO AS, Billingstad, Norway). The results obtained (mean SD) were 56.1 5.7 μg/L for L-1 (Lot 1,103,128,N 205), compared to the recommended level of 59 (35–83) μg/L, and116 1.5 μg/L for L-2 (Lot 1,103,129, N 205), compared to the recommended level of 112 (66–158) μg/L (Caspersen et al., 2019). Seleniumwas analysed in whole blood, since this reflects both current statusand uptake, while plasma/serum selenium levels only reflect shortterm status (Thomson, 2004). Plasma selenium concentrations above80 μg/L ( 100 μg/L whole-blood selenium) are considered to be adequate (Thomson, 2004) while Hurst et al. found that in non-pregnantwomen and men a plasma selenium concentrations up to 124 μg/Lstill maximizes selenoprotein P concentrations (Hurst et al., 2010).The statistical analyses were performed using IBM SPSS Statisticsversion 25.0. All p-values were two-sided and values 0.003 were considered significant after Bonferroni correction (0.05 / (4 exposures 4outcomes)). Differences in median selenium intake from diet and selenium status according to maternal characteristics were studied withthe Kruskal-Wallis test. Differences in frequency of seleniumcontaining supplement intake (both inorganic and organic) were studied with the Chi-Square test. Prior to analysis, selenium intake and selenium status variables were standardised. The logistic regression modelswith standardised selenium intake from diet and supplements as exposures were adjusted for the following pre-defined covariates: maternalage and BMI, nausea, smoking during pregnancy, alcohol consumption,total energy intake, fibre intake, iodine intake and mutual adjustmentfor selenium source (diet, organic supplements, inorganic supplements). Analyses were performed in the whole group as well as in primiparas only. To test for possible threshold effects and non-linearassociations, a sensitivity analysis with exposure in the form of decilesof dietary selenium intake was performed. The decile closest to the recommended daily intake (RDI) of 60 μg/day, decile seven with a mean selenium intake of 62 μg/day, was used as the reference group. Thethreshold model was adjusted for the same covariates as the modelfor selenium intake as a continuous variable. The regression modelswith standardised selenium concentration in whole blood as exposurewere adjusted for maternal age and BMI, nausea, smoking during pregnancy and alcohol consumption.3. Results2.4. Pregnancy-induced hypertensive disorders3.1. Sources of dietary selenium intakeThe main outcome was PIHD, with sub-analyses for PIH and preeclampsia. PIHD data were obtained from the MBRN, based on formscompleted by midwives after birth. The form has tick-boxes for PIH, asfollows; mild preeclampsia; severe preeclampsia; preeclampsia withonset before gestational week 34; the haemolysis, elevated liver enzymes and low platelet count (HELLP) syndrome; and eclampsia. PIHDwas defined as any of these diagnoses. Sub-analyses were performedfor women diagnosed with PIH, mild preeclampsia, and severe preeclampsia (including any of the following: severe preeclampsia, earlyonset preeclampsia, HELLP syndrome or eclampsia).Fig. 2 shows dietary selenium sources. Bread, pasta, rice, cereals andgrains accounted together for around one-third (34%) of the selenium2.5. CovariatesMaternal age at delivery was obtained from the MBRN and categorized as 25, 25–29, 30–34 or 34 years. Pre-pregnancy body massindex (BMI) was calculated from height and weight self-reported inQ1 and categorized according to the WHO classification as underweight( 18.5 kg/m2), normal-weight (18.5 to 24.9 kg/m2), overweight (25 to29.9 kg/m2) or obese ( 30 kg/m2). Maternal education reported in Q1was categorized into three categories ( 12 years, 13–16 years, 17years). Smoking habits during pregnancy reported in Q1 were categorized as non-smoker, occasional smoker or daily smoker. Alcohol consumption and persistent nausea reported in the FFQ were bothapplied as dichotomized variables (yes or no). Previous studies inMoBa and elsewhere have shown that healthy dietary patterns are associated with lower risk of preeclampsia (Brantsaeter et al., 2009); we adjusted for dietary fibre intake (g/day) as an overall marker of healthydiet. Finally, we adjusted for total energy intake. Selenium and iodine intakes from food correlate (rho 0.57) and iodine intake has been foundto be non-linearly associated with preeclampsia (Abel et al., 2020).Fig. 2. Sources of selenium from food, percent of contribution.5

E. Holmquist, A.L. Brantsæter, H.M. Meltzer et al.Science of the Total Environment 798 (2021) 149271intake. Other important sources were fish and seafood, contributing 23%of the total intake, as well as eggs and dairy products. No foods are fortified with selenium in Norway.Previous RCTs on selenium supplementation during pregnancy havegenerally been small (see Table 1). Rayman et al. (2014) studied the effect of selenium supplementation on risk markers for preeclampsia in230 UK pregnant women, randomised to 60 μg/day of selenium(enriched yeast) or placebo from 12 to 14 weeks of pregnancy until delivery. In participants with low selenium status at baseline, the medianserum concentration of soluble vascular endothelial growth factorreceptor-1 (sFlt-1) was significantly lower in the selenium-treatedgroup than in the placebo group at gestational week 35. sFlt-1 is ananti-angiogenic factor, linked to increased risk of preeclampsia. Whenthe analyses were adjusted for baseline selenium concentration andhaematocrit, selenium treatment significantly reduced the odds of either preeclampsia or PIH (OR 0.350, 95% CI 0.126, 0.974; 0 0.044)(Rayman et al., 2015; Rayman et al., 2014). Another RCT including 100pregnant women in China with high risk of PIH, randomised to 100 μgselenium/day or placebo, found that selenium supplementation wassignificantly associated with decreased PIH incidence (Han and Zhou,1994). Contrary to these findings, a RCT in 166 pregnant Iranianwomen randomised to 100 μg selenium/day or placebo did not showany difference in preeclampsia incidence between the groups (Taraet al., 2010). One hypothesis is that selenium supplementation only influences preeclampsia risk in populations with low selenium status(Rayman, 2000). In the UK, where several of the previous supplementation studies have been performed, population selenium levels are low(Ghaemi et al., 2013; Rayman et al., 2015; Tara et al., 2010), whichmight explain the difference in findings between those studies andthe current study (Tables 1 and 2). Another explanation might be thesource of dietary selenium. The availability of dietary selenium dependson the conversi

relation to preeclampsia and pregnancy-induced hypertension in a large Norwegian Pregnancy Cohort Study Downloaded from: https://research.chalmers.se, 2021-12-16 01:15 UTC . tween higher selenium levels during pregnancy and decreased incidence of preeclampsia (Atamer et al., 2005; Ghaemi et al., 2013;