Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Aug 26.
Published in final edited form as: J Affect Disord. 2020 Jan 7;265:255–262. doi: 10.1016/j.jad.2020.01.005

Maternal vitamin D status during pregnancy and offspring risk of childhood/adolescent depression: Results from the Avon Longitudinal Study of Parents and Children (ALSPAC)

Min-Jung Wang 1, Erin C Dunn 2,3,4,5, Olivia I Okereke 1,3,6, Peter Kraft 1,7, Yiwen Zhu 2, Jordan W Smoller 1,2,3,4
PMCID: PMC7448808  NIHMSID: NIHMS1611896  PMID: 32090749

INTRODUCTION

Vitamin D deficiency among pregnant women is highly prevalent globally, and particularly in geographical regions at higher latitudes, such as Northern Europe and North America (Bodnar, Simhan, et al., 2007; Eggemoen et al., 2016), where it affects up to 80% of dark-skinned women and 47% of light-skinned women (Bodnar, Catov, et al., 2007; Dror & Allen, 2010; Eggemoen et al., 2016; Holmes, Barnes, Alexander, McFaul, & Wallace, 2009). Maintaining sufficient vitamin D levels during pregnancy is important, given that it is the sole source of fetal vitamin D, which may play a crucial role in fetal brain development. Since vitamin D levels can be easily increased by lifestyle modifications and supplementation, understanding the influence of maternal vitamin D levels on offspring brain health is needed to guide the design and implementation of public health efforts aimed at promoting brain health and possibly even prevent brain-related diseases as early on in the lifespan as possible.

There are several possible mechanisms linking gestational vitamin D exposure to offspring neurodevelopment, including pathways linking vitamin D to neuronal growth and both the signaling and regulation of endocrine functions (D. W. Eyles et al., 2009). Several experimental studies have reported that vitamin D deficiency during pregnancy results in alterations in brain morphology and functioning of rat offspring with long-lasting effects on behavior spanning into adulthood (D. Eyles, Brown, Mackay-Sim, McGrath, & Feron, 2003; D. W. Eyles et al., 2009; O’Loan et al., 2007; Pet & Brouwer-Brolsma, 2016). Observational research in humans has provided some evidence that maternal vitamin D status during gestation, defined by circulating levels of 25-hydroxyvitamin D [25(OH)D], could influence offspring neurocognitive and mental health outcomes (Darling et al., 2017; Keim, Bodnar, & Klebanoff, 2014; Morales et al., 2012; Strom et al., 2014), including increasing risk for schizophrenia, autism, and attention deficit/hyperactivity disorder (ADHD) (Agarwal, Kovilam, & Agrawal, 2018; D. W. Eyles et al., 2018; Magnusson et al., 2016; Morales et al., 2015; Pet & Brouwer-Brolsma, 2016). Though largely untested, early life vitamin D deficiency in offspring may also be a risk factor for subsequent onset of depression. To our knowledge, only one study has assessed the effect of low gestational vitamin D levels on the risk of offspring depression (Strom et al., 2014). However, the relatively small sample size of 850 mother-child pairs and relatively crude definition for depression rendered findings inconclusive. Studies examining the relationship between maternal 25(OH)D levels during gestation and offspring depression are needed.

Furthermore, because genetic factors are known to contribute to depression risk in both children and young adults (Rice, 2010; Xia & Yao, 2015), efforts to understand the interactions between maternal vitamin D status during pregnancy and polygenic risk scores (PRS) for depression are required. Insights generated from such studies would have potentially important public implications as the presence or absence of gene-environment interactions (GxE) could inform whether preventive strategies adopt a targeted or universal approach (Carey & Crammond, 2017).

In the current study, we used data from the Avon Longitudinal Study of Parents and Children study (ALSPAC) (Boyd et al., 2013), a birth cohort study based in South West England, to examine the prospective association between maternal vitamin D status during pregnancy and offspring depression in childhood and adolescence, and to investigate whether the effects of maternal vitamin D status were modified by offspring genetic risk for depression.

PARTICIPANTS AND METHODS

Study Population

ALSPAC is an ongoing population-based birth cohort from South West England. Detailed information about ALSPAC has been described previously (Boyd et al., 2013; Fraser et al., 2013) and is available on the study website (http://www.bristol.ac.uk/alspac), which also includes a searchable data dictionary and variable search tool (http://www.bristol.ac.uk/alspac/researchers/our-data/). In brief, the study enrolled 14,541 pregnant women with estimated delivery dates between April 1, 1991 and December 31, 1992, which resulted in 14,062 live births (Boyd et al., 2013; Fraser et al., 2013). Informed consent for the use of data collected via questionnaires and clinics was obtained from participants following the recommendations of the ALSPAC Ethics and Law Committee at the time. Consent for biological samples has been collected in accordance with the Human Tissue Act (2004). From the age of 7 years onwards, children attended annual assessment clinics during which they participated in physical tests and in-person interviews. To address loss to follow-up through attrition, further phases of recruitment were carried out after age 7, resulting in an additional 617 participants enrolling by 16 years of age (Boyd et al., 2013).

The eligible study sample for these analyses included mother-offspring pairs with singleton births and complete data on maternal 25(OH)D levels during pregnancy, offspring genetic data, and offspring depressive symptoms collected during childhood and/or adolescence. Since less than 3% of the ALSPAC cohort were non-European (Matijasevich et al., 2012), only white participants were included. Mothers were excluded if their total 25(OH)D levels were three standard deviations (SD) above or below the study population mean (n=39). This yielded analytic sample sizes of 2,938 (child subsample) and 2,485 (adolescent subsample) (Figure 1).

Figure 1.

Figure 1.

Open in a new tab

Flow chart of participants in the Avon Longitudinal Study of Parents and Children sample included in the analytic samples. ALSPAC: Avon Longitudinal Study of Parents and Children; SMFQ: Short Moods and Feelings Questionnaire; 25(OH)D: 25-hydroxyvitamin D

Offspring Depression

Depressive symptoms were measured using the Short Mood and Feelings Questionnaire (SMFQ) by a trained interviewer during childhood (mean: 10.62; SD: 0.25) and adolescence (mean: 13.83; SD: 0.21). The SMFQ consists of 13 items that capture depressive symptoms in the past two weeks. Total SMFQ scores were obtained by summing across all items (possible score range: 0-26), with higher scores corresponding to higher depressive symptoms. The SMFQ has been validated in adolescents (ages 6-18), and correlates highly with both Children’s Depression Inventory scores and past-year Diagnostic Interview for Children depression scores (Angold, Erkanli, Silberg, Eaves, & Costello, 2002). The SMFQ was dichotomized, with depression defined as SMFQ≥11, since the raw SMFQ scores were positively skewed (skewness: 1.36 for childhood subsample, 1.46 for adolescent sample) and the dichotomized score provided greater clinical interpretability. This threshold has been shown to have high sensitivity and specificity for depression defined by the revised Diagnostic and Statistical Manual of Mental Disorders, Third Edition (DSM-III-R) (Thapar & McGuffin, 1998), and has been used in prior ALSPAC studies of depression (Joinson, Kounali, & Lewis, 2017; Pearson et al., 2015).

Serum 25(OH)D Measurements

Maternal serum 25(OH)D2 and 25(OH)D3 were measured on non-fasting blood samples taken for routine pregnancy tests throughout pregnancy. Blood samples were assayed using high-performance liquid chromatography tandem mass spectrometry (HPLC/MS) in accordance with Vitamin D External Quality Assessment Scheme requirements. Details about sampling, storage, and processing are described in detail elsewhere (Lawlor et al., 2013). For the few mothers who had multiple measurements taken during pregnancy (4.8%) (Wills et al., 2013), the last result available was used, in line with similar ALSPAC studies that have assessed the influence of maternal 25(OH)D concentrations during pregnancy on child health outcomes (Lawlor et al., 2013; Williams et al., 2013). Total 25(OH)D was calculated by summing 25(OH)D2 and 25(OH)D3 levels.

As total maternal 25(OH)D levels displayed sinusoidal seasonal variation (Figure S1), a seasonality-adjusted 25(OH)D variable was derived to represent annual mean levels of 25(OH)D. Similar trigonometric models have been previously applied to adjust for seasonal variation in 25(OH)D in ALSPAC (Lawlor et al., 2013; Williams et al., 2013), which have been shown to have good validity (Sachs et al., 2013). Further details of the method are described in the Supplementary Materials. Briefly, 25(OH)D was modeled in a linear regression against sine and cosine transformations of the month when blood sampling occurred. The annual mean 25(OH)D concentration for each mother was then estimated using the residuals of the model. The seasonality-adjusted and unadjusted 25(OH)D levels were strongly correlated (Spearman’s r=0.88).

Genetic Data

9,912 children in the ALSPAC cohort were genotyped on the Illumina HumanHap550 Quad chip (Illumina Inc., San Diego, CA). Standard quality control (QC) measures were performed to exclude individuals on the basis of gender mismatch, minimal or excessive heterozygosity, individual genotyping call rates<97%, cryptic relatedness (IBD>10%), and non-European ancestry (assessed using multidimensional scaling analysis and compared to HapMap II). Single nucleotide polymorphisms (SNPs) were excluded based on the following criteria: minor allele frequency (MAF)<1%; missing rate>5%; and significant deviation from Hardy-Weinberg Equilibrium (HWE) (p<5x10−7). 500,527 directly genotyped SNPs and 8,365 children remained after quality control (QC). Imputation was conducted using Impute V2.2.2 (Marchini, Howie, Myers, McVean, & Donnelly, 2007) against the 1000 genomes reference panel (Phase 1, version 3) (Genomes Project et al., 2012), with 2186 reference haplotypes (including non-Europeans).

Polygenic Risk Scores (PRS)

PRS for depression were generated in PLINK v.1.90 using methods described by Purcell el al. (Purcell et al., 2007). The Psychiatric Genomics Consortium wave 2 (PGC-MDD2) (Wray et al., 2018) was used as the discovery sample (meta-analytic subsample excluding 23andMe: n=173,005) and the ALSPAC cohort was used as the training sample. 23andMe data was excluded due to delays in negotiations between our institution and 23andMe for data access. Additional details about the data sources and QC procedures implemented prior to PRS construction are provided in the Supplementary Materials. We generated a single PRS for each child based on a threshold of p<0.05, which was the threshold that maximized the variance explained in depression in a previous discovery sample (Howard et al., 2019). PRS was generated by summing the risk alleles (0, 1, or 2) for each SNP, weighted by the natural log-transformed odds ratio (OR) of its association with major depressive disorder in the PGC-MDD2. PRS scores were then standardized using the z-transformation and categorized into three risk groups (low risk: quartile 1; medium risk: quartiles 2 and 3; high risk: quartile 4) (Figure S2).

Missing Data

In the analytic samples, a substantial proportion of participants had missing data on one or more of the covariates. For any single covariate, the amount of missingness ranged between 0.3% and 20.9%; 71.2% and 72.9% of participants had complete covariate data for the child and adolescent subsamples, respectively. Missing covariate information was imputed using multiple imputation by chained equations (MICE) in the mice (version 2.30) R package (van Buuren, 2011) (Supplementary Materials). Covariate distribution after imputation was similar to that in the observed data (Table S1). All primary analyses were performed in the imputed data sets and effect estimates were pooled across the 20 generated imputed data sets using Rubin’s rule (Marshall, Altman, Holder, & Royston, 2009).

Statistical Analysis

Comparisons of covariate distributions were made across categories of maternal 25(OH)D during pregnancy using the analysis of variance (ANOVA), Kruskal-Wallis rank sum test, or Chi-squared test. Similar comparisons were made to compare characteristics of the eligible sample to the excluded sample (with maternal 25(OH)D and offspring genetic data, but missing offspring depression measures).

25(OH)D was modeled as a categorical variable, consistent with clinical cut-offs defined by the Endocrine Society (i.e., deficient <20 ng/mL; insufficient 20-29.9 ng/mL; normal ≥30 ng/mL) (Dror & Allen, 2010), and also to account for the non-linearity of continuous 25(OH)D levels. PRS was modeled categorically by polygenic risk groups, as described above, with the reference group set as low PRS. Logistic regression models were used to assess: (1) the main effects of PRS and maternal 25(OH)D on offspring depression during childhood and/or adolescence, separately (Models 1-2) and together (Model 3); and (2) the interaction effect between PRS and maternal 25(OH)D on offspring depression (Model 4). All models were adjusted for gestational age at 25(OH)D measurement, maternal age at delivery, maternal pre-pregnancy BMI, maternal education, maternal occupation, parity, smoking during the first trimester, maternal depression during pregnancy, and child sex (Supplementary Materials). Models including PRS were additionally adjusted for the top three principal components (PCs); additional details about the population structure analysis are provided in the Supplementary Materials. Models with multiplicative interaction terms were further adjusted for all significant covariate*maternal 25(OH)D and covariate*PRS interactions, based on recommendations by Keller et al. (Keller, 2014). Effect estimates were presented as ORs and 95% confidence intervals (CI). Linear trend tests were performed to evaluate the dose-response relationship between vitamin D status and depression, and results were presented as p for trend (p-trend). Multiplicative interaction terms between maternal 25(OH)D and PRS were tested separately using the Wald test and globally using the likelihood ratio test. Tests for interaction were performed even in absence of significant main effects as there may be crossover interactions (i.e., the effect of maternal vitamin D is dependent on offspring PRS, or vices versa). Statistical significance was set at two-tailed p<0.05 for all analyses. All analyses were performed in R (version 3.3.1).

Based on the power calculations (Supplementary Materials), the study had 80% power to detect OR=1.4-1.7 for the association between vitamin D deficiency and child/adolescent depression, and sufficient power to detect GxE effects of magnitudes greater than ORgxe=1.7.

Sensitivity and Secondary Analyses

We conducted several sensitivity analyses to assess the robustness of the findings. First, we fit logistic regression models to further adjust for covariates that may be affected by or affect 25(OH)D levels during pregnancy, including vitamin D intake, calcium intake, vitamin D supplementation, and oily fish intake at 32 weeks of gestation. As the cut-offs for vitamin D deficiency remain controversial, we additionally conducted analyses using more conservative cut-offs for deficiency set by the Institute of Medicine (i.e., deficient <10 ng/mL; insufficient 10-19.9 ng/mL; normal ≥20 ng/mL) (Ross & Institute of Medicine (US), 2011). Also, given that vitamin D may have different effects on fetal brain development at different stages of gestation, we repeated the analyses stratified by the trimester of 25(OH)D measurement. Finally, we performed a complete case analysis on the subsample of participants with complete observed data on all variables (child subsample: n=2,091, adolescent subsample: n=1,812).

As a secondary analysis, we also examined the prospective association between maternal vitamin D status and the risk of offspring depressive symptoms using negative binomial regression models. Negative binomial regression allows for the modeling of the highly skewed and over-dispersed counts of depressive symptoms and could potentially provide greater statistical power to detect associations (Agresti, 2013). Estimates were presented as rate ratios (RR) and 95% CI, representing the association between maternal vitamin D status, PRS, and the risk for offspring depressive symptoms.

RESULTS

Among the eligible mother-offspring pairs (n=3,173), the mean maternal 25(OH)D was 27.38 ng/mL (SD=11.97), and 63.5% of mothers were vitamin D insufficient or deficient during pregnancy based on clinical cut-offs defined by the Endocrine Society. The proportion of mothers who had their 25(OH)D measurements taken during pregnancy was similar across seasons (Winter: 23.9%; Spring: 28.7%; Summer: 24.6%; Fall: 22.8%). The median gestational week at 25(OH)D measurement was 29.4 weeks (IQR: 12.7, 33.1), with most available measurements collected in the third trimester (58.5%). Table 1 displays the distribution of mother and offspring characteristics by maternal vitamin D status during pregnancy. Higher maternal vitamin D levels was also positively associated with maternal age at delivery, non-manual occupations, having more than one child, non-smoking during the first trimester, vitamin D intake during pregnancy, vitamin D supplementation intake, oily fish intake during pregnancy, and breastfeeding. 6.2% and 12.9% of the offspring were classified as experiencing depression during childhood and adolescence, respectively.

Table 1.

Maternal and offspring characteristics by maternal vitamin D status during pregnancy, among eligible mother-offspring pairs (defined by complete data on maternal 25(OH)D, offspring genetic data, and at least one SMFQ measure)

All (n=3173) Maternal 25(OH)D1
<20 ng/mL
(n=988)		 20-29.9 ng/mL
(n=1027)		 ≥30 ng/mL
(n=1158)		 p2
Maternal age at delivery, n (%)
 15-19 46 (1.4) 18 (39.1) 18 (39.1) 10 (21.7) 0.02
 20-35 2857 (90.0) 906 (31.7) 915 (32.0) 1036 (36.3)
>35 270 (8.5) 64 (23.7) 94 (34.8) 112 (41.5)
Maternal BMI (kg/m2), median [IQR] 22.18 [20.53, 24.38] 22.18 [20.53, 24.52] 22.35 [20.53, 24.51] 22.02 [20.47, 23.88]
Maternal education, n (%)
 Lower than O-levels 577 (19.3) 203 (52.3) 180 (31.2) 194 (33.8) 0.09
 O-levels 1080 (36.1) 331 (30.6) 359 (33.2) 390 (36.1)
 Higher than O-levels 1332 (44.6) 390 (29.3) 428 (32.1) 514 (38.6)
Maternal occupation, n (%)
 Manual 427 (16.0) 154 (36.1) 131 (30.7) 142 (33.2) 0.03
 Non-manual 2237 (84.0) 659 (29.5) 747 (33.4) 831 (37.1)
Parity, n (%)
 0 1446 (46.8) 482 (33.3) 477 (33.0) 487 (33.7) 0.007
 ≥1 1647 (53.2) 483 (29.3) 524 (31.8) 640 (38.9)
Gestational week of 25(OH)D measurement, median [IQR] 29.43 [12.71, 33.14] 28.29 [11.29, 32.86] 28.71 [11.43, 32.86] 31.14 [18.75, 33.29] <0.001
Season of 25(OH)D measurement, n (%)
 Winter 758 (23.9) 343 (45.3) 256 (33.8) 159 (20.9) <0.001
 Spring 912 (28.7) 412 (45.2) 297 (32.6) 203 (22.3)
 Summer 781 (24.6) 100 (12.8) 218 (27.9) 463 (59.3)
 Fall 722 (22.8) 133 (18.4) 256 (35.5) 333 (46.1)
Tobacco use during 1st trimester, n (%)
 No 2569 (82.2) 752 (29.3) 831 (32.4) 984 (38.3) <0.001
 Yes 557 (17.8) 221 (39.7) 181 (32.5) 155 (27.8)
Oily fish intake at 32 weeks, n (%)
 ≥1 times/week 813 (26.8) 214 (26.3) 261 (32.1) 338 (41.6) <0.001
<1 times/week 1045 (34.5) 318 (30.4) 341 (32.6) 386 (37.0)
 Never/rarely 1173 (38.7) 408 (34.8) 379 (32.3) 386 (32.9)
Vitamin D supplementation at 32 weeks, n (%)
 No 3035 (95.7) 961 (31.7) 985 (32.4) 1089 (35.9) 0.001
 Yes 138 (4.3) 27 (19.6) 42 (30.4) 69 (50.0)
Vitamin D intake at 32 weeks (ug), median [IQR] 3.43 [2.44, 5.40] 3.21 [2.28, 4.84] 3.39 [2.46, 5.46] 3.69 [2.59, 5.72] <0.001
Calcium intake at 32 weeks (mg), median [IQR] 938.54 [762.25, 1131.15] 935.50 [762.92, 1129.98] 940.01 [763.07, 1129.33] 939.08 [759.17, 1134.02] 0.87
Maternal depression during pregnancy3, n (%)
 No 2066 (75.6) 637 (30.8) 642 (31.1) 787 (38.1) 0.10
 Yes 667 (24.4) 211 (31.6) 231 (34.6) 225 (33.8)
Breastfeeding, n (%)
 No 1067 (34.9) 366 (34.3) 356 (33.4) 345 (32.3) 0.001
 Yes 1994 (65.1) 585 (29.4) 633 (31.7) 776 (38.9)
Child sex, n (%)
 Male 1592 (50.2) 483 (30.3) 520 (32.7) 589 (37.0) 0.62
 Female 1581 (49.8) 505 (31.9) 507 (32.1) 569 (36.0)
Open in a new tab
1

Unadjusted 25(OH)D levels

2

P-value calculated using the Chi-squared test for categorical variables and the ANOVA or Kruskal-Wallis rank sum test for normally and non-normally distributed continuous variables, respectively

3

Maternal depression at 18 and/or 32 weeks of gestation

Compared to the excluded participants, the eligible sample comprised mothers with higher 25(OH)D levels during pregnancy and offspring with lower PRS (Table S2). Further, mothers in the eligible sample also tended to be older when they gave birth, had higher educational attainment, were in non-manual occupations, had no previous pregnancies, were non-smokers during the first trimester, and showed no evidence of depression during pregnancy. However, the absolute difference between maternal 25(OH)D (i.e., 1.3 ng/mL higher in the eligible sample) and offspring PRS (i.e., 0.07 standardized units lower in the eligible sample) in the eligible and excluded samples was small, suggesting that the results are unlikely to be affected by selection bias.

Maternal Vitamin D Status and PRS on Offspring Depression

Associations between maternal 25(OH)D during pregnancy, PRS, and offspring depression are presented in Table 2. There were no associations between maternal 25(OH)D and offspring depression during childhood. However, there was suggestive evidence that offspring exposed to deficient vitamin D levels during gestation had higher odds of depression in adolescence, compared to offspring exposed to normal vitamin D levels (OR=1.32; 95% CI: 0.98, 1.79; p=0.07).

Table 2.

Association between maternal vitamin D status during pregnancy, offspring polygenic risk scores (PRS), and offspring depression during childhood or adolescence

Childhood depression (n=2938) Adolescent depression (n=2485)
OR 95% CI p OR 95% CI p
Model 1: Maternal 25(OH)D only
 Normal (≥30 ng/mL) ref ref ref ref ref ref
 Insufficient (20-29.9 ng/mL) 1.02 0.71, 1.48 0.90 1.12 0.84, 1.51 0.44
 Deficient (<20 ng/mL) 1.07 0.73, 1.58 0.72 1.32 0.98, 1.79 0.07
Model 2: PRS only
 PRS-Low ref ref ref ref ref ref
 PRS-Intermediate 1.37 0.90, 2.09 0.14 0.99 0.73, 1.34 0.93
 PRS-High 1.94 1.24, 3.03 0.004 1.34 0.96, 1.87 0.09
Model 3: PRS and Maternal 25(OH)D (Main effects)
 Normal (≥30 ng/mL) ref ref ref ref ref ref
 Insufficient (20-30 ng/mL) 1.03 0.71, 1.49 0.87 1.13 0.84, 1.51 0.43
 Deficient (<20 ng/mL) 1.08 0.73, 1.59 0.70 1.33 0.98, 1.81 0.06
 PRS-Low ref ref ref ref ref ref
 PRS-Intermediate 1.36 0.90, 2.09 0.15 0.98 0.73, 1.33 0.92
 PRS-High 1.94 1.24, 3.03 0.004 1.34 0.96, 1.87 0.09
Childhood depression (n=2938) Adolescent depression (n=2485)
OR 95% CI p p-inter. OR 95% CI p p-inter.
Model 4: PRS and Maternal 25(OH)D (Interaction)
 Normal (≥30 ng/mL) ref ref ref ref - ref ref -
 Insufficient (20-29.9 ng/mL) 0.65 0.29, 1.45 0.29 1.15 0.62, 2.11 0.66
 Deficient (<20 ng/mL) 0.38 0.13, 1.06 0.06 1.45 0.78, 2.69 0.24
 PRS-Low ref ref ref - ref ref ref -
 PRS-Intermediate 0.74 0.63, 1.46 0.38 1.21 0.71, 2.06 0.49
 PRS-High 1.26 0.37, 2.53 0.52 1.06 0.58, 1.96 0.84
 Insufficient*PRS-Intermediate 1.97 0.74, 5.26 0.18 0.19 0.80 0.38, 1.66 0.54 0.36
 Deficient*PRS-Intermediate 4.09 1.26, 13.3 0.02 0.68 0.32, 1.45 0.32
 Insufficient*PRS-High 1.65 0.59, 4.58 0.34 1.38 0.61, 3.15 0.44
 Deficient*PRS-High 2.92 0.86, 10.0 0.09 1.39 0.60, 3.22 0.44
Open in a new tab

Abbreviations: 25(OH)D: 25-hydroxyvitamin D; CI: confidence intervals; OR: odds ratio; p-inter.: p for interaction

Model 1: Adjusted for gestational age of 25(OH)D measurement, maternal age, maternal pre-pregnancy BMI, maternal education, maternal occupation, parity, smoking during 1st trimester, maternal depression during pregnancy, child sex

Model 2: Model 1 confounders + 3 PCs

Model 3: Model 2 confounders

Model 4: Model 3 confounders + significant PRS*covariate, maternal 25(OH)D*covariate interactions

Offspring PRS was positively associated with the odds of depression during childhood (p-trend=0.003), that is, offspring with high PRS had approximately 2-fold higher odds of childhood depression compared to offspring with low PRS (OR=1.94; 95% CI: 1.24, 3.03; p=0.0004); higher PRS was associated with non-significantly higher odds of depression during adolescence (p-trend=0.06) (Model 2). When modeled together (Model 3), the effect estimates for both maternal vitamin D status and PRS were almost identical to those yielded from their separate models, suggesting that they have independent associations with offspring depression. In Model 4, there was no evidence for significant interactions between maternal vitamin D status and PRS on childhood (p=0.19) or adolescent (p=0.36) depression.

Sensitivity and Secondary Analyses

Further adjustments for nutritional intake at 32 weeks of gestation (Model 5) (Table S3) yielded estimates similar to those from the main analyses. Using more conservative cut-offs for vitamin D deficiency created large sample size imbalances across categories, resulting in a small number of mothers classified as vitamin D deficient during pregnancy (n=55-71). Nonetheless, the direction and magnitude of effects were generally similar to those using higher cut-offs for deficiency (Table S4). The results did not change substantially when stratified by trimester of 25(OH)D measurement, although some of the effect directions were reversed in the small subsamples with measurements in the second trimester (n=381-438) (Table S5). Complete case analysis also showed similar trends to those identified from the primary analyses using the imputed data (Table S6), but the associations between PRS and childhood/adolescent depression were attenuated to non-significance, likely due to large reductions in sample sizes.

Results from the negative binomial regression models showed similar associational patterns as those produced from the logistic regression models (Table S7). There were no significant main effects of maternal 25(OH)D nor interactions with PRS on the risk of offspring depressive symptoms during childhood or adolescence. The only notable difference was the presence of a statistically significant positive association between PRS and the risk of depressive symptoms during adolescence (p-trend=0.03).

DISCUSSION

In this large, prospective birth cohort, we found little evidence for an association between maternal pregnancy vitamin D status (i.e., serum 25(OH)D level) and offspring depression in childhood or adolescence. In this sample, PRS was positively associated with risk for offspring depression at both time points but did not interact with maternal 25(OH)D. These findings were robust to adjustments for a range of potential confounders and different cut-offs for vitamin D insufficiency/deficiency.

Although several studies have reported the influence of maternal and/or cord blood 25(OH)D on the offspring’s risk for neuropsychiatric or neurodevelopmental disorders/traits, including schizophrenia (D. W. Eyles et al., 2018), ADHD (Morales et al., 2015), and autism-related traits (Vinkhuyzen et al., 2018), the same effects do not appear to be present for depression. Results from the current study are consistent with the only existing study, to our knowledge, that assessed the association between maternal 25(OH)D during pregnancy and offspring depression. In this Strom et al. study (Strom et al., 2014), maternal vitamin D deficiency during pregnancy was unassociated with offspring risk for depression over 22 years of follow-up. Other studies that assessed features similar to depression in childhood, such as emotional problems and/or internalizing symptoms, also did not report significant associations with maternal 25(OH)D (Gale et al., 2008; Keim et al., 2014; Whitehouse et al., 2012). Taken together, these findings do not support a role of maternal vitamin D status during pregnancy on offspring depression or related symptoms in childhood or adolescence.

One possible explanation for these lack of findings is that there may be sensitive developmental periods during which the effects of fetal 25(OH)D on neurodevelopment are particularly salient (Pet & Brouwer-Brolsma, 2016). The precise timing of the neurodevelopmental effects of vitamin D is unclear, but may have importance at any stage of pregnancy: early gestation, when structures important for behavioral regulation and dopaminergic neurons develop (Almqvist et al., 1996; Gale et al., 2008); mid-gestation, which is characterized by neural circuit formation and myelination (Keunen, Counsell, & Benders, 2017; Kinney, Brody, Kloman, & Gilles, 1988); or late gestation, when prolific brain development and growth occurs (Bouyssi-Kobar et al., 2016; Clouchoux, Guizard, Evans, du Plessis, & Limperopoulos, 2012; Kostovic & Jovanov-Milosevic, 2006). Yet, when we stratified our analysis by trimester of 25(OH)D measurement, we did not observe any meaningful differences. However, the stratified analyses had limited statistical power given the small sample sizes within some of the subgroups. Hence, the possibility for sensitive periods for vitamin D effects remains, and longitudinal studies sampling 25(OH)D in all three trimesters are needed to establish stronger conclusions; the Strom et al. study also did not explore these relationships. Alternatively, the lack of gestational 25(OH)D effects on child depression may be due to critical window(s) of vulnerability to vitamin D exposure occurring postnatally. In the Tolpannen et al. study (Tolppanen et al., 2012), similarly based on children in the ALSPAC cohort, childhood 25(OH)D3 levels were found to be associated with adolescent depressive symptoms. It is plausible that the protective effects of vitamin D on depression and mental health are exerted through neuroprotective actions, such as moderating inflammatory processes or modulating neurotrophic factors, after birth. Further studies are necessary to characterize the precise mechanisms of vitamin D on the developing brain.

Our study has several strengths. First, this study was conducted in a large, population-based sample that was more than three times larger than the previous study examining the same associations. Second, the longitudinal study design allowed us to prospectively investigate the effects of maternal 25(OH)D on offspring depression measured during both childhood and adolescence, while adjusting for a wide range of potential confounding factors.

The study also has some limitations. First, only one measurement of 25(OH)D, taken at any time during pregnancy, was available for all mothers in the study sample, which may not have been representative of average 25(OH)D levels for the duration of the pregnancy. However, studies have shown that single measurements may be reasonable proxies for vitamin D status throughout pregnancy, given the strong correlation in 25(OH)D concentrations over time (Hofmann, Yu, Horst, Hayes, & Purdue, 2010; Major et al., 2013). Second, the SMFQ only captures depressive symptoms in the past two weeks, which may not represent depression over longer periods. This may have led to outcome misclassification, particularly for adolescent depression, which has been shown to be highly episodic in nature (Holsen, 2000); such misclassification may have biased the results towards the null. Third, it is possible that an association between low maternal vitamin D levels and offspring depression is only detectable at very low levels of maternal 25(OH)D, and that the distribution of 25(OH)D levels in this population was not sufficiently wide to capture this. Compared to prior studies that detected significant associations (D. W. Eyles et al., 2018; Vinkhuyzen et al., 2018), mothers in the current study were less diverse ethnically/racially and/or socioeconomically, and on average had higher 25(OH)D levels (27.38 ng/mL compared to 15.3-23.6 ng/mL in other studies) with only a very small proportion having 25(OH)D levels <10 ng/mL (2%). Fourth, as with most longitudinal studies, there was considerable attrition over time, which could introduce bias when both maternal 25(OH)D and offspring depression were associated with loss to follow-up. Since the difference between the average maternal 25(OH)D levels in the excluded and included samples was small in magnitude, we do not expect attrition in this case to substantially influence the results. Fifth, the study was based in a white population, thus the results may not be generalizable to other racial/ethnic groups or populations with different prevalence of vitamin D deficiency and/or depression. Finally, we may have lacked power to detect weaker effects of maternal 25(OH)D on offspring depression and had even lower power to detect GxE. Larger sample sizes are needed to definitively rule out the presence of smaller associations and gene-environment interactions.

In sum, our findings do not support an association between maternal vitamin D status during pregnancy and offspring depression during childhood and adolescence. This suggests that interventions aimed at increasing 25(OH)D levels during pregnancy may not substantially reduce the risk of offspring depression during childhood or adolescence, although such efforts may still be beneficial for other child health outcomes (Wagner & Hollis, 2018). Larger prospective studies in more racially/ethnically diverse populations with lower average values and/or broader ranges of levels of 25(OH)D during pregnancy are needed to confirm the current findings.

Supplementary Material

Supplementary Material

Acknowledgments

We are extremely grateful to all the families who took part in this study, the midwives for their help in recruiting them, and the whole ALSPAC team, which includes interviewers, computer and laboratory technicians, clerical workers, research scientists, volunteers, managers, receptionists and nurses. ALSPAC children were genotyped using the Illumina HumanHap550 quad chip genotyping platforms by Sample Logistics and Genotyping Facilities at Wellcome Sanger Institute and LabCorp (Laboratory Corporation of America), using support from 23andMe. The UK Medical Research Council and Wellcome (Grant ref: 102215/2/13/2) and the University of Bristol provide core support for ALSPAC. A comprehensive list of grants funding is available on the ALSPAC website (http://www.bristol.ac.uk/alspac/external/documents/grant-acknowledgements.pdf). This research was specifically funded by: The UK Medical Research Council provided funds to ALSPAC for completion of the 25(OH)D assays used in this paper (grant no. G0701603).

REFERENCES

  1. Agarwal S, Kovilam O, & Agrawal DK (2018). Vitamin D and its impact on maternal-fetal outcomes in pregnancy: A critical review. Crit Rev Food Sci Nutr, 58(5), 755–769. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/27558700. doi: 10.1080/10408398.2016.1220915 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Agresti A (2013). Categorical data analysis (3rd ed.). Hoboken, NJ: Wiley. [Google Scholar]
  3. Almqvist PM, Akesson E, Wahlberg LU, Pschera H, Seiger A, & Sundstrom E (1996). First trimester development of the human nigrostriatal dopamine system. Exp Neurol, 139(2), 227–237. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/8654525. doi: 10.1006/exnr.1996.0096 [DOI] [PubMed] [Google Scholar]
  4. Angold A, Erkanli A, Silberg J, Eaves L, & Costello EJ (2002). Depression scale scores in 8-17-year-olds: effects of age and gender. J Child Psychol Psychiatry, 43(8), 1052–1063. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/12455926. [DOI] [PubMed] [Google Scholar]
  5. Bodnar LM, Catov JM, Simhan HN, Holick MF, Powers RW, & Roberts JM (2007). Maternal vitamin D deficiency increases the risk of preeclampsia. J Clin Endocrinol Metab, 92(9), 3517–3522. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/17535985. doi: 10.1210/jc.2007-0718 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bodnar LM, Simhan HN, Powers RW, Frank MP, Cooperstein E, & Roberts JM (2007). High prevalence of vitamin D insufficiency in black and white pregnant women residing in the northern United States and their neonates. J Nutr, 137(2), 447–452. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/17237325. doi: 10.1093/jn/137.2.447 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bouyssi-Kobar M, du Plessis AJ, McCarter R, Brossard-Racine M, Murnick J, Tinkleman L, … Limperopoulos C (2016). Third Trimester Brain Growth in Preterm Infants Compared With In Utero Healthy Fetuses. Pediatrics, 138(5). Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/27940782. doi: 10.1542/peds.2016-1640 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Boyd A, Golding J, Macleod J, Lawlor DA, Fraser A, Henderson J, … Davey Smith G (2013). Cohort Profile: the 'children of the 90s'--the index offspring of the Avon Longitudinal Study of Parents and Children. Int J Epidemiol, 42(1), 111–127. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/22507743. doi: 10.1093/ije/dys064 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Carey G, & Crammond B (2017). A glossary of policy frameworks: the many forms of 'universalism' and policy 'targeting'. J Epidemiol Community Health, 71(3), 303–307. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/25294894. doi: 10.1136/jech-2014-204311 [DOI] [PubMed] [Google Scholar]
  10. Clouchoux C, Guizard N, Evans AC, du Plessis AJ, & Limperopoulos C (2012). Normative fetal brain growth by quantitative in vivo magnetic resonance imaging. Am J Obstet Gynecol, 206(2), 173 e171–178. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/22055336. doi: 10.1016/j.ajog.2011.10.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Darling AL, Rayman MP, Steer CD, Golding J, Lanham-New SA, & Bath SC (2017). Association between maternal vitamin D status in pregnancy and neurodevelopmental outcomes in childhood: results from the Avon Longitudinal Study of Parents and Children (ALSPAC). Br J Nutr, 117(12), 1682–1692. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/28697816. doi: 10.1017/S0007114517001398 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Dror DK, & Allen LH (2010). Vitamin D inadequacy in pregnancy: biology, outcomes, and interventions. Nutr Rev, 68(8), 465–477. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/20646224. doi: 10.1111/j.1753-4887.2010.00306.x [DOI] [PubMed] [Google Scholar]
  13. Eggemoen AR, Falk RS, Knutsen KV, Lagerlov P, Sletner L, Birkeland KI, & Jenum AK (2016). Vitamin D deficiency and supplementation in pregnancy in a multiethnic population-based cohort. BMC Pregnancy Childbirth, 16, 7 Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/26785795. doi: 10.1186/s12884-016-0796-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Eyles D, Brown J, Mackay-Sim A, McGrath J, & Feron F (2003). Vitamin D3 and brain development. Neuroscience, 118(3), 641–653. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/12710973. [DOI] [PubMed] [Google Scholar]
  15. Eyles DW, Feron F, Cui X, Kesby JP, Harms LH, Ko P, … Burne TH (2009). Developmental vitamin D deficiency causes abnormal brain development. Psychoneuroendocrinology, 34 Suppl 1, S247–257. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/19500914. doi: 10.1016/j.psyneuen.2009.04.015 [DOI] [PubMed] [Google Scholar]
  16. Eyles DW, Trzaskowski M, Vinkhuyzen AAE, Mattheisen M, Meier S, Gooch H, … McGrath JJ (2018). The association between neonatal vitamin D status and risk of schizophrenia. Sci Rep, 8(1), 17692 Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/30523285. doi: 10.1038/s41598-018-35418-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fraser A, Macdonald-Wallis C, Tilling K, Boyd A, Golding J, Davey Smith G, … Lawlor DA (2013). Cohort Profile: the Avon Longitudinal Study of Parents and Children: ALSPAC mothers cohort. Int J Epidemiol, 42(1), 97–110. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/22507742. doi: 10.1093/ije/dys066 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gale CR, Robinson SM, Harvey NC, Javaid MK, Jiang B, Martyn CN, … Princess Anne Hospital Study, G. (2008). Maternal vitamin D status during pregnancy and child outcomes. Eur J Clin Nutr, 62(1), 68–77. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/17311057. doi: 10.1038/sj.ejcn.1602680 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Genomes Project C, Abecasis GR, Auton A, Brooks LD, DePristo MA, Durbin RM, … McVean GA (2012). An integrated map of genetic variation from 1,092 human genomes. Nature, 491(7422), 56–65. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/23128226. doi: 10.1038/nature11632 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hofmann JN, Yu K, Horst RL, Hayes RB, & Purdue MP (2010). Long-term variation in serum 25-hydroxyvitamin D concentration among participants in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial. Cancer Epidemiol Biomarkers Prev, 19(4), 927–931. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/20332255. doi: 10.1158/1055-9965.EPI-09-1121 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Holmes VA, Barnes MS, Alexander HD, McFaul P, & Wallace JM (2009). Vitamin D deficiency and insufficiency in pregnant women: a longitudinal study. Br J Nutr, 102(6), 876–881. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/19331703. doi: 10.1017/S0007114509297236 [DOI] [PubMed] [Google Scholar]
  22. Holsen I K. P, ; Vitterso J (2000). Stability in depressed mood in adolescence: Results from a 6-year longitudinal panel study. Journal of Youth and Adolescence, 29, 61–78. [Google Scholar]
  23. Howard DM, Adams MJ, Clarke TK, Hafferty JD, Gibson J, Shirali M, … McIntosh AM (2019). Genome-wide meta-analysis of depression identifies 102 independent variants and highlights the importance of the prefrontal brain regions. Nat Neurosci, 22(3), 343–352. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/30718901. doi: 10.1038/s41593-018-0326-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Joinson C, Kounali D, & Lewis G (2017). Family socioeconomic position in early life and onset of depressive symptoms and depression: a prospective cohort study. Soc Psychiatry Psychiatr Epidemiol, 52(1), 95–103. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/27837235. doi: 10.1007/s00127-016-1308-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Keim SA, Bodnar LM, & Klebanoff MA (2014). Maternal and cord blood 25(OH)-vitamin D concentrations in relation to child development and behaviour. Paediatr Perinat Epidemiol, 28(5), 434–444. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/24938425. doi: 10.1111/ppe.12135 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Keller MC (2014). Gene x environment interaction studies have not properly controlled for potential confounders: the problem and the (simple) solution. Biol Psychiatry, 75(1), 18–24. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/24135711. doi: 10.1016/j.biopsych.2013.09.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Keunen K, Counsell SJ, & Benders M (2017). The emergence of functional architecture during early brain development. Neuroimage, 160, 2–14. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/28111188. doi: 10.1016/j.neuroimage.2017.01.047 [DOI] [PubMed] [Google Scholar]
  28. Kinney HC, Brody BA, Kloman AS, & Gilles FH (1988). Sequence of central nervous system myelination in human infancy. II. Patterns of myelination in autopsied infants. J Neuropathol Exp Neurol, 47(3), 217–234. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/3367155. [DOI] [PubMed] [Google Scholar]
  29. Kostovic I, & Jovanov-Milosevic N (2006). The development of cerebral connections during the first 20-45 weeks' gestation. Semin Fetal Neonatal Med, 11(6), 415–422. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/16962836. doi: 10.1016/j.siny.2006.07.001 [DOI] [PubMed] [Google Scholar]
  30. Lawlor DA, Wills AK, Fraser A, Sayers A, Fraser WD, & Tobias JH (2013). Association of maternal vitamin D status during pregnancy with bone-mineral content in offspring: a prospective cohort study. Lancet, 381(9884), 2176–2183. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/23518316. doi: 10.1016/S0140-6736(12)62203-X [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Magnusson C, Lundberg M, Lee BK, Rai D, Karlsson H, Gardner R, … Dalman C (2016). Maternal vitamin D deficiency and the risk of autism spectrum disorders: population-based study. BJPsych Open, 2(2), 170–172. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/27703770. doi: 10.1192/bjpo.bp.116.002675 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Major JM, Graubard BI, Dodd KW, Iwan A, Alexander BH, Linet MS, & Freedman DM (2013). Variability and reproducibility of circulating vitamin D in a nationwide U.S. population. J Clin Endocrinol Metab, 98(1), 97–104. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/23144464. doi: 10.1210/jc.2012-2643 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Marchini J, Howie B, Myers S, McVean G, & Donnelly P (2007). A new multipoint method for genome-wide association studies by imputation of genotypes. Nat Genet, 39(7), 906–913. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/17572673. doi: 10.1038/ng2088 [DOI] [PubMed] [Google Scholar]
  34. Marshall A, Altman DG, Holder RL, & Royston P (2009). Combining estimates of interest in prognostic modelling studies after multiple imputation: current practice and guidelines. BMC Med Res Methodol, 9, 57 Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/19638200. doi: 10.1186/1471-2288-9-57 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Matijasevich A, Victora CG, Lawlor DA, Golding J, Menezes AM, Araujo CL, … Smith GD (2012). Association of socioeconomic position with maternal pregnancy and infant health outcomes in birth cohort studies from Brazil and the UK. J Epidemiol Community Health, 66(2), 127–135. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/20628081. doi: 10.1136/jech.2010.108605 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Morales E, Guxens M, Llop S, Rodriguez-Bernal CL, Tardon A, Riano I, … Project, I. (2012). Circulating 25-hydroxyvitamin D3 in pregnancy and infant neuropsychological development. Pediatrics, 130(4), e913–920. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/22987876. doi: 10.1542/peds.2011-3289 [DOI] [PubMed] [Google Scholar]
  37. Morales E, Julvez J, Torrent M, Ballester F, Rodriguez-Bernal CL, Andiarena A, … Sunyer J (2015). Vitamin D in Pregnancy and Attention Deficit Hyperactivity Disorder-like Symptoms in Childhood. Epidemiology, 26(4), 458–465. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/25867115. doi: 10.1097/EDE.0000000000000292 [DOI] [PubMed] [Google Scholar]
  38. O'Loan J, Eyles DW, Kesby J, Ko P, McGrath JJ, & Burne TH (2007). Vitamin D deficiency during various stages of pregnancy in the rat; its impact on development and behaviour in adult offspring. Psychoneuroendocrinology, 32(3), 227–234. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/17276604. doi: 10.1016/j.psyneuen.2006.12.006 [DOI] [PubMed] [Google Scholar]
  39. Pearson RM, Heron J, Button K, Bentall RP, Fernyhough C, Mahedy L, … Lewis G (2015). Cognitive styles and future depressed mood in early adulthood: the importance of global attributions. J Affect Disord, 171, 60–67. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/25285900. doi: 10.1016/j.jad.2014.08.057 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Pet MA, & Brouwer-Brolsma EM (2016). The Impact of Maternal Vitamin D Status on Offspring Brain Development and Function: a Systematic Review. Adv Nutr, 7(4), 665–678. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/27422502. doi: 10.3945/an.115.010330 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, … Sham PC (2007). PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet, 81(3), 559–575. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/17701901. doi: 10.1086/519795 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Rice F (2010). Genetics of childhood and adolescent depression: insights into etiological heterogeneity and challenges for future genomic research. Genome Med, 2(9), 68 Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/20860851. doi: 10.1186/gm189 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Ross AC, & Institute of Medicine (US). (2011). Dietary reference intakes : calcium, vitamin D. Washington, DC: National Academies Press. [PubMed] [Google Scholar]
  44. Sachs MC, Shoben A, Levin GP, Robinson-Cohen C, Hoofnagle AN, Swords-Jenny N, … de Boer IH (2013). Estimating mean annual 25-hydroxyvitamin D concentrations from single measurements: the Multi-Ethnic Study of Atherosclerosis. Am J Clin Nutr, 97(6), 1243–1251. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/23615830. doi: 10.3945/ajcn.112.054502 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Strom M, Halldorsson TI, Hansen S, Granstrom C, Maslova E, Petersen SB, … Olsen SF (2014). Vitamin D measured in maternal serum and offspring neurodevelopmental outcomes: a prospective study with long-term follow-up. Ann Nutr Metab, 64(3-4), 254–261. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/25300268. doi: 10.1159/000365030 [DOI] [PubMed] [Google Scholar]
  46. Thapar A, & McGuffin P (1998). Validity of the shortened Mood and Feelings Questionnaire in a community sample of children and adolescents: a preliminary research note. Psychiatry Res, 81(2), 259–268. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/9858042. [DOI] [PubMed] [Google Scholar]
  47. Tolppanen AM, Sayers A, Fraser WD, Lewis G, Zammit S, & Lawlor DA (2012). The association of serum 25-hydroxyvitamin D3 and D2 with depressive symptoms in childhood--a prospective cohort study. J Child Psychol Psychiatry, 53(7), 757–766. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/22211693. doi: 10.1111/j.1469-7610.2011.02518.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. van Buuren S G.-O. K (2011). mice: Multivariate Imputation by Chained Equations in R. Journal of Statistical Software, 45(3), 1–67. Retrieved from https://www.jstatsoft.org/v45/i03/. [Google Scholar]
  49. Vinkhuyzen AAE, Eyles DW, Burne THJ, Blanken LME, Kruithof CJ, Verhulst F, … McGrath JJ (2018). Gestational vitamin D deficiency and autism-related traits: the Generation R Study. Mol Psychiatry, 23(2), 240–246. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/27895322. doi: 10.1038/mp.2016.213 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Wagner CL, & Hollis BW (2018). The Implications of Vitamin D Status During Pregnancy on Mother and her Developing Child. Front Endocrinol (Lausanne), 9, 500 Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/30233496. doi: 10.3389/fendo.2018.00500 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Whitehouse AJ, Holt BJ, Serralha M, Holt PG, Kusel MM, & Hart PH (2012). Maternal serum vitamin D levels during pregnancy and offspring neurocognitive development. Pediatrics, 129(3), 485–493. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/22331333. doi: 10.1542/peds.2011-2644 [DOI] [PubMed] [Google Scholar]
  52. Williams DM, Fraser A, Fraser WD, Hypponen E, Davey Smith G, Deanfield J, … Lawlor DA (2013). Associations of maternal 25-hydroxyvitamin D in pregnancy with offspring cardiovascular risk factors in childhood and adolescence: findings from the Avon Longitudinal Study of Parents and Children. Heart, 99(24), 1849–1856. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/24125739. doi: 10.1136/heartjnl-2013-303678 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Wills AK, Shaheen SO, Granell R, Henderson AJ, Fraser WD, & Lawlor DA (2013). Maternal 25-hydroxyvitamin D and its association with childhood atopic outcomes and lung function. Clin Exp Allergy, 43(10), 1180–1188. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/24074336. doi: 10.1111/cea.12172 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Wray NR, Ripke S, Mattheisen M, Trzaskowski M, Byrne EM, Abdellaoui A, … Major Depressive Disorder Working Group of the Psychiatric Genomics, C. (2018). Genome-wide association analyses identify 44 risk variants and refine the genetic architecture of major depression. Nat Genet, 50(5), 668–681. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/29700475. doi: 10.1038/s41588-018-0090-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Xia L, & Yao S (2015). The Involvement of Genes in Adolescent Depression: A Systematic Review. Front Behav Neurosci, 9, 329 Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/26733829. doi: 10.3389/fnbeh.2015.00329 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material
NIHMS1611896-supplement-Supplementary_Material.docx (780KB, docx)

RESOURCES