Vitamin D Deficiency and Antenatal and Postpartum Depression: A Systematic Review

Vitamin D has been implicated in antenatal depression (AD) and postpartum depression (PPD) in many studies; however, results have been inconsistent due to the complexity of this association. We searched the MEDLINE, Embase, PsycINFO, and Maternity and Infant Care databases for literature addressing associations between vitamin D and AD and PPD. Two independent authors reviewed titles and abstracts of the search results and selected studies for full review. Data were extracted, and a quality rating was done using the Newcastle–Ottawa Scale (NOS) on the selected studies. A total of 239 studies were identified; 14 were included in the review. The quality assessment of the included studies ranged from moderate to high. Of the studies on PPD, five of nine (55%) showed a significant association between vitamin D and PPD. Five of seven (71%) studies on AD showed a significant association with vitamin D status. As the included studies used different effect estimates and statistical analyses to report the association, it was not possible to transform the existing data into one single effect measure to employ meta-analytic techniques. While results of this systematic review vary, they indicate a significant association between vitamin D status and AD and PD.


Introduction
Depressive symptoms during and after pregnancy are the leading cause of disease-related disability among women [1][2][3]. Antenatal depression (AD) and postpartum depression (PPD) are common, with a prevalence of 18% and 19%, respectively [4]. Depressive symptoms during and after pregnancy are associated with unfavourable outcomes for mothers and their infants [5]; AD and PPD have been associated with poor cognitive development, behavioural outcomes, and mental and physical health in children [5].
Although biological, psychological, and environmental theories of depression have been advanced, the underlying pathophysiology of depression remains unknown, and it is probable that several different mechanisms are involved [6]. Studies have shown the beneficial effect of dietary factors on depressive symptoms during pregnancy [7][8][9]. Vitamin D is one of those dietary factors and has been suggested to beneficially affect depression in adults [10][11][12]. It has been hypothesized that vitamin D may act as a neuroactive hormone [13][14][15]. Several studies have shown that vitamin D receptors are broadly distributed throughout the human brain [16] and its deficiency alters neuro-transmitters that are known to be involved in depressive symptoms [17]. Most recently, it has been postulated that vitamin D modulates levels of neuronal calcium ions (Ca 2+ ) that are responsible for the onset of depressive symptoms. Conversely, a deficiency of vitamin D may lead to an increase in neuronal Ca 2+ , thus increasing depression [13]. In addition, vitamin D may also play a role in neuro-immunomodulation and neuro-plasticity, both of which are proposed mechanisms for the observed effect on mood [17].

Materials and Methods
This systematic review was conducted following the Cochrane Collaboration methodology for systematic reviews [33] and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) reporting guidelines [34].

Search Strategy and Study Selection
In accordance with a protocol developed a priori, we identified all relevant articles regardless of language by searching the MEDLINE, Embase, PsycINFO and Maternity and Infant Care databases from inception to February 2017 in consultation with a senior research librarian. We also scanned the bibliographies of identified articles. To allow a systematic review of all studies assessing the association between serum 25(OH)D concentration and depression during and after pregnancy outcomes, our initial search was not limited to any type of study. The search was broken into three themes: (1) to identify vitamin D, a Boolean search was performed using the term "or" to explode (search by subject heading) and map (search by keyword) the following MeSH headings: "vitamin D" OR "vitamin D2" OR "vitamin D3" OR "ergocalciferol" OR "cholecalciferol" OR "25-Hydroxyvitamin D" OR "25(OH)D" OR "25(OH)D2" OR "25(OH)D3" OR "3-epi-25 hydroxyvitamin D" or "D2" OR "D3" OR "vitamin D Deficiency" OR "hypovitaminosis D"; (2) to identify pregnancy-related outcomes, a second Boolean search was performed using the term "or" to explode (search by subject heading) and map (search by keyword) the following MeSH headings: "pregnancy" OR "prenatal" OR "antenatal"; (3) to identify depression, a third Boolean search was performed using the term "or" to explode (search by subject heading) and map (search by keyword) the following MeSH headings: "depression" OR "postpartum depression" OR "puerperal depression".
Themes 1, 2, and 3 were combined using the Boolean operator "and" to answer the focus questions (association between vitamin D status and AD and PD). The search results were compiled using citation management software (EndNote version X7; Clarivate Analytics, Philadelphia, PA, USA).

Eligibility Criteria
We determined screening criteria a priori, with the aim of identifying fully published articles that assessed the association of 25(OH)D and AD and PPD. Two reviewers (F.A. and N.L.) screened abstracts and titles to identify articles for further review. Articles were considered for inclusion if they reported on original data from an original study, included outcome of interest (prenatal and postpartum depression), and utilized blood samples during pregnancy that assessed serum 25(OH)D levels. Two independent reviewers (F.A. and N.L.) searched related articles and links. Additional articles were identified by manual search of the references from the key articles selected. Although the search was not limited by language, all articles included in the review were published in English. Two independent reviewers (F.A. and N.L.) independently assessed the abstracts for potential inclusion, fully reviewed the selected articles, and selected the final articles for the systematic review. Studies that were only published as abstracts were excluded. Disagreements were resolved by means of meeting and discussion among all authors to establish a consensus.

Data Extraction and Management
We developed a data extraction form to collect key indicators of each study, including study design, definition of 25(OH)D cut-off levels, 25(OH)D assessment method, gestational age at serum sampling, location and latitude of population, scale and the cut-off used to assess the outcome of interest (AD and PPD), and whether the studies adjusted for potential confounders in their statistical analysis. Articles were categorized on the basis of the outcomes of interest, antenatal depression and PPD. We collected data on definition of these outcomes as reported in the articles. Articles reporting on multiple outcomes were included. Two reviewers (N.M. and N.C.) independently extracted information from each article and compared findings; any discrepancies were resolved by consensus in meetings with all authors. Attempts were made to contact authors of studies with unclear data.

Assessment of Methodological Quality of Included Studies
Two reviewers (F.A. and N.L.) assessed the methodologic quality of included studies using the Newcastle-Ottawa Scale (NOS) for observational studies [35]. This scale rates studies on three major domains: selection, comparability, and the ascertainment of outcome of interest. We identified high-quality choices by answering "Yes" to the questions in each domain. The more Yeses allocated to a study (to a maximum of nine Yeses), the better the quality it was. Although one of the main sources of bias in these studies was confounders, the definition of cases (participants with or without depression) and the outcome assessment (studies using different scales and different cut-off of the scale to measure depression) was also important. At any point, any disagreement between reviewers was resolved by means of meeting and discussion among all authors to establish a consensus.

Data Synthesis
We extracted data from the included studies and prepared the data in table format. The study outcomes are described in the Results section. The included studies used different estimates and statistical analyses to report associations, thus making it impossible to transform the existing data into one single effect measure. As there was huge heterogeneity among the studies, we were unable to employ meta-analytic techniques.

Methodologic Quality of Included Reviews
When using the NOS for cohort/cross-sectional studies, the quality assessment of the included studies ranged from low scores of 4 [24] and moderate scores of 5-6 [20,25,27,28,30] to high scores of 8-9 [19,22,23,25,29]. For the one case-control study, the NOS score was high [21]. Table 3 shows the NOS quality scores of the included reviews. The quality of the included randomized clinical trial (RCT) [31] was assessed using the Jadad scoring system [50] and was moderate (score of 3 out of 5).

Association between Vitamin D Status and AD
Five of the seven studies (71%) that examined AD as an outcome showed a significant association between lower levels of vitamin D and AD [26,[28][29][30][31], while one study showed no association [27]. In contrast, one study found that higher (as opposed to lower) vitamin D concentration was associated with a higher risk of AD at 16 weeks (AOR 1.14, 95% CI 1.02-1.27) [32]. Higher vitamin D was associated with reduced odds of AD at 18.5 weeks (AOR 0.51, 95% CI 0.31-0.72) [30], and 21 weeks gestation (AOR 0.54, 95% CI 0.29-0.99) [29], as well as reduced symptoms of depression at 38-40 weeks gestation (lower EPDS scores, p < 0.001) [31]. In one study, vitamin D concentrations <29 nmol/L at 16 weeks of pregnancy was associated with higher odds of AD (AOR 1.48, 95% CI 1.13-1.95) [28]. Of the studies with significant associations, two were considered high quality [28,29]. Significant associations were found with Caucasian [28,32], African American [29,32], Japanese [30], and Iranian [31] samples. As the included studies used different effect estimates and statistical analyses to report associations, it was not possible to transform the existing data into one single effect measure to employ meta-analytic techniques. See Tables 1 and 2 for details.

Association between Vitamin D Status and PPD
Five of the nine studies (55%) that examined PPD as an outcome showed a significant association between vitamin D in pregnancy and PPD [22][23][24][25]31], while four studies showed no association [19][20][21]26]. Lower vitamin D concentration (<47 nmol/L) was associated with increased odds of PPD at three days postpartum (AOR 2.72, 95% CI 1. 42-5.22) [25], and higher levels of vitamin D was associated with decreased odds of PPD three months postpartum (AOR 0.81, 95% CI 0.70-0.92) [22], as well as reduced symptoms at one week, six weeks, and six months postpartum (r = 0.2, p = 0.02; r = 0.2, p = 0.01; r = 0.3, p = 0.01, respectively) [23], four to eight weeks postpartum (p < 0.001) [31], and one to seven months postpartum (p < 0.02) [24]. One study found an association between lower vitamin D concentration and higher depressive symptom score, but not with diagnosis of major depressive disorders [26]. Of the studies that found significant associations, three were of high quality [22,23,26]. Significant associations were found with Caucasian (American and Australian) [24][25][26], Chinese [22], Iranian [31], and Turkish [23] samples. As the included studies used different effect estimates and statistical analyses to report the association, it was not possible to transform the existing data into one single effect measure to employ meta-analytic techniques. See Tables 1 and 2 for details.

Discussion
Available data provide evidence on the potential association between vitamin D status during pregnancy and AD and PPD. We systematically assessed evidence of association between serum 25(OH)D concentration and the risk of AD and PPD, including observational studies and one RCT. Variations in study characteristics, analytical methods, and methodological qualities were identified. To the best of our knowledge, this is the first systemic review examining the association between vitamin D and both AD and PPD. We were not able to employ meta-analytic techniques to this review due to different effect estimates and statistical analyses that these studies used to report the association. This systemic review overall showed a trend toward significant association between vitamin D status during pregnancy and AD and PPD; however, the causality from these findings cannot be inferred due to the observational nature of the studies included in this systematic review.
These findings are of concern, particularly given recent evidence suggesting that 25(OH)D deficiency or insufficiency is common during pregnancy, especially among high-risk groups [51][52][53]. Vitamin D supplementation may be a simple way to reduce the risk of these adverse outcomes. In fact, one RCT [31] showed that supplementation with 2000 IU/day of vitamin D during pregnancy reduces the risk of antenatal depression. This suggests that low levels of 25(OH)D may be a modifiable risk factor in pregnancy, and health care providers should at least be encouraging pregnant women to follow current guidelines on recommended daily allowances for vitamin D [18].
Several mechanisms may explain the observed association between 25(OH)D concentration and risk of depression in pregnancy. Depression is associated with dysregulated hypothalamic-pituitaryadrenal axis function, overactivity of the sympatho-adrenal system, and increased level of inflammatory markers [54]. Vitamin D has been shown to down-regulate inflammatory mediators that have been linked to sickness behaviour, psychosocial stress, and depression [55]. Studies in pregnant women suggest that the effect of low 25(OH)D on PPD symptoms may be potentiated in the presence of inflammation [20]. Vitamin D may also have direct neuroregulatory activity. Vitamin D receptor (VDR) gene polymorphisms in humans have been associated with cognitive impairment and depressive symptoms [56]. In addition, systematic reviews of epidemiological studies have shown that lower concentration of 25(OH)D is correlated with increased risk of depression in adults [10,57].

Limitations
This systematic review is limited by a lack of or limited adjustment for confounding factors in some of the included studies. Studies have used different assays to measure vitamin D concentrations. A few systematic reviews of vitamin D and disease-related outcomes have found that the method of 25(OH)D measurement was an important determinant of heterogeneity [58,59]. The method of vitamin D measurement is an important factor, as DEQAS (International Vitamin D External Quality Assessment Scheme) has reported a range of inter-method variability for identical blood samples [60]. Currently, the best assay to measure different types of vitamin D including epimers is LC-MS/MS (liquid chromatography-tandem mass spectrometry) [61]. In addition to the inherent flaws in the assay methodology, the studies varied in their definition of cut-offs for 25(OH)D deficiency (>50 nmol/L) and (>75 nmol/L). Moreover, there are other confounding factors that may affect vitamin D status, including exposure to UVB (Ultraviolet B), latitude, season, ethnicity, nutritional status, BMI, and VDR (vitamin D receptor) genotype [62]. Furthermore, there are multiple factors affecting depression during and after pregnancy: age, other comorbidities, education and living situations, social support, and income [4]. These parameters were absent or partially evaluated in most studies included in this systematic review. Finally, in the assessment of antenatal and postnatal depression, discrepancies exist among the studies on the type of assessment tools and the cut-offs they used to describe antenatal or postnatal depression.

Conclusions
In summary, based on the systemic evaluation of the previous literature in the field, there may be an association between lower vitamin D status and increased risk of depressive symptoms during and after pregnancy. While the quality of the available evidence was not always optimal due to lower methodologic quality of the studies, this review provides an analysis of the methodological issue that future supplementation studies need to consider in their research design.

Acknowledgments:
The authors would like to acknowledge Debbie Booth, Medical Librarian, The University of Newcastle, Australia.
Author Contributions: F.A. designed the study, searched databases, performed the selection of studies, critically evaluated the studies, and wrote the manuscript. N.L. performed the selection of studies, critically evaluated the review, and wrote the manuscript. N.M. abstracted the data, contributed to writing the manuscript, commented on it, and approved the last version. N.C. abstracted the data, contributed to writing the manuscript, commented on it, and approved the last version. G.G. contributed to data abstraction and critical evaluation of the data, contributed to writing the manuscript, commented on it, and approved the last version. All authors reviewed and approved the final manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.