This systematic review was performed by considering the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) criteria (23) (Supplementary Table S1). The protocol of the current study was approved and registered by the Research Vice-Chancellor of Tabriz University of Medical Sciences (ethical code: IR.TBZMED.REC.1401.258). Electronic databases of the Web of Sciences, PubMed, ScienceDirect, Scopus, and Google Scholar were searched without date restrictions until December 2022 by the MESH terms and the following keywords: maternal OR pregnancy (mesh) OR “pregnant women” (mesh) OR prenatal OR growth OR mother AND “vitamin D” (mesh) OR Calciferol OR “1,25(OH)2D” OR “1,25-dihydroxyvitamin D” AND “gastrointestinal microbiome” (mesh) OR microbiota (mesh) OR “human microbiome” (mesh) OR microbiome (mesh) OR microflora OR “gut flora” AND infants OR infancy OR neonate OR newborn OR baby OR babies OR offspring. To ensure the inclusion of all eligible studies, a separate search was conducted through Google, and the references of included studies were reviewed. Supplementary Table S2 displays the strategy of search.

The extracted studies were saved in EndNote software, and duplicate studies were removed. The remaining titles and abstracts were screened to identify studies with the correct scope for the current review. Then, the original full-text English-language articles were selected among the screened articles and critically and separately analyzed for eligibility. All peer-reviewed animal and human studies that addressed the findings regarding the effects or association of maternal prenatal vitamin D supplementation on/with an infant's gut microbiota were eligible for inclusion in this study. Since routine maternal vitamin D supplementation is done during pregnancy and lactation, the cohort studies that evaluated the association between maternal vitamin D levels in pregnancy and the infant's microbiome met the eligibility criteria for inclusion in the present review. The PICOS (Population, Intervention, Comparison, Outcome, and Study Design) principles were used to structure the study's inclusion and exclusion criteria (Table 1), and discrepancies in selecting the included studies were resolved by discussion. PICO criteria for this review were as follows: Population: pregnant individuals and their infants under 2 years old; Intervention: maternal prenatal vitamin D supplementation (different doses); Comparison: placebo or low doses of vitamin D; and Outcome: gut microbiota composition of infants such as Firmicutes, Bacteroidetes, Clostridales, and Lachnobacterium abundance.

Reviews, abstracts, conference papers, editorials, book chapters, posters, letters, theses, and generic studies were not included. Studies that measured the effects of infants' vitamin D supplementation on their gut microbiota were excluded. Studies that the full texts were not available were also excluded.

The extracted data from the eligible studies were the following: the authors' name, publication year, study design and location, the number of participants, mean age of infants and mothers, infant sex, maternal supplementation, breastfeeding status, confounders considered in the analysis, and findings concerning the effects of maternal supplementation with vitamin D during pregnancy on infant's gut microbiota (diversity and abundance of bacterial taxa such as Bacteroidetes, Firmicutes, Actinobacteria, and Proteobacteria). Moreover, the method and duration of vitamin D treatment were extracted from the included studies.

The quality of the mice studies, cohort studies, and randomized controlled trials (RCTs) was evaluated by the Office of Health Assessment and Translation (OHAT) Handbook (24), the Newcastle–Ottawa scale, and the Cochrane Handbook, respectively. For each risk of bias question, the OHAT tool offers the following response options: definitely a high risk of bias (score = −2), probably a high risk of bias (score = −1), probably a low risk of bias (score =  +1), or definitely a low risk of bias (score =  +2). After answering questions, numerical values were given to each question from −2 to +2, and according to the calculated average for each included study, studies were tiered to three levels of high, moderate, and low risk of bias, using the OHAT recommendations. Based on Approach 1, Tier 1 studies had probably low or low risk of bias, Tier 3 studies had probably high or high risk of bias, and Tier 2 studies had none of Tier 1 or Tier 3 criteria. Approach 2 calculates the overall average ratings of all questions, then using a mean rating per the study, each study is assigned to one of the three tiers based on the cutoffs of more than 0.7, 0.7 to −0.6, and less than −0.6 for Tiers 1–3, respectively (25, 26).

Assessment using the Newcastle–Ottawa scale was divided into three sections: selection, comparability, and outcome or exposure evaluation. Its score for cohort studies had a maximum of 9 points, and the study with an overall score between 6 and 9 points (≥3 points for the selection, 1 point for the comparability, and ≥2 points for the outcome or exposure section) was of good quality (27, 28). The Cochrane Collaboration's tool includes six domains of bias and three risk-of-bias response options: “low,” “high,” and “unclear” (25). Each included study was considered to have a “low risk of bias” if it had a low risk of bias in all three key domains (selection, detection, and performance domains), an “unclear risk of bias” if it had an unclear risk of bias in one or more than one key domain, and a “high risk of bias” if it had a high risk of bias in one or more than one key domain (25, 29).

The process of search and study selection (PRISMA diagram) for this systematic review is shown in Figure 1. One hundred ninety-seven potential articles were retrieved through searching in PubMed (n = 34), Web of Science (n = 64), Scopus (n = 85), ScienceDirect (n = 7), and Google Scholar (n = 7) databases. After the elimination of duplicate articles, 118 studies remained for further screening. Of these, 104 studies were excluded based on the title and abstract screening of the articles in the first stage. During critical analysis, 16 articles were screened, of which 8 articles were excluded because these were studying the effects of infants' supplementation with vitamin D on their gut microbiome (n = 2) (30, 31), studying the association between maternal diet and breastmilk microbiota (n = 1) (32), studying the impacts of maternal vitamin D supplementation on their gut microbiome (n = 1) (33), studying the association of maternal diets' vitamin D and infants' gut (n = 1) (34), were opinion studies (n = 1) (35), were published in other languages (n = 1) (36), and were abstract only (n = 1) (37). Finally, eight articles (two mice, three RCTs, and two cohorts) were included in this systematic review.

Tables 2, 3 present the main characteristics of the included studies. As shown in Table 2, two mice studies (19, 38) included 64 female C57BL/6 J mice that received 5,000 IU (high) and 25 IU (low) oral vitamin D during pregnancy. Three included RCTs were conducted from 2017 to 2020 and involved 1,236 mother–infant pairs. Pregnant individuals in intervention groups received 2,400, 4,000, and 4,400 IU vitamin D daily during week 24 pregnancy to 1 week postpartum (39), second and third trimesters (22), and all pregnancy time (40), respectively, while pregnant individuals in other groups received 400 IU routine vitamin D intake during these times. All women were instructed to continue 400 IU of vitamin D supplementation during pregnancy. In two RCT infants, stool samples were obtained at age 3–6 months (22, 40), and in one case, stool samples were collected on 1 week, 1 month, and 1 year after birth (39). Both sexes had been included in both three RCTs. Exclusively breastfeeding percentage of infants was 17% and 29.41% in two RCTs, and one case has not reported the breastfeeding status of included infants.

A total of 3,328 mother–child pairs participated in three included cohort studies. All the cohort studies included both sexes of infants. In included cohort studies, infants' stool samples were obtained at ages 3–4 months (41), 1 and 6 months (42), and 1 month after birth (43). The breastfeeding percentage of infants was 74.81% (41) and 64% (43) in vitamin D-supplemented infants, and one case has not reported the percentage of infants’ breastfeeding. Drall et al. (41) created four categories for maternal supplementation with vitamin D according to the Dietary Reference Intakes and Recommended Dietary Allowances of 400 and 600 IU/day vitamin D supplementation during pregnancy and breastfeeding, respectively (44, 45). Their four maternal supplementations of vitamin D categories are the following: no or less than 400 IU/day supplementation, only prenatal supplementation of ≥400 IU/day, only postnatal supplementation of ≥400 IU/day, and prenatal and postnatal supplementation with ≥400 IU/day vitamin D. In another cohort study also based on maternal multivitamin supplementation containing vitamin D, three categories for maternal vitamin D supplementation were considered, including none supplementation, <400, and ≥400 IU/day (43). All RCTs and cohort studies considered some possible confounders in their analysis.

Supplementary Table S3 shows the mouse studies' risk of bias assessment and tier classifications. Based on the OHAT's tier system, two mouse studies were classified as tier 1. One of the mice studies has not reported using of randomization in the mice's allocation into the groups and the blindness of outcome assessors (19), and other mice studies provided unclear data about adequately concealed allocation to study groups (38).

The included RCTs' risk of bias assessment details are described in Supplementary Table S4. Based on the Cochrane collaboration's criteria, all included RCTs had a low risk of bias. The personnel and participants were not blind in one of the RCTs (40), and others have not reported clear data about allocation concealment (39) and blinding of the outcome's assessor (22). The quality assessment of cohort studies resulted in the mean scores of 8 (41, 42) and 9 (43) as good quality (Supplementary Tables S5).

Table 3 summarizes the findings about vitamin D supplementation effects on offspring's gut flora composition. Moreover, a synoptic view of the main changes in the offspring's gut microbiome after maternal vitamin D administration is presented in Supplementary Table S6. Of eight included studies (two mice, three RCTs, and three cohorts), two mice and four humans (one RCT and three cohorts) studies (75%) demonstrated that maternal vitamin D supplementation during pregnancy was associated with the infant's gut microbiome and could impact gut microbiota composition such as Firmicutes and Bacteroidetes phyla. The description of the included studies is as follows.

Four included studies (one micee, one RCT, and two cohorts) revealed that maternal prenatal supplementation with vitamin D was associated with Firmicutes phylum abundance (19, 22, 41, 42). Li et al. indicated that vitamin D deficiency in the maternal significantly aggravated the dysbiosis of the mice's gut microbiota, with a relatively high abundance of Firmicutes (Lachnoclostridium, Lactobacillus, Ruminiclostridium_9, and Romboutsia), and high Firmicutes/Bacteroidetes ratio (19). Moreover, Sordillo et al. suggested that vitamin D levels in utero could impact the colonized microbial community of the infant's gut. They observed that the levels of Lachnospiraceae/U. Clostridales were higher in infants with higher vitamin D levels in cord blood, and multivariate analysis showed an increased Lachnobacterium abundance and a decreased Lactococcus abundance (22). Drall et al. also showed that maternal prenatal supplementation with vitamin D was related to lower Bilophila and Lachnospiraceae abundance but increased Haemophilus abundance in exclusively breastfed offspring (41). Another cohort study reported that prenatal 25(OH)D levels were positively associated with the microbiota composition of 1-month-aged infants including Acinetobacter, Corynebacterium, and Ruminococcus gnavus (42).

Two mouse studies mentioned above also reported that maternal prenatal vitamin D supplementation had an effect on the Bacteroidetes (Bacteroides, Alliprevotella, and Akkermansia) abundant in offspring mice gut microbiome (19, 38). Moreover, Talsness et al., in a cohort study, revealed similar findings. They reported a positive relationship between maternal 25(OH)D quintiles and Bacteroides fragilis group counts in the gut microbiota of 1-month-old infants. In addition, they indicated a negative linear trend between maternal supplementation of vitamin D and maternal 25(OH)D quintiles with Bifidobacterium spp. counts as a genus of Actinobacteria phylum. Their results suggested that vitamin D was associated with several key bacterial taxa abundance within the infants' microbiota (43).

Finally, two RCT studies reported no significant findings (39, 40). Hjelmsø et al. (39) demonstrated that the vitamin D interventions did not affect the infant's fecal microbiota, and Savage et al. (40) indicated a week association between the diet of maternal during pregnancy and the composition of the infant's gut microbiome. Overall, the included RCTs revealed mixed results for the effect of maternal prenatal vitamin D supplementation on the gut microbiome of infants (Table 3).

The inclusion of both animal and human studies in this review makes this study a comprehensive systematic review. However, there were some limitations to this study. The small number of included animal, RCT, and cohort studies was the main limitation of this systematic review. The included RCTs also used different doses of vitamin D. Furthermore, due to the heterogeneity of the results reported among the included studies, it was not possible to perform a meta-analysis. In addition, stool samples were collected at different time points; future studies should collect more samples over the first year of life. In the included studies, infant microbiota was not sampled right after birth. This would have allowed us to detect a possible fecal bacterial transfer event during birth.