Gestational diabetes mellitus (GDM) is a common complication of pregnancy, named ‘any degree of glucose intolerance with onset or first recognition during pregnancy’(Reference Metzger and Gabbe1). With the increasing trend during the past two decades, it has become a major public health problem(Reference Metzger and Gabbe1). According to data released by the International Diabetes Federation in 2021, the global prevalence of women with hyperglycaemia during pregnancy is 16·7 %, of which GDM accounts for 80·3 %, and the number of live births affected is as high as 21·1 million(2). GDM can cause pre-eclampsia and may also increase the incidence of type 2 diabetes mellitus 3–6 years later after pregnancy(Reference Sun, Kim and Lee3,Reference Montoro, Kjos and Chandler4) . In addition, it can also lead to adverse pregnancy outcomes (e.g. macrosomia and preterm birth) and increase the risk of fetal adulthood diseases (e.g. obesity and CVD)(Reference Gou, Guan and Bi5–Reference Aviram, Guy and Ashwal7). Therefore, exploring the aetiology of GDM and seeking effective prevention strategies have realistic and long-term significance for improving maternal and infant outcomes.
Vitamin B12, the only water-soluble vitamin that contains metal ions, also known as cobalamin, plays an important role in pregnancy and is essential for DNA synthesis, cell division and amino acid metabolism. Total vitamin B12 decreases in concentration during the course of pregnancy, whereas holotranscobalamin (holoTC) remains relatively stable(Reference Greibe, Andreasen and Lildballe8,Reference Morkbak, Hvas and Milman9) . Two studies reported that lower serum total B12 concentrations were associated with a higher homeostatic model assessment for insulin resistance (HOMA-IR) index during the third trimester of pregnancy(Reference Sobczynska-Malefora, Yajnik and Harrington10), and B12-deficient pregnant women had higher BMI, the sum of skinfold thickness and insulin resistance than non-deficient women(Reference Krishnaveni, Hill and Veena11). Therefore, the relationship between vitamin B12 status during pregnancy and GDM has received widespread attention, but the findings were inconsistent. On the one hand, two studies found that vitamin B12 levels were inversely correlated with fasting blood glucose concentrations (r = -0·44, P = 0·0009; r = -0·29, P = 0·004)(Reference Radzicka, Ziolkowska and Zaborowski12,Reference Butt, Malik and Waheed13) . Studies reported that vitamin B12 levels were significantly lower in GDM groups in the second and third trimesters of pregnancy, especially among Asians(Reference Butt, Malik and Waheed13–Reference Li, Hou and Yan17). Some studies revealed vitamin B12 deficiency as a risk factor for GDM(Reference Krishnaveni, Hill and Veena11,Reference Butt, Malik and Waheed13–Reference Sukumar, Venkataraman and Wilson15,Reference Li, Hou and Yan17) . On the other hand, studies showed no association between vitamin B12 levels in the second and third trimesters of pregnancy and GDM, and the finding was also supported by others(Reference van Weelden, Seed and Antoun18–Reference Guven, Kilinc and Batukan20). And vitamin B12 deficiency was not associated with GDM(Reference van Weelden, Seed and Antoun18). Even more, a Chinese cohort study demonstrated higher vitamin B12 concentrations in the first-trimester GDM group than in the control group (421·00 v. 364·00 pg/ml, P < 0·002)(Reference Chen, Zhang and Chen21).
The aim of this study is to conduct a systematic review and meta-analysis on the association of vitamin B12 with GDM and to clarify whether vitamin B12 deficiency is associated with GDM. Recent studies have shown that high concentrations of folate interfere with GDM, and the article further explores the relationship between high folate:low vitamin B12 ratio and GDM.
Methods
Search strategy
The study was conducted by the Preferred Reporting Items of Systematic Reviews and Meta-Analyses (PRISMA) guidelines recommendations and searches performed in PubMed, Web of Science, EMBASE and Cochrane to identify all relevant publications updated before 22 September 2021. This review evaluated the effect of vitamin B12 concentration during pregnancy on GDM. Databases were searched by terms: vitamin B12; B12, Vitamin; Vitamin B12; B12, Vitamin; Cyanocobalamin; Cobalamins; Cobalamin; Eritron; Diabetes, Gestational; Diabetes, Pregnancy-Induced; Diabetes, Pregnancy Induced; Pregnancy-Induced Diabetes; Gestational Diabetes; Diabetes Mellitus, Gestational; Gestational Diabetes Mellitus. The full search strategy for the PubMed database is listed (online Supplementary Table S1).
Inclusion criteria
For this paper, the association of vitamin B12 and GDM was eligible for inclusion as follows: (1) studies showed the comparison of vitamin B12 concentrations in GDM and the non-GDM (control group) women; or (2) studies reported the incidence or prevalence of GDM based on the concentration of vitamin B12 or its deficiency levels, or the folic acid:vitamin B12 ratio.
Exclusion criteria
The following studies were excluded: (1) title/abstract/full text due to non-relevance; or (2) reviews, meta-analyses, conferences, letters, guidelines or English Literature; or (3) unable to get the full text; or (4) no in vitro or in vivo experiments; or (5) no relevant value for vitamin B12 or study variables.
Study selection process
Two independent authors (Xue Chen and Yushan Du) first independently screened the titles and abstracts, then identified that the relevant variables of the full-text articles were eligible. Finally, this paper included fifteen studies that fulfilled the inclusion criteria (Fig. 1).
Fig. 1. PRISMA flow diagram for reviews of selection process. PRISMA, Preferred Reporting Items of Systematic Reviews and Meta-Analyses.
Data extraction and quality assessment
Two reviewers independently extracted data from the study according to the pre-designed Excel tables. The data include the first author, publication year, participant nationality, study design, GDM diagnostic criteria, period of oral glucose tolerance test and vitamin B12 test, and the sample size. The quality assessment used the Newcastle–Ottawa Scale. The third reviewer will be joined the assessment if there are disagreements on the trial’s risk of bias.
Data synthesis and analysis
Meta-analysis was performed using R 4.1.0 software. Results are presented as mean differences (MD) for continuous variables (concentrations of vitamin B12) and the differences in risk ratios (adjustment of vitamin B12 deficiency) between GDM and non-GDM. Given that different periods for vitamin B12 and oral glucose tolerance tests are evaluated, we expect a large amount of heterogeneity in the results. The I 2 statistic is used to assess the heterogeneity between studies. I 2 values >50 % are considered to indicate a large heterogeneity, and further subgroup analyses explored the cause. Funnel charts with Egger’s tests were used to measure publication bias.
Result
Study characteristics
A total of 304 studies were identified, and fifteen studies were finally included. Details of the included studies are shown in Table 1. Of the fifteen studies, nine of them were from Europe (including the UK, Poland, Italy and Turkey) and six of them were from Asia (including China, Singapore, Pakistan and India). In all fifteen studies, the periods of oral glucose tolerance test, that is, the diagnosis time of GDM was between 24 and 32 weeks of pregnancy. However, the periods for vitamin B12 are inconsistent: four of the fifteen studies were detected during the first trimester (<13 weeks), ten were during the second or third trimester (>13 weeks), and one was not listed.
Table 1. Characteristics of the inclusion study (n 15)
GDM, gestational diabetes mellitus; OGTT, oral glucose tolerance test; IADPSG, International Association of the Diabetes and Pregnancy Study Groups; NICE, National Institute for Health and Clinical Excellence; ADA, American Diabetes Association; C&C, Carpenter and Coustan.
Effects of vitamin B12 and gestational diabetes mellitus
The effect of vitamin B12 concentration (pmol/l) during pregnancy on GDM is shown in Fig. 2. Compared with the non-GDM group, the vitamin B12 concentration in the GDM group was lower (MD = −6·83; 95 % CI (−17·56, 3·89)), although there is no significant statistical difference.
Fig. 2. Effect of vitamin B12 concentration (pmol/l) during pregnancy on GDM. GDM, gestational diabetes mellitus.
In subgroup analysis (Fig. 2), there was no association between maternal vitamin B12 concentration with GDM during the first trimester (MD = 15·75; 95 % CI (-20·05, 51·54)). However, the vitamin B12 concentration in the GDM group was remarkably lower than that in the non-GDM group (MD = -10·79; 95 % CI (-21·37, -0·21)) during the second or third trimester. We further analysed ethnicity as there is large heterogeneity (I 2 = 78 %, P < 0·01). The results illustrated that the vitamin B12 concentration of Asians dropped dramatically in the GDM group during the second or third trimester (MD = -25·78; 95 % CI (-38·56, -12·99)) (Fig. 3). There is no publication bias in the analysis (online Supplementary Fig. S1), and Egger’s tests were not significant (P = 0·52).
Fig. 3. Effect of vitamin B12 concentration (pmol/l) on GDM in the second or third trimester. GDM, gestational diabetes mellitus.
Effects of vitamin B12 deficiency and gestational diabetes mellitus
Six of the fifteen included studies examined the effect of vitamin B12 deficiency in the second or third trimester on GDM. The results suggested that vitamin B12 deficiency was associated with 1·59 times higher risk of GDM than the non-GDM group (OR: 1·59; 95 % CI (1·10, 2·29)) (Fig. 4). Further steps were taken to analyse these results due to heterogeneity (I 2 = 77 %, P < 0·01). There was a higher risk of vitamin B12 deficiency in GDM in Asians (OR: 2·08; 95 % CI (1·47, 2·96)), but no effects were observed among Europeans (Fig. 4). The publication bias was acceptable (online Supplementary Fig. S2), and Egger’s tests were not significant (P = 0·09).
Fig. 4. Effect of vitamin B12 deficiency on GDM in the second or third trimester. GDM, gestational diabetes mellitus.
Ratio of folate and vitamin B12 on gestational diabetes mellitus
Five of the fifteen studies assessed the risk of GDM with high folate:low vitamin B12 ratio (Fig. 5). An increased maternal high folate:low vitamin B12 ratio during the second or third trimester was associated with an 87 % increased risk of GDM compared with non-GDM (OR: 1·87; 95 % CI (1·46, 2·41)). The publication bias was acceptable (online Supplementary Fig. S3), and Egger’s test was not significant (P = 0·70).
Fig. 5. Effect of the high folate:low vitamin B12 ratio on GDM. GDM, gestational diabetes mellitus.
Discussion
This meta-analysis showed that vitamin B12 concentrations were lower in the GDM group, particularly in the second or third trimester, which was more remarkable for Asians. Some biochemical studies have shown decrease in circulating cobalamin levels during pregnancy(Reference Greibe, Andreasen and Lildballe8,Reference Morkbak, Hvas and Milman9,Reference Behere, Deshmukh and Otiv22) . This was associated with a physiological decline in maternal vitamin B12. The level of vitamin B12 was influenced by the concentration of the two plasma-binding proteins, transcobalamin and haptocorrin. Some studies had reported that total vitamin B12 levels in the mother gradually decreased during pregnancy with the decrease in holohaptocorrin (holo HC), and total B12-binding capacities had less saturation, although the levels of holotranscobalamin (holoTC) remain relatively stable(Reference Greibe, Andreasen and Lildballe8,Reference Morkbak, Hvas and Milman9,Reference Koebnick, Heins and Dagnelie23) . Thus, in addition to vitamin B12 transferred to the fetus and hemodilution, physiological fall in maternal B12 was also associated with changes in B12-binding proteins and biomarkers(Reference Greibe, Andreasen and Lildballe8,Reference Allen24,Reference Milman, Byg and Bergholt25) . Since there were only two studies on ethnic groups of Asian(Reference Chen, Zhang and Chen21,Reference Liu, Liu and Ma26) , the results may need further studies to verify.
Intracellularly, there were two vitamin B12-dependent enzymatic reactions. And vitamin B12 deficiency led to impaired methylation and impaired metabolism of methylmalonate. In the absence of vitamin B12, 5-methyltetrahydrofolate cannot be used for the formation of methionine and tetrahydrofolate, resulting in a state of pseudo-folate deficiency (methyl-trap). DNA synthesis based on methyl-traps was impaired, and alterations in mitochondrial content or function may progressively lead to the development of insulin resistance(Reference Shin, Beuhring and Stokstad27,Reference Maher and Sobczynska-Malefora28) . High homocysteine (Hcy) can induce cellular stress, apoptosis, and endothelial and DNA damage, which associate with CVD, pre-eclampsia and type 2 diabetes mellitus (Reference Selhub29,Reference Huang, Ren and Huang30) . Additionally, vitamin B12 deficiency leads to an elevated methylmalonic acid (MMA) concentration, a product of hydrolysis of the excessive concentration of methylmalonyl-CoA. MMA-CoA accumulation can inhibit fatty acids oxidation and hence increase adipogenesis and IR(Reference Maher and Sobczynska-Malefora28). This paper meta-analysed the risk of vitamin B12 deficiency during pregnancy on GDM and found that the risk of vitamin B12 deficiency during pregnancy in the GDM group was 1·59 times higher than that in the non-GDM group (OR: 1·59; 95 % CI (1·10, 2·29)), and the risk for Asians could reach 2·08 times (OR: 2·08; 95 % CI (1·47, 2·96)). Therefore, we speculated that vitamin B12 deficiency during pregnancy may induce GDM through IR. Unfortunately, four of the six studies defined vitamin B12 deficiency concentrations as <150 pmol/l, while the other two were well below this standard. This was unlikely to provide a GDM risk threshold based on vitamin B12 deficiency levels in the second or third trimester.
It was known that folate played an important role in the prevention of neural tube defects, and it was recommended for women of childbearing age to take folic acid supplements during the first trimester(Reference Estevez-Ordonez, Davis and Hopson31,Reference Osterhues, Ali and Michels32) . Some studies had shown that adequate folate intake before pregnancy could reduce the risk of GDM(Reference Li, Li and Chavarro33); however, some had also shown that high folate intake and low levels of vitamin B12 significantly increased the risk of GDM(Reference Chen, Zhang and Chen21,Reference Lai, Pang and Cai34) . Therefore, this study tried to clarify the correlation between the high folate to low B12 and GDM and showed that in the second or third trimester, higher folate:vitamin B12 ratios increased the risk of GDM by 87 % compared with non-GDM (OR: 1·87; 95 % CI (1·46, 2·41)).
A recent systematic review of the relationship between vitamin B12 status, pregnancy outcomes and offspring outcomes in India showed that the prevalence of vitamin B12 deficiency in Indian women during pregnancy had 40–70 %, and low maternal B12 and low vitamin B12:high folate ratio were associated with a higher risk for GDM(Reference Behere, Deshmukh and Otiv22). Unlike it, this study targets a wider population and is a multi-ethnic analysis of all countries, not just the Indian population, so the findings of the study are more comprehensive. In addition, this study also discussed the status of vitamin B12 at different stages of pregnancy, and the risk of vitamin B12 deficiency on GDM, which provided a basis for further analysis of the reasonable timing of vitamin B12 supplementation during pregnancy and a more accurate exploration of its mechanism.
We should be considered the limitations of the study. First, the lack of vitamin B12 biomarkers makes the assessment of GDM relatively simple. Some studies suggested that holoTC might be a better marker of vitamin B12 status during pregnancy than total vitamin B12 as it was less affected by hormonal changes and by the decrease in the levels of haptocorrin during pregnancy(Reference Nexo and Hoffmann-Lucke35,Reference Herzlich and Herbert36) . Hcy and MMA levels increased in the third trimester compared with the first trimester, which might be indicative of a degree of metabolic intracellular vitamin B12 depletion, even though both Hcy and MMA were lower than the established cut-off levels defining deficiency in non-pregnant women(Reference Greibe, Andreasen and Lildballe8,Reference Green, Allen and Bjorke-Monsen37) . Most markers of vitamin B12 status (circulating levels of total vitamin B12, holoTC, MMA and Hcy) are physiologically affected at low levels during pregnancy, complicating the assessment of vitamin B12. Whether the reference values for vitamin B12 status in non-pregnant women apply to pregnant women is also debatable. Second, the limited number of studies available for meta-analysis, particularly the lack of data on maternal vitamin B12 concentrations in the first trimester (only two studies were inconsistent), hindered accurate assessment of the association between vitamin B12 and GDM throughout pregnancy. Third, there were differences in the adjustment confounding factors of GDM risk in the report, so it is uncertain whether the above results will still occur in the second or third trimester.
Conclusions
Taken together, this study revealed that vitamin B12 concentrations were lower in the GDM group in the second or third trimester, and vitamin B12 deficiency increased the risk of GDM, which was more significant in Asians. An increased maternal high folate:low vitamin B12 ratio during the second or third trimester also increased the risk of GDM. These results suggest that more vitamin B12 needs to be provided during pregnancy. To better promote the results and consider whether vitamin B12 supplementation is needed during pregnancy, we need more studies on folate and vitamin B12 associated with GDM or glycolipid metabolism. It can provide a theoretical clue for the prevention strategy of GDM and further study on the mechanism of vitamin B12 and GDM.
Acknowledgements
This research was supported by the Chinese Nutrition Society-Zhen Dong Tizhi and Health Fund (CNS-ZD2020-115). We also would like to express our sincere gratitude to the editor Matthew Woodcock (Dr.) and three anonymous reviewers for their valuable comments, which have greatly improved this paper.
X. C.: conceptualisation, data curation, formal analysis, methodology, software and writing original draft preparation. Y. D.: data curation. S. X.: methodology and writing – review. Z. L.: writing – review. J. L.: methodology, supervision, writing – review and editing.
The authors declare no conflict of interest.
Supplementary material
For supplementary material/s referred to in this article, please visit https://doi.org/10.1017/S000711452200246X
