Association of maternal serum magnesium with pre-eclampsia in African pregnant women: a systematic review and meta-analysis

Abstract Pre-eclampsia (PE) is a pregnancy-related disorder characterized by hypertension and proteinuria occurring after 20 weeks of gestation. Several studies have been performed to determine the serum magnesium (Mg) level in PE, but most report inconclusive results. Consequently, this study was designed to resolve this controversy among African women. PubMed, Hinari, Google Scholar and African Journals Online electronic databases were searched for studies published in English. The qualities of included articles were appraised using the Newcastle–Ottawa quality assessment tool. Stata 14 software was utilized for analysis and serum Mg levels in cases and normotensive controls were compared through mean and standardized mean difference (SMD) at the 95% confidence interval (CI). In this review, we found that the mean serum Mg level was significantly reduced in cases (0.910±0.762 mmol/L) vs controls (1.167±1.060 mmol/L). The pooled SMD of serum Mg was significantly lower in cases (−1.20 [95% CI −1.64 to −0.75]). Therefore, since serum Mg is reduced in cases vs controls, we propose that Mg is involved in the pathophysiology of PE. Nevertheless, to know the exact mechanisms of Mg in PE development will require large-scale prospective studies.


Introduction
Magnesium (Mg) is the fourth most abundant cation, after calcium, potassium and sodium, and it is the second most prevalent intracellular cation after potassium in human tissues. 1,2The human body contains approximately 21-28 g of magnesium.4][5] Another 40-50% of the total Mg is in the intracellular fluid and only 1% of Mg is found in the extracellular environment. 3,4In serum, Mg exists in three forms.The free ionized form is biologically active and accounts for 60%, the second form is protein bounded and accounts for 30% and the remaining 10% is found in complexes with anions such as phosphate and citrate. 4,5g is an essential mineral naturally found in many food stuffs and serves as a cofactor for >300 enzymatic reactions in energy production, protein synthesis, nucleic acid synthesis, mitochondrial functions, neuromuscular activity, bone formation, blood pressure (BP) regulation, heart rhythm regulation, genome integrity and immune system competence. 2,6There are sev-eral reports that revealed the association between serum Mg and several chronic diseases, including diabetes mellitus, hypertension, pre-eclampsia (PE), hyperlipidaemia, cardiac arrhythmia and bronchial asthma, and Mg is used for the treatment of these diseases. 7However, a recent randomized clinical trial showed that oral Mg supplementation did not reduce the risk of PE. 8 PE is a pregnancy-related disorder of hypertension and proteinuria occurring after 20 weeks of pregnancy. 9Despite the remarkable progress in the understanding of the pathophysiology of PE, the exact cause of PE is not fully understood, but evidence from animal and human studies so far have shown that failure of spiral artery remodelling causes placental ischaemia, leading to maternal syndromes. 10Several factors increase the risk of PE, including maternal micronutrient deficiency 11 and maternal genetic susceptibility, 12 and foetal factors also play a significant role in the pathogenesis of PE. 13 The molecular mechanism of how Mg deficiency causes PE is not fully understood, but it is believed that Mg ion decreases BP by acting as a calcium antagonist, leading to reduced peripheral vascular resistance. 14Likewise, Mg deficiency causes a decrease in the production of nitric oxide and endothelial prostaglandin I 2 , which results in vasoconstriction and high BP. 15,16There are readily available tests to determine the serum Mg status, although the serum Mg level has little correlation with the total body Mg level or with Mg in specific tissues. 4,5Many studies have tried to determine the serum Mg level in pre-eclamptic and normotensive pregnant women, but most of these have reported controversial results.3][24][25] Therefore, this review is intended to resolve this discrepancy in African pre-eclamptic and normotensive pregnant women.

Protocol registration
The protocol of the current study is recorded at the National Institute for Health Research (PROSPERO registration CRD42020192856).

Study design and search strategy
A systematic review was performed on available articles to compare maternal serum Mg levels in pre-eclamptic and normotensive pregnant African women.We searched the follow-ing databases: PubMed, Hinari, African Journals Online (AJOL) and Google Scholar.The search was done by using the following Medical Subject Heading (MeSH) terms: 'Electrolytes OR Magnesium (Mg) AND pre-eclampsia AND Africa' separately or in combination.All published and unpublished articles through 30 September 2021 were retrieved and assessed for their eligibility.The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline was utilized to conduct this systematic review and meta-analysis.

Eligibility criteria
Inclusion criteria were case-control and cross-sectional studies performed in pregnant African women, published and unpublished studies in English reporting serum Mg in mean±standard deviation (SD) and community and health-institution-based studies.Studies presented in conferences, editorials, reviews and clinical trials were excluded.

Study selection and screening
All documents recognized by our search were transferred into EndNote X9 (Clarivate, Philadelphia, PA, USA) and identical studies were excluded.The titles and abstracts of recognized studies were evaluated by two assessors and eligible articles were included for additional review.All recorded studies were carefully reviewed before data abstraction to ensure their appropriateness.When discrepancies occurred, a third evaluator was consulted.Our search strategy is presented in a PRISMA flow chart (Figure 1). 26. Tesfa et al.

Outcome of interest and terms
The primary objective of the current study was to evaluate the association between maternal serum Mg levels in pre-eclamptic and normotensive African pregnant women.Hypertension was defined as measured BP ≥140/90 mmHg a minimum of two times at 4-h intervals.Gestational hypertension was defined as measured BP ≥140/90 mmHg after 20 weeks of pregnancy.Preeclampsia was defined as measured BP ≥140/90 mmHg after 20 weeks of pregnancy plus proteinuria or the presence of severe features of the disease.Proteinuria was defined as urinary protein excretion ≥300 mg/24-h urine sample or ≥1+ on a presumptive dipstick investigation or a total protein:creatinine ratio ≥30 mg/mmol.Eclampsia was defined as seizures in the mother with high BP that could not have another cause. 27

Quality assessment
The quality of the case-control studies was appraised using the Newcastle-Ottawa Scale (NOS).A modified form of the NOS quality assessment tool was utilized to appraise the quality of cross-sectional studies. 28In the NOS, there are three criteria to a maximum score of nine.The quality of each article was evaluated by the following procedure: a score ≥7 was 'good', 4-6 was 'fair' and ≤3 was 'poor'.To increase the strength of our study, we included articles with good and fair quality scores. 28

Data extraction process
The data for this review were extracted using the Joanna Briggs Institute Reviewers' Manual data collection instrument. 29The abstracts and full texts were studied by two different evaluators.Data abstraction consisted of the name of the author, year of publication, state, study strategy, number of participants, maternal age, weeks of pregnancy, BMI, laboratory determination methods, systolic and diastolic BP and mean serum Mg level.Studies reporting the mean serum Mg level in mg/dl were converted into mmol/L by multiplying by 0.4114.

Data analysis
The data were entered into Excel (Microsoft, Redmond, WA, USA) and the meta-analysis was performed using Stata 14 software (StataCorp, College Station, TX, USA) and the paired sample t-test was performed using SPSS version 20 software (IBM, Armonk, NY, USA).A forest plot of the standardized mean difference (SMD) was used to compare the maternal serum Mg level among preeclamptic and normotensive pregnant women at the 95% CI.The SMD is the ratio of the mean difference to the pooled SD.The standard error of the mean (SEM) was calculated as SEM=SD/Ýn.Variables such as mean maternal age, mean gestational age, mean BMI, mean systolic BP, mean diastolic BP and mean Mg were evaluated using a paired sample t-test for their difference between cases and controls.

Heterogeneity and publication bias
Heterogeneity was assessed using Cochran's Q, the I 2 statistic and p-value.An I 2 statistic value <25%, 25-50% and ≥50% indicated little, average and high heterogeneity, respectively.In this study we utilized the random effects model (REM) for data analysis.Sensitivity tests were performed to determine the influence of independent variables and publication bias was judged using funnel plots and Egger's test.

Study selection
In this study, 246 studies were identified in four databases using a combination of search terms and 170 studies were found to be suitable for title and abstract evaluation after excluding 76 identical studies.A total of 140 articles were excluded by their title and abstract assessment.A total of 30 articles underwent full-text assessment for eligibility and 9 studies were excluded (5 were done outside the study area, 2 studies did not' report the result as mean±SD and the others were repeated publications and a review article).0][21][22][23][24][25][30][31][32][33][34][35][36][37][38][39][40][41]

Study characteristics
33]36,37 Studies that were conducted in Africa and published through 30 September 2021 were included.Fifteen studies were conducted in Nigeria, three in South Africa, one in Ghana and one each in Sudan and the Democratic Republic of Congo.The current study included 2398 pregnant women (1276 cases and 1122 controls) (Table 1).

Association of independent variables with pre-eclampsia
In this review, we computed the mean values of variables in pre-eclamptic and normotensive African pregnant mothers.Nonsignificant association has been observed among the study groups in respect to the mean values of maternal age, gestational age, and body mass index.However, significant association has been observed in the mean values of systolic and diastolic blood pressure with serum Mg level between the cases and controls (Table 2).

Sensitivity test and publication bias
In this review we performed sensitivity test by removing a single study at a time to know the consistency of the findings.There has no substantial variation in the combined SMD once removing one of the studies at 95% CI (Figure 3).This indicates that there is no individual study that extremely influences the combined values of serum Mg concentration among the two groups.Additionally, the funnel plot didn't show any evidence of publication bias between the serum Mg level in pre-eclamptic and normotensive pregnant women.Besides, Egger's test didn't show evidence of publication bias (p-value = 0.53).

Discussion
This study is the first systematic review and meta-analysis in Africa which provides baseline information on the association of maternal serum Mg level in pre-eclamptic and normotensive pregnant women.In this study, maternal age, gestational age and BMI were comparable between the two groups and these variables didn't showed statistical difference between the two groups.The mean systolic and diastolic blood pressure readings were significantly higher in pre-eclamptic women as compared to those normotensive pregnant women with the p-value of <0.001, as it is defined in the case definitions of this review.In this study, we found that the mean maternal serum Mg level was significantly reduced in pre-eclamptic women as compared with normotensive pregnant women (0.910 ± 0.762 vs 1.167 ± 1.06 mmol/L) with the overall SMD of Mg (SMD = −1.20,95% CI: −1.64, −0.75).Similar finding were reported in the studies conducted in Nigeria and China. 18,42This low serum Mg level might be attributed to an inadequate dietary intake or an increased Mg loss from the body.Even though, decreased serum Mg level does not always indicate the actual nutritional status of the women, because of the low percentage of Mg in the extracellular environment. 43There are reports which failed to clearly show oral Mg supplementation and risk of PE. 8,44 Mg sulfate is the preferred pharmacological intervention for the prevention and treatment of seizures in the women with severe PE, even though, the mechanism of action of Mg sulfate in eclampsia prophylaxis and treatment remains poorly understood. 45However, it is believed that dilation of cerebral blood vessels reduces cerebral ischemia and its action related to of Mg acting as calcium receptor blocker of muscle contraction. 46Mg sulfate causes peripheral vasodilation by blocking calcium receptor via inhibiting N-methyl-D aspartate receptors in the brain.It also alters the neuromuscular transmission via competitively blocking the entry of calcium into synaptic endings. 46Mg sulfate treatment showed a significant reduction in the risk of convulsion by 52% and 67% as compared to diazepam and phenytoin treatment, respectively.Mg sulfate treatment for severe pre-eclampsia reduces the risk of progression to eclampsia by more than half and also reduces maternal mortality.As well, Mg sulfate treatment showed a significant improvement in the neonatal outcomes as compared to phenytoin treatment. 43 is an essential micronutrient that is required in every cell and organism as an activator of several enzymes, bone formation, boosting immune system, mitochondrial functions, neuromuscular activity, proteins synthesis and DNA synthesis. 6,47In the blood and serum, Mg is mostly bound to serum albumin and stored in muscle fibers and in bones.The biologically active form is the ionized which is measured in the plasma. 47The serum level of Mg is maintained by the interaction between gastro-intestinal Mg absorption and renal Mg excretion. 47The intra-cellular Mg and bone Mg level did not play an active role in the regulation of blood Mg concentration; while a major role is played by the renal tubule, which adapts to match the urinary magnesium excretion and net entry of magnesium into extracellular fluid. 47he molecular mechanism of pre-eclampsia is not fully understood, but poor placentation cause's placental ischemia and maternal syndrome which resulted in hypertension and proteinuria. 48Maternal micronutrient deficiency and maternal genetic susceptibility play a substantial role in the etiopathogenesis of PE. 49,50 It is believed that Mg ion decrease the blood pressure by blocking calcium channel and by reducing peripheral vascular resistance. 14Mg ion affects blood pressure by modulating vascular tone and structure through different biochemical reactions that control vascular contraction/dilation, growth/apoptosis, differentiation and inflammation.It also stimulates production of vasodilator prostacyclin and nitric oxide and it alters vascular responses to vasoconstrictor agents. 15,51he serum Mg is tightly regulated by the interaction of intestine, kidney, bone and parathyroid hormone. 5The measured serum Mg concentration may be increased in patients with severe acute or chronic renal failure.In this case ionized and total Mg concentrations occasionally increased in individuals with E. Tesfa et al.   renal failure.Low serum Mg concentration may be observed in nutritional deficiency and in chronic diseases such as: hypertension, diabetes mellitus, coronary heart disease, and osteoporosis. 52In this review, we found that the mean serum Mg level was significantly low in pre-eclamptic women as compared with normotensive pregnant women that is might be due to nutritional deficiency.

Strengths and Limitations
This systematic review and meta-analysis generates pooled data that compared the serum Mg level in pre-eclamptic and normotensive African pregnant women.In addition, this review serves as the baseline information for further study.The study is not free from some form of limitations specially, the search strategy was limited to articles published in English, and this could lead to reporting bias.Most of the included studies originated from one country Nigeria, and this may influence its generalizability to African pregnant women.Besides, presence of high statistical heterogeneity among the included studies would decrease the generated evidence of this review.

Conclusions
The mean maternal serum Mg level was significantly reduced in pre-eclamptic women as compared with normotensive pregnant women.The overall pooled SMD of serum Mg level was also significantly decreased in cases than controls.Thus, we conclude that Mg could play certain roles in the etiopathogenesis of PE.However, concrete evidence on the functions of Mg and risk of PE pathogenesis in African pregnant women would require large scale prospective studies.

Figure 1 .
Figure 1.PRISMA flow chart showing studies conducted in African pregnant women.

Figure 2 .
Figure 2. Forest plot of SMD of serum Mg in pre-eclamptic and normotensive pregnant women.

Figure 3 .
Figure 3. Influence analysis of serum Mg in pre-eclamptic and normotensive pregnant women.

Table 1 .
Characteristics of research articles included in this review (N = 21)

Table 2 .
Paired sample t-test of variables associated with pre-eclampsiaSerum level of Mg in pre-eclamptic and normotensive pregnant women.