Calcium supplementation commenced before pregnancy for preventing hypertensive disorders and related problems

Abstract

Rationale

Hypertensive disorders of pregnancy are a leading cause of maternal and perinatal morbidity and mortality. Calcium supplementation commenced before pregnancy may prevent the development of these disorders. This is an update of a review last published in 2019.

Objectives

To assess the effects of calcium supplementation commenced before pregnancy on hypertensive disorders of pregnancy and related maternal and neonatal outcomes.

Search methods

We searched CENTRAL, MEDLINE, Embase, CINAHL, Portal Regional BVS‐Lilacs, Scopus, Web of Science, WHO ICTRP and ClinicalTrials.gov on 7 January 2025 and searched reference lists of retrieved trials and relevant systematic reviews.

Eligibility criteria

We included randomised controlled trials (RCTs) that compared calcium supplementation commenced before pregnancy with placebo or standard care. Trials conducted after 2010 needed to be prospectively registered. We applied a trustworthiness checklist.

Outcomes

Critical outcomes for women were pre‐eclampsia or pregnancy loss, and pre‐eclampsia. The critical outcome for children was perinatal loss.

Our main important outcomes for women were pregnancy loss at any gestational age, maternal death, maternal death or severe morbidity, and adverse effects. Our main important outcomes for children were preterm delivery before 37 weeks, neonatal death or severe morbidity, stillbirth, neonatal death, and early neonatal death.

Risk of bias

We used version 2 of the Cochrane tool for assessing risk of bias in randomised trials (RoB 2).

Synthesis methods

Two review authors independently selected trials, extracted data, and assessed risk of bias and trustworthiness. We pooled data using random‐effects meta‐analysis. We assessed the certainty of the evidence using GRADE. Because conception and pregnancy loss are intermediate outcomes potentially on the causal pathway to pre‐eclampsia, we chose the composite outcome 'pre‐eclampsia or pregnancy loss' as the first critical outcome (and most relevant to pregnant women), and included sensitivity analyses including only women who conceived during the trial.

Included studies

We included one multicentre, double‐blind, randomised, placebo‐controlled trial. It included 1355 parous women, whose most recent planned pregnancy had been complicated by pre‐eclampsia or eclampsia.

Synthesis of results

Pre‐eclampsia or pregnancy loss at any gestational age

Compared to placebo, calcium may result in little to no difference in pre‐eclampsia or pregnancy loss at any gestational age (risk ratio (RR) 0.85, 95% confidence interval (CI) 0.67 to 1.07; 1 RCT, 1355 women; risk difference (RD) 28/1000 fewer, 95% CI 61 fewer to 13 more; low‐certainty evidence).

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.82 (95% CI 0.66 to 1.00; 1 RCT, 633 women).

Pre‐eclampsia

Compared to placebo, calcium may result in little to no difference in pre‐eclampsia (RR 0.84, 95% CI 0.62 to 1.14; 1 RCT, 1355 women; RD 19/1000 fewer, 95% CI 46 fewer to 17 more; low‐certainty evidence).

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.81 (95% CI 0.61 to 1.07; 1 RCT, 633 women).

Pregnancy loss at any gestational age

Compared to placebo, calcium may result in little to no difference in pregnancy loss at any gestational age (RR 0.92, 95% CI 0.69 to 1.24; 1 RCT, 1355 women; RD 9/1000 fewer, 95% CI 37 fewer to 28 more; low‐certainty evidence).

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.89 (95% CI 0.67 to 1.17; 1 RCT, 633 women).

Maternal death

The evidence is very uncertain about the effect of calcium compared to placebo on maternal death (RR 1.00, 95% CI 0.14 to 7.07; 1 RCT, 1355 women; RD 0/10,000 fewer, 95% CI 25 fewer to 179 more; very low‐certainty evidence).

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.96 (95% CI 0.14 to 6.77; 1 RCT, 633 women).

Maternal death or severe morbidity

The evidence is very uncertain about the effect of calcium, compared to placebo, on maternal death or severe morbidity (RR 0.97, 95% CI 0.70 to 1.35; 1 RCT, 1355 women; RD 3/1000 fewer, 95% CI 29 fewer to 34 more; very low‐certainty evidence).

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.93 (95% CI 0.68 to 1.27; 1 RCT, 633 women).

Preterm delivery before 37 weeks

Compared to placebo, calcium may result in little to no difference in preterm delivery before 37 weeks (RR 0.94, 95% CI 0.74 to 1.19; 1 RCT, 1355 women; RD 11/1000 fewer, 95% CI 46 fewer to 33 more; low‐certainty evidence).

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.90 (95% CI 0.74 to 1.11; 1 RCT, 633 women).

Stillbirth

The evidence is very uncertain about the effect of calcium compared to placebo on stillbirth (RR 0.82, 95% CI 0.50 to 1.34; 1 RCT, 1355 women; RD 9/1000 fewer, 95% CI 24 fewer to 17 more; very low‐certainty evidence).

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.79 (95% CI 0.48 to 1.27; 1 RCT, 633 women).

The included trial did not measure perinatal loss, maternal adverse effects, neonatal death or severe morbidity, neonatal death, or early neonatal death.

Authors' conclusions

When all randomised women are considered, calcium commenced before pregnancy may result in little to no difference in pre‐eclampsia or pregnancy loss, and pre‐eclampsia. When only pregnant women are considered, calcium may result in little to no difference in pre‐eclampsia but may result in a slight reduction in pre‐eclampsia or pregnancy loss. No trials measured perinatal loss.

The evidence is drawn from one trial of calcium supplementation that commenced before and continued into the first half of pregnancy. Current evidence neither supports nor refutes the routine use of calcium supplementation commencing before conception.

Funding

This review was funded in part by the World Health Organization.

Registration

Updated protocol (2025): PROSPERO: CRD420250649571

Review update (2019): DOI: 10.1002/14651858.CD011192.pub3

Original review (2017): DOI: 10.1002/14651858.CD011192.pub2

Plain language summary

Do calcium supplements started before pregnancy help to prevent high blood pressure and its complications?

Key messages

• Evidence from one study (1355 women) shows that extra calcium before pregnancy may result in little to no difference to pre‐eclampsia (a complication of pregnancy) or pregnancy loss, and pre‐eclampsia. Evidence from 633 women who got pregnant during the study showed similar results except for a slight reduction in pre‐eclampsia or pregnancy loss.

• The study did not provide evidence for unwanted effects of calcium or death of the baby during pregnancy and early life.

• We need more research to decide whether calcium supplements started before pregnancy help women avoid high blood pressure, pre‐eclampsia and other complications. Research should also address how acceptable women find calcium supplements.

Why is high blood pressure a problem in pregnancy, and how could calcium help?

High blood pressure in pregnancy is a leading cause of death and severe illness in mothers and babies. Pre‐eclampsia is the most serious complication. It is diagnosed when the pregnant woman develops high blood pressure with signs of damage to the placenta and other organs such as the kidneys or liver. There is currently no treatment for pre‐eclampsia apart from delivering the baby. Calcium may help to lower blood pressure. Many people, especially in low‐income countries, have insufficient calcium in their diets, so giving calcium supplements before pregnancy may prevent pre‐eclampsia and could save many mothers and babies.

What did we want to find out?

We wanted to know if giving calcium supplements to women before they get pregnant could help prevent high blood pressure, pre‐eclampsia and other serious health problems for the mother and baby. We were also interested in any unwanted effects of calcium supplements.

What did we do?

We searched for studies that compared giving calcium supplements before pregnancy with placebo (a dummy medicine) or normal care. Women in the studies had to be planning a pregnancy. They could be any age and live anywhere in the world. Their normal diets could include any amount of calcium, and they could be at any risk of high blood pressure. Studies could give any dose of calcium for any length of time, as long as supplements started before pregnancy. We looked at the evidence for all the women in the study, and only for women who became pregnant during the study.

We compared and summarised the studies' results, and rated our confidence in the evidence, based on factors such as study methods and sizes.

What did we find?

We found one study with 1355 women. They had all had pre‐eclampsia in their most recent pregnancy, so were at high risk of getting it again. They were given extra calcium or a placebo tablet daily until 20 weeks of pregnancy. After 20 weeks, all women were switched to extra daily calcium until they gave birth. They lived in Argentina, South Africa, and Zimbabwe.

Calcium supplements before pregnancy may make little to no difference to:

• pre‐eclampsia or pregnancy loss;

• pre‐eclampsia;

• pregnancy loss; and

• delivery of the baby before 37 weeks.

We are not sure about the effect of extra calcium before pregnancy on death of the mother, death or severe illness of the mother and stillbirth.

The results for 633 women who got pregnant during the study are very similar, except that there may be a slight reduction in pre‐eclampsia or pregnancy loss.

There was no evidence about unwanted effects, death of the baby during pregnancy and early life, death of the newborn baby, death of newborns in the first 7 days of life, or death or severe illness of newborns.

What are the limitations of the evidence?

Our confidence in the evidence from this study is limited because nearly a third of the women included were not followed up or withdrew from the study. We do not know whether these women became pregnant and what their outcomes were.

How up to date is this evidence?

The evidence is current to January 2025.

Summary of findings

Summary of findings 1. Summary of findings table ‐ Calcium supplementation commenced before pregnancy compared to placebo or standard care for hypertensive disorders of pregnancy.

Calcium supplementation commenced before pregnancy compared to placebo or standard care for hypertensive disorders of pregnancy
Patient or population: women planning to become pregnant Setting: hospitals in South Africa, Zimbabwe and Argentina Intervention: calcium supplementation commenced before pregnancy Comparison: placebo or standard care
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo or standard care	Risk with calcium supplementation commenced before pregnancy
Pre‐eclampsia or pregnancy loss at any gestational age	186 per 1000	158 per 1000 (125 to 199)	RR 0.85 (0.67 to 1.07)	1355 (1 RCT)	⊕⊕⊝⊝ Lowa,b	Calcium may result in little to no difference in pre‐eclampsia or pregnancy loss at any gestational age.
Pre‐eclampsia	121 per 1000	102 per 1000 (75 to 138)	RR 0.84 (0.62 to 1.14)	1355 (1 RCT)	⊕⊕⊝⊝ Lowa,b	Calcium may result in little to no difference in pre‐eclampsia.
Pregnancy loss at any gestational age	118 per 1000	109 per 1000 (82 to 147)	RR 0.92 (0.69 to 1.24)	1355 (1 RCT)	⊕⊕⊝⊝ Lowa,b	Calcium may result in little to no difference in pregnancy loss at any gestational age.
Maternal death	30 per 10,000	30 per 10,000 (4 to 209)	RR 1.00 (0.14 to 7.07)	1355 (1 RCT)	⊕⊝⊝⊝ Very lowa,c	The evidence is very uncertain about the effect of calcium on maternal death.
Maternal death or severe morbidity	96 per 1000	93 per 1000 (67 to 130)	RR 0.97 (0.70 to 1.35)	1355 (1 RCT)	⊕⊝⊝⊝ Very lowa,d	The evidence is very uncertain about the effect of calcium on severe maternal morbidity and mortality.
Adverse effects	Not pooled	Not pooled	Not pooled	(0 studies)	‐	No trials measured this outcome.
Perinatal loss	Not pooled	Not pooled	Not pooled	(0 studies)	‐	No trials measured this outcome.
Preterm delivery before 37 weeks	176 per 1000	165 per 1000 (130 to 209)	RR 0.94 (0.74 to 1.19)	1355 (1 RCT)	⊕⊕⊝⊝ Lowa,b	Calcium may result in little to no difference in preterm birth before 37 weeks.
Neonatal death or severe morbidity	Not pooled	Not pooled	Not pooled	(0 studies)	‐	No trials measured this outcome.
Stillbirth	49 per 1000	40 per 1000 (24 to 65)	RR 0.82 (0.50 to 1.34)	1355 (1 RCT)	⊕⊝⊝⊝ Very lowa,d	The evidence is very uncertain about the effect of calcium on stillbirth.
Neonatal death	Not pooled	Not pooled	Not pooled	(0 studies)	‐	No trials measured this outcome.
Early neonatal death	Not pooled	Not pooled	Not pooled	(0 studies)	‐	No trials measured this outcome.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_458521133664360774.

^a Downgraded by one level for risk of attrition bias. In the calcium group, 29% of women were lost to follow‐up or withdrew. in the placebo group, 28% of women were lost to follow‐up or withdrew.
^b Downgraded by one level for serious imprecision: the 95% confidence interval includes both appreciable harms and benefits.
^c Downgraded by three levels for extremely serious imprecision: the 95% confidence interval includes both appreciable harms and benefits, the confidence interval is very wide and events were rare.
^d Downgraded by two levels for very serious imprecision: the 95% confidence interval includes both appreciable harms and benefits and the confidence interval is wide.

Background

Hypertensive disorders of pregnancy are a leading cause of maternal and perinatal morbidity and mortality. They are characterised by high blood pressure during pregnancy. The main types include chronic hypertension, gestational hypertension, pre‐eclampsia, and pre‐eclampsia superimposed on chronic hypertension. Of these, pre‐eclampsia is the most severe [1]. Pre‐eclampsia is a multi‐organ, pregnancy‐specific condition characterised by hypertension with end‐organ injury. It is associated with varying degrees of impaired placental perfusion, which results in the release of placental factors into the maternal circulation, causing widespread endothelial injury [1].

Description of the condition

Pre‐eclampsia is estimated to complicate about 3% to 8% of all pregnancies [2, 3], and is estimated to cause at least 42,000 maternal deaths and more than half a million perinatal deaths every year [2, 4, 5]. It is a progressive disease of pregnancy. The definition has evolved over time and the majority of international guidelines have moved to a broader classification [6]. Most agree that pre‐eclampsia can be diagnosed as new‐onset hypertension (sustained systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg) with proteinuria (1+ by dipstick testing, ≥ 300 mg in 24 hours, or ≥ 30 mg/dL), or end‐organ dysfunction that develops after 20 weeks of gestation [6]. Superimposed pre‐eclampsia is diagnosed when there is loss of blood pressure control with worsening proteinuria, or end‐organ involvement in women with chronic hypertension. End‐organ involvement includes eclampsia (generalised seizures), pulmonary oedema, haemolysis, elevated liver enzymes and low platelet (HELLP) syndrome, liver dysfunction or rupture, renal impairment and uteroplacental involvement. There is currently no treatment for pre‐eclampsia apart from delivery.

Description of the intervention and how it might work

Calcium supplementation is an oral dietary supplement, usually produced in the form of calcium carbonate or calcium gluconate. The dose is expressed in terms of the amount of elemental calcium in the preparation.

The daily dietary calcium requirement is 1000 mg to 1200 mg; 50% of the global population are at risk of calcium deficiency, particularly in Africa and Asia [7]. Calcium supplementation reduces blood pressure in normotensive people [8], and in non‐pregnant people with previous pre‐eclampsia [9].

An inverse relationship between calcium intake and hypertensive disorders of pregnancy was first described in 1980 [10]. This was based on the observation that Mayan Indians in Guatemala, who traditionally soaked their corn in lime before cooking, had a high calcium intake and a low incidence of pre‐eclampsia and eclampsia. A very low prevalence of pre‐eclampsia had been reported from Ethiopia, where the diet, among other features, contained high levels of calcium [11]. These observations were supported by other epidemiological and clinical trials, leading to the hypothesis that an increase in calcium intake during pregnancy may reduce the incidence of pre‐eclampsia among women with low calcium intake [12, 13, 14, 15, 16, 17].

Low calcium intake may cause hypertension by stimulating either parathyroid hormone or renin release, thereby increasing intracellular calcium in vascular smooth muscle, leading to vasoconstriction [12]. A possible mode of action for calcium supplementation is that it reduces parathyroid release and intracellular calcium, and so reduces smooth muscle contractility. By a similar mechanism, calcium supplementation could also reduce uterine smooth muscle contractility and prevent preterm labour and delivery [18]. Calcium might also have an indirect effect on smooth muscle function by increasing magnesium levels [19]. There is evolving basic scientific evidence that calcium is potentially important in spermiogenesis, oocyte maturation, decidualisation and placentation, deficiencies of which may contribute to impaired fertility, pregnancy loss and pre‐eclampsia [20].

Calcium supplementation is attractive as a potential intervention to reduce the risk of developing pre‐eclampsia as it is readily available, and is likely to be safe.

Why it is important to do this review

Most research has focused on calcium supplementation during pregnancy to prevent pregnancy‐induced hypertensive disorders [21]. This may be too late to interrupt early pregnancy events that are precursors of pre‐eclampsia. It is therefore important to assess the data to see if calcium supplementation commenced before pregnancy can prevent hypertensive disorders of pregnancy and related complications.

Objectives

To assess the effects of calcium supplementation commenced before pregnancy on hypertensive disorders of pregnancy and related maternal and child outcomes.

Methods

We followed the Methodological Expectations for Cochrane Intervention Reviews (MECIR) when conducting the review [22], and PRISMA for reporting the review [23].

Although this is an update of a previous review published in 2019 [24], we developed a new protocol prior to beginning the update [25]. We developed a new search strategy and searched databases and trials registries from inception to January 2025, made amendments to critical and important outcomes, screened all trials from the new search output, assessed eligible trials for trustworthiness, extracted data and assessed risk of bias using the Risk of Bias 2 (RoB 2) tool for all included trials [26]. The previous 2019 review version defined the population as "women of child‐bearing age but not yet pregnant." We redefined the population as "women before pregnancy at any age."

The previous 2019 review had three primary outcomes for women [24]. We have retained two of these as 'critical' outcomes, 'pre‐eclampsia', and 'pre‐eclampsia or pregnancy loss'. The composite outcome, 'severe maternal morbidity and mortality', is now an 'important' outcome.

We also removed the following outcomes, which were listed as secondary outcomes in the 2019 review [24].

• Maternal outcomes: no conception during study period, pregnancy loss before 20 weeks' gestational age', severe gestational hypertension, severe pre‐eclampsia, uric acid > reference values for gestational age, renal failure, liver failure, mother’s hospital stay seven days or more after birth, caesarean section, and severe pre‐eclamptic complications index

• Neonatal outcomes: birthweight < 2500 g, early preterm birth (< 32 weeks’ gestation), death or admission to neonatal ICU (intensive care unit) for 24 hours or more, and pregnancy loss, stillbirth, or neonatal death before discharge.

We added several outcomes to this updated review, namely: stroke, acute kidney injury, liver capsule haematoma or rupture, intubation and mechanical ventilation (not for childbirth), maternal satisfaction, maternal well‐being, postpartum haemorrhage, cortical blindness, retinal detachment, hypertension, recurrent eclampsia, raised liver enzymes, low platelets, infant loss up to six weeks after the due date, early preterm delivery at before 34 weeks, extremely preterm delivery at before 28 weeks, necrotising enterocolitis, intraventricular haemorrhage (IVH) grade 3 or 4, respiratory distress syndrome (RDS) 3 or 4, neonatal sepsis, neonatal encephalopathy, birthweight (g), admission to neonatal intensive care unit (NICU) or high care unit (HCU), childhood hypertension, neonatal seizures, Apgar score less than 7 at five minutes after birth, adverse effects to the neonate as a result of maternal treatment interventions, and gestational age at birth in weeks.

We also amended the definition of the composite outcome, 'maternal death or severe morbidity', based on outcomes in a Delphi consensus [27]. Our definition includes eclampsia, stroke, pulmonary oedema, heart failure, renal failure, liver capsule haematoma or rupture, placental abruption, HELLP syndrome. We specified in the protocol that postpartum haemorrhage and ICU admission would not be part of the composite definition, despite being part of the core outcome set [27].

In the 2019 review, we presented the following outcomes in summary of findings tables.

• Maternal outcomes: pre‐eclampsia; pre‐eclampsia or pregnancy loss, or stillbirth (or combination) at any gestational age; severe maternal morbidity and mortality index; pregnancy loss or stillbirth at any gestational age; caesarean section

• Neonatal outcomes: birthweight < 2500 g; Apgar score less than 7 at five minutes; perinatal death or NICU admission for > 24 hours (or both); pregnancy loss, stillbirth or neonatal death before discharge

In this updated review, we present the following outcomes in a summary of findings table.

• For women: pre‐eclampsia or pregnancy loss at any gestational age; pre‐eclampsia; pregnancy loss at any gestational age; maternal death; maternal death or severe morbidity; adverse effects

• For children; perinatal loss; preterm delivery before 37 weeks; neonatal death or severe morbidity; stillbirth; neonatal death; early neonatal death

Criteria for considering studies for this review

Types of studies

We included parallel‐group randomised controlled trials (RCTs) with individual or cluster randomisation. We excluded cross‐over RCTs.

We included trials irrespective of their publication status and language of publication.

Trials conducted after 2010 needed to be prospectively registered in a clinical trials registry.

Types of participants

We searched for trials that recruited women of any age, before pregnancy, regardless of the risk of hypertensive disorders of pregnancy, within any setting of care. We included women who had low or adequate dietary calcium intake.

Types of interventions

We included trials that assessed the effects of calcium supplementation in any dose, alone or in combination with co‐interventions, compared with placebo, no intervention, or standard treatment alone, starting before pregnancy and administered for any period of time. Eligible comparators were placebo, no intervention, standard treatment alone or low‐dose calcium supplementation. We only included trials with co‐interventions if they were administered in both groups.

We were particularly interested in the following comparisons:

• calcium supplementation compared with no calcium supplementation;

• high‐dose (≥ 1 g elemental calcium daily) compared with low‐dose (< 1 g elemental calcium daily) calcium supplementation.

Outcome measures

Outcomes were based on a core outcome set on pre‐eclampsia research published in 2020 [27].

Because pregnancy loss is an intermediate outcome potentially on the causal pathway to pre‐eclampsia [28], we chose the composite outcome, 'pre‐eclampsia or pregnancy loss at any gestational age', as our first critical outcome (and the outcome most relevant to pregnant women).

Critical outcomes

For women

• Pre‐eclampsia or pregnancy loss at any gestational age (including super‐imposed pre‐eclampsia as defined by trial authors)

• Pre‐eclampsia (including super‐imposed pre‐eclampsia as defined by trial authors)

Time point: as defined by trial authors.

For children

• Perinatal loss: death of a foetus or baby between 22 weeks and 7 days postpartum or as defined by trial authors.

Important outcomes

For women

The time point for all outcomes is as defined by the trial authors.

• Pregnancy loss at any gestational age

• Maternal death

• Maternal death or severe morbidity (eclampsia, stroke, pulmonary oedema, heart failure, renal failure, liver capsule haematoma or rupture, placental abruption, HELLP syndrome, or as defined by trial authors)

• Adverse effects of interventions as defined by trial authors

• Eclampsia

• HELLP syndrome

• ICU admission

• Stroke

• Pulmonary oedema

• Acute kidney injury

• Liver capsule haematoma or rupture

• Placental abruption

• Intubation and mechanical ventilation (not for childbirth)

• Maternal satisfaction as defined by trial authors

• Maternal well‐being as defined by trial authors

• Postpartum haemorrhage

• Cortical blindness

• Retinal detachment

• Hypertension

• Gestational hypertension

• Recurrent eclampsia

• Severe hypertension (systolic blood pressure of ≥ 160/110 mmHg, or as defined by trial authors)

• Miscarriage as defined by trial authors

• Proteinuria

• Raised liver enzymes

• Low platelets

• Gestational age at diagnosis of pre‐eclampsia in weeks

• Development of pre‐eclampsia before 34 weeks

• Development of pre‐eclampsia before 37 weeks

For children

• Preterm delivery before 37 weeks

• Neonatal death or severe morbidity (necrotising enterocolitis, IVH grade 3 or 4, RDS 3 or 4, neonatal sepsis, or as defined by the trial authors)

• Stillbirth: defined as death of a foetus before or during birth after 22 weeks of gestation, or as defined by the trial authors

• Neonatal death: defined as death of a baby within the first 28 days of life, or as defined by the trial authors

• Early neonatal death: defined as death of a baby within the first 7 days of life, or as defined by the trial authors

• Infant loss up to 6 weeks after the due date

• Early preterm delivery before 34 weeks

• Extremely preterm delivery before 28 weeks

• Necrotising enterocolitis

• IVH grade 3 or 4

• RDS 3 or 4

• Neonatal sepsis

• Neonatal encephalopathy

• Birthweight (g)

• Small‐for‐gestational‐age

• Admission to NICU or HCU

• Childhood hypertension

• Neonatal seizures

• Dental caries in childhood (one or more decayed, missing or filled teeth, or as defined by trial authors)

• Apgar score less than 7, five minutes after birth

• Adverse effects on the neonate as a result of maternal treatment interventions

• Gestational age at birth in weeks

Search methods for identification of studies

The Information Specialist searched for all published, unpublished, and ongoing trials, without restrictions on language or publication status to 7 January 2025. They used a standardised randomised trial filter in MEDLINE [29]. The updated search adheres to current Cochrane methods and reporting standards. The search strategies for each database are available in Supplementary material 1.

Electronic searches

We searched for published, unpublished, and ongoing trials in the following databases.

• Cochrane Central Register of Controlled Trials (CENTRAL; 2024, Issue 12) via Ovid EBM Reviews

• MEDLINE ALL via PubMed (from 1946 to 7 January 2025)

• Embase (including Embase Classic) via Elsevier (from 1947 to 7 January 2025)

• CINAHL via EBSCO (from 1977 to 7 January 2025)

• Scopus via Elsevier (from 1823 to 7 January 2025)

• Web of Science (from 1900 to 7 January 2025), depending on the index (SCIE, SSCI, AHCI, ESCI)

• Portal Regional BVS (LILACS) (from 2003 to 7 January 2025)

We also searched the following trials registries.

• WHO International Clinical Trials Registry Platform (ICTRP): trialsearch.who.int/

• National Institutes of Health ClinicalTrials.gov: clinicaltrials.gov/

Searching other resources

We checked the bibliographies of the included trials and any relevant systematic reviews for further references to relevant trials. We contacted experts and organisations in the field to obtain additional information on relevant trials.

Data collection and analysis

Selection of studies

We downloaded all titles and abstracts retrieved by electronic searching to a reference management database and removed duplicates [30]. The review authors used the Covidence tool to remove any further reports that were not RCTs. One review author (MRT) additionally screened these titles and abstracts, and we are confident that no trials were missed. Two review authors (CAC, CR, ACR, MRT) screened titles and abstracts independently and in duplicate using Covidence software [30], excluding records that were clearly ineligible. We retrieved the full‐text articles of potentially eligible trials, and two review authors (CAC, CR, ACR, MRT) assessed them against our inclusion criteria independently and in duplicate. We resolved discrepancies through discussion or by consulting a third review author. We recorded reasons for the exclusion of the ineligible trials at this stage in Supplementary material 3. We collated multiple reports of the same trial so that each trial rather than each report was the unit of interest in the review. We also recorded any available information about ongoing trials and trials awaiting classification. The selection process was recorded in sufficient detail to complete a PRISMA flow diagram [31].

We applied the Checklist to assess Trustworthiness in Randomised Controlled Trials (TRACT) [32]. This checklist contains 19 items categorised into seven domains: governance, author group, plausibility of intervention usage, timeframe, dropout rates, baseline characteristics, and outcomes.

• Domain 1. Governance: included checking for absent or retrospective registration of RCTs commencing after 2010, as well as absent or vague descriptions of research ethics or other apparent concerns regarding ethics.

• Domain 2. Author group: included checking whether authors' other trials have been retracted not at their request, whether there are a large number of RCTs published in a short time frame by one author or by one institute, and whether the number of authors is three or less or there is a low author‐to‐trial‐size ratio.

• Domain 3. Plausibility of intervention usage: included checking whether there were insufficient or implausible descriptions, or unnecessary or illogical descriptions of methodological standards.

• Domain 4. Time frame: included checking for fast recruitment of participants within the trial time, or short or impossible time frames between ending recruitment and follow‐up and submission of the paper. The recruitment time frame is from the date of the first recruited participant to the date of the last recruited participant.

• Domain 5. Dropout rates: included checking for zero participants lost to follow‐up, or no reasons given for loss to follow‐up, especially in trials with long follow‐up, as well as the ideal number of losses to follow‐up, resulting in a perfectly rounded number in each group.

• Domain 6. Baseline characteristics: included checking for baseline characteristics and whether no or fewer than five baseline characteristics were presented, or implausible participant characteristics judging from common sense, the literature and local data. We checked for perfect balance for multiple baseline characteristics or significant or large differences between baseline characteristics, as well as whether important prognostic factors were not reported as baseline characteristics.

• Domain 7. Outcomes: included checking for effect sizes that were much larger than in other RCTs on the same topic, or conflicting information between outcomes, as well as a change in primary outcome from registration to publication.

Two review authors (CC and CR) applied the checklist to each eligible RCT independently and in duplicate. Answers for each item were either no concerns, some concerns or no information, or major concerns. All review authors then reviewed these decisions. Where review authors deemed trials to have 'some concerns' or 'major concerns' in trustworthiness, reasonable efforts were made to contact trial authors via email to clarify these specific concerns. We planned to use only full texts to assess eligibility and where only an abstract or insufficient trial information was available, we would have contacted trial authors via email to request further information. These trials would have been placed in ‘awaiting classification’ until we received a response with sufficient information to determine inclusion or exclusion. We would have clearly documented this process and any concerns we had in the table of trials awaiting classification. Trials that were conducted after 2010 were excluded if they were not prospectively registered. Disagreements in trustworthiness assessments were discussed among the author team until consensus was reached.

One review author (JH) is the trial author for the included trial in this review. He did not participate in the selection of trials, data extraction, or risk of bias or trustworthiness assessments.

Data extraction and management

We created a data extraction form in Covidence [30], and after piloting, two review authors (CAC, MRT) extracted data from the included trial independently and in duplicate. We resolved disagreements through discussion between all the review authors. We extracted the following trial characteristics.

• Methods: trial design, number of trial centres and location, trial setting, withdrawals, date of trial, follow‐up, approach to adjustment for design effects or confounding

• Participants: number, mean age, age range, risk of pre‐eclampsia, dietary calcium intake, diagnostic criteria, inclusion criteria, exclusion criteria, other relevant characteristics

• Interventions: intervention components, dose and regimen, comparison, any co‐interventions

• Outcomes: events, means, relative effects, time points reported, adjusted effect estimates and information about the confounders and design effects accounted for, details on how the outcome was defined and measured, what unit of measurement was used and if a scale was used, the upper and lower limits and details on whether high or low scores are good

• Notes: funding for trial, notable conflicts of interest of trial authors and ethics.

Two review authors independently extracted outcome data from the included trials. We noted in Supplementary material 2 whether outcome data were reported in an unusable way. We contacted trial authors where data were missing or needed clarification. One review author entered data into Review Manager and a second review author checked for accuracy [33].

Risk of bias assessment in included studies

Our review aimed to quantify the effect of assignment to the intervention, and our risk of bias evaluations aligned with this aim. Two review authors (CAC, MRT) assessed the risk of bias for each outcome presented in the summary of findings table, independently and in duplicate, using RoB 2 [26]. We resolved any disagreements by discussion with all the review authors. Our effect of interest was the effect of assignment, also known as the intention‐to‐treat (ITT) effect. We assessed the risk of bias for outcomes that are included in the summary of findings table.

• For women

  • Pre‐eclampsia or pregnancy loss at any gestational age

  • Pre‐eclampsia

  • Pregnancy loss at any gestational age

  • Maternal death

  • Maternal death or severe morbidity

  • Adverse effects of interventions as defined by trial authors

• For children

  • Perinatal loss

  • Preterm delivery before 37 weeks

  • Neonatal death or severe morbidity

  • Stillbirth

  • Neonatal death

  • Early neonatal death

RoB 2 assigns risk of bias judgements based on the answers to signalling questions that fall under the following domains.

• Bias arising from the randomisation process

• Bias due to deviations from intended interventions

• Bias due to missing outcome data

• Bias in measurement of the outcome

• Bias in selection of the reported result

For each outcome, we used the signalling questions to reach one of three judgements: 'low risk of bias', 'some concerns', or 'high risk of bias'. We used the RoB Excel tool to record answers to the signalling questions [34], and our answers are available on reasonable request. We summarised the risk of bias judgements across different trials for each of the domains for each prespecified outcome.

The tool derives an overall judgement for each trial as follows.

• Low risk of bias: all domains have a 'low risk' judgement

• Some concerns: there are 'some concerns' in at least one domain, but no domains have a 'high risk' judgement

• High risk of bias: at least one domain has a 'high risk' judgement, or there are 'some concerns' in multiple domains

Measures of treatment effect

We reported the effects of interventions for dichotomous outcomes as risk ratios (RRs) with 95% confidence intervals (CIs). For continuous data, we reported mean differences (MDs) with 95% CIs when trials measured outcomes using the same scale. Where results were reported in a format that could not be entered into a meta‐analysis, we used guidance provided in Chapter 6 of the Cochrane Handbook for Systematic Reviews of Interventions to convert the data to the necessary format [35].

For future updates of this review, we will use standardised mean differences (SMDs) for continuous outcomes if trials measured the same outcome using different scales. Furthermore, where possible, we will use outcomes adjusted for confounders or design effects (e.g. clustering) in meta‐analyses. Where trials report count data (the number of events rather than the number of people who experienced an event), we will use the number of events and number of person‐years to calculate rate ratios (Cochrane Handbook section 6.7 [35]).

Unit of analysis issues

Multi‐arm trials

In future updates, if outcomes are reported in multi‐armed trials, we will combine all relevant intervention groups of the trial into a single group, and all relevant control intervention groups into a single control group. We will combine both the sample sizes and the numbers of people with events from all groups for dichotomous outcomes. For continuous outcomes, we will combine means and standard deviations (SDs) as per Chapter 10 of the Cochrane Handbook [36]. If we consider one of the arms irrelevant, we will exclude it from the analysis.

Cluster‐randomised trials

We have not identified any cluster‐randomised trials to date. For future update(s) of this review, we will include any eligible cluster‐randomised trials in the analyses along with individually randomised trials. We will adjust their sample sizes using the methods described in the Cochrane Handbook using an estimate of the intracluster correlation co‐efficient (ICC) derived from the trial (if possible), from a similar trial or from a trial of a similar population [36]. If we use ICCs from other sources, we will report this and conduct sensitivity analyses to investigate the effect of variation in the ICC. If we identify cluster‐randomised trials in addition to the individually randomised trials, we plan to synthesise the relevant information. We will consider it reasonable to combine the results from both if there is little heterogeneity between the trial designs and the interaction between the effect of intervention and the choice of randomisation unit is considered to be unlikely.

Cross‐over trials

Cross‐over trials were not eligible for inclusion in this review.

Dealing with missing data

We would have contacted investigators or trial sponsors to verify key trial characteristics and obtain missing numerical outcome data for trials identified as abstract only. We reported all relevant data in Supplementary material 2. If we had made any assumptions about missing data, we would have reported the potential impact in the 'Discussion' section of the review.

For all outcomes, we carried out analyses, to the greatest degree possible, on an ITT basis, that is, we attempted to include all participants randomised to each group in the analyses, and to analyse all participants in the group to which they had been allocated, regardless of whether they received the allocated intervention. If data had been missing because the outcome was not measured in all participants, the reason for missing data was unrelated to the outcome, and missing data were balanced between groups, the denominator for each outcome in each trial would have been the number randomised minus any participants whose outcomes were known to be missing (modified ITT analysis). If no explanation had been given for missing outcome data, missing data were not balanced between groups, and the missing data were suspected to be related to the outcome, we would have contacted the trial authors and conducted sensitivity analyses.

Reporting bias assessment

We did not have sufficient trials in the meta‐analysis to create funnel plots, but we cannot rule out the possibility of publication bias or small‐trial effects.

In future updates of the review, if we include at least 10 trials in a meta‐analysis of a primary outcome, we will create and visually inspect funnel plots and run a formal statistical test for asymmetry to evaluate publication bias. If we observe funnel plot asymmetry, we will discuss the implications of reporting and non‐reporting biases (e.g. small‐trial effects) for our results.

Synthesis methods

As this review only has one included trial, meta‐analysis was not possible. For future updates of the review, we will undertake meta‐analysis only where this will be meaningful based on the comparability of the trials available (i.e. if the treatments, participants and the underlying clinical question are similar enough for pooling to make sense). In future reviews, we will base our choice of meta‐analysis approach on an evaluation of clinical, methodological and statistical heterogeneity. We will generate pooled analyses of outcomes with sufficient data using the DerSimonian and Laird random‐effects technique [36]. The 95% CI will be calculated using the Wald‐type method [36]. For random‐effects analyses, we presented the results as the average treatment effect with 95% CIs, and the estimates of Tau² and I² statistics. We presented the effect estimates and 95% CIs for the individual trial effects and the pooled effect in forest plots.

For future updates of this review, we will pool data using the inverse variance method where trials report adjusted estimates that are appropriate to include in a meta‐analysis. Where meta‐analysis is not possible, we will present results in a narrative synthesis and create appropriate tables and figures, following the Synthesis Without Meta‐analysis (SWiM) reporting guideline [37].

Investigation of heterogeneity and subgroup analysis

As there was only one included trial, there was no heterogeneity. In future updates, we will assess statistical heterogeneity by visually inspecting forest plots and describing the direction and magnitude of effects and the degree of 95% CI overlap. We would also consider the statistics generated in forest plots that measure statistical heterogeneity and use the I² statistic to quantify inconsistency amongst the trials in each analysis [38]. We would also consider the P value from the Chi² test to assess whether this heterogeneity is significant (P < 0.1). If we identified substantial heterogeneity, we would report the finding and explore possible explanatory factors using prespecified subgroup analysis. To interpret the I² value, we would use the following rough thresholds [36].

• 0% to 40%: might not be important

• 30% to 60%: may represent moderate heterogeneity

• 50% to 90%: may represent substantial heterogeneity

• 75% to 100%: considerable heterogeneity

We originally planned to conduct subgroup analyses according to high or low risk of pre‐eclampsia, high or low dose of daily calcium supplementation, high‐ or low‐intake populations and also by duration of calcium supplementation. However, these analyses were not possible for this review version as there is only one included trial.

We also planned to pursue subgroup analysis according to women becoming pregnant compared to women who did not become pregnant for outcomes that were not dependent on pregnancy, such as adverse events, but did not have any data for these outcomes.

Equity‐related assessment

We did not investigate health inequity in this review.

Sensitivity analysis

We planned to perform sensitivity analysis to exclude small trials where the data showed the possibility of a small‐trial effect, but as there was only one included trial, we did not. As we could not find a definition of 'small trial' in the literature, we contacted biostatisticians and discussed this within our author team. We defined a 'small trial' as one with fewer than 500 participants. We would have shown the results of the sensitivity analyses in separate forest plots and assessed their certainty of evidence.

We also planned to perform sensitivity analyses for aspects of the review that might have affected the results, for example, where there was a risk of bias associated with the quality of some of the included trials, where there were high levels of missing data, and the difference between fixed‐ and random‐effects meta‐analysis, but this was not applicable as there is only one included trial.

Because conception is an intermediate outcome potentially on the causal pathway to pre‐eclampsia [28], we performed sensitivity analysis on important outcomes that are dependent on pregnancy to show the results when considering only those participants who became pregnant.

Certainty of the evidence assessment

Two review authors (CR, ACR) independently evaluated the certainty of the evidence according to the five GRADE domains (trial limitations, consistency of effect, imprecision, indirectness and publication bias) using GRADEpro GDT software [39, 40], and following the methods and recommendations provided in Chapter 14 of the Cochrane Handbook [41]. When judging imprecision, we considered the range of the confidence interval of the relative effect as well as the total cumulative trial population for that outcome and the total number of events per outcome. Where there were very few events (fewer than 30) and the confidence interval was very wide, we downgraded for imprecision. The threshold for suggested appreciable benefit for the relative effect was 0.75 and the threshold for suggested appreciable harm was 1.25. Where the confidence interval crossed these thresholds, we downgraded the certainty of the evidence by one, two, or three levels depending on the width of the confidence interval.

We resolved any disagreements by discussion with all authors.

We created a summary of findings table for the comparison of calcium compared with placebo for the population of all randomised women.

In the summary of findings table, we provided results for the following key outcomes for women and children and justified our decisions to downgrade the certainty of the evidence in footnotes.

For women

• Pre‐eclampsia or pregnancy loss at any gestational age

• Pre‐eclampsia

• Pregnancy loss at any gestational age

• Maternal death

• Maternal death or severe morbidity

• Adverse effects

For children

• Perinatal loss

• Preterm delivery before 37 weeks

• Neonatal death or severe morbidity

• Stillbirth

• Neonatal death

• Early neonatal death

Consumer involvement

Consumers were not involved in this review due to limited resources, although the review authors did use core outcome sets for the outcomes in this protocol, which were developed with consumer involvement.

Results

Description of studies

Results of the search

From the 14,903 records identified through database and register searches, we excluded 4154 duplicates. After screening by one review author (MRT), we also excluded 3325 records that were identified as clearly irrelevant (i.e. not trials) by the Covidence automation tool.

Two review authors screened the titles and abstracts of the remaining 7424 records, and selected five for full‐text review. Of these, we included one RCT (4 references; see Supplementary material 2), and excluded one trial (1 reference) because it had not been prospectively registered (see Supplementary material 3).

We planned to apply an assessment of trustworthiness to each potentially relevant trial, and exclude trials without prospective registration if they were conducted after 2010 or, in cases where information about registration or other information was required, we would contact the trial authors via email and categorise these trials as awaiting classification until an adequate response was received, and an eligibility decision could be made. However, we did not identify any such trials.

We did not identify any ongoing trials.

This updated review thus comprises one included trial and one excluded trial. The PRISMA study flow diagram in Figure shows the results of the search process.

1.

PRISMA flow diagram

Included studies

The characteristics of the included trial are summarised in Supplementary material 2.

Study design

Hofmeyr 2019 [42, 43, 44, 45] was a multicountry, double‐blind, randomised, placebo‐controlled trial.

Settings

Hofmeyr 2019 was conducted in nine hospitals in South Africa (4 hospitals), Zimbabwe (2 hospitals), and Argentina (3 hospitals).

In Zimbabwe, the two hospitals were state secondary or tertiary referral hospitals with large obstetric services (4000‐20,000 births/year) providing comprehensive obstetric care, serving urban and rural lower‐income populations. In Argentina, of the three maternity hospitals, one was a public referral hospital with approximately 9000 deliveries a year and the other two were private, third‐level maternity hospitals, each with approximately 2000 deliveries a year. No information was available about the characteristics of the South African hospital.

Participants

Hofmeyr 2019 enrolled parous women whose most recent pregnancy had been complicated by pre‐eclampsia or eclampsia and who were intending to become pregnant. The trial excluded women if they were younger than 18 years old; were already pregnant; were taking calcium supplementation; had chronic hypertension with persistent proteinuria; had a history or symptoms of urolithiasis, renal disease, or parathyroid disease; were not in a sexual relationship; were using long‐term contraception (e.g. hormonal injections or implant, intrauterine contraceptive device, or sterilisation); or were unwilling to give informed consent. At trial entry, a total of 1355 women were randomised.

Of the 678 women randomised to the calcium group and 677 women randomised to the placebo group, 347 in the calcium group, and 357 in the placebo group did not conceive, were lost to follow‐up before conception or withdrew before conception.

Of the 331 women in the calcium group and 320 women in the placebo group who conceived during the trial, 33 versus 37 were excluded because they were lost to follow‐up before 20 weeks' gestation, had pregnancy loss before 20 weeks, requested an abortion before 20 weeks, or discontinued the trial medication while pregnant. Of the 298 women in the calcium group versus 283 women in the placebo group with follow‐up data after 20 weeks, two versus none were lost to follow‐up, leaving 296 women in the calcium group versus 283 in the placebo group to be analysed.

Daily calcium intake

Hofmeyr 2019 included populations with a low baseline calcium intake. The baseline dietary calcium intake (mg/day) overall in the different countries was as follows: Argentina 481 mg (median), South Africa 441 mg (SD 87.7) and Zimbabwe 360.5 mg (SD 171.4).

Risk profile

Hofmeyr 2019 included women at a high risk for pre‐eclampsia. Women had had a previous pregnancy complicated with pre‐eclampsia or eclampsia.

Interventions and comparisons

In Hofmeyr 2019, women received 500 mg of calcium daily or placebo from enrolment until 20 weeks' gestation. After 20 weeks, all women received 1.5 g of calcium daily until delivery, regardless of their allocation to the intervention or the placebo group.

Route and dose of calcium

Hofmeyr 2019 administered calcium and placebo through chewable tablets. Participants received one 500 mg tablet daily until 20 weeks' gestation and thereafter received three 1500 mg tablets daily.

Timing of calcium

Hofmeyr 2019 started women on the trial regimen before they became pregnant and continued with their allocated trial regimen until 20 weeks' gestation. Thereafter, all women received 1.5 g of calcium daily until delivery, regardless of their allocation to the intervention or the placebo.

Adherence

Hofmeyr 2019 presented compliance data only for women who had a pregnancy of 20 weeks’ gestation or longer. Of these women, data were only available for 213/298 in the calcium group when compliance was measured from randomisation up to the last visit before pregnancy and 274/298 when measured from the last visit before pregnancy to the 20‐week visit. Data were only available for 208/283 in the placebo group when compliance was measured from randomisation up to the last visit before pregnancy and 269/283 when measured from the last visit before pregnancy to the 20‐week visit.

• In the calcium group, the data presented showed that:

  • 47% of women had a compliance of more than 80% from randomisation up to the last visit before pregnancy. Of these, 53% of women had a compliance of more than 80% from the last visit before pregnancy to the 20‐week visit.

  • 81% of women had a compliance of more than 50% from randomisation up to the last visit before pregnancy. Of these, 77% of women had a compliance of more than 50% from the last visit before pregnancy to the 20‐week visit.

• In the placebo group, the data presented showed that:

  • 48% of women had a compliance of more than 80% from randomisation up to the last visit before pregnancy. Of these, 55% of women had a compliance of more than 80% from the last visit before pregnancy to the 20‐week visit.

  • 90% of women had a compliance of more than 50% from randomisation up to the last visit before pregnancy. Of these, 81% of women had a compliance of more than 50% from the last visit before pregnancy to the 20‐week visit.

Outcomes

Hofmeyr 2019 reported most of our prespecified review outcomes, except perinatal loss, maternal adverse effects, neonatal death or severe morbidity, neonatal death, and early neonatal death (see Outcome measures).

Sources of trial funding

Hofmeyr 2019 reported funding from the University of British Columbia, a grantee of the Bill & Melinda Gates Foundation; UNDP/UNFPA/UNICEF/WHO/World Bank Special Programme of Research, Development and Research Training in Human Reproduction, World Health Organization; the Argentina Fund for Horizontal Cooperation of the Argentinean Ministry of Foreign Affairs; and the Centre for Intervention Science in Maternal and Child Health (CISMAC) and a research grant from The International Development Research Centre, Canada.

Declarations of interest

Hofmeyr 2019 reported: "We declare no competing interests. Midway through the study, the study team was approached by Alternative Discovery & Development, GlaxoSmithKline (GSK) Medicines Research Centre, UK, who partnered with us to collect blood samples from a subgroup of participants in our trial for an independent, open, innovation pre‐eclampsia biomarker study, following a separate protocol, which was approved by the trial ethics committee. Apart from direct funding to the largest site (Chris Hani Baragwanath Hospital), specifically for the costs of this blood sample collection, GSK provided no funding to the main trial, and did not participate in any aspect of the main trial."

Contact with study authors

We contacted the authors of Hofmeyr 2019 for additional information and clarification, which they were able to provide.

Excluded studies

We excluded one trial identified from the search as it was not prospectively registered (Singh 2023 [46]). We attempted to contact the trial authors by email requesting information about the registration but did not receive an answer.

Risk of bias in included studies

One of the review authors (JH) was an author of Hofmeyr 2019 and was not involved in assessments of risk of bias or data collection, which were independently conducted by two other review authors. We judged risk of bias for each critical and some important outcomes. See the risk of bias summary (Figure) and Supplementary material 4 for details.

2.

Risk of bias summary: review authors' judgements about each risk of bias item for the included study

Random sequence generation and allocation concealment

We judged Hofmeyr 2019 to be at low risk of bias for both the generation of the randomisation sequence and allocation concealment in all outcomes.

Deviations from the intended interventions

We judged Hofmeyr 2019 to be at low risk of bias in this domain in all outcomes.

Incomplete outcome data

We judged Hofmeyr 2019 to have 'some concerns' in this domain in all outcomes, as 29% of the women randomised to the calcium group and 28% of women in the placebo group were lost to follow‐up or withdrew from the trial before or during pregnancy.

Selective reporting

We judged Hofmeyr 2019 to be at low risk of bias in this domain in all outcomes.

Other bias

We had no concerns regarding any other aspects of the trial.

Overall bias

Our overall risk of bias judgement of Hofmeyr 2019 was 'some concerns' due to the high loss to follow‐up.

Synthesis of results

We included one trial (Hofmeyr 2019), which addressed one comparison (calcium supplementation compared with placebo) and, therefore, we did not conduct a meta‐analysis. All analyses are available in Supplementary material 5. We performed sensitivity analyses on the important outcomes. A summary of these analyses can be found in Table and further details in Supplementary material 7.

1. Sensitivity analyses.

Outcome	Pooled effect with all randomised women	Pooled effect with only pregnant women
Pre‐eclampsia or pregnancy loss	RR 0.85 (95% CI 0.68 to 1.07) P = 1.07	RR 0.82 (95% CI 0.66 to 1.00) P = 0.05
Pre‐eclampsia	RR 0.84 (95% CI 0.62 to 1.14) P = 0.26	RR 0.81 (95% CI 0.61 to 1.07) P = 0.13
Pregnancy loss	RR 0.92 (95% CI 0.69 to 1.24) P = 0.60	RR 0.89 (95% CI 0.67 to 1.17) P = 0.40
Maternal death	RR 1.00 (95% CI 0.14 to 7.07) P = 1.00	RR 0.96 (95% CI 0.14 to 6.74) P = 0.96
Maternal death or severe morbidity	RR 0.97 (95% CI 0.70 to 1.35) P = 0.85	RR 0.93 (95% CI 0.68 to 1.27) P = 0.65
Adverse events	No trial measured this outcome.
Perinatal loss	No trial measured this outcome.
Preterm delivery before 37 weeks	RR 0.94 (95% CI 0.74 to 1.19) P = 0.60	RR 0.90 (95% CI 0.74 to 1.11) P = 0.33
Neonatal death or severe morbidity	No trial measured this outcome.
Stillbirth	RR 0.82 (95% CI 0.50 to 1.34) P = 0.43	RR 0.79 (95% CI 0.48 to 1.27) P = 0.33
Neonatal death	No trial measured this outcome.
CI: confidence interval; RR: risk ratio

Critical outcomes for women

Pre‐eclampsia or pregnancy loss at any gestational age

Calcium may result in little to no difference in pre‐eclampsia or pregnancy loss at any gestational age compared to placebo (RR 0.85, 95% CI 0.68 to 1.07; P = 0.17; 1 RCT, 1355 women; low‐certainty evidence) resulting in 28 fewer per 1000 women with pre‐eclampsia or pregnancy loss (from 61 fewer to 13 more) when all women randomised are considered, regardless of whether they had a pregnancy or not (Analysis 1.1 in Supplementary material 5). A forest plot is shown in Figure.

3.

Forest plot: Pre‐eclampsia or pregnancy loss at any time after conception

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.82 (95% CI 0.66 to 1.00; P = 0.05; 1 RCT, 633 women).

Pre‐eclampsia

Calcium may result in little to no difference in pre‐eclampsia compared to placebo (RR 0.84, 95% CI 0.62 to 1.14; P = 0.26; 1 RCT, 1355 women; low‐certainty evidence), resulting in 19 fewer per 1000 women with pre‐eclampsia (from 46 fewer to 17 more) when all women randomised are considered, regardless of whether they had a pregnancy or not (Analysis 1.2 in Supplementary material 5). A forest plot is shown in Figure.

4.

Forest plot: Pre‐eclampsia)

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.81 (95% CI 0.61 to 1.07 ; P = 0.13; 1 RCT, 633 women).

Important outcomes for women

Pregnancy loss at any gestational age

Calcium may result in little to no difference in pregnancy loss at any gestational age (RR 0.92, 95% CI 0.69 to 1.24; P = 0.60; 1 RCT, 1355 women; low‐certainty evidence), resulting in 9 fewer per 1000 women with pregnancy loss at any gestational age (from 37 fewer to 28 more) when all women randomised are considered, regardless of whether they had a pregnancy or not (Analysis 1.3 in Supplementary material 5). A forest plot is shown in Figure.

5.

Forest plot: Pregnancy loss at any time after conception

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.89 (95% CI 0.67 to 1.17 ; P = 0.40; 1 RCT, 633 women).

Maternal death

The evidence is very uncertain about the effect of calcium on maternal death compared to placebo (RR 1.00, 95% CI 0.14 to 7.07; P = 1.00; 1 RCT, 1355 women; very low‐certainty evidence), resulting in 0 fewer per 10,000 women dying (from 25 fewer to 179 more) when all women randomised are considered, regardless of whether they had a pregnancy or not (Analysis 1.4 in Supplementary material 5). A forest plot is shown in Figure.

6.

Forest plot: Maternal death

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.96 (95% CI 0.14 to 6.74 ; P = 0.96; 1 RCT, 633 women).

Maternal death or severe morbidity

Severe morbidity includes eclampsia, stroke, pulmonary oedema, heart failure, renal failure, liver capsule haematoma or rupture, placental abruption, HELLP syndrome or may be defined by the trial authors. The evidence is very uncertain about the effect of calcium on severe maternal morbidity and mortality compared to placebo (RR 0.97, 95% CI 0.70 to 1.35; P = 0.85; 1 RCT, 1355 women; very low‐certainty evidence), resulting in 3 fewer per 1000 women with severe morbidity (from 29 fewer to 34 more) when all women randomised are considered, regardless of whether they had a pregnancy or not (Analysis 1.5 in Supplementary material 5). A forest plot is shown in Figure.

7.

Forest plot: Maternal death or severe morbidity

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.93 (95% CI 0.68 to 1.27; P = 0.65; 1 RCT, 633 women).

Adverse effects

The included trial did not measure this outcome.

Eclampsia

Events were rare, with four events reported in the calcium group and five in the placebo group (RR 0.80, 95% CI 0.22 to 2.96 ; P = 0.74; 1 RCT, 1355 women; Analysis 1.7 in Supplementary material 5).

HELLP syndrome

Events were rare, with 10 events reported in the calcium group and seven in the placebo group (RR 1.43, 95% CI 0.55 to 3.73; P = 0.47; 1 RCT, 1355 women; Analysis 1.8 in Supplementary material 5).

Intensive care unit admission

Events were rare, with two events reported in the calcium group and three in the placebo group (RR 0.67, 95% CI 0.11 to 3.97; P = 0.66; 1 RCT, 1355 women; Analysis 1.9 in Supplementary material 5).

Stroke

No events were reported in either group.

Pulmonary oedema

Events were rare, with no events reported in the calcium group and one in the placebo group (RR 0.33, 95% CI 0.01 to 8.16; P = 0.50; 1 RCT, 1355 women; Analysis 1.11 in Supplementary material 5).

Acute kidney injury

Events were rare, with seven events reported in the calcium group and five in the placebo group (RR 1.40, 95% CI 0.45 to 4.38 ; P = 0.57; 1 RCT, 1355 women; Analysis 1.12 in Supplementary material 5).

Placental abruption

Events were rare, with nine events reported in the calcium group and five in the placebo group (RR 1.80, 95% CI 0.61 to 5.34 ; P = 0.29; 1 RCT, 1355 women; Analysis 1.13 in Supplementary material 5).

Hypertension

One trial contributed to this analysis with a risk ratio of 0.98 (95% CI 0.83 to 1.16; P = 0.84; 1 RCT, 1355 women; Analysis 1.14 in Supplementary material 5).

Gestational hypertension

One trial contributed to this analysis with a risk ratio of 1.09 (95% CI 0.86 to 1.37; P = 0.48; 1 RCT, 1355 women; Analysis 1.15 in Supplementary material 5).

Severe hypertension

One trial contributed to this analysis, defined as systolic blood pressure of 160/110 mmHg or as defined by the trial authors, with a risk ratio of 0.94 (95% CI 0.77 to 1.15; P = 0.55; 1 RCT, 1355 women; Analysis 1.16 in Supplementary material 5).

Proteinuria

One trial contributed to this analysis with a risk ratio of 0.87 (95% CI 0.65 to 1.15 ; P = 0.32; 1 RCT, 1355 women; Analysis 1.17 in Supplementary material 5).

Raised liver enzymes

One trial contributed to this analysis with a risk ratio of 1.14 (95% CI 0.42 to 3.13 ; P = 0.80; 1 RCT, 1355 women; Analysis 1.18 in Supplementary material 5).

Low platelets

One trial contributed to this analysis with a risk ratio of 1.08 (95% CI 0.50 to 2.35 ; P = 0.84; 1 RCT, 1355 women; Analysis 1.19 in Supplementary material 5).

Remaining outcomes for women

The included trial did not measure the following important outcomes for women.

• Liver capsule haematoma or rupture

• Intubation and mechanical ventilation (not for childbirth)

• Maternal satisfaction as defined by trial authors

• Maternal well‐being as defined by trial authors

• Postpartum haemorrhage

• Cortical blindness

• Retinal detachment

• Recurrent eclampsia

• Miscarriage

• Gestational age at diagnosis of pre‐eclampsia diagnosis in weeks

• Development of pre‐eclampsia before 34 weeks

• Development of pre‐eclampsia before 37 weeks

Critical outcomes for children

Perinatal loss

The included trial did not measure this outcome.

Important outcomes for children

Preterm delivery before 37 weeks

Calcium may result in little to no difference in preterm delivery before 37 weeks (RR 0.94, 95% CI 0.74 to 1.19 ; P = 0.60; 1 RCT, 1355 women; low‐certainty evidence), resulting in 11 fewer per 1000 women delivering before 37 weeks (from 46 fewer to 33 more) when all women randomised are considered, regardless of whether they had a pregnancy or not. (Analysis 1.21 in Supplementary material 5). A forest plot is shown in Figure.

8.

Forest plot: Preterm delivery before 37 weeks

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.90 (95% CI 0.74 to 1.11; P = 0.33; 1 RCT, 633 women).

Stillbirth

The evidence is very uncertain about the effect of calcium on stillbirth compared to placebo (RR 0.82, 95% CI 0.50 to 1.34 ; P = 0.43; 1 RCT, 1355 women; very low‐certainty evidence), resulting in 9 fewer per 1000 women having stillborn babies (from 24 fewer to 17 more) when all women randomised are considered, regardless of whether they had a pregnancy or not (Analysis 1.23 in Supplementary material 5). A forest plot is shown in Figure.

9.

Forest plot: Stillbirth

Sensitivity analysis including only women who conceived during the trial yielded a risk ratio of 0.79 (95% CI 0.48 to 1.27 ; P = 0.33; 1 RCT, 633 women).

Early preterm delivery before 34 weeks

One trial contributed to this analysis with a risk ratio of 0.82 (95% CI 0.57 to 1.19; P = 0.30; 1 RCT, 1355 women; Analysis 1.26 in Supplementary material 5).

Apgar score less than 7 at five minutes after birth

Events were rare, with five reported in the calcium group and 11 in the placebo group. One trial contributed to this analysis with a risk ratio of 0.43 (95% CI 0.15 to 1.21; P = 0.11; 1 RCT, 1355 women; Analysis 1.27 in Supplementary material 5).

Remaining important outcomes for children

The included trial did not measure the following important outcomes for children.

• Neonatal death or severe morbidity

• Neonatal death

• Early neonatal death

• Infant loss up to 6 weeks after the due date

• Extremely preterm delivery before 28 weeks

• Necrotising enterocolitis

• Intraventricular haemorrhage (IVH) grade 3 or 4

• Respiratory distress syndrome (RDS) 3 or 4

• Neonatal sepsis

• Neonatal encephalopathy

• Birthweight (g)

• Small‐for‐gestational‐age

• Admission to neonatal intensive care unit (NICU) or high care unit (HCU)

• Childhood hypertension

• Neonatal seizures

• Dental caries in childhood

• Adverse effects on the neonate as a result of maternal treatment interventions

• Gestational age at birth in weeks

Discussion

Summary of main results

The findings of this Cochrane review are based on one trial.

When all randomised women are considered, regardless of whether they conceived during the course of the trial or not, the results show that calcium may result in little to no difference in the composite outcome of pre‐eclampsia or pregnancy loss at any gestational age, pre‐eclampsia, pregnancy loss at any gestational age, and preterm delivery before 37 weeks. The evidence is very uncertain about the effect of calcium on maternal death, the composite outcome of maternal death or severe morbidity, and stillbirth. No trials measured adverse effects, perinatal loss, neonatal death, early neonatal death or the composite of outcome of neonatal death or severe morbidity.

When only randomised women who became pregnant during the course of the trial are considered in the analysis, the results show that calcium may result in a slight reduction in the composite outcome of pre‐eclampsia or pregnancy loss at any gestational age. Calcium may result in little to no difference in pre‐eclampsia, pregnancy loss at any gestational age, the composite outcome of maternal death or severe morbidity, and in preterm delivery before 37 weeks. The evidence is very uncertain about the effect of calcium on maternal death, the composite outcome of maternal death or severe morbidity, and stillbirth.

Limitations of the evidence included in the review

The results of this review are based on one trial. We assessed Hofmeyr 2019 at low risk of bias arising from the randomisation process and low risk of bias due to deviations from the intended interventions. However, we judged it to have 'some concerns' in the risk of bias due to missing data, since nearly a quarter of the women were lost to follow‐up preconception (Figure).

We assessed the risk of bias in measurement of the outcome and risk of bias in selection of the reported result to be low. As a result, we judged the overall risk of bias to be at 'some concerns'.

Results should be interpreted with caution, as for some outcomes, assessments to measure these outcomes were not routine for all women and were only performed if there was clinical indication.

Women were randomised to either calcium or placebo in this trial, but it should be noted that all women received calcium supplementation after 20 weeks' gestation, in line with the guidance on calcium supplementation.

There was no difference in the result of the outcome of pre‐eclampsia whether all randomised women or only pregnant randomised women were considered. When all randomised women were considered for the outcome 'pre‐eclampsia or pregnancy loss', results show that calcium may result in little to no difference, but when only pregnant women were considered, calcium may result in a slight reduction in this outcome. It is important to note that interpretation of this sensitivity analysis is limited as a substantial proportion of the women who were initially enrolled did not become pregnant.

A limitation of this trial was that the main outcomes were conditional upon intermediate outcomes. For instance, for pre‐eclampsia, the reported results were conditional upon women reaching the intermediate outcomes of conception, and pregnancy reaching 20 weeks' gestation. This raises two concerns. First, if calcium supplementation had an effect on conception or early pregnancy loss, this may have introduced bias for the outcome 'pre‐eclampsia'. This suggests that the outcome least susceptible to bias is 'pre‐eclampsia or early pregnancy loss', since this composite avoids the possible confounding effect of early calcium supplementation on pregnancy loss. Second, it is also possible that there may have been differences between the groups in the women who didn't conceive or who had early pregnancy losses. These differences would then introduce bias into analyses including only women who conceived or reached 20 weeks' gestation.

We used the GRADE approach to assess the certainty of evidence. We downgraded all outcomes for risk of bias, due to the high loss to follow‐up in the trial. Certainty was most commonly downgraded due to imprecision, when there were very few events (fewer than 30) and if the confidence interval around the risk ratio was very wide.

The threshold for suggested appreciable benefit for the relative effect was 0.75 and the threshold for suggested appreciable harm was 1.25. Where the confidence interval crossed these thresholds, we downgraded by one, two, or three levels depending on the width of the confidence interval and included reasons for imprecision judgements in the footnotes of Table.

The evidence comes from only one trial of calcium supplementation (Hofmeyr 2019). Although this trial directly addressed the review question, and reached its own prespecified sample size, it is possible that the trial may not have included enough women to yield compelling evidence for the efficacy of the intervention.

A limitation of the included trial was suboptimal compliance. The possibility that this may have contributed to the under‐estimation of the effect of calcium supplementation is suggested by the report of reduced pre‐eclampsia in a compliant subgroup (RR 0.66, 95% CI 0.44 to 0.98; Hofmeyr 2019).

The single included trial recruited women who were at increased risk of pre‐eclampsia, on the basis of a history of pre‐eclampsia in the most recent, previous pregnancy. Therefore, this trial did not include primigravid women during their first pregnancy, a group that is also recognised to be at increased risk of pre‐eclampsia. It is possible that calcium may have different (i.e. either more or less beneficial) effects for women who are pregnant for the first time. If so, these could not be identified by this trial.

This review did not include evidence on the acceptability for women of taking calcium supplements before pregnancy. The single included trial included data on adherence to the intervention, and whilst compliance was suboptimal, it was similar between the intervention and control groups. Future updates of this review could fruitfully report on outcomes that explicitly address both adherence and women's views about the intervention.

Limitations of the review processes

One review author (JH) is an author of the included trial. He did not participate in decisions regarding the inclusion, data extraction or risk of bias assessment of this trial.

The review process limits the number of outcomes prespecified as critical and only these outcomes were assessed using the GRADE approach and presented in the Summary of findings table. We did not assess the certainty of evidence for individual outcomes representing rare but severe outcomes like eclampsia, HELLP syndrome, placental abruption or stroke, but combined them in a composite outcome of 'maternal death or severe maternal morbidity'. Severe maternal morbidity was defined as eclampsia, stroke, pulmonary oedema, heart failure, renal failure, liver capsule haematoma or rupture, placental abruption, HELLP syndrome or as defined by trial authors.

Agreements and disagreements with other studies or reviews

Review authors are not aware of any other trials that assess the effect of calcium supplementation limited to the pre‐and early pregnancy periods.

The results of this update are similar to those in the previous version of the review [24].

Authors' conclusions

Implications for practice

When all randomised women planning a pregnancy are considered, the results show that calcium may result in little to no difference in both pre‐eclampsia or pregnancy loss, and pre‐eclampsia. When only pregnant women are considered, calcium may result in little to no difference in pre‐eclampsia but may result in a slight reduction in pre‐eclampsia or pregnancy loss. The included trial did not measure perinatal loss.

The evidence is drawn from one trial of calcium supplementation that commenced before conception and continued into the first half of pregnancy. Current evidence neither supports nor refutes the routine use of calcium supplementation commencing before conception and continuing into early pregnancy.

Implications for research

Further research is needed to confirm whether calcium supplementation before or early in pregnancy is associated with a reduction in adverse pregnancy outcomes, such as pre‐eclampsia and pregnancy loss. Such trials should be adequately powered, limited to calcium supplementation, and be placebo‐controlled, and include the outcomes chosen for this review, with the addition of outcomes that assess the acceptability of the intervention to women.

Further research is needed, including research to determine the overall effect of calcium supplementation commenced before pregnancy and continued throughout pregnancy versus placebo, and to address the limitation of suboptimal compliance in the included trial.

Although we did not assess equity in this updated review, we recognise the difficulties in including populations from low‐ and middle‐income countries (LMICs) due to limited access to prenatal healthcare services and rates of unplanned pregnancies. We also recognise the barriers to investigating the effects of calcium supplementation as a prophylactic intervention in real‐world settings other than the use of calcium as a population‐wide staple food fortification.

Supporting Information

Supplementary materials are available with the online version of this article: 10.1002/14651858.CD011192.pub4.

Supplementary materials are published alongside the article and contain additional data and information that support or enhance the article. Supplementary materials may not be subject to the same editorial scrutiny as the content of the article and Cochrane has not copyedited, typeset or proofread these materials. The material in these sections has been supplied by the author(s) for publication under a Licence for Publication and the author(s) are solely responsible for the material. Cochrane accordingly gives no representations or warranties of any kind in relation to, and accepts no liability for any reliance on or use of, such material.

Supplementary material 1 Search strategies

Supplementary material 2 Characteristics of included studies

Supplementary material 3 Characteristics of excluded studies

Supplementary material 4 Risk of bias

Supplementary material 5 Analyses

Supplementary material 6 Data package

Supplementary material 7 Summary of findings ‐ sensitivity analysis

New search for studies and content updated (no change to conclusions)

Additional information

Contributions of authors

In this review, authors C Cluver, C Rohwer, A Rohwer, MR Torloni and GJ Hofmeyr prepared and reviewed the protocol. M Eduarda dos Santos Puga prepared and ran the search. CC, CR, AR and MRT performed study selection, data extraction, risk of bias and trustworthiness assessments. CC, MRT and GJH provided clinical perspective to the review. CR, AR and CC conducted the analyses, prepared the summary of findings table, and drafted the review. All authors contributed to the final draft of the review and approved it for submission.

Declarations of interest

GJ Hofmeyr: no relevant interests; involved in Hofmeyr 2019 and did not make study eligibility or trustworthiness decisions about, extract data from, carry out the risk of bias assessment for, or perform GRADE assessments of this study.

CA Cluver: no relevant interests; works as a maternal fetal medicine specialist.

CI Rohwer: none known

AC Rohwer: none known

MES Puga: none known

MR Torloni: none known

Sources of support

Internal sources

• Eastern Cape Department of Health, South Africa

  Salary support (GJH)

External sources

• No sources of support provided

Registration and protocol

The published protocol and updates to the review can be accessed:

Protocol (2025) DOI: Prospero CRD420250649571

Original Review (2014): DOI: 10.1002/14651858.CD011192

Review Update (2019): DOI: 10.1002/14651858.CD011192.pub3

Data, code and other materials

As part of the published Cochrane Review, the following is made available for download for users of the Cochrane Library: Full search strategies for each database; full citations of each unique report for all trials included, ongoing or awaiting classification, or excluded at the full text screen, in the final review; study data, including study information, study arms, and study results or test data and analysis data, including overall estimates and settings, subgroup estimates , and individual data rows. Appropriate permissions have been obtained for such use. Analyses and data management were conducted within Cochrane’s authoring tool, Review Manager, using the inbuilt computation methods. Template data extraction forms from Covidence and consensus risk of bias assessments are available from the authors on reasonable request.

All data available in Supplementary material 6.

What's new

Date	Event	Description
18 September 2025	New citation required but conclusions have not changed	We have not identified any new studies for this latest updated review and thus conclusions remain similar to those in the previous review version.
18 September 2025	New search has been performed	Searches updated to 7 January 2025. No new studies are added.

History

Protocol first published: Issue 8, 2014
Review first published: Issue 9, 2017

Date	Event	Description
13 July 2018	New citation required and conclusions have changed	New conclusions regarding the effect of pre‐pregnancy and early pregnancy calcium supplementation (new trial).
13 July 2018	New search has been performed	Review search updated. One new trial added (Hofmeyr 2019), which was ongoing in the previous version of the review. One previously included trial, Rumiris 2006, was excluded in this version of the review.