Prepregnancy and early pregnancy calcium supplementation among women at high risk of pre-eclampsia: a multicentre, double-blind, randomised, placebo-controlled trial

Background: Reducing deaths from hypertensive disorders of pregnancy is a global priority. Low dietary calcium might account for the high prevalence of pre-eclampsia and eclampsia in low-income countries. Calcium supplementation in the second half of pregnancy is known to reduce the serious consequences of pre-eclampsia; however, the effect of calcium supplementation during placentation is not known. We aimed to test the hypothesis that calcium supplementation before and in early pregnancy (up to 20 weeks’ gestation) prevents the development of pre-eclampsia Methods: We did a multicountry, parallel arm, double-blind, randomised, placebo-controlled trial in South Africa, Zimbabwe, and Argentina. Participants with previous pre-eclampsia and eclampsia received 500 mg calcium or placebo daily from enrolment prepregnancy until 20 weeks’ gestation. Participants were parous women whose most recent pregnancy had been complicated by pre-eclampsia or eclampsia and who were intending to become pregnant. All participants received unblinded calcium 1·5 g daily after 20 weeks’ gestation. The allocation sequence (1:1 ratio) used computer-generated random numbers in balanced blocks of variable size. The primary outcome was pre-eclampsia, defined as gestational hypertension and proteinuria. The trial is registered with the Pan-African Clinical Trials Registry, number PACTR201105000267371. The trial closed on Oct 31, 2017. Findings: Between July 12, 2011, and Sept 8, 2016, we randomly allocated 1355 women to receive calcium or placebo; 331 of 678 participants in the calcium group versus 320 of 677 in the placebo group became pregnant, and 298 of 678 versus 283 of 677 had pregnancies beyond 20 weeks’ gestation. Pre-eclampsia occurred in 69 (23%) of 296 participants in the calcium group versus 82 (29%) of 283 participants in the placebo group with pregnancies beyond 20 weeks’ gestation (risk ratio [RR] 0·80, 95% CI 0·61–1·06; p=0·121). For participants with compliance of more than 80% from the last visit before pregnancy to 20 weeks’ gestation, the pre-eclampsia risk was 30 (21%) of 144 versus 47 (32%) of 149 (RR 0·66, CI 0·44–0·98; p=0·037). There were no serious adverse effects of calcium reported. Interpretation: Calcium supplementation that commenced before pregnancy until 20 weeks’ gestation, compared with placebo, did not show a significant reduction in recurrent pre-eclampsia. As the trial was powered to detect a large effect size, we cannot rule out a small to moderate effect of this intervention. Funding: 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, WHO; the Argentina Fund for Horizontal Cooperation of the Argentinean Ministry of Foreign Affairs; and the Centre for Intervention Science in Maternal and Child Health. This was a multicentre, parallel arm, double blind, randomised, placebo-controlled trial, done in South Africa, Zimbabwe, and Argentina. In South Africa (Frere and Cecilia Makiwane Hospitals in East London; Chris Hani Baragwanath Hospital Johannesburg; Groote Schuur, Mowbray Maternity and Tygerberg Hospitals in Cape Town) and Zimbabwe (Harare Maternity and Mbuya Nehanda Maternity Hospitals, Harare) sites were state secondary or tertiary referral hospitals with large obstetric services (4000 to 20 000 births per year) providing comprehensive obstetric care, serving urban and rural lower-income populations. The sites in Argentina comprised three maternity hospitals, one in the province of Tucumán and two in Buenos Aires: the Institute of Maternity and Gynecology of Our Lady of Mercedes is located in the capital city of Tucumán province and it is the public referral maternity hospital of northwest Argentina with around 9000 deliveries each year; Hospital Italiano and CEMIC are private third-level maternity hospitals in Buenos Aires, each with around 2000 deliveries per year. The protocol of this study was approved by the Research Project Review Panel of the UNDP–UNFPA–UNICEF–WHO–World Bank Special Programme of Research, Development and Research Training in Human Reproduction at the Department of Reproductive Health and Research of WHO, and the WHO Research Ethics Review Committee, Geneva, Switzerland. Ethical approval was obtained from the appropriate national and institutional ethics review bodies applicable to each study site before the start of the trial. Data management procedures were compliant with good clinical practice. The protocol of this trial is published online in The Lancet (protocol 11PRT/4028).19 We used several approaches to identify potentially eligible women: direct searching of hospital records; prospective identification of women with, or after, pregnancies complicated by pre-eclampsia or eclampsia to recruit them at a future date; and raising community awareness through posters, community outreach visits, and radio interviews. Details of recruitment and retention strategies are reported elsewhere.20 Participants were parous women whose most recent pregnancy had been complicated by pre-eclampsia or eclampsia and who were intending to become pregnant. All participants provided informed written consent. Women were not eligible for the trial 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 (eg, hormonal injections or implant, intrauterine contraceptive device, or sterilisation); or were unwilling to give informed consent. The random allocation sequence was generated centrally at WHO headquarters in Geneva, Switzerland, using computer-generated random numbers in a ratio of 1:1 and in balanced blocks of variable size, stratified by site. Participants, care providers, and outcome assessors were all masked to group allocation. Calcium and placebo tablets were packed in identical 12-week treatment bottles, each containing 84 tablets. Participants were randomly assigned to a calcium supplementation or placebo group. The calcium group received one chewable tablet containing 500 mg elemental calcium (as calcium carbonate) daily from prepregnancy randomisation until 20 weeks’ gestation. The placebo tablet was indistinguishable from the calcium tablet in appearance and taste. At enrolment, allocation was done using an online service hosted by WHO, which allocated the next available treatment pack number from the site’s supply. At subsequent visits, the same online service was used to allocate an appropriate pack number, which ensured continuation of the allocated treatment without revealing group allocation. After 20 weeks’ gestation, all participants received open-label calcium tablets. Three bottles of 84 tablets each were supplied at a time, which was sufficient for a 12-week period (ie, from 20 weeks’ to 32 weeks’ gestation and from 32 weeks’ gestation until birth). The trial methods built on those used in the WHO trial of calcium supplementation during the second half of pregnancy that was done among pregnant women in similar settings,16 incorporating experiences and lessons learnt. Participants were asked to chew their allocated tablet in the evening at a time distant from taking food or iron supplements. They were encouraged not to take any additional calcium supplements. For those participants who needed analgesics, paracetamol was recommended, and for those needing antacids, a non-calcium-based antacid was recommended. The rationale for the calcium dosage selected was based on a dietary calcium survey done to inform the design of the WHO trial of calcium supplementation during the second half of pregnancy,16 which showed that pregnant primiparous women in Argentina had a median daily dietary calcium intake of 481 mg and in South Africa a median daily dietary calcium intake of 567 mg.20 Supplementation with 500 mg of calcium per day could thus achieve a median daily intake of about 1000 mg. This amount would approximate the daily calcium intake among pregnant women in high-income countries.21 Moreover, 500 mg is a level of supplementation that is achievable with food fortification.22 In addition to receiving the allocated calcium or placebo tablets until 20 weeks’ gestation, all participants also received calcium supplementation as recommended by WHO (1·5 g elemental calcium daily) from 20 weeks’ gestation until childbirth.23 Once recruited, participants were asked to return to the clinic-based research office every 12 weeks until pregnant, and then every 12 weeks until childbirth for follow-up visits. The visits during pregnancy were scheduled to take place as close as possible to 8 weeks’, 20 weeks’, and 32 weeks’ gestation, alongside the routine antenatal care visits. Contact was maintained by 4-weekly telephone calls. After participants had given birth, members of the research team extracted their outcome data from hospital records. At 6 weeks after childbirth, participants were telephoned to enquire about any complications since discharge from hospital. Participants who entered the trial and did not become pregnant were followed up for 1 month after cessation of calcium supplements. Data were recorded on case report forms specifically designed for the trial and entered in duplicate in a web-based data management system (OpenClinica). Blood pressure was recorded using mercury sphygmomanometers at the African sites, and with electronic blood pressure monitors in Argentina. All members of the research team were trained and assessed on their ability to measure blood pressure, with the British Hypertension Society training materials. Blood pressure measurements were standardised in the following way: participants were at rest, having been seated for 5 min or in bed with their abdomen tilted at least 30° to the left; the cuff was placed on the right arm at the level of the heart; two blood pressure measurements (systolic and diastolic) were taken at 3-min intervals; diastolic blood pressure was measured at the fifth Korotkoff sound (disappearance of the sounds); and the average of the two readings was recorded. This was a pragmatic trial and, therefore, pre-eclampsia was recorded as an outcome if it was diagnosed by a health-care provider and recorded in the participant’s medical records or if we identified both high blood pressure and proteinuria, as per the trial definitions of pre-eclampsia, in a participant’s medical records. Urine was collected routinely from all participants during trial antenatal visits, on admission in labour, or before elective caesarean section, to detect the presence of protein with a dipstick. This process was in accordance with routine clinical practice at the study sites. Results of relevant laboratory investigations during the provision of routine care, such as blood haemoglobin concentrations, were recorded in the case report forms, and for those who developed pre-eclampsia, these usually included a blood platelet count, serum urate, urea, creatinine, and liver function tests (ie, lactate dehydrogenase, alanine aminotransferase and aspartate aminotransferase, and a urine protein:creatinine ratio, urine protein, or 24 h urinary protein investigation, or both). Where several tests had been done, the last test before labour or caesarean section was used or, if none had been done previously, the first test after the start of labour or caesarean section was used. In a substudy done at all the African sites, data on dietary intake were obtained from participants at 20 weeks’ gestation, using a triple pass 24 h dietary recall questionnaire.24 A detailed description of the methodology and analysis of these findings has been published elsewhere.25 Diet assessment at 20 weeks’ gestation showed a mean daily calcium intake of 441 mg per day (SD 87·7; n=224) in South African women and 360·5 mg per day (SD 171·4; n=88) in Zimbabwean women, confirming the very low dietary calcium intake among trial participants. Every 12 weeks during trial follow-up visits both before and during pregnancy, all returned, unused tablets were counted, and the number assumed to have been taken was calculated and recorded. Treatment compliance was calculated by dividing the number of used tablets by the total number of tablets that should have been taken since the last count. A new bottle of 84 tablets was supplied to participants and returned tablets were stored until the end of the trial. The original protocol registered with the Pan-African Clinical Trials Registry identified three primary outcomes: pre-eclampsia, pre-eclampsia or pregnancy loss before labour at any gestation, and severe maternal morbidity and mortality index: defined as one or more of the specified secondary outcomes. The reason for the second primary outcome was to account for possible confounding of the pre-eclampsia outcome by pregnancy loss, which might be causally related. Before trial commencement, we revised the protocol to specify only the first primary outcome. Our primary outcome was therefore pre-eclampsia, defined as gestational hypertension and proteinuria (see the panel for definitions). A full list of secondary outcomes can be found in the panel. Primary outcome: Secondary outcomes: HELLP syndrome=haemolysis, elevated liver enzymes, and low platelet count. *Severe pre-eclamptic complications index: severe pre-eclampsia, early-onset pre-eclampsia (<32 weeks' gestation), eclampsia, HELLP syndrome, placental abruption, or severe gestational hypertension. †Severe maternal morbidity and mortality index: maternal admission to intensive care, eclampsia, severe pre-eclampsia, placental abruption, HELLP syndrome, renal failure, or death. The sample size calculation was informed by the WHO trial of calcium supplementation in the second half of pregnancy, where hypertension occurred in 14% of the relatively low-risk participants who received calcium supplementation from 20 weeks' gestation.16 The risk of recurrent pre-eclampsia for women with previous pre-eclampsia or eclampsia is estimated to be at least 25%.26 We calculated that, to show a reduction in pre-eclampsia from 25% to 15%, we would need 540 participants with pregnancies continuing beyond 20 weeks' gestation (0·05 α level, 80% power; calculated using Epi Info software (CDC, Atlanta, GA, USA). We anticipated that 50% of participants recruited would become pregnant during the trial. Allowing for a miscarriage rate of 15% and loss to follow-up of 10%, we needed to recruit approximately 1440 participants who were not pregnant. Enrolment would be stopped once we were able to predict reaching our primary sample size (ie, the 540 participants with pregnancies continuing beyond 20 weeks' gestation). Analyses of the effects of the intervention were intention to treat (ITT), irrespective of supplementation compliance. We prepared a table of baseline characteristics of the study groups for all participants enrolled and also for those in the final study sample (participants ≥20 weeks' gestation) to make sure that exclusion of those who did not have a pregnancy until 20 weeks did not compromise randomisation. Among those lost to follow-up, the number of participants lost, and their baseline characteristics, were compared to exclude any imbalances between study groups (data not shown). We also did a prespecified per-protocol analysis for participants for whom compliance with calcium supplementation was more than 80%. Other prespecified subgroup analyses are still to be done and will be reported in a separate paper. For categorical variables, we calculated the number of participants, number of missing values and percentages, and for continuous variables, we calculated the number of participants, number of missing values, minima, maxima, means, and SDs. Categorical variables were compared as risk ratios (RR) with 95% CIs. In this study, randomisation took place before conception, yet exposure to the primary outcome was limited to those who conceived and carried a pregnancy continuing beyond 20 weeks' gestation during the study period. Therefore, we anticipated potential confounders by first comparing the rates of pregnancy between study groups, for which we used a χ2 test. Had a difference in pregnancy rates been found, all subsequent analyses would have been done with the total enrolment numbers for each group as denominators, as well as with the numbers who became pregnant as denominators, and the effect of the differential rates of pregnancy on the results explored. A more likely confounder anticipated was early pregnancy loss, as the same pathology might cause both early pregnancy loss and pre-eclampsia. Any reduction in early pregnancy loss owing to calcium supplementation might increase the proportion of participants diagnosed with pre-eclampsia. For this reason we included the composite secondary outcome, pre-eclampsia or pregnancy loss, or both. SPSS (version 22) and R (version 3.3.0) were used for all statistical analyses. All tests were done using the usual two-sided, 5% significance levels. The study was overseen by the Data Safety and Monitoring Committee. The trial was registered with the Pan-African Clinical Trials Registry on Dec 6, 2010, registration number PACTR201105000267371. The funders of the trial had no role in the trial design, data collection, data analysis, data interpretation, or writing of the report. The primary author, GJH, has full access to all the data in the trial and had the final responsibility for the decision to submit for publication. Midway through the study, the study team was approached by Alternative Discovery & Development, GlaxoSmithKline Medicines Research Centre, UK, who partnered with us to collect blood samples from a sub-group 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.