BMJ Open Is Committed to Open Peer Review. As Part of This Commitment We Make the Peer Review History of Every Article We Publish Publicly Available

BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay- per-view fees (http://bmjopen.bmj.com).

If you have any questions on BMJ Open’s open peer review process please email [email protected]

http://bmjopen.bmj.com/ on September 25, 2021 by guest. Protected copyright. BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open

Cohort profile: Effect of malaria in early pregnancy on fetal growth in Benin (RECIPAL pre-conceptional cohort).

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2017-019014

Article Type: Cohort profile

Date Submitted by the Author: 07-Aug-2017

Complete List of Authors: Accrombessi, Manfred; French National Research Institute for Sustainable Development (IRD), Université Paris Descartes, UMR216-MERIT; Centre d'Etude et de Recherche sur le Paludisme Associé à la Grossesse et à l'Enfance (CERPAGE), Epidemiology Yovo, Emmanuel; Centre d'Etude et de Recherche sur le Paludisme Associé à la Grossesse et à l'Enfance (CERPAGE), Epidemiology Cottrell, Gilles; French National Research Institute for Sustainable Development (IRD), Université Paris Descartes, UMR216-MERIT Agbota, Gino; French National Research Institute for Sustainable Development (IRD), Université Paris Descartes, UMR216-MERIT; Centre d'Etude et de Recherche sur le Paludisme Associé à la Grossesse et à l'Enfance (CERPAGE), Epidemiology http://bmjopen.bmj.com/ Gartner, Agnès; French National Research Institute for Sustainable Development (IRD), Université de Montpellier, UMR204-Nutripass Martin-Prevel, Yves; French National Research Institute for Sustainable Development (IRD), Université de Montpellier, UMR204-Nutripass Fanou-Fogny, Nadia; Ecole de Nutrition et des Sciences et Technologies Alimentaires (ENSTA), Faculté des Sciences Agronomiques, Université d’Abomey-Calavi Djossinou, Diane; French National Research Institute for Sustainable Development (IRD), Université de Montpellier, UMR204-Nutripass; Ecole de Nutrition et des Sciences et Technologies Alimentaires (ENSTA), Faculté on September 25, 2021 by guest. Protected copyright. des Sciences Agronomiques, Université d’Abomey-Calavi Zeitlin, Jennifer; National Institute for Health and Medical Research (INSERM), Université Paris Descartes, Obstetrical, Perinatal, and Pediatric Epidemiology Research Team, Centre for Epidemiology and Biostatistics (U1153-EPOPé) Tuikue-Ndam, Nicaise; French National Research Institute for Sustainable Development (IRD), Université Paris Descartes, UMR216-MERIT Bodeau-Livinec, Florence; National Institute for Health and Medical Research (INSERM), Université Paris Descartes, Obstetrical, Perinatal, and Pediatric Epidemiology Research Team, Centre for Epidemiology and Biostatistics (U1153-EPOPé); Ecole des Hautes Etudes en Santé Publique (EHESP), Université de Rennes, Département Méthodes Quantitatives en Santé Publique Houzé, Sandrine ; French National Research Institute for Sustainable Development (IRD), Université Paris Descartes, UMR216-MERIT; CNR du Paludisme, AP-HP, Hôpital Bichat, Laboratoire de Parasitologie Jackson, Nicola; Oxford University, United Kingdom Ayemonna, Paul; Centre Hospitalier Universitaire d’Abomey-Calavi, Abomey-Calavi, Service de gynécologie et obstétrique

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 1 of 35 BMJ Open

1 2 3 4 Massougbodji, Achille; Centre d'Etude et de Recherche sur le Paludisme 5 Associé à la Grossesse et à l'Enfance (CERPAGE), Parasitology Cot, Michel; French National Research Institute for Sustainable 6 Development (IRD), Université Paris Descartes, UMR216-MERIT 7 Fievet, Nadine; French National Research Institute for Sustainable 8 Development (IRD), Université Paris Descartes, UMR216-MERIT 9 Briand, Valerie ; French National Research Institute for Sustainable 10 Development (IRD), Université Paris Descartes, UMR216-MERIT 11 Primary Subject 12 Epidemiology Heading: 13 14 Global health, Infectious diseases, Obstetrics and gynaecology, Paediatrics, Secondary Subject Heading: 15 For peerPublic health review only 16 Longitudinal pre-conceptional cohort, Cohort profile, Malaria, Nutritional 17 Keywords: 18 status, Fetal growth, Africa 19 20 21 22 23 24 25 26 27 28 29 30 31 http://bmjopen.bmj.com/ 32 33 34 35 36 37 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 2 of 35

1 2 3 Title 4 5 6 7 Cohort profile: Effect of malaria in early pregnancy on fetal growth in Benin (RECIPAL pre 8 9 conceptional cohort). 10 11 12 Authors 13 14 15 Manfred Accrombessi,For1,2 peer Emmanuel Yovo, review2 Gilles Cottrell,1 Ginoonly Agbota,1,2 Agnès Gartner,3 16 17 Yves MartinPrevel,3 Nadia FanouFogny,4 Diane Djossinou,3,4 Jennifer Zeitlin,5 Nicaise 18 19 1 5,6 1,7 8 20 TuikueNdam, Florence BodeauLivinec, Sandrine Houzé, Nicola Jackson, Paul 21 22 Ayemonna,9 Achille Massougbodji,2 Michel Cot,1 Nadine Fievet,1 ValérieBriand1* 23 24 25 Affiliations 26 27 28 1 UMR216MERIT, French National Research Institute for Sustainable Development (IRD), 29 30 31 Université Paris Descartes, Paris, France. http://bmjopen.bmj.com/ 32 33 2 Centre d'Etude et de Recherche sur le Paludisme Associé à la Grossesse et à l'Enfance 34 35 36 (CERPAGE), Cotonou, Benin. 37 38 3 UMR204Nutripass, French National Research Institute for Sustainable Development (IRD), 39 on September 25, 2021 by guest. Protected copyright. 40 Université de Montpellier, SupAgro, Montpellier, France. 41 42 43 4 Ecole de Nutrition et des Sciences et Technologies Alimentaires (ENSTA), Faculté des 44 45 Sciences Agronomiques, Université d’AbomeyCalavi, Benin. 46 47 48 5 Obstetrical, Perinatal, and Pediatric Epidemiology Research Team, Centre for Epidemiology 49 50 and Biostatistics (U1153EPOPé), National Institute for Health and Medical Research 51 52 53 (INSERM), Université Paris Descartes, Paris, France. 54 55 6 Département Méthodes Quantitatives en Santé Publique, Ecole des Hautes Etudes en Santé 56 57 Publique (EHESP), Université de Rennes, France. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 3 of 35 BMJ Open

1 2 7 3 Laboratoire de Parasitologie, CNR du Paludisme, APHP, Hôpital Bichat, Paris, France. 4 5 8 6 Oxford University, United Kingdom. 7 8 9 Service de gynécologie et obstétrique, Centre Hospitalier Universitaire d’AbomeyCalavi, 9 10 11 AbomeyCalavi, Benin. 12 13 14 * Correspondingauthor:[email protected] 15 For peer review only 16 17 Dr Valérie Briand 18 19 MERITUMR216/IRD « Mother and child face to tropical infections » 20 21 Faculté des sciences pharmaceutiques Paris Descartes 22 4 avenue de l'Observatoire, 75006 Paris 23 24 Email : [email protected] 25 26 Tel:+33153731527 27 28 29 30 31 http://bmjopen.bmj.com/ 32 33 34 35 36 37 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 4 of 35

1 2 3 Abstract 4 5 6 7 Purpose: RECIPAL is an original preconceptional cohort designed to assess the 8 9 consequences of malaria during the first trimester of pregnancy, which is a poorly 10 11 investigated period in Africa and during which malaria may be detrimental to the fetus. 12 13 14 Participants: For this purpose, a total of 1214 women of reproductive age living in SôAva 15 For peer review only 16 and Akassato districts (south Benin) were followedup monthly from June 2014 to December 17 18 19 2016 until 411 of them became pregnant. A large range of health determinants was collected 20 21 both before and during pregnancy from the first weeks of gestation to delivery. Five Doppler 22 23 ultrasound scans were performed for early dating of the pregnancy and longitudinal fetal 24 25 growth assessment. 26 27 28 Findings to date: Pregnant women were identified at a mean of 6.9 weeks of gestation (wg). 29 30 Preliminary results confirmed the high prevalence of malaria in the first trimester of 31 http://bmjopen.bmj.com/ 32 33 pregnancy, with more than 25.4% of women presenting at least one microscopic malarial 34 35 infection during this period. Most infections occurred before 6 wg. The prevalence of low 36 37 birthweight, smallbirthweightforgestational age (according to INTERGROWTH21st 38 39

charts) and preterm birth was 9.3%, 18.3%, and 12.6%, respectively. on September 25, 2021 by guest. Protected copyright. 40 41 42 Future plans: RECIPAL represents at this time a unique resource that will provide 43 44 45 information on multiple infectious (including malaria), biological, nutritional and 46 47 environmental determinants in relation to health outcomes in women of reproductive age, 48 49 pregnant women and their newborns. It will contribute to better define future 50 51 recommendations for the prevention of malaria in early pregnancy and maternal malnutrition 52 53 in Africa. It confirms that it is possible to constitute a preconceptional pregnancy cohort in 54 55 56 Africa and provides valuable information for researchers starting cohorts in the future. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 5 of 35 BMJ Open

1 2 3 Strengths and limitations of this study 4 5 6 • A unique cohort combining highly detailed, highquality health 7 related information on the pregnancies of ≈ 400 women recruited 8 9 in the preconception period, with a substantial related biobank 10 of plasmas, placental and other samples. 11 12 13 • Preconceptional design allowing the early identification and 14 followup of pregnant women: screening of women for both 15 For peer review only 16 microscopic and submicroscopic malaria from the first weeks of 17 pregnancy; accurate estimation of gestational age by using early 18 19 obstetrical ultrasound scan and collection of some important 20 factors influencing fetal growth such as prepregnancy nutritional 21 22 status and gestational hypertensive disorders. 23 24 • Collection of valuable information for the implementation of 25 26 future preconceptional cohorts in Africa. 27 28 • High attrition in the preconceptional cohort, with a possible 29 30 impact on the external validity of some findings. 31 http://bmjopen.bmj.com/ 32 • Reduced sample size at delivery due to a high proportion of 33 34 spontaneous abortion, with a possible lack of power for analyses 35 on birth outcomes. 36 37 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 6 of 35

1 2 3 Introduction 4 5 6 7 Malaria in pregnancy is associated with a wide range of deleterious effects in women and 8 9 their offspring. In SubSahara African (SSA) countries, preventive strategies are based on 10 11 longlasting insecticide treated bed nets (LLITNs) and intermittent preventive treatment in 12 13 pregnancy (IPTp). IPTp consists in the administration of sulfadoxinepyrimethamine (SP) at 14 15 each antenatalFor care (ANC) peer visit from the review 2nd trimester of pregnancy only onwards [1]. LLITN are 16 17 18 distributed at the first ANC visit, which generally occurs around 45 months of pregnancy [2]. 19 20 Therefore, pregnant women remain unprotected or insufficiently protected during the first 21 22 months of pregnancy, and particularly during the first trimester. However, this period may be 23 24 a highrisk period for the fetus if pregnancyassociated malaria parasites accumulate into the 25 26 27 placenta during trophoblast differentiation and vascular remodelling of the uterus [3,4]. 28 29 Impaired placentation may then contribute to fetal growth restriction. 30 31 http://bmjopen.bmj.com/ 32 Few studies have investigated malaria in early pregnancy and while some have shown an 33 34 adverse effect on birth outcomes, this has not been consistent [5]. In Burkina Faso and in 35 36 Benin, malaria infection in the first trimester [6] or before 45 months of pregnancy [7,8] has 37 38 39 been associated with a higher risk of low birthweight (LBW) or a decrease in birthweight. 40 on September 25, 2021 by guest. Protected copyright. 41 Conversely, such association was not reported in Uganda and Malawi [9,10] but the number 42 43 of women screened in the first trimester was low in these two studies. One of the problems in 44 45 interpreting this existing literature relates to the challenges of recruiting women early in 46 47 pregnancy; most of the studies were not designed to assess the consequences of malaria early 48 49 50 in pregnancy and therefore women were generally recruited late in the first trimester, which 51 52 may lead to misclassification errors and underestimation of exposure. 53 54 55 The RECIPAL study aimed to assess the effect of malaria (both microscopic and 56 57 submicroscopic) in the first trimester of pregnancy on both the mother and the fetus by 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 7 of 35 BMJ Open

1 2 3 following a cohort of pregnant women recruited before conception. In addition, it aimed to 4 5 assess the influence of the woman’s nutritional status in the relationship between malaria in 6 7 early pregnancy and birth outcomes. 8 9 10 11 Cohort description 12 13 14 Study setting 15 For peer review only 16 17 The study cohort was conducted in the districts of SôAva and AbomeyCalavi, located in 18 19 20 Southern Benin (Figure 1). In the area, malaria is hyperendemic with a mean entomological 21 22 inoculation rate of 2.1 infected anopheles bites/person/100 nights [11]. The main mosquito 23 24 vectors of malaria are Anopheles gambiae ss and Anopheles funestus [12]. Four subdistricts 25 26 (SôAva, Vekky and Houedo in SôAva District, and Akassato in AbomeyCalavi District) 27 28 29 were selected for the study according to the population density and mean number of ANC 30 31 visits and deliveries per month in the main public maternity clinic. Thirtyfive out of a total of http://bmjopen.bmj.com/ 32 33 36 villages were selected for the study according to their proximity to the maternity clinics. 34 35 36 Study design, subject identification, recruitment and enrolment procedures 37 38 39 Briefly, women of reproductive age (WRA) were recruited at communitylevel and followed 40 on September 25, 2021 by guest. Protected copyright. 41 monthly for a maximum period of 24 months until becoming pregnant, they constituted the 42 43 44 RECIPAL initial cohort. The subsample of women who became pregnant was then followed 45 46 up monthly at the maternity clinic from early pregnancy to delivery; they constituted the 47 48 RECIPAL final cohort. 49 50 51 A sample size of 466 births was estimated sufficient to evidence a 2fold odds ratio for the 52 53 association between malaria and poor birth outcomes [4], assuming a prevalence of 25% of 54 55 women with microscopic malaria in the 1st trimester of pregnancy and 30% of poor birth 56 57 58 outcomes (LBW, smallforgestational age (SGA), preterm birth (PTB) or stillbirth) [8,13,14] 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 8 of 35

1 2 3 with a 80% power and significant level at 0.05. We estimated that following 2,000 WRA 4 5 (initial cohort) would have allowed identifying 510 pregnant women (final cohort) during the 6 7 time of the study (based on a total fertility rate of 200/1,000/year in Benin 8 9 10 (http://esa.un.org/unpd/wpp/index.htm) and allowing 15% of loss to followup). 11 12 13 After explanation of the study to the local political and administrative authorities, three 14 15 concomitant Forprocedures werepeer used to recruit review WRA. First, repeated only sensitization events were 16 17 organized in each of the 35 selected villages for presenting the study to all inhabitants. If 18 19 interested, women were invited to register with the head of the community and were visited at 20 21 home the day after for their inclusion. That procedure constituted the main way of recruitment 22 23 24 throughout the study. Second, leaders of the community, religious authorities and massmedia 25 26 were involved in order to help us mobilizing women during public meetings (masses, 27 28 women’s association meetings, etc.). The last way of recruitment consisted of going doorto 29 30 door to identify eligible and interested women with the assistance a network of community 31 http://bmjopen.bmj.com/ 32 33 health workers. 34 35 36 To be included, women had met the following criteria: negative urinary pregnancy test, 18 to 37 38 45 years old, no current contraception, no previous fecundity issues, willingness to become 39 40 pregnant, no planned travel for more than 2 months within the next 18 months, acceptance of on September 25, 2021 by guest. Protected copyright. 41 42 RECIPAL protocol and signed written informed consent. Both oral and written 43 44 45 communication was used for providing information on the project to the women and their 46 47 husbands. The consent form was translated into the local language for women who were 48 49 illiterate. 50 51 52 From June 2014 to December 2016, 1302 WRA were willing to participate in the study. 53 54 Among them, 1214 (93.2%) met the inclusion criteria and were recruited, 88 (6.8%) were not 55 56 included mainly because of a positive urinary pregnancy test at inclusion (63.6%) or past 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 9 of 35 BMJ Open

1 2 3 infertility issues (14.8%). Based on both the 2013 national census and 2011 Demographic 4 5 Health Survey in Benin, we estimated that women included in the project represented 6% of 6 7 the total number of WRA living in the study area. 8 9 10 RECIPAL study was approved by the ethics committees of the Institut des Sciences 11 12 13 Biomédicales Appliquées (ISBA) in Benin and from the IRD in France as well as by the 14 15 Ministry of HealthFor in Benin. peer review only 16 17 18 Cohort follow-up 19 20 21 Pre-conceptional follow-up 22 23 24 After enrolment, WRA were visited at home monthly by specifically trainedinvestigators 25 26 assisted by local communityhealth workers. Followup consisted mainly of recording the first 27 28 day of last menstrual period (LMP) and performing a urinary pregnancy test every month. To 29 30 minimize the risk of loss of confidentiality during the tracking of women for followup, the 31 http://bmjopen.bmj.com/ 32 study team was instructed to speak only to people—including family members—to whom 33 34 35 women have granted them permission to speak. After 12 months, women who were still 36 37 followed received medical advices to help conceive and they were invited to attend the 38 39 maternity clinic for a medical examination in case of symptoms of genital infection. After 24 40 on September 25, 2021 by guest. Protected copyright. 41 months, the followup ended and women who did not conceive were invited to attend a 42 43 44 gynaecologist for fertility issues assessment. 45 46 47 Gestational follow-up 48 49 Once pregnant, women were followedup at the maternity clinic for monthly ANC visits. In 50 51 52 addition, they were encouraged to attend the maternity clinic any time in case of any 53 54 symptoms. Both the maternity staff (for usual ANC followup and clinical examination) and 55 56 study investigators (for all specific issues related to RECIPAL) were involved in women’s 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 10 of 35

1 2 3 followup. Throughout the study, women were offered free transport to the maternity clinic; in 4 5 SôAva District a free river shuttle was set up. At 37 weeks gestation, women were visited at 6 7 home weekly and transportation to the study maternity clinic was scheduled for delivery. 8 9 10 According to Beninese national recommendations, a kit including iron and folic acid tablets 11 12 st 13 for one month, mebendazole for deworming, the 1 dose of SP for IPTp and a LLITN was 14 15 given to eachFor pregnant womanpeer at the first review ANC visit. The maternity only staff was encouraged to 16 17 administer at least three doses of IPTp from the 2nd trimester onwards as recommended by the 18 19 Beninese National Malaria Control Program, as well as iron and folic acid supplementation 20 21 throughout the pregnancy. RECIPAL women benefited from free management of any diseases 22 23 24 related to the pregnancy—detected either as part of RECIPAL followup or during 25 26 unscheduled visits, free ANC visits and delivery. Women infected with malaria were treated 27 28 with quinine in the first trimester of pregnancy and arthemeterlumefantrine from the 2nd 29 30 trimester onwards. Management of preterm or growthrestricted newborns were also in charge 31 http://bmjopen.bmj.com/ 32 33 of the project. 34 35 36 Cohort attrition 37 38 Figure 2 presents the flowchart diagram of the study. Among the 1214 WRA enrolled in the 39 40 on September 25, 2021 by guest. Protected copyright. 41 initial cohort, 411 (33.9%) became pregnant, 359 (29.6%) completed the 24 monthfollowup 42 43 without conceiving and 444 women (36.6%) did not complete the study mainly because of 44 45 informed consent withdrawal (272/444) or migration outside the study area (138/444) that 46 47 occurred after 5.2 and 6.3 months of followup in average, respectively. 48 49 50 Among the 411 pregnant women, 207 (50.4%) have already delivered, 16.5% (68/411) had a 51 52 spontaneous abortion (n=68) or a nonviable pregnancy (n=2), 52 did not complete the study 53 54 55 because of informed consent withdrawal (37/411, 9%) or migration (15/411, 3.6%), and 72 56 57 women are still followed (Figure 2). Spontaneous abortions, informed consent withdrawals 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 11 of 35 BMJ Open

1 2 3 and migrations occurred at 8.9 (± 3.9), 13.2 (± 8.1) and 14.2 (± 8.6) weeks of gestation (wg) 4 5 in average (SD), respectively. 6 7 8 Data collection 9 10 11 Pre-conceptional data 12 13 14 Demographic, socioeconomic and household characteristics were collected at inclusion (Table 15 For peer review only 16 1). At this occasion, screening for malaria, urinary schistosomiasis, Chlamydia trachomatis 17 18 and Neisseria gonorrhoeae was performed. Haemoglobin (Hb) and markers of inflammation 19 20 21 levels were determined. The first day of LMP was recorded, and a urinary pregnancy test 22 23 (OneStep®, International Holding Corp, Germany) was performed each month. 24 25 Anthropometric measurements were collected every three months and dietary intakes were 26 27 assessed twice during followup (see below for details). After 12 months, for those women 28 29 who were still followed, they were screened for urinary schistosomiasis again in case of 30 31 http://bmjopen.bmj.com/ 32 macroscopic haematuria. 33 34 35 Gestational data 36 37 At each ANC visit, clinical, anthropometric and obstetrical data, as well as LLITN use, IPTp 38 39 40 administration, alcohol consumption, and iron, folic acid and other micronutrient on September 25, 2021 by guest. Protected copyright. 41 42 supplementation intake were collected (Table 2). In addition, malaria screening was 43 44 performed, and proteinuria, glycosuria and urinary infection were detected using a urine 45 46 dipstick test (Combiscreen®, Analyticon, Germany). Dietary intakes were assessed once 47 48 49 every trimester. Five Dopplerultrasound scans (US) were performed (see below for details).

50 st rd 51 A sample of venous blood was collected in the 1 and 3 trimesters for Hb, markers of 52 53 inflammation and lead levels determination. 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 12 of 35

1 2 3 At delivery, newborns were weighed within 1 hour after birth on an electronic digital scale 4 5 with an accuracy of 2 grams (SECA 757, SECA Ltd., Germany). Newborn’s length (SECA 6 7 infantometer 416, Germany) and head circumference (SECA 201, Germany) were recorded to 8 9 10 the nearest millimetre. Malaria screening was performed on maternal venous blood, placental 11 12 and cord blood (Table 2). A placental biopsy was performed for malaria histopathology. 13 14 15 The types andFor origins ofpeer the biological review samples collected before only and during pregnancy are 16 17 listed in Table 3. Preparation and storage are performed by two laboratory technicians under 18 19 the supervision of a senior biologist. Total blood, serum, plasma urine samples are stored in 20 21 20°C or 80°C freezers with a daily control of the temperature. 22 23 24 25 Specifics components 26 27 Malaria diagnosis. Before conception and during pregnancy, malaria screening was 28 29 performed using both microscopy (thick blood smear) and polymerase chain reaction (PCR). 30 31 http://bmjopen.bmj.com/ 32 Blood smears were stained with Giemsa and parasitaemia was quantified by the Lambaréné 33 34 method [15]. A molecular diagnostic approach using a combination of realtime PCR assays 35 36 comprising genusspecific primers and probes for the gene encoding the small (18S) of 37 38 Plasmodiumr RNA and an ultrasensitive P. falciparumspecific detection system [16] were 39 40 on September 25, 2021 by guest. Protected copyright. 41 used to screen samples containing Plasmodium parasites. In addition, any time during 42 43 pregnancy, a rapid diagnostic test (Pf + pan rapid test SD Bioline Ag®, IDA Foundation, 44 45 Netherlands; BioSynex®, France) was performed in case of symptoms suggestive of malaria 46 47 for immediate diagnosis and treatment. 48 49 50 Ultrasound follow-up. The first US for dating the pregnancy was performed between 9 and 51 52 13 wg based on LMP recorded during the preconceptional followup (supplementary file S1). 53 54 55 The gestational age (GA) was estimated in days based on crownrump length measurement 56 57 using the Robinson’s chart [17]. At the end, GA estimation was based either on LMP if the 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 13 of 35 BMJ Open

1 2 3 difference between the two measurements (LMP/US) was less than 7 days or on US if the 4 5 difference is > 7 days. Following INTERGROWTH21st methodology, four additional scans 6 7 were performed for fetal biometry assessment, and uterine and umbilical blood flow 8 9 10 measurements [18]. Each parameter was measured twice and then averaged; a third measure 11 12 was performed in case of discrepancy between the two first measurements. A quality control 13 14 was performed on 10% of the pictures. 15 For peer review only 16 17 Anthropometric measurements and dietary intake assessment. Anthropometric 18 19 measurements were collected using standard procedures (Tables 1 and 2) [19]. During 20 21 household visits, women’s body weight was measured with a 200 g precision with calibrated 22 23 24 electronic scales (Tefal, France). At the facilitylevel, women were weighed with a Tanita 25 26 MC780 body composition device (Tanita Corporation, Tokyo, Japan). Height was measured 27 28 to the nearest millimeter with a SECA 206 (Hamburg, Germany) gauge. The left midupper 29 30 armcircumference (MUAC) was measured with a SECA 201 (Hamburg, Germany) 31 http://bmjopen.bmj.com/ 32 33 ergonomic circumference measuring tape. Skinfold thickness measurements were made to the 34 35 nearest millimeter at the triceps and biceps sites using a Holtain (Crymyeh, UK) skinfold 36 37 caliper. The set of measurements was repeated twice and then the measurements were 38 39 averaged. Body composition, assessed by bioelectrical impedance analysis method, was 40 on September 25, 2021 by guest. Protected copyright. 41 measured by the multifrequency body composition analyser Tanita MC780 with a four 42 43 44 electrode arrangement paired on hands and feet. 45 46 47 Dietary intakes were assessed by 24hrecalls using the standard multiplepass methodology. A 48 49 comprehensive list of the most common food items and recipes was developed from a 50 51 preliminary study of the diet habits of the study area. This list was used to identify foods 52 53 mentioned by the women as consumed during the interview. Amounts of food consumed were 54 55 56 estimated by household measures consisted of precalibrated utensils, food pictures of 57 58 different portion sizes, monetary value of known food portions sold on the study area markets. 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 14 of 35

1 2 3 A food composition table and recipes database are under construction for conversion of 4 5 household units to gram, and to determine macro and micronutrients intake by the women. 6 7 8 Characteristics of the study populations 9 10 11 At enrolment, WRA were 27 years old in average (Table 4). Nulligravidae represented 8.8% 12 13 of the cohort. More than half of the women were anaemic, one quarter was infected with 14 15 For peer review only 16 schistosomiasis and 38% had an abnormal BMI. The prevalence of microscopic malaria was 17 18 5.7%, with two geographical clusters in SôAva and Vekky subdistricts (Figure 3). The 19 20 median duration of followup in the initial cohort was 5.9 months (mean=7.1, range: 0.49– 21 22 23.7). Compared to women included in the initial cohort, women who did not complete the 23 24 25 24month followup were significantly different in terms of residence area, ethnicity, 26 27 household density and schistosomiasis status (Table 4). 28 29 30 Pregnant women were identified at a mean of 6.9 wg and benefited from 7.5 scheduled ANC 31 http://bmjopen.bmj.com/ 32 visits in average (Table 4). The prevalence of microscopic malaria infection was 25.4%, 33 34 19.4% and 16.1% during the 1st, 2nd and 3rd trimester of pregnancy, respectively. The 35 36 prevalence of SGA was 11.5% and 18.3% using Schmiegelow’s and INTERGROWTH 21st 37 38 39 charts, respectively (Table 5). The proportion of PTB and LBW was 12.6% and 9.3%, 40 on September 25, 2021 by guest. Protected copyright. 41 respectively. Overall, 29.5% (61/207) of newborns presented at least one poor birth outcome 42 43 defined as LBW, PTB, SGA or stillbirth. 44 45 46 47 Findings to date 48 49 50 Fertility and time-to-pregnancy 51 52 53 The median timetopregnancy was 12.3 months (Figure 4). The fertility rate was 5.4 54 55 pregnancies/100 personsmonth (95% confidence interval [CI] = 4.95.9). The probability 56 57 (95% CI) of conceiving after 5, 10, 15 and 20 months of followup were 24.5% (21.927.4), 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 15 of 35 BMJ Open

1 2 3 42.1% (38.645.9), 56.6% (52.161.1) and 62.8% (57.668), respectively. Using multivariate 4 5 survival analysis, factors associated with a lower risk of conceiving were maternal age more 6 7 than 35 years, a low education level, being polygamous, living in Vekky subdistrict, long 8 9 10 term couple (> 8 years), infrequent sexual activity, overweight and urinary schistosomiasis 11 12 infection (Yovo, personal communication). 13 14 15 Malaria in theFor first trimester peer of pregnancy review only 16 17 18 The incidence rate of malaria infection during the first trimester was 27.5 cases per 100 19 20 personsmonth. Using a multilevel logistic regression, we showed that women infected with 21 22 malaria before pregnancy were significantly more likely to be infected with malaria during the 23 24 1st trimester of pregnancy (adjusted relative risk (aRR)= 2.72, 95% CI: 1.265.85). Besides, 25 26 malaria was significantly more frequent in the first weeks of pregnancy (≤ 6 wg) than later in 27 28 29 the first trimester (aRR= 1.77, 95% CI: 1.072.94) (Accrombessi, personal communication). 30 31 http://bmjopen.bmj.com/ 32 Strengths and limitations 33 34 35 36 The preconceptional design—and early identification of pregnant women—allowed (i) the 37 38 screening of women for malaria from the very beginning of pregnancy; (ii) the accurate 39 40 estimation of GA by using US before 14 wg; (iii) the determination of prepregnancy on September 25, 2021 by guest. Protected copyright. 41 42 nutritional status[20] and hypertensive disorders screening [21], which highly influence fetal 43 44 45 growth; and (iv) the minimization of selection bias embedded in the design of women 46 47 recruited at facilitylevel. Besides, PTB and IUGR could be accurately assessed thanks to the 48 49 precise estimation of GA and longitudinal fetal growth assessment. 50 51 52 In addition to malaria, RECIPAL will provide valuable information on WRA in Africa in 53 54 terms of fertility and nutrition, which are seldom evaluated and may help to define new pre 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 16 of 35

1 2 3 conceptional interventions for improving maternal and child health [22]. Also, it yields 4 5 logistical and ethical information for the implementation of future preconceptional cohorts. 6 7 8 Because of constraints related to the study design (long duration of the preconceptional 9 10 followup and weariness of the women) and African sociocultural realities (rumour about 11 12 13 blood collection, suspicion regarding medical care provided free of charge, etc.), a noticeable 14 15 proportion (22%)For of women peer withdrew their review informed consent beforeonly the end of the 24 month 16 17 followup. Overall, the initial cohort attrition was 37%. Since some baseline characteristics in 18 19 women who completed the preconceptional followup and those who did not were different, 20 21 this might impact the external validity of some of results. During pregnancy, there was a high 22 23 24 proportion (16.5%) of spontaneous abortions, which contributed to the final cohort attrition. 25 26 At the end, birth outcomes could be evaluated in only 68% of women, with a possible lack of 27 28 power for some upcoming analyses. 29 30 31 http://bmjopen.bmj.com/ 32 Conclusion 33 34 35 If the deleterious effects of malaria in the first trimester of pregnancy are confirmed, our 36 37 results may argue in favor of starting preventive measures from the very beginning of 38 39 40 pregnancy or even before pregnancy. Dihydroartemisininpiperaquine has been shown to be on September 25, 2021 by guest. Protected copyright. 41 42 safe during the first trimester of pregnancy and, therefore, may be a good option for the 43 44 replacement of SP [23]. Besides, a vaccine against pregnancyassociated malaria parasites, 45 46 which could elicit protective immunity prior to pregnancy to best protect pregnant women 47 48 49 during early pregnancy may be proposed as a complementary strategy. Such a vaccine is 50 51 currently under evaluation [24]. 52 53 54 Collaboration 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 17 of 35 BMJ Open

1 2 3 The “Mother and child face to tropical infections” research unit (MERIT), French National 4 5 Research Institute for Sustainable Development (IRD)/Paris Descartes University was the 6 7 promoter of the study. RECIPAL was a collaborative project between MERIT, another IRD 8 9 10 research unit (Nutripass, Montpellier), EPOPé research team (Inserm), the Ecole des Hautes 11 12 Etudes en Santé Publique (EHESP) and two Beninese collaborators from AbomeyCalavi 13 14 University (the Faculty of Health Sciences and the Faculty of Agronomy Sciences). 15 For peer review only 16 17 The researchers associated with this study are open for the sharing and reuse of the data but an 18 19 end user data use agreement will be required for accessing the data. Collaborations are 20 21 encouraged, although data sets are not currently publicly available. Any researcher interested 22 23 24 in exploring RECIPAL data should contact directly the principal investigator, Valerie Briand 25 26 (email: [email protected]). 27 28 29 30 31

http://bmjopen.bmj.com/ 32 33 34 35 36 37 38 39

on September 25, 2021 by guest. Protected copyright. 40 41

42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from

‡

Page 18 of 35

‡

‡ http://bmjopen.bmj.com/

‡ BMJ Open

‡

‡ on September 25, 2021 by guest. Protected copyright.

‡

¤ For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

- - † † ------† ------† † - - † - - - - - † - - † - - - - † - - - † † † - - - † - † ------† - - - - - † † † - - - † - - † - - - - - † - - † - † ------† - † - - - - † ------† - † ------† - - † - - - † ------† ------† † ------† - - - † ------† - † - - † ------† ------† - - - † ------† ------† - - † ------† ------† - † † † † † † † † † † † † † † † ‡ † For† † - - - - - peer ------review ------only ------Inclusion D1 M1 M2,M3 M4 M5,M6 M7 M8,M9 M10 M11,M12 M13 M14,M15 M16 M17,M18 M19 M20,M21 M22

Clinical, nutritional and biological data collected during the preconceptional followup. The RECIPAL study, study, 20142017. The RECIPAL followup. preconceptional the collected during data and biological nutritional Clinical,

(BIA)

Table 1. Table 1.

h recall h

demographic characteristics * C. trachomatis and N. gonorrhoeae screening (using PCR and serology), as well as schistosomiasis screening (Nytrel® filter) started 9 and 12 months after study initiation,respectively. study and filter)12after (Nytrel® started 9 months screening schistosomiasis asas PCR serology), (using well and screening C. gonorrhoeae trachomatisN.* and storage furtherforUrineanalysis ** reaction; sTfR: chain Soluble polymerase smear;PCR: blood TBS: Thick period; menstrual Last analysis; LMP: BIA: Bio impedance MUAC:M: MidUpperArmCircumference; Months; Abbreviations: day; D: Socio characteristicsHousing composition Body Dietary 24 CRP,AGP, levels sTfR Chlamydiatrachomatis Urine storage** Place of visit Eligibility criteria Urinary schistosomiasis* specific attimepoint. () (†) collected/assessed each or not Information clinic; (‡)at visit; study maternity Household the performed (¤) Visit Hb: Haemoglobin. Glycoprotein; protein; AGP: alpha1Acid CRP:CReactive Transferrin Receptor; General General characteristics Economic characteristics Clinicalnutritional and data Obstetricalhistory Height Weight, MUAC, skinfold thickness Amenorrhea/LMP Biological data Urinary testpregnancy Malaria (TBS,screening PCR) Hb, folic ferritin,acidlevels Informed consent GPS location Neisseriagonorrhoeae* 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from

- - - †

- † † †

------† - - - - - † - - - † - † - - † † † † † † † † † † † † † † † † † † † † † † † - † † † † - † † † † - † † † † † † † † ------† † - - - † † - - -

------† † † † † † † † † † † † † † † † † † †

------† † † † † † † † † † † † † † † † † † † †

------† † † † † † † † † † † † † † † † † http://bmjopen.bmj.com/ BMJ Open

------† † † † † † † † † † † † † † † † † † † † † † † on September 25, 2021 by guest. Protected copyright.

------† † † † † † † † † † † † † † † 6.8 ± 2.5 6.8 ± 11.62.9 ±2.9 16.7 22.1 3.7 ± ± 27.6 3.8 ± 32.7 3.5 ± 36.2 2.5 ± 38.4 1.4 ± 39.1 3.0 ANCv 1 ANCv 2 ANCv 3 ANCv 4 ANCv 5 ANCv 6 ANCv 7 ANCv ANCv8 Delivery For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For peer review only Clinical, nutritional and biological data collected during pregnancy. The RECIPAL study, 20142017. 20142017. study, The RECIPAL pregnancy. collected during and data biological nutritional Clinical,

‡ ) or not () at each specific timepoint. specific timepoint. ()each )at not or † Table Table 2.

Information collected/assessed ( collected/assessed Information Abbreviations: Abbreviations: ANCv: Antenatal care visit; ITN: bed Insecticidetreated net; IPTp: treatmentpreventive Intermittent in pregnancy; LMP: Last menstrual period; TBS: Thick blood smear; PCR: polymerase chain Hb: Haemoglobin. AGP:protein; alpha1Acid Glycoprotein, Receptor; CReactive Transferrin sTfR: CRP: reaction; Soluble Weight Anthropometric Anthropometric measurements Gestational age (based on Gestational (based age on LMPfirst US)or Blood pressure Blood Axillary temperatureAxillary Use of ITN Use night the of thebefore visit Iron and folate Iron intake lastin 24 the hours Alcohol consumptionAlcohol in lastthe 24 hours Peripheral Peripheral malaria (TBS,PCR) malaria Placental(TBS, PCR, histology) Biological data Biological Body composition Body (Bio impendanceanalysis) Past medical Pastmedical and history* obstetrical screeningHIV 1 schistosomiasisUrinary Gestational age Gestational (weeks),age mean (SD) MidUpperArmCircumference recall Dietary 24h and rhesusBlood group Serum sTfR,ferritin, CRP, acidAGP,levels folic intestinalHelminthic infection(PCR) disorders. orhypertensive of delivery preterm history affection; other chronicmedical or Sickledisease, cell diabetes * at collected delivery. pregnancy during istory smoking of cigarette ‡H Cord blood Cord blood PCR) malaria (TBS, Clinical and nutritional data data nutritional and Clinical Lead Lead level Skinfold thickness (bicipitalandtricipital) Hb levelHb Urinary infection Urinary and proteinuria test)(dipstick

Page 19 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 20 of 35

detection

phenotyping phenotyping

N. gonorrhoeae Vitamin A Vitamin Metals level Metals level and and

s, cytokines and chemokines cytokines s, and chemokines cytokines s, Pollutant/pesticide Pollutant/pesticide A A Helminthic infection Helminthic Planned analyses analyses Planned Erythrocytic folic acid Erythrocytic folic Immunological markers Immunological markers Immunological markers Immunological markers Immunological markers Immunological markers Plasmodium Ferritin, CRP, sTfR, AGP AGP sTfR,Ferritin, CRP, Histopathology for malaria Histopathology RN RN detection (PCR) Plasmodium detection Micro Micro C. trachomatis C.

cord cord

placenta placenta Placenta Placenta Placenta Placenta Placenta Placenta Delivery/Birth Delivery/Birth Motherplacentacord Motherplacentacord Motherplacentacord Immunological markers Mother Mother

http://bmjopen.bmj.com/ BMJ Open

‡ ‡

Mother (x8) Motherplacentacord Mother (x2) Motherplacentacord Mother (x2) Mother (x2) Mother (x2) Motherplacentacord Mother (x2) Mother (x3) Mother (x2) Mother (x2) Mother (x8) Mother (x8) on September 25, 2021 by guest. Protected copyright. trimester of pregnancy), three times (during each trimester), or eight times (during each ANC visit). visit). ANC each (during times eight or trimester), each times (during three pregnancy), of trimester rd

and 3 and st

Mother (x1) Mother Mother (x1) Mother Mother (x1) Mother Mother (x1) Mother (x1) Mother (x1) Mother Mother (x1) Mother (x2) Mother (x1) Mother For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Pre-pregnancy Pre-pregnancy Pregnancy For peer review only Biological samples stored before conception, during pregnancy and at birth. The RECIPAL study, 20142018 RECIPAL study, The at birth. and pregnancy during conception, before samples stored Biological Table Table 3.

†

TBS: Thick blood smear; sTfR: Soluble Transferrin Receptor; CRP: CReactive protein; AGP: alpha1Acid Glycoprotein, PCR: Polymerase Chain Reaction. Reaction. Chain Polymerase PCR: Glycoprotein, alpha1Acid AGP: protein; CReactive CRP: Receptor; Soluble Transferrin sTfR: smear; Thick blood TBS:

† †

80°C 20°C (1.8 mL) (80°C)

During pregnancy, samples were collected twice (during the1 (during twice collected were samples pregnancy, During In case of unscheduled visit, an additional blood spot was collected. collected. was spot blood additional an visit, unscheduled caseIn of respectively. µl, 600 600800, was200, delivery and at pregnancy during before, stored plasma of Quantity Biopsy with formalin with Biopsy Biopsy 20°C 20°C Biopsy Later RNA with Biopsy Blood spot +4°C (50 µL) (50µL) +4°C spot Blood Total blood µL)(500 Total blood Plasma Plasma (100µL) Serum Total blood µL)(200 Total blood Plasma/serum (400 µL) Serum Urine Stool Placenta brackets. in is indicated sample collection of The frequency $ ‡ † Abbreviations: Total blood blood Total

Blood spot 20°C (100 µL) µL) (10020°C spot Blood Red cells (Trizol, 100 µL)100 (Trizol,cells Red Buffy coat PAXgene (300 µL) PAXgene coat Buffy Serum (100µL) Serum 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from

P value P value (411 vs 359) (411359) vs

0.04 0.03

0.10 0.002

< 0.001

< 0.001 0.51 0.65

0.03

5.9 5.9 9.7 9.7 6.1 6.1 7.8 7.8 7.8 7.8 4.9 4.9 6.3 6.3 8.0 8.0 8.0 8.0 27.0 27.0 54.2 14.4 14.4 33.8 33.8 28.7 28.7 65.2 65.2 (N=411) (N=411) Final cohort Pregnant women women Pregnant 70.6 0.41

92.0 25.8 64.5 0.10

72.9

32.1

26.3

3.4 3.4 8.1 8.1 0.3 0.3 8.6 8.6 8.4 8.4 7.5 7.5 8.9 8.9 73.3 73.3 87.1 87.1 32.7 32.7 59.2 59.2 22.3 22.3 60.7 60.7 15.9 15.9 12.5 12.5 42.9 42.9 33.4 33.4 56.8 56.8 19.6 19.6 46.5 38.2 38.2 12.9 12.9 follow-up but no no but follow-up pregnancy (N=359) pregnancy WRA with completewith WRA

P value P value (1214 vs 444) (1214444) vs 0.06 0.62

0.46

http://bmjopen.bmj.com/ 7.2 7.2 7.5 7.5 6.1 6.1 6.9 6.9 3.1 3.1 5.2 5.2 4.3 4.3 7.4 7.4 BMJ Open WRA with with WRA up (N=444) up incomplete follow- incomplete

77.9 12.6 0.001 51.4 < 0.001 23.0 38.9

57.9 0.45 10.8 13.1 0.17 5.7 5.7 6.7 6.7 8.2 8.2 8.8 8.8 3.3 3.3 5.6 5.6 7.1 7.1 73.2 73.2 75.7 0.14 13.8 13.8 77.4 77.4 76.1 30.0 30.0 31.0 61.8 61.8 61.1 0.36 71.2 71.2 36.7 36.7 16.1 16.1 60.0 60.0 34.8 34.8 15.3 15.3 24.7 24.7 52.5 31.2 56.3 0.03 0.07 26.9 26.9 18.2 23.0 23.0 on September 25, 2021 by guest. Protected copyright. 2.8 (± 2.0) 2.0) 2.8 (± 2.0) 2.6 (± 0.07 2.1) 2.9 (± 1.9) 2.7 (± 0.16 5.7 (± 3.0) 3.0) 5.7 (± 3.4) 6.0 (± 0.009 5.2 (±2.3) 5.9 (±3.0) 0.001 27.8 (± 5.5) 5.5) (± 27.8 5.6) (± 28.0 0.31 5.6) (± 28.8 5.0) (± 26.8 < 0.001 Initial cohortInitial (N=1214 WRA) WRA) (N=1214

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Yes Yes Yes Yes Yes Fon Fon Aizo Aizo Toffin Toffin Vekky Vekky Others Others SôAva SôAva Houedo Houedo < 23 cm cm 23 < Illiterate Illiterate Akassato Akassato Polygamy Polygamy ≥ 25kg/m² ≥ 25kg/m² Monogamy Monogamy < 18.5kg/m² < 18.5kg/m² Cohabitation Cohabitation Mean (± SD) (± SD) Mean Mean (± SD) (± SD) Mean Mean (± SD) (± SD) Mean Nulligravidae Nulligravidae 18.5‒24kg/m² 18.5‒24kg/m² Primigravidae Primigravidae For peer review Multigravidae only Maternal length < 150 cm cm 150 < length Maternal 8.0 Characteristics of women included in the initial and final RECIPAL cohorts, Southern Benin, 20142018. 20142018. Benin, cohorts, Southern final and RECIPAL initial in the women included of Characteristics

£ Table Table 4.

Body mass index (%) (%) mass index Body Schistosomiasis infection (%) (%) infection Schistosomiasis

Subdistrict (%) (%) Subdistrict (%) groupEthnic Pre-conceptionalcharacteristics General characteristics General

Anaemia (%) Anaemia (%) malaria Microscopic

Age (years) (years) Age Marital status (%) (%) status Marital

Gravidity (%) (%) Gravidity Household density density Household Education (%) (%) Education

Short stature (%) stature (%) Short (%) MidUpperArmCircumference

Number of live birth(s) birth(s) live of Number Page 21 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 22 of 35 P P value* (411 vs 359) (411359) vs

). 5.9) −

8.0 8.0 1.7 1.7 7.8 92.0 92.0 (N=411) (N=411) continued 5.4 (4.9 Final cohort Pregnant women women Pregnant

48.5 25.4 19.4 57.0 16.1 68.6 44.9

65.7 12.3

1.1) 7.5 (± 1.3) 2.1 (± 2.5) 6.9(± 57.9 12.1

follow-up but no no but follow-up pregnancy (N=359) pregnancy WRA with completewith WRA

P P value* (1214 vs 444) (1214444) vs

http://bmjopen.bmj.com/ BMJ Open WRA with with WRA up (N=444) up incomplete follow- incomplete

on September 25, 2021 by guest. Protected copyright. Initial cohortInitial (N=1214 WRA) WRA) (N=1214

trimester trimester trimester trimester trimester trimester trimester trimester st st rd rd For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml ≥ 2 nd Yes Yes Yes Yes Yes Mean (± SD) SD) (± Mean Mean (± SD) SD) (± Mean Mean (± SD) SD) (± Mean Primigravidae Primigravidae Multigravidae Multigravidae During pregnancy pregnancy During During pregnancy pregnancy During In the 1 In In the 1 In In the 3 In In the 3 In For peer review the 2 In only

†

† Characteristics of women included in the initial and final RECIPAL cohorts, Southern Benin, 20142018 ( 20142018 Benin, cohorts, Southern RECIPAL final and initial in the of women Characteristics included

†

Table Table 4. ANC visit (weeks) (weeks) visit ANC SD) (± Mean st

† § § †

Incidence rate defined as the number of pregnancies / personsmonth at risk. risk. at personsmonth / pregnancies of number the as defined rate Incidence

Gestational age at 1 at age Gestational HIV status (%) HIV

† Results based on the 207 women who have already delivered. delivered. already have who 207 women the on based Results † care. Antenatal ANC: deviation; Standard age; SD: reproductive of women WRA: Abbreviations: Inadequate gestational weight gain (%) (%) gain weight gestational Inadequate Microscopic malaria (%) (%) malaria Microscopic (%) infection malaria Placental

Gestational hypertension (%) (%) hypertension Gestational respectively. variables, categorical and continuous comparing for used were test chi2 and Tstudent * women (n=718). of subsample a on based Results £ pregnant). for and 11 g/dL women nonpregnant for (12 g/dL thresholds to WHO § According Anaemia (%) (%) Anaemia

Median time to pregnancy (months) (months) pregnancy to time Median µ Fertility incidence rate(95% CI) incidence rate(95% Fertility Gestational Gestational characteristics Gravidity (%) (%) Gravidity

Number of ANC visits (scheduled) (scheduled) visits ANC of Number Number of IPTp doses (%) IPTp of Number Number of ANC visits (unscheduled) (unscheduled) visits ANC of Number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 23 of 35 BMJ Open

1 2 3 Table 5.Preliminary data description of newborns at birth (subsample of the cohort). 4 The RECIPAL study (20142018). 5 6 Mean (± SD) or Newborn characteristics† N 7 Proportion (%) 8 9 Gender Male 205 55.2 10 11 Stillbirth Per 1000 live births 206 14.8 12 Preterm birth (< 37 weeks) Yes 207 12.6 13 Smallbirthweightforgestational age ‡§ Yes 191 11.5 14 ¥ § 15 SmallbirthweightforgestationalFor peer age reviewYes only191 18.3 16 Birthweight (g)§ 193 3043.9 (± 432.8) 17 < 2500 193 9.3 18 19 Poor birth outcome £,§ Yes 207 29.5 20 21 † Data based on the first 207 deliveries. §Twins excluded 22 ‡ Smallbirthweightforgestational age: < 10th percentile of birthweight for gestational age using Schmiegelow's charts. 23 ¥ Smallbirthweightforgestational age: < 10th percentile of birthweight for gestational age using; 24 £ Stillbirth, preterm birth, smallbirthweightforgestational age (using Intergrowth 21stcharts) or low birthweight 25 26 27 28

29 30 31 http://bmjopen.bmj.com/ 32 33 34 35

36 37 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 24 of 35

1 2 3 Funding 4 5 6 7 This study wassupported by the French National Research Agency (ANR, ANR13JSV1 8 9 0004, grant 2013) and the Fondation de France (n°00074147, grant 2017). 10 11 12 Author’s contribution 13 14 15 For peer review only 16 Project management: MA, NF, AM, MC and VB (principal investigator). 17 18 19 Field, Epidemiology and data collection: MA, EY, GC, GA, PA, NF and VB. 20 21 Nutritional data collection: YMP, AG, NFF, DD, MA and EY. 22 23 24 Biology and molecular analyses: NF, NTN and SH. 25 26 27 Obstetrical data collection: MA, EY, NJ and VB. 28 29 Statistical analysis: MA, GC, MC, JZ, FBL, YMP, AG, NFF, DD and VB 30 31 http://bmjopen.bmj.com/ 32 Manuscript writing: MA, GA, GC, AG, YMP, NFF, DD, JZ, NTM, FBL, SH, MC, NF and 33 34 VB. 35 36 37 38 Acknowledgements 39 40 on September 25, 2021 by guest. Protected copyright. 41 We are extremely grateful to all families who took part in this study, the midwifes, nurses and 42 43 44 communityhealth workers for recruiting and following them, and the whole RECIPAL team, 45 46 including research scientists, engineers, technicians, managers. We also thank SeunEgbinola 47 48 for study area mapping. MA was funded by the Réseau doctoral de l'Ecole des Hautes Etudes 49 50 en Santé Publique(EHESP) for PhD scholarship and received a prize from the Fondation des 51 52 53 Treilles (http://www.lestreilles.com/en/). 54 55 56 Conflict of interest.None declared 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 25 of 35 BMJ Open

1 2 3 References 4 5 6 1 WHO Malaria Policy Advisory Committee and Secretariat. Malaria Policy Advisory 7 8 Committee to the WHO: conclusions and recommendations of eighth biannual meeting 9 10 (September 2015). Malar J 2016;15:117. doi:10.1186/s129360161169x 11 12 13 14 2 Crawley J, Hill J, Yartey J, et al. From evidence to action? Challenges to policy change 15 For peer review only 16 and programme delivery for malaria in pregnancy. Lancet Infect Dis 2007;7:145–55. 17 18 doi:10.1016/S14733099(07)700269 19 20 21 3 Umbers AJ, Aitken EH, Rogerson SJ. Malaria in pregnancy: small babies, big problem. 22 23 Trends Parasitol 2011;27:168–75. doi:10.1016/j.pt.2011.01.007 24 25 26 27 4 Griffin JB, Lokomba V, Landis SH, et al.Plasmodium falciparum parasitaemia in the first 28 29 half of pregnancy, uterine and umbilical artery blood flow, and foetal growth: a 30 31

longitudinal Doppler ultrasound study. Malar J 2012;11:319. doi:10.1186/1475287511 http://bmjopen.bmj.com/ 32 33 319 34 35 36 37 5 Huynh BT, Cottrell G, Cot M, et al.Burden of malaria in early pregnancy: a neglected 38 39 problem? Clin Infect Dis 2015;60:598–604. doi:10.1093/cid/ciu848 40 on September 25, 2021 by guest. Protected copyright. 41 42 6 Valea I, Tinto H, Drabo MK, et al. An analysis of timing and frequency of malaria 43 44 infection during pregnancy in relation to the risk of low birth weight, anaemia and 45 46 perinatal mortality in Burkina Faso. Malar J 2012;11:71. doi:10.1186/147528751171 47 48 49 50 7 Cottrell G, Mary JY, Barro D, et al.The importance of the period of malarial infection 51 52 during pregnancy on birth weight in tropical Africa. Am J Trop Med Hyg 2007;76:849– 53 54 54. 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 26 of 35

1 2 3 8 Huynh BT, Fievet N, Gbaguidi G, et al. Influence of the timing of malaria infection 4 5 during pregnancy on birth weight and on maternal anemia in Benin. Am J Trop Med Hyg 6 7 2011;85:214–20. doi:10.4269/ajtmh.2011.110103 8 9 10 11 9 De Beaudrap P, Turyakira E, White LJ, et al. Impact of malaria during pregnancy on 12 13 pregnancy outcomes in a Ugandan prospective cohort with intensive malaria screening 14 15 and promptFor treatment. peer Malar J 2013; review12:139. doi:10.1186/1475287512139 only 16 17 18 10 KalilaniPhiri L, Thesing PC, Nyirenda OM, et al.Timing of malaria infection during 19 20 pregnancy has characteristic maternal, infant and placental outcomes. PloS One 21 22 23 2013;8:e74643. doi:10.1371/journal.pone.0074643 24 25 26 11 Djènontin A, BioBangana S, Moiroux N, et al. Culicidae diversity, malaria transmission 27 28 and insecticide resistance alleles in malaria vectors in OuidahKpomasseTori district 29 30 from Benin (West Africa): A preintervention study. Parasit Vectors 2010;3:83. 31 http://bmjopen.bmj.com/ 32 doi:10.1186/17563305383 33 34 35 36 12 Corbel V, N’Guessan R, Brengues C, et al.Multiple insecticide resistance mechanisms in 37 38 Anopheles gambiae and Culex quinquefasciatus from Benin, West Africa. Acta Trop 39 40 2007;101:207–16. doi:10.1016/j.actatropica.2007.01.005 on September 25, 2021 by guest. Protected copyright. 41 42 43 13 Briand V, Saal J, Ghafari C, et al.Fetal Growth Restriction Is Associated With Malaria in 44 45 46 Pregnancy: A Prospective Longitudinal Study in Benin. J Infect Dis 2016;214:417–25. 47 48 doi:10.1093/infdis/jiw158 49 50 51 14 Unger HW, OmeKaius M, Karl S, et al. Factors associated with ultrasoundaided 52 53 detection of suboptimal fetal growth in a malariaendemic area in Papua New Guinea. 54 55 BMC Pregnancy Childbirth 2015;15:83. doi:10.1186/s1288401505116 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 27 of 35 BMJ Open

1 2 3 15 Planche T, Krishna S, Kombila M, et al. Comparison of methods for the rapid laboratory 4 5 assessment of children with malaria. Am J Trop Med Hyg 2001;65:599–602. 6 7 8 16 Hofmann N, Mwingira F, Shekalaghe S, et al. Ultrasensitive detection of Plasmodium 9 10 11 falciparum by amplification of multicopy subtelomeric targets. PLoS Med 12 13 2015;12:e1001788. doi:10.1371/journal.pmed.1001788 14 15 For peer review only 16 17 Robinson HP, Fleming JE. A critical evaluation of sonar “crownrump length” 17 18 measurements. Br J Obstet Gynaecol 1975;82:702–10. 19 20 21 18 Papageorghiou AT, Sarris I, Ioannou C, et al. Ultrasound methodology used to construct 22 23 24 the fetal growth standards in the INTERGROWTH21st Project. BJOG Int J Obstet 25 26 Gynaecol 2013;120 Suppl 2:27–32, v. doi:10.1111/14710528.12313 27 28 29 19 Lohman T, Roche A, Martorell R. Anthropometric Standardisation Reference 30 31 Manual.Champaign, IL: Human Kinetics. 1998. http://bmjopen.bmj.com/ 32 33 34 20 Kramer MS. The epidemiology of adverse pregnancy outcomes: an overview. J Nutr 35 36 37 2003;133:1592S–1596S. 38 39 40 21 Ndao CT, Dumont A, Fievet N, et al. Placental malarial infection as a risk factor for on September 25, 2021 by guest. Protected copyright. 41 42 hypertensive disorders during pregnancy in Africa: a casecontrol study in an urban area 43 44 of Senegal, West Africa. Am J Epidemiol 2009;170:847–53. doi:10.1093/aje/kwp207 45 46 47 22 Persson LÅ, Arifeen S, Ekström EC, et al. Effects of prenatal micronutrient and early 48 49 50 food supplementation on maternal hemoglobin, birth weight, and infant mortality among 51 52 children in Bangladesh: the MINIMat randomized trial. JAMA 2012;307:2050–9. 53 54 doi:10.1001/jama.2012.4061 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 28 of 35

1 2 3 23 Moore KA, Simpson JA, Paw MK, et al.Safety of artemisinins in first trimester of 4 5 prospectively followed pregnancies: an observational study. Lancet Infect Dis 6 7 2016;16:576–83. doi:10.1016/S14733099(15)005472 8 9 10 11 24 TuikueNdam N, Deloron P. Developing vaccines to prevent malaria in pregnant women. 12 13 Expert Opin Biol Ther 2015;15:1173–82. doi:10.1517/14712598.2015.1049595 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 http://bmjopen.bmj.com/ 32 33 34 35 36 37 38 39 on September 25, 2021 by guest. Protected copyright. 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 29 of 35 BMJ Open

1 2 3 Figure legends 4 5 6 Figure 1. Geographical location of SôAva and AbomeyCalavi Districts, and of the 35 7 8 villages of the RECIPAL study, Southern Benin, 20142018. 9 10 11 Figure 2. Flowchart diagram of RECIPAL study. 12 13 * 24month followup without conceiving 14 15 For peer review only 16 Figure 3. Malaria status (infected vs. non infected) in women of reproductive age at inclusion 17 18 19 in RECIPAL (initial cohort), Southern Benin, 20142018. 20 21 22 Figure 4. KaplanMeier failure estimate of the probability of conceiving; 1214 women of 23 24 reproductiveage included in the initial RECIPAL cohort, Southern Benin, 20142018. 25 26 Probability of conceiving (solid line) and its 95% confidence interval (grey lines).Number of 27 28 pregnancy events are in brackets. Number of censored women at 5, 10, 15, 20 and 24 months 29 30 31 were : 223, 260, 88, 50 and 30, respectively. http://bmjopen.bmj.com/ 32 33 34 Supporting information 35 36 37 38 S1Fig. Ultrasound follow-up (RECIPAL study) 39 on September 25, 2021 by guest. Protected copyright. 40 HC: Head circumference; AC: Abdominal circumference; FL: Femur length; BPD: Biparietal 41 42 43 diameter; OFD: occipito frontal diameter; ATD: Abdominal transverse diameter; APAD: 44 45 Anteroposterior abdominal diameter. 46 47 * Uterine, and umbilical blood flows were measured at 2125 wg and from 28 wg, 48 49 respectively 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 30 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 http://bmjopen.bmj.com/ 32 33 34 35 36 Figure 1. Geographical location of Sô-Ava and Abomey-Calavi Districts, and of the 35 villages of the RECIPAL 37 study, Southern Benin, 2014-2018. 38 39 172x150mm (96 x 96 DPI) 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 31 of 35 BMJ Open

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Figure 2. Flowchart diagram of RECIPAL study. 30 * 24-month follow-up without conceiving 31 http://bmjopen.bmj.com/ 32 194x127mm (96 x 96 DPI) 33 34 35 36 37 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 32 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 http://bmjopen.bmj.com/ 32 33 34 35 Figure 3. Malaria status (infected vs. non infected) in women of reproductive age at inclusion in RECIPAL 36 (initial cohort), Southern Benin, 2014-2018. 37 154x128mm (96 x 96 DPI) 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 33 of 35 BMJ Open

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 http://bmjopen.bmj.com/ 32 33 Figure 4. Kaplan-Meier failure estimate of the probability of conceiving; 1214 women of reproductiveage 34 included in the initial RECIPAL cohort, Southern Benin, 2014-2018. Probability of conceiving (solid line) and its 95% confidence interval (grey lines).Number of pregnancy 35 events are in brackets. Number of censored women at 5, 10, 15, 20 and 24 months were : 223, 260, 88, 50 36 and 30, respectively. 37 38 152x117mm (96 x 96 DPI) 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 34 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 202x112mm (96 x 96 DPI) 29 30 31 http://bmjopen.bmj.com/ 32 33 34 35 36 37 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 35 of 35 BMJ Open

1 2 STROBE Statement—checklist of items that should be included in reports of observational studies 3 Item 4 No Recommendation 5 Title and abstract 1 (a) Indicate the study’s design with a commonly used term in the title or the abstract 6 7 (page 2) 8 (b) Provide in the abstract an informative and balanced summary of what was done 9 and what was found (Page 4) 10 11 Introduction 12 Background/rationale 2 Explain the scientific background and rationale for the investigation being reported 13 (page 6) 14 Objectives 3 State specific objectives, including any pre-specified hypotheses (page 6) 15 For peer review only 16 Methods 17 Study design 4 Present key elements of study design early in the paper (page 7) 18 Setting 5 Describe the setting, locations, and relevant dates, including periods of recruitment, 19 exposure, follow-up, and data collection (page 7, 8, 9, 10, 11) 20 21 Participants 6 (a) Cohort study—Give the eligibility criteria, and the sources and methods of 22 selection of participants. Describe methods of follow-up (page 8) 23 Case-control study—Give the eligibility criteria, and the sources and methods of 24 case ascertainment and control selection. Give the rationale for the choice of cases 25 26 and controls (Not applicable) 27 Cross-sectional study—Give the eligibility criteria, and the sources and methods of 28 selection of participants (Not applicable) 29 (b) Cohort study—For matched studies, give matching criteria and number of 30 exposed and unexposed (Not applicable) 31 http://bmjopen.bmj.com/ 32 Case-control study—For matched studies, give matching criteria and the number of 33 controls per case (Not applicable) 34 Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, and effect 35 modifiers. Give diagnostic criteria, if applicable (page 12, 13) 36 37 Data sources/ 8* For each variable of interest, give sources of data and details of methods of 38 measurement assessment (measurement). Describe comparability of assessment methods if there 39 is more than one group (Page 12, 13) on September 25, 2021 by guest. Protected copyright. 40 Bias 9 Describe any efforts to address potential sources of bias (page 15) 41 Study size 10 Explain how the study size was arrived at (page 7, 8) 42 43 Quantitative variables 11 Explain how quantitative variables were handled in the analyses. If applicable, 44 describe which groupings were chosen and why 45 (Not applicable for cohort profiles publication) 46 Statistical methods 12 (a) Describe all statistical methods, including those used to control for confounding 47 48 (Not applicable for cohort profiles publication) 49 (b) Describe any methods used to examine subgroups and interactions 50 (Not applicable for cohort profiles publication) 51 (c) Explain how missing data were addressed 52 53 (d) Cohort study—If applicable, explain how loss to follow-up was addressed 54 (page 10 and 14) 55 Case-control study—If applicable, explain how matching of cases and controls was 56 addressed (Not applicable) 57 Cross-sectional study—If applicable, describe analytical methods taking account of 58 59 sampling strategy (Not applicable) 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml1 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 36 of 35

1 2 (e) Describe any sensitivity analyses (Not applicable) 3 Results 4 5 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers potentially eligible, 6 examined for eligibility, confirmed eligible, included in the study, completing follow-up, and 7 analysed (page 10) 8 (b) Give reasons for non-participation at each stage (page 10) 9 (c) Consider use of a flow diagram (Figure 2) 10 11 Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, social) and 12 information on exposures and potential confounders (Table 4 ) 13 (b) Indicate number of participants with missing data for each variable of interest (page 10) 14 (c) Cohort study—Summarise follow-up time (eg, average and total amount) (Table 1 and 2) 15 For peer review only 16 Outcome data 15* Cohort study—Report numbers of outcome events or summary measures over time (page 14) 17 Case-control study—Report numbers in each exposure category, or summary measures of 18 exposure (Not applicable) 19 Cross-sectional study—Report numbers of outcome events or summary measures 20 21 (Not applicable) 22 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates and their 23 precision (eg, 95% confidence interval). Make clear which confounders were adjusted for 24 and why they were included (page 14, 15) 25 26 (b) Report category boundaries when continuous variables were categorized 27 (c) If relevant, consider translating estimates of relative risk into absolute risk for a 28 meaningful time period 29 Other analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and sensitivity 30 analyses (Not applicable) 31 http://bmjopen.bmj.com/ 32 Discussion (Not applicable for cohort profiles publication) 33 Key results 18 Summarise key results with reference to study objectives 34 Limitations 19 Discuss limitations of the study, taking into account sources of potential bias or imprecision. 35 36 Discuss both direction and magnitude of any potential bias 37 Interpretation 20 Give a cautious overall interpretation of results considering objectives, limitations, 38 multiplicity of analyses, results from similar studies, and other relevant evidence 39

on September 25, 2021 by guest. Protected copyright. 40 41 Generalisability 21 Discuss the generalisability (external validity) of the study results 42 43 Other information 44 Funding 22 Give the source of funding and the role of the funders for the present study and, if applicable, 45 46 for the original study on which the present article is based (page 24) 47 48 *Give information separately for cases and controls in case-control studies and, if applicable, for exposed and 49 unexposed groups in cohort and cross-sectional studies. 50 51 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and 52 published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely 53 available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 54 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is 55 available at www.strobe-statement.org. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml2 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open

Cohort profile: Effect of malaria in early pregnancy on fetal growth in Benin (RECIPAL pre-conceptional cohort).

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2017-019014.R1

Article Type: Cohort profile

Date Submitted by the Author: 25-Oct-2017

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 1 of 43 BMJ Open

1 2 3 Title 4 5 6 7 Cohort profile: Effect of malaria in early pregnancy on fetal growth in Benin (RECIPAL pre 8 9 conceptional cohort). 10 11 12 Authors 13 14 15 Manfred Accrombessi,For1,2 peer Emmanuel Yovo, review2 Gilles Cottrell,1 Ginoonly Agbota,1,2 Agnès Gartner,3 16 17 Yves MartinPrevel,3 Nadia FanouFogny,4 Diane Djossinou,3,4 JenniferZeitlin,5 18 19 1 5,6 1,7 8 20 NicaiseTuikueNdam, Florence BodeauLivinec, Sandrine Houzé, Nicola Jackson, Paul 21 22 Ayemonna,9 Achille Massougbodji,2 Michel Cot,1 Nadine Fievet,1 ValérieBriand1* 23 24 25 Affiliations 26 27 28 1 UMR216MERIT, French National Research Institute for Sustainable Development (IRD), 29 30 31 Université Paris Descartes, Paris, France. http://bmjopen.bmj.com/ 32 33 2 Centre d'Etude et de Recherche sur le Paludisme Associé à la Grossesse et à l'Enfance 34 35 36 (CERPAGE), Cotonou, Benin. 37 38 3 UMR204Nutripass, French National Research Institute for Sustainable Development (IRD), 39 on September 25, 2021 by guest. Protected copyright. 40 Université de Montpellier, SupAgro, Montpellier, France. 41 42 43 4 Ecole de Nutrition et des Sciences et Technologies Alimentaires (ENSTA), Faculté des 44 45 Sciences Agronomiques, Université d’AbomeyCalavi, Benin. 46 47 48 5 Obstetrical, Perinatal and Pediatric Epidemiology Research Team, Centre for Epidemiology 49 50 and Biostatistics (U1153EPOPé), National Institute for Health and Medical Research 51 52 53 (INSERM), Université Paris Descartes, Paris, France. 54 55 6 Département Méthodes Quantitatives en Santé Publique, Ecole des Hautes Etudes en Santé 56 57 Publique (EHESP), Université de Rennes, France. 58 59 60 1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 3 of 43 BMJ Open

1 2 7 3 Laboratoire de Parasitologie, CNR du Paludisme, APHP, Hôpital Bichat, Paris, France. 4 5 8 6 Oxford University, United Kingdom. 7 8 9 Service de gynécologie et obstétrique, Centre Hospitalier Universitaire d’AbomeyCalavi, 9 10 11 AbomeyCalavi, Benin. 12 13 14 * Correspondingauthor:[email protected] 15 For peer review only 16 17 Dr Valérie Briand 18 19 MERITUMR216/IRD « Mother and child face to tropical infections » 20 21 Faculté des sciences pharmaceutiques Paris Descartes 22 4 avenue de l'Observatoire, 75006 Paris 23 24 Email : [email protected] 25 26 Tel:+33153731527 27 28 29 30 31 http://bmjopen.bmj.com/ 32 33 34 35 36 37 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 4 of 43

charts) and preterm birth was 9.3%, 18.3%, and 12.6%, respectively. on September 25, 2021 by guest. Protected copyright. 40 41 42 Future plans: RECIPALrepresents at this time a unique resource that will provide 43 44 45 information on multiple infectious (including malaria), biological, nutritional and 46 47 environmental determinants in relation to health outcomes in women of reproductive age, 48 49 pregnant women and their newborns. It will contribute to better define future 50 51 recommendations for the prevention of malaria in early pregnancy and maternal malnutrition 52 53 in Africa. It confirms that it is possible to constitute a preconceptional pregnancy cohort in 54 55 56 Africa and provides valuable information for researchers starting cohorts in the future. 57 58 59 60 3 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 5 of 43 BMJ Open

1 2 3 4 5 6 Strengths and limitations of this study 7 8 9 • A unique cohort combining highly detailed, highquality health 10 related information on the pregnancies of ≈ 400 women recruited 11 12 in the preconception period, with a substantial related biobank 13 of plasmas, placental and other samples. 14 15 For peer review only 16 • Preconceptional design allowing the early identification and 17 followup of pregnant women: screening of women for both 18 19 microscopic and submicroscopic malaria from the first weeks of 20 pregnancy;accurate estimation of gestational ageby using early 21 22 obstetrical ultrasound scan andcollectionofsome important factors 23 influencing fetal growth such as prepregnancy nutritional status 24 25 and gestational hypertensive disorders. 26 27 • Collection of valuable information for the implementation of 28 29 future preconceptional cohorts in Africa. 30 31 • High attrition in the preconceptional cohort, with a possible http://bmjopen.bmj.com/ 32 33 impact on the external validity of some findings. 34 35 • Reduced sample size at delivery due to a high proportion of 36 37 spontaneous abortion, with a possible lack of power for analyses 38 on birth outcomes. 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 6 of 43

(LBW). http://bmjopen.bmj.com/ 32 33 34 Few studies have investigated malaria in early pregnancy and while some have shown an 35 36 adverse effect on birth outcomes, this has not been consistent [5]. In Burkina Faso and in 37 38 39 Benin, malaria infection in the first trimester [6] or before 45 months of pregnancy [7,8] has 40 on September 25, 2021 by guest. Protected copyright. 41 been associated with a higher risk of LBW or a decrease in birthweight. Conversely, such 42 43 association was not reported in Uganda and Malawi [9,10] but the number of women screened 44 45 in the first trimester was low in these two studies. One of the problems in interpreting this 46 47 existing literature relates to the challenges of recruiting women early in pregnancy; most of 48 49 50 the studies were not designed to assess the consequences of malaria early in pregnancy and 51 52 therefore women were generally recruited late in the first trimester, which may lead to 53 54 misclassification errors and underestimation of exposure. Besides, using LBW as a proxy for 55 56 fetal growth restriction (FGR) can lead to the overestimation of FGR prevalence and possible 57 58 59 60 5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 7 of 43 BMJ Open

1 2 3 bias when estimating the effect of malaria. Finally, maternal nutrition is one of the main 4 5 factors influencing fetal growth in developing countries [11]. Therefore, adjusting for this 6 7 factor when assessing the effect of early malaria on FGR is important. As well, a recent study 8 9 10 has suggested a higher risk of FGR due to malaria in undernourished women [12]. 11 12 13 The RECIPAL study aimed to assess the effect of malaria (both microscopic and 14 15 submicroscopic)For in the firstpeer trimester of reviewpregnancy on both the only mother and the fetus, with a 16 17 focus on fetal growth, by following a cohort of pregnant women recruited before conception. 18 19 Moreover, it also aimed at assessing the influence of the woman’s nutritional status in the 20 21 relationship between malaria in early pregnancy and birth outcomes. 22 23 24 25 Cohort description 26 27 28 29 Study setting 30 31 http://bmjopen.bmj.com/ 32 The study cohort is currently conducted in the districts of SôAva and AbomeyCalavi, 33 34 located in Southern Benin (Figure 1). Women's recruitment has been completed by December 35 36 2016, and pregnancy followup is still ongoing. In the area, malaria is hyperendemic with a 37 38 mean entomological inoculation rate of 2.1 infected anopheles bites/person/100 nights [13]. 39 40 on September 25, 2021 by guest. Protected copyright. The main mosquito vectors of malaria are Anopheles gambiae ss and Anopheles funestus [14]. 41 42 43 Four subdistricts (SôAva, Vekky and Houedo in SôAva District, and Akassato in Abomey 44 45 Calavi District) were selected for the study according to the population density and mean 46 47 number of ANC visits and deliveries per month in the four main public maternity clinics of 48 49 SôAva and AbomeyCalavi districts. In these four subdistricts, the population size was 50 51 52 estimated to 124,994 people in 2013, including 17.4% (21,779) women of reproductive age 53 54 (WRA) [15]. Thirtyfive out of a total of 36 villages were then selected for the study 55 56 according to their proximity to the maternity clinics. 57 58 59 60 6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 8 of 43

1 2 3 Study design, subject identification, recruitment and enrolment procedures 4 5 6 Briefly, WRAs were recruited at communitylevel and followed monthly for a maximum 7 8 period of 24 months until becoming pregnant, they constituted the RECIPAL initial cohort. 9 10 The subsample of women who became pregnant was then followedup monthly at the 11 12 13 maternity clinic from early pregnancy to delivery; they constituted the RECIPAL final cohort. 14 15 For peer review only 16 A sample size of 466 births was estimated sufficient to evidence a 2fold odds ratio for the 17 18 association between malaria and poor birth outcomes [4], assuming a prevalence of 25% of 19 20 women with microscopic malaria in the 1st trimester of pregnancy and 30% of poor birth 21 22 outcomes (LBW, smallforgestational age (SGA), preterm birth (PTB) or stillbirth) [8,16,17] 23 24 25 with a 80% power, significant level at 0.05. We estimated that following 2,000 WRA (initial 26 27 cohort) would have allowed identifying 510 pregnant women (final cohort) during a planned 28 29 18 monthperiod for recruitment in the initial cohort (based on a total fertility rate of 30 31 200/1,000/year in Benin (http://esa.un.org/unpd/wpp/index.htm) and allowing 15% of loss to http://bmjopen.bmj.com/ 32 33 followup after recruitment in the initial cohort. 34 35 36 After explanation of the study to the local political and administrative authorities, three 37 38 39 concomitant procedures were used to recruit WRA. First, repeated sensitization events were 40 on September 25, 2021 by guest. Protected copyright. 41 organized in each of the 35 selected villages for presenting the study to all inhabitants. If 42 43 interested, women were invited to register with the head of the community and were visited at 44 45 home the day after for their inclusion. That procedure constituted the main way of recruitment 46 47 48 throughout the study. Second, leaders of the community, religious authorities and massmedia 49 50 were involved in order to help us mobilizing women during public meetings (masses, 51 52 women’s association meetings, etc.). The last way of recruitment consisted of going doorto 53 54 door to identify eligible and interested women with the assistance a network of community 55 56 health workers (more details are provided in Supplemental file S1). 57 58 59 60 7 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 9 of 43 BMJ Open

1 2 3 To be included, women had met the following criteria: negative urinary pregnancy test, 18 to 4 5 45 years old, no current contraception, no previous fecundity issues, willingness to become 6 7 pregnant, no planned travel for more than 2 months within the next 18 months, acceptance of 8 9 10 RECIPAL protocol and signed written informed consent. Both oral and written 11 12 communication was used for providing information on the project to the women and their 13 14 husbands. The consent form was translated into the local language for women who were 15 For peer review only 16 illiterate. 17 18 19 From June 2014 to December 2016, 1302 WRA were willing to participate in the study. 20 21 Among them, 1214 (93.2%) met the inclusion criteria and were recruited, 88 (6.8%) were not 22 23 24 included mainly because of a positive urinary pregnancy test at inclusion (63.6%) or past 25 26 infertility issues (14.8%). Based on both the 2013 national census and the 20112012 27 28 Demographic Health Survey (DHSIV) in Benin, we estimated that women included in the 29 30 project represented 6% of the total number of WRA living in the study area. Compared with 31 http://bmjopen.bmj.com/ 32 33 women included in Dansou et al. study [18], which used individual data of a representative 34 35 sample of WRA included in the Beninese DHSIV, WRA included in RECIPAL had a higher 36 37 level of poverty. Age, education level, household wealth, gravidity and media exposure were, 38 39 however, similar (Supplemental file S2). 40 on September 25, 2021 by guest. Protected copyright. 41 42 RECIPAL study was approved by the ethics committees of the Institut des Sciences 43 44 45 Biomédicales Appliquées (ISBA) in Benin and from the IRD in France as well as by the 46 47 Ministry of Health in Benin. 48 49 50 Cohort follow-up 51 52 53 Pre-conceptional follow-up 54 55 56 After enrolment, WRAs were visited at home monthly by specifically trainedinvestigators 57 58 assisted by local communityhealth workers. Followup consisted mainly of recording the first 59 60 8 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 10 of 43

1 2 3 day of last menstrual period (LMP) and performing a urinary pregnancy test every month. To 4 5 minimize the risk of loss of confidentiality during the tracking of women for followup, the 6 7 study team was instructed to speak only to people—including family members—to whom 8 9 10 women have granted them permission to speak. After 12 months, women who were still 11 12 followed received medical advices to help conceive and they were invited to attend the 13 14 maternity clinic for a medical examination in case of symptoms of genital infection. After 24 15 For peer review only 16 months, the followup ended and women who did not conceive were invited to attend a 17 18 gynaecologist for fertility issues assessment. More details on the followup procedures have 19 20 21 been provided in Supplemental File S3. 22 23 24 Gestational follow-up 25 26 Once pregnant, women were followedup at the maternity clinic for monthly ANC visits. In 27 28 29 addition, they were encouraged to attend the maternity clinic any time in case of any 30 31 symptoms. Both the maternity staff (for usual ANC followup and clinical examination) and http://bmjopen.bmj.com/ 32 33 study investigators (for all specific issues related to RECIPAL) were involved in women’s 34 35 followup. Throughout the study, women were offered free transport to the maternity clinic; in 36 37 38 SôAva District a free river shuttle was set up. At 37 weeks gestation, women were visited at 39 40 home weekly and transportation to the study maternity clinic was scheduled for delivery. on September 25, 2021 by guest. Protected copyright. 41 42 43 According to Beninese national recommendations, a kit including iron and folic acid tablets 44 45 for one month, mebendazole for deworming, the 1st dose of SP for IPTp and a LLITN was 46 47 given to each pregnant woman at the first ANC visit. The maternity staff was encouraged to 48 49 administer at least three doses of IPTp from the 2nd trimester onwards as recommended by the 50 51 52 Beninese National Malaria Control Program, as well as iron and folic acid supplementation 53 54 throughout the pregnancy. RECIPAL women benefited from free management of any diseases 55 56 related to the pregnancy—detected either as part of RECIPAL followup or during 57 58 59 60 9 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 11 of 43 BMJ Open

1 2 3 unscheduled visits, free ANC visits and delivery. Women infected with malaria were treated 4 5 with quinine in the first trimester of pregnancy and artemetherlumefantrine from the 2nd 6 7 trimester. 8 9 10 Infant follow-up 11 12 13 The infant followup was not initially planned as part of the RECIPAL project. It is currently 14 15 carried out asFor part of the peerSEPSIS project review (bioMérieuxfunding, only UMR216/9198 coordination), 16 17 18 which aims to assess immune dysfunctions associated with the risk of sepsis in preterm and 19 20 growthrestricted newborns as well as in those exposed to malaria in utero. All infants born 21 22 from the RECIPAL women from April 2016 have been included and followed until 3 months 23 24 of age. 25 26 27 Cohort attrition 28 29 30 Figure 2 presents the flowchart diagram of the study. Among the 1214 WRA enrolled in the 31 http://bmjopen.bmj.com/ 32 initial cohort, 411 (33.9%) became pregnant, 359 (29.6%) completed the followup without 33 34 conceiving and 444 women (36.6%) did not complete the study mainly because of informed 35 36 37 consent withdrawal (272/444) or migration outside the study area (138/444) that occurred 38 39 after 5.2 and 6.3 months of followup in average, respectively. 40 on September 25, 2021 by guest. Protected copyright. 41 42 Among the 411 pregnant women, 207 (50.4%) have already delivered, 16.5% (68/411) had a 43 44 spontaneous abortion (n=68) or a nonviable pregnancy (n=2), 62 did not complete the study 45 46 because of informed consent withdrawal (45/411, 10.9%) or migration (17/411, 4.1%), and 72 47 48 49 women are still followed (Figure 2). Spontaneous abortions, informed consent withdrawals 50 51 and migrations occurred at 8.9 (± 3.9), 13.2 (± 8.1) and 14.2 (± 8.6) weeks of gestation (wg) 52 53 in average (SD), respectively. More details on the reasons for cohort attrition are provided in 54 55 Supplemental file S4. 56 57 58 59 60 10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 12 of 43

1 2 3 Data collection 4 5 6 Pre-conceptional data 7 8 9 Demographic, socioeconomic and household characteristics were collected at inclusion (Table 10 11 1). At this occasion, screening for malaria, urinary schistosomiasis, Chlamydia trachomatis 12 13 and Neisseria gonorrhoeae was performed. Haemoglobin (Hb) and markers of inflammation 14 15 levels were Fordetermined. peerThe first day ofreview LMP was recorded, only and a urinary pregnancy test 16 17 18 (OneStep®, International Holding Corp, Germany) was performed each month. 19 20 Anthropometric measurements were collected every three months and dietary intakes were 21 22 assessed twice during followup. 23 24 25 Gestational data 26 27 28 At each ANC visit, clinical, anthropometric and obstetrical data, as well as LLITN use, IPTp 29 30 administration, alcohol consumption, and iron, folic acid and other micronutrient 31 http://bmjopen.bmj.com/ 32 supplementation intake were collected (Table 2). In addition, malaria screening was 33 34 performed, and proteinuria, glycosuria and urinary infection were detected using a urine 35 36 37 dipstick test (Combiscreen®, Analyticon, Germany). Dietary intakes were assessed once 38 39 every trimester. Five Dopplerultrasound scans (US) were performed. A sample of venous 40 on September 25, 2021 by guest. Protected copyright. 41 blood was collected in the 1st and 3rd trimesters for Hb, markers of inflammation and lead 42 43 levels determination. 44 45 46 At delivery, newborns were weighed within 1 hour after birth on an electronic digital scale 47 48 49 with an accuracy of 2 grams (SECA 757, SECA Ltd., Germany). Newborn’s length (SECA 50 51 infantometer 416, Germany) and head circumference (SECA 201, Germany) were recorded to 52 53 the nearest millimetre. Malaria screening was performed on maternal venous blood, placental 54 55 and cord blood (Table 2). A placental biopsy was performed for malaria histopathology. 56 57 Followup and quality control procedures are provided in Supplemental file S3. 58 59 60 11 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 13 of 43 BMJ Open

1 2 3 The types and origins of the biological samples collected before and during pregnancy are 4 5 listed in Table 3. Total blood, serum, plasma urine samples are stored in 20°C or 80°C 6 7 freezers with a daily monitoring of the temperature. 8 9 10 Specifics components 11 12 13 Malaria diagnosis. Before conception and during pregnancy, malaria screening was 14 15 performed usingFor both microscopy peer (thick reviewblood smear) and polymerase only chain reaction (PCR). 16 17 18 Blood smears were stained with Giemsa and parasitaemia was quantified by the Lambaréné 19 20 method [19]. A molecular diagnostic approach using a combination of realtime PCR assays 21 22 comprising genusspecific primers and probes for the gene encoding the small (18S) of 23 24 Plasmodium rRNA and an ultrasensitive P. falciparumspecific detection system [20] were 25 26 used to screen samples containing Plasmodium parasites. In addition, any time during 27 28 29 pregnancy, a rapid diagnostic test (Pf + pan rapid test SD Bioline Ag®, IDA Foundation, 30 31 Netherlands; BioSynex®, France) was performed in case of symptoms suggestive of malaria http://bmjopen.bmj.com/ 32 33 for immediate diagnosis and treatment. 34 35 36 Ultrasound follow-up. The first US for dating the pregnancy was performed between 9 and 37 38 13 wg based on LMP recorded during the preconceptional followup (supplementary file S5). 39 40 on September 25, 2021 by guest. Protected copyright. 41 The gestational age (GA) was estimated in days based on crownrump length measurement 42 43 using the Robinson’s chart [21]. At the end, GA estimation was based either on LMP if the 44 45 difference between the two measurements (LMP/US) was less than 7 days or on US if the 46 47 difference is > 7 days. Following INTERGROWTH21st methodology, four additional scans 48 49 were performed for fetal biometry assessment, and uterine and umbilical blood flow 50 51 52 measurements [22]. Each parameter was measured twice and then averaged; a third measure 53 54 was performed in case of discrepancy between the two first measurements. A quality control 55 56 was performed on 10% of the pictures. 57 58 59 60 12 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 14 of 43

1 2 3 Anthropometric measurements and dietary intake assessment. Anthropometric 4 5 measurements were collected using standard procedures (Tables 1 and 2) [23]. During 6 7 household visits, women’s body weight was measured with a 200 g precision with calibrated 8 9 10 electronic scales (Tefal, France). At the facilitylevel, women were weighed with a Tanita 11 12 MC780 body composition device (Tanita Corporation, Tokyo, Japan). Height was measured 13 14 to the nearest millimeter with a SECA 206 (Hamburg, Germany) gauge. The left midupper 15 For peer review only 16 armcircumference was measured with a SECA 201 (Hamburg, Germany) ergonomic 17 18 circumference measuring tape. Skinfold thickness measurements were made to the nearest 19 20 21 millimeter at the triceps and biceps sites using a Holtain (Crymyeh, UK) skinfold caliper. The 22 23 set of measurements was repeated twiceby the same field investigatorand then the 24 25 measurements were averaged. Body composition, assessed by bioelectrical impedance 26 27 analysis method, was measured by the multifrequency body composition analyser Tanita 28 29 30 MC780 with a fourelectrode arrangement paired on hands and feet. A quality control of 31 http://bmjopen.bmj.com/ 32 anthropometric data was performed periodically by a senior research scientist in nutrition. 33 34 35 Dietary intakes were assessed by 24hrecalls using the standard multiplepass methodology 36 37 [24]. A comprehensive list of the most common food items and recipes was developed from a 38 39

preliminary study of the diet habits of the study area. This list was used to identify foods on September 25, 2021 by guest. Protected copyright. 40 41 42 mentioned by the women as consumed during the interview. Amounts of food consumed were 43 44 estimated by household measures consisted of precalibrated utensils, food pictures of 45 46 different portion sizes, monetary value of known food portions sold on the study area markets. 47 48 A food composition table and recipes database are under construction for conversion of 49 50 51 household units to gram, and to determine macro and micronutrients intake by the women. 52 53 54 Characteristics of the study populations 55 56 57 58 59 60 13 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 15 of 43 BMJ Open

1 2 3 At enrolment, WRA were 27 years old in average (Table 4). Nulligravidae represented 8.8% 4 5 of the cohort. More than half of the women were anaemic, one quarter was infected with 6 7 schistosomiasis and 38% had an abnormal BMI. The prevalence of microscopic malaria was 8 9 10 5.7%, with two geographical clusters in SôAva and Vekky subdistricts (Figure 3). The 11 12 median duration of followup in the initial cohort was 5.9 months (mean=7.1, range: 0.49– 13 14 23.7). Compared to women included in the initial cohort, women who did not complete the 15 For peer review only 16 followup were significantly different in terms of residence area, ethnicity, household density 17 18 and schistosomiasis status (Table 4). 19 20 21 Pregnant women were identified at a mean of 6.9 wg and benefited from 7.5 scheduled ANC 22 23 24 visits in average (Table 4). The prevalence of microscopic malaria infection was 25.4%, 25 26 19.4% and 16.1% during the 1st, 2nd and 3rd trimester of pregnancy, respectively. The 27 28 prevalence of SGA was 11.5% and 18.3% using Schmiegelow’s and INTERGROWTH 21st 29 30 charts, respectively (Table 5). The proportion of PTB and LBW was 12.6% and 9.3%, 31 http://bmjopen.bmj.com/ 32 33 respectively. Overall, 29.5% (61/207) of newborns presented at least one poor birth outcome 34 35 defined as LBW, PTB, SGA or stillbirth. 36 37 38 Findings to date 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 Fertility and time-to-pregnancy 43 44 45 The median timetopregnancy was 12.3 months (Figure 4). The fertility rate was 5.4 46 47 pregnancies/100 personsmonth (95% confidence interval [CI] =4.95.9). The probability 48 49 (95% CI) of conceiving after 5, 10, 15 and 20 months of followup were 24.5% (21.927.4), 50 51 42.1% (38.645.9), 56.6% (52.161.1) and 62.8% (57.668), respectively. Using multivariate 52 53 survival analysis, factors associated with a lower risk of conceiving were maternal age more 54 55 56 than 35 years, a low education level, being polygamous, living in Vekky subdistrict, long 57 58 59 60 14 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 16 of 43

1 2 3 term couple (> 8 years), infrequent sexual activity, overweight and urinary schistosomiasis 4 5 infection (Yovo, personal communication). 6 7 8 Malaria in the first trimester of pregnancy 9 10 11 The incidence rate of malaria infection during the first trimester was 19.7 cases per 100 12 13 personsmonth (95% CI 15.824.5). Using a multilevel logistic regression, we showed that 14 15 women infectedFor with malaria peer before conception review were more likely only to be infected during the 1st 16 17 18 trimester (aOR: 2.68, 95% CI 1.24, 5.78). Gestational age was also negatively correlated with 19 20 malaria infection (aOR: 0.64, 95% CI 0.41, 0.98) (Accrombessi, personal communication). 21 22 23 Strengths and limitations 24 25 26 27 The preconceptional design—and early identification of pregnant women—allowed (i) the 28 29 screening of women for malaria from the very beginning of pregnancy; (ii) the accurate 30 31 estimation of GA by using US before 14 wg; (iii) the determination of prepregnancy http://bmjopen.bmj.com/ 32 33 nutritional status [25] and hypertensive disorders screening [26], which highly influence fetal 34 35 36 growth; and (iv) the minimization of selection bias embedded in the design of women 37 38 recruited at facilitylevel. Besides, PTB and IUGR could be accurately assessed thanks to the 39 40 precise estimation of GA and longitudinal fetal growth assessment. on September 25, 2021 by guest. Protected copyright. 41 42 43 In addition to malaria, RECIPAL will provide valuable information on WRA in Africa in 44 45 terms of fertility and nutrition, which are seldom evaluated and may help to define new pre 46 47 48 conceptional interventions for improving maternal and child health [27]. Also, it yields 49 50 logistical and ethical information for the implementation of future preconceptional cohorts. 51 52 53 Only107 (8%) nulligravidae were included in the study. The main reason is that RECIPAL 54 55 was implemented shortly after a study, which has recruited nulligravid WRAs living in the 56 57 same area. This study aimed at assessing maternal immune responses in early pregnancy in 58 59 60 15 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 17 of 43 BMJ Open

1 2 3 order to help designing a phase 1 vaccine trial against malaria. It included 278 nulligravid 4 5 WRAs, of whom 60 became pregnant [28]. Both studies were conducted concomitantly for 6 7 several months and women already included in the first study could not be eligible for 8 9 10 RECIPAL. Because primigravidae were underrepresented in the final RECIPAL cohort, it is 11 12 likely that the prevalence of maternal conditions related to primigravidity were 13 14 underestimated. Although we cannot completely exclude selection bias, it should not have 15 For peer review only 16 strongly influenced the association between malaria in early pregnancy and FGR. 17 18 19 Finally, because of constraints related to the study design (long duration of the pre 20 21 conceptional followup and weariness of the women) and African sociocultural realities 22 23 24 (rumour about blood collection, suspicion regarding medical care provided free of charge, 25 26 etc.), a noticeable proportion (22%) of women withdrew their informed consent before the 27 28 end of the 24 monthfollowup. Overall, the initial cohort attrition was 37%. Since some 29 30 baseline characteristics in women who completed the preconceptional followup and those 31 http://bmjopen.bmj.com/ 32 33 who did not were different, this might impact the external validity of some of results. During 34 35 pregnancy, there was a high proportion (16.5%) of spontaneous abortions, which contributed 36 37 to the final cohort attrition. At the end, birth outcomes could be evaluated in only 68% of 38 39 women, with a possible lack of power for some upcoming analyses. 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 Conclusion 44 45 46 If the deleterious effects of malaria in the first trimester of pregnancy are confirmed, our 47 48 49 results may argue in favor of starting preventive measures from the very beginning of 50 51 pregnancy or even before pregnancy. Artemisininbased Combination Therapies (ACTs) have 52 53 been shown to be safe during the first trimester of pregnancy and, therefore, may be a good 54 55 option for the replacement of SP [29]. Besides, a vaccine against pregnancyassociated 56 57 malaria parasites, which could elicit protective immunity prior to pregnancy to best protect 58 59 60 16 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 18 of 43

1 2 3 pregnant women during early pregnancy may be proposed as a complementary strategy. Such 4 5 a vaccine is currently under evaluation [30]. 6 7 8 9 Collaboration 10 11 12 The “Mother and child face to tropical infections” research unit (MERIT), French National 13 14 Research Institute for Sustainable Development (IRD)/Paris Descartes University was the 15 For peer review only 16 17 promoter of the study. RECIPAL was a collaborative project between MERIT, another IRD 18 19 research unit (Nutripass, Montpellier), EPOPé research team (Inserm), the Ecole des Hautes 20 21 Etudes en Santé Publique (EHESP) and two Beninese collaborators from Abomey 22 23 CalaviUniversity (the Faculty of Health Sciences and the Faculty of Agronomy Sciences). 24 25 26 The researchers associated with this study are open for the sharing and reuse of the data but an 27 28 29 end user data use agreement will be required for accessing the data. Collaborations are 30 31 encouraged, although data sets are not currently publicly available. Any researcher interested http://bmjopen.bmj.com/ 32 33 in exploring RECIPAL data should contact directly the principal investigator, Valerie Briand 34 35 (email: [email protected]). 36 37 38 39 Funding 40 on September 25, 2021 by guest. Protected copyright. 41 42 This study wassupportedby the French National Research Agency (ANR, ANR13JSV1 43 44 45 0004, grant 2013) and the Fondation de France (n°00074147, grant 2017). 46 47 48 Author’s contribution 49 50 51 52 Project management: MA, NF, AM, MC and VB (principal investigator). 53 54 Field, Epidemiology and data collection: MA, EY, GC, GA, PA, NF and VB. 55 56 57 Nutritional data collection: YMP, AG, NFF, DD, MA and EY. 58 59 60 17 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 19 of 43 BMJ Open

1 2 3 Biology and molecular analyses: NF, NTN and SH. 4 5 6 Obstetrical data collection: MA, EY, NJ and VB. 7 8 Statistical analysis: MA, GC, MC, JZ, FBL, YMP, AG, NFF, DD and VB 9 10 11 Manuscript writing: MA, GA, GC, AG, YMP, NFF, DD, JZ, NTM, FBL, SH, MC, NF and 12 13 VB. 14 15 For peer review only 16 17 Acknowledgements 18 19 20 We are extremely grateful to all families who took part in this study, the midwifes, nurses and 21 22 23 communityhealth workers for recruiting and following them, and the whole RECIPAL team, 24 25 including research scientists, engineers, technicians, managers. We also thank Seun Egbinola 26 27 for study area mapping. MA was funded by the Réseau doctoral de l'Ecole des Hautes Etudes 28 29 en Santé Publique (EHESP) for PhD scholarship and received a prize from the Fondation des 30 31 Treilles (http://www.lestreilles.com/en/). http://bmjopen.bmj.com/ 32 33 34 35 Conflict of interest. None declared 36 37 38 39 References 40 on September 25, 2021 by guest. Protected copyright. 41 42 1 WHO Malaria Policy Advisory Committee and Secretariat. Malaria Policy Advisory 43 44 Committee to the WHO: conclusions and recommendations of eighth biannual meeting 45 46 (September 2015). Malar J 2016;15:117. doi:10.1186/s129360161169x 47 48 49 2 Crawley J, Hill J, Yartey J, et al. From evidence to action? Challenges to policy change 50 51 and programme delivery for malaria in pregnancy. Lancet Infect Dis 2007;7:145–55. 52 53 doi:10.1016/S14733099(07)700269 54 55 56 3 Umbers AJ, Aitken EH, Rogerson SJ. Malaria in pregnancy: small babies, big problem. 57 58 Trends Parasitol 2011;27:168–75. doi:10.1016/j.pt.2011.01.007 59 60 18 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 20 of 43

1 2 3 4 Griffin JB, Lokomba V, Landis SH, et al.Plasmodium falciparum parasitaemia in the first 4 5 half of pregnancy, uterine and umbilical artery blood flow, and foetal growth: a 6 7 longitudinal Doppler ultrasound study. Malar J 2012;11:319. doi:10.1186/1475287511 8 9 10 319 11 12 5 Huynh BT, Cottrell G, Cot M, et al.Burden of malaria in early pregnancy: a neglected 13 14 problem? Clin Infect Dis 2015;60:598–604. doi:10.1093/cid/ciu848 15 For peer review only 16 17 6 Valea I, Tinto H, Drabo MK, et al. An analysis of timing and frequency of malaria 18 19 infection during pregnancy in relation to the risk of low birth weight, anaemia and 20 21 22 perinatal mortality in Burkina Faso. Malar J 2012;11:71. doi:10.1186/147528751171 23 24 7 Cottrell G, Mary JY, Barro D, et al.The importance of the period of malarial infection 25 26 27 during pregnancy on birth weight in tropical Africa. Am J Trop Med Hyg 2007;76:849– 28 29 54. 30 31 http://bmjopen.bmj.com/ 32 8 Huynh BT, Fievet N, Gbaguidi G, et al. Influence of the timing of malaria infection 33 34 during pregnancy on birth weight and on maternal anemia in Benin. Am J Trop Med Hyg 35 36 2011;85:214–20. doi:10.4269/ajtmh.2011.110103 37 38 39 9 De Beaudrap P, Turyakira E, White LJ, et al. Impact of malaria during pregnancy on 40 on September 25, 2021 by guest. Protected copyright. 41 pregnancy outcomes in a Ugandan prospective cohort with intensive malaria screening 42 43 and prompt treatment. Malar J 2013;12:139. doi:10.1186/1475287512139 44 45 46 10 KalilaniPhiri L, Thesing PC, Nyirenda OM, et al.Timing of malaria infection during 47 48 pregnancy has characteristic maternal, infant and placental outcomes. PloS One 49 50 2013;8:e74643. doi:10.1371/journal.pone.0074643 51 52 53 11 Osrin D, de L Costello AM. Maternal nutrition and fetal growth: practical issues in 54 55 international health. Semin Neonatol SN 2000;5:209–19. doi:10.1053/siny.2000.0024 56 57 58 59 60 19 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 21 of 43 BMJ Open

1 2 3 12 Cates JE, Unger HW, Briand V, et al. Malaria, malnutrition, and birthweight: A meta 4 5 analysis using individual participant data. PLoS Med 2017;14:e1002373. 6 7 doi:10.1371/journal.pmed.1002373 8 9 10 13 Djènontin A, BioBangana S, Moiroux N, et al. Culicidae diversity, malaria transmission 11 12 and insecticide resistance alleles in malaria vectors in OuidahKpomasseTori district 13 14 from Benin (West Africa): A preintervention study. Parasit Vectors 2010;3:83. 15 For peer review only 16 17 doi:10.1186/17563305383 18 19 14 Corbel V, N’Guessan R, Brengues C, et al.Multiple insecticide resistance mechanisms in 20 21 22 Anopheles gambiae and Culex quinquefasciatus from Benin, West Africa. Acta Trop 23 24 2007;101:207–16. doi:10.1016/j.actatropica.2007.01.005 25 26 27 15 Institut National de la Statistique et de l’Analyse Economique. Fourth (2013) General 28 29 Census of Population and Habitation in Benin, 2015. 30 31 http://bmjopen.bmj.com/ 32 16 Briand V, Saal J, Ghafari C, et al. Fetal Growth Restriction Is Associated With Malaria in 33 34 Pregnancy: A Prospective Longitudinal Study in Benin. J Infect Dis 2016;214:417–25. 35 36 doi:10.1093/infdis/jiw158 37 38 39 17 Unger HW, OmeKaius M, Karl S, et al. Factors associated with ultrasoundaided 40 on September 25, 2021 by guest. Protected copyright. 41 detection of suboptimal fetal growth in a malariaendemic area in Papua New Guinea. 42 43 BMC Pregnancy Childbirth 2015;15:83. doi:10.1186/s1288401505116 44 45 46 18 Dansou J, Adekunle AO, Arowojolu AO. Factors associated with antenatal care services 47 48 utilisation patterns amongst reproductive age women in Benin Republic: An analysis of 49 50 2011/2012 benin republic’s demographic and health survey data. Niger Postgrad Med J 51 52 2017;24:67–74. doi:10.4103/npmj.npmj_16_17 53 54 55 19 Planche T, Krishna S, Kombila M, et al. Comparison of methods for the rapid laboratory 56 57 assessment of children with malaria. Am J Trop Med Hyg 2001;65:599–602. 58 59 60 20 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 22 of 43

1 2 3 20 Hofmann N, Mwingira F, Shekalaghe S, et al. Ultrasensitive detection of Plasmodium 4 5 falciparum by amplification of multicopy subtelomeric targets. PLoS Med 6 7 2015;12:e1001788. doi:10.1371/journal.pmed.1001788 8 9 10 21 Robinson HP, Fleming JE. A critical evaluation of sonar “crownrump length” 11 12 measurements. Br J Obstet Gynaecol 1975;82:702–10. 13 14 15 22 PapageorghiouFor AT, Sarrispeer I, Ioannou review C, et al. Ultrasound methodologyonly used to construct 16 17 the fetal growth standards in the INTERGROWTH21st Project. BJOG Int J Obstet 18 19 Gynaecol 2013;120 Suppl 2:27–32, v. doi:10.1111/14710528.12313 20 21 22 23 Lohman T, Roche A, Martorell R. Anthropometric Standardisation Reference 23 24 Manual.Champaign, IL: Human Kinetics. 1998. 25 26 27 24 Gibson R. Gibson RS. Principles of nutritional assessment (second edition). Oxford, UK: 28 29 Oxford University Press; 2005. 30 31 http://bmjopen.bmj.com/ 32 25 Kramer MS. The epidemiology of adverse pregnancy outcomes: an overview. J Nutr 33 34 2003;133:1592S–1596S. 35 36 37 26 Ndao CT, Dumont A, Fievet N, et al. Placental malarial infection as a risk factor for 38 39 hypertensive disorders during pregnancy in Africa: a casecontrol study in an urban area 40 on September 25, 2021 by guest. Protected copyright. 41 of Senegal, West Africa. Am J Epidemiol 2009;170:847–53. doi:10.1093/aje/kwp207 42 43 44 27 Persson LÅ, Arifeen S, Ekström EC, et al. Effects of prenatal micronutrient and early 45 46 food supplementation on maternal hemoglobin, birth weight, and infant mortality among 47 48 49 children in Bangladesh: the MINIMat randomized trial. JAMA 2012;307:2050–9. 50 51 doi:10.1001/jama.2012.4061 52 53 28 Gbédandé K, Fievet N, Viwami F, et al.Clinical development of a VAR2CSAbased 54 55 56 placental malaria vaccine PAMVAC: Quantifying vaccine antigenspecific memory B & 57 58 59 60 21 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 23 of 43 BMJ Open

1 2 3 T cell activity in Beninese primigravidae. Vaccine 2017;35:3474–81. 4 5 doi:10.1016/j.vaccine.2017.05.027 6 7 8 29 Moore KA, Simpson JA, Paw MK, et al. Safety of artemisinins in first trimester of 9 10 prospectively followed pregnancies: an observational study. Lancet Infect Dis 11 12 2016;16:576–83. doi:10.1016/S14733099(15)005472 13 14 15 30 TuikueNdamFor N, Deloron peer P. Developing review vaccines to prevent only malaria in pregnant women. 16 17 Expert Opin Biol Ther 2015;15:1173–82. doi:10.1517/14712598.2015.1049595 18 19 20 21 22 23 24 25 26 27 28 29 30 31 http://bmjopen.bmj.com/ 32 33 34 35 36 37 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 22 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from

‡ Page 24 of 43

23 ‡

‡

‡ http://bmjopen.bmj.com/

‡ BMJ Open

‡

‡ on September 25, 2021 by guest. Protected copyright.

‡

¤ For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

- - - - † † - - † - - - - - † - - † - - - - † - - - † † † - - - † - † ------† - - - - - † † † - - † - - † - - - - - † - - - † ------† † † - - - - † - - - - - † - - † - † ------† - † - - - † ------† ------† † ------† - - - † ------† - - † - - † - - - - - † - - - † ------† - - - - † ------† ------† - - † ------† ------† ------† † ------† - † † † † † † † † † † † † † † † ‡ For† - peer - - - review - - - only ------Inclusion D1 M1 M2,M3 M4 M5,M6 M7 M8,M9 M10 M11,M12 M13 M14,M15 M16 M17,M18 M19 M20,M21 M22

Clinical, nutritional and biological data collected during the preconceptional followup. The RECIPAL study, study, 20142017. The RECIPAL followup. the preconceptional collected during data and biological nutritional Clinical,

(BIA) Table 1. Table 1.

histosomiasis

c demographic characteristics * C. trachomatis and N. gonorrhoeae screening (using PCR and serology), as well as schistosomiasis screening (Nytrel® filter) started 9 and 12 months after study initiation,respectively. study and filter)12after (Nytrel® started 9 months screening schistosomiasis asas PCR serology), (using well and screening C. gonorrhoeae trachomatisN.* and storage furtherforUrineanalysis ** reaction; sTfR: chain Soluble polymerase smear;PCR: blood TBS: Thick period; menstrual Last analysis; LMP: BIA: Bio impedance MUAC:M: MidUpperArmCircumference; Months; Abbreviations: day; D: GPS location Socio Weight, MUAC, skinfold thickness composition Body Hb, folic ferritin,acidlevels CRP,AGP, levels sTfR Urinary s Urine storage** Place of visit Eligibility criteria specific attimepoint. () (†) collected/assessed each or not Information clinic; (‡)at visit; study maternity Household the performed (¤) Visit Hb: Haemoglobin. Glycoprotein; protein; AGP: alpha1Acid CRP:CReactive Transferrin Receptor; Informed consent General General characteristics Housing characteristicsHousing Economic characteristics Clinicalnutritional and data Obstetricalhistory Height Dietary 24h recall Amenorrhea/LMP Biological data Urinary testpregnancy Malaria (TBS,screening PCR) Chlamydiatrachomatis* Neisseriagonorrhoeae* 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from

† 24

†

------† ------† - † - - - † - - † † † † † † † † † † † † † † † - † † † † - † † † † - † † † † - † † † † † † † † † - † † - - - † † - - - † † - - - † † - - - † † - - - † † - - -

------† † † † † † † † † † † † † † † † † † †

------† † † † † † † † † † † † † † † † † † † † † † †

------† † † † † † † † † † † † † † † † † http://bmjopen.bmj.com/ BMJ Open

------† † † † † † † † † † † † † † † † † † † † † † † † † † † on September 25, 2021 by guest. Protected copyright.

------† † † † † † † † † † † † † † 6.8 ± 2.5 6.8 ± 11.62.9 ±2.9 16.7 22.1 ± 3.7 ± 27.6 3.8 ± 32.7 3.5 ± 36.2 2.5 ± 38.4 1.4 ± 39.1 3.0 ANCv 1 ANCv 2 ANCv 3 ANCv 4 ANCv 5 ANCv 6 ANCv 7 ANCv ANCv8 Delivery For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For peer review only Clinical, nutritional and biological data collected during pregnancy. The RECIPAL study, 20142017. 20142017. study, The RECIPAL pregnancy. collected during and data biological nutritional Clinical,

‡ ) or not () at each specific timepoint. specific timepoint. ()each )at not or † Table Table 2.

iological data iological Information collected/assessed ( collected/assessed Information Placental malaria malaria Placental(TBS, PCR, histology) Peripheral Peripheral malaria (TBS,PCR) HIV screeningHIV 1 Gestational age (based on Gestational (based age on LMPfirst US) or Axillary temperature,Axillary blood pressure Use of ITN Use night the of thebefore visit Blood and rhesusBlood group Cord blood Cord blood PCR) malaria (TBS, Iron and folate Iron intake lastin 24 the hours Alcohol consumptionAlcohol in lastthe 24 hours Hb levelHb Serum ferritin, Serum sTfR,ferritin, CRP, acidAGP,levels folic Urinary schistosomiasisUrinary Lead Lead level Urinary infection Urinary and proteinuria test)(dipstick intestinalHelminthic infection(PCR) storage**Urine MidUpperArmCircumference Past medical Pastmedical and history* obstetrical Weight Gestational age Gestational (weeks),age mean (SD) Anthropometric measurements disorders. orhypertensive of delivery preterm history affection; other chronicmedical or Sickledisease, cell diabetes * furtherUrineanalysis storage for at during** collected delivery; pregnancy of H istory smoking ‡ cigarette Clinical and nutritional data data nutritional and Clinical Skinfold thickness (bicipitalandtricipital) recall Dietary 24h Body composition Body (Bio impendanceanalysis) Abbreviations: Abbreviations: ANCv: Antenatal care visit; ITN: bed Insecticidetreated net; IPTp: treatmentpreventive Intermittent in pregnancy; LMP: Last menstrual period; TBS: Thick blood smear; PCR: polymerase chain Hb: Haemoglobin. AGP:protein; alpha1Acid Glycoprotein, Receptor; CReactive Transferrin sTfR: CRP: reaction; Soluble B Page 25 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 26 of 43 25

detection detection phenotyping phenotyping

N. gonorrhoeae Vitamin A Vitamin Metals level Metals level and and Pollutant/pesticide Pollutant/pesticide Helminthic infection Helminthic Planned analyses analyses Planned Erythrocytic folic acid Erythrocytic folic Immunological markers Immunological markers Immunological markers Immunological markers Immunological markers Immunological markers Plasmodium Plasmodium Ferritin, CRP, sTfR, AGP AGP CRP,sTfR,Ferritin, Histopathology for malaria Histopathology detection (PCR) Plasmodium detection Micro RNAs, cytokines and chemokines and chemokines cytokines Micro RNAs, and chemokines cytokines Micro RNAs, C. trachomatis C.

Placenta Placenta Placenta Placenta Placenta Placenta Delivery/Birth Delivery/Birth Motherplacentacord Motherplacentacord Motherplacentacord Immunological markers

http://bmjopen.bmj.com/ BMJ Open

$ ‡ ‡

Mother (x8) Motherplacentacord Mother (x2) Motherplacentacord Mother (x2) Mother (x2) Motherplacentacord Mother (x2) Mother (x2) Mother (x2) Motherplacentacord Motherplacentacord Mother (x2) Mother (x3) Mother (x8) Mother (x8) on September 25, 2021 by guest. Protected copyright. trimester of pregnancy), three times (during each trimester), or eight times (during each ANC visit). visit). ANC each (during times eight or trimester), each times (during three pregnancy), of trimester rd and 3 and st

Mother (x1) Mother Mother (x1) Mother Mother (x1) Mother Mother (x1) Mother Mother (x1) Mother (x1) Mother Mother (x1) Mother (x2) Mother (x1) Mother For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Pre-pregnancy Pre-pregnancy Pregnancy For peer review only Biological samples stored before conception, during pregnancy and at birth. The RECIPAL study, 20142017 study, The RECIPAL at birth. and pregnancy during conception, before samples stored Biological Table Table 3.

† TBS: Thick blood smear; sTfR: Soluble Transferrin Receptor; CRP: CReactive protein; AGP: alpha1Acid Glycoprotein, PCR: Polymerase Chain Reaction. Reaction. Chain Polymerase PCR: Glycoprotein, alpha1Acid AGP: protein; CReactive CRP: Receptor; Soluble Transferrin sTfR: smear; Thick blood TBS:

† †

(1.8 mL) (80°C)

During pregnancy, samples were collected twice (during the the 1 (during twice collected were samples pregnancy, During In case of unscheduled visit, an additional blood spot was collected. collected. was spot blood additional an visit, unscheduled caseIn of respectively. µl, 600 600800, was200, delivery and at pregnancy during before, stored plasma of Quantity Biopsy with formalin with Biopsy Biopsy 20°C 20°C Biopsy Later RNA with Biopsy Blood spot +4°C (50 µL) (50µL) +4°C spot Blood Total blood µL)(500 Total blood Total blood µL)(200 Total blood Plasma/serum Urine Stool Placenta brackets. in is indicated sample collection of The frequency $ ‡ † Abbreviations: Total blood blood Total

Plasma20°C Plasma20°C Serum (400 µL) Serum Plasma 80°C Plasma (100µL) Serum Buffy coat PAXgene (300 µL) PAXgene coat Buffy Blood spot 20°C (100 µL) µL) (10020°C spot Blood Red cells (Trizol, 100 µL)100 (Trizol,cells Red Serum (100µL) Serum 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from

P value P value 26 (411 vs 359) (411359) vs

0.03 0.04

0.002

< 0.001 < 0.001 0.22

32 (7.8) 32 (7.8) 33 (8.0) 33 (8.0) 64 (27.0) 64 (27.0) 59 (14.4) 59 (14.4) (N=411) (N=411) 221 (53.8) 221 (53.8) 132 (32.1) 132 (32.1) 118 (28.7) 118 (28.7) 138 (33.8) 138 (33.8) 268 (65.2) 268 (65.2) Final cohort Pregnant women women Pregnant

follow-up but no no but follow-up pregnancy (N=359) (N=359) pregnancy WRA with completewith WRA

P value P value (1214 vs 444) (1214 444) vs

http://bmjopen.bmj.com/ 26 (6.9) 26 (6.9) 29 (8.1) 40 (9.7) 14 (3.1) 14 (3.1) 1 (0.3) 25 (6.1) 23 (5.2) 23 (5.2) 31 (8.6) 33 (7.4) 33 (7.4) 120 (33.4) 33 (7.8) 19 (4.3) 19 (4.3) 30 (8.4) 20 (4.9) 28 (7.5) 28 (7.5) 0.06 12 (3.4) 24 (5.9) 0.10 23 (5.2) 23 (5.2) 0.62 27 (7.5) 26 (6.3) 0.51 27 (6.1) 27 (6.1) 0.46 32 (8.9) 33 (8.0) 0.65 BMJ Open follow-up (N=444) (N=444) follow-up WRA with incomplete incomplete with WRA

102 (23.0) 45 (12.5)

173 (38.9) 154 (42.9) 228 (51.4) < 0.001 57 (15.9) 346 (77.9) 56 (12.6) 0.001 80 (22.3) 218 (60.7) 300 (72.9) 257 (57.9) 0.45 204 (56.8)

58 (13.1) 0.17 43 (12.0) 48 (10.8) 26 (7.2) on September 25, 2021 by guest. Protected copyright. 76 (6.3) 76 (6.3) 95 (8.3) 95 (8.3) 40 (3.3) 40 (3.3) 69 (5.6) 69 (5.6) 64 (5.7) 64 (5.7) 86 (7.1) 86 (7.1) 107 (8.8) 107 (8.8) 889 (73.2) 889 (73.2) 336 (75.7) 0.14 263 (73.3) 290 (70.6) 0.41 167 (13.8) 167 (13.8) 940 (77.4) 940 (77.4) 338 (76.1) 290 (80.8) 378 (92.0) 343 (30.0) 343 (30.0) 120 (32.0) 117 (32.7) 106 (25.8) 706 (61.7) 706 (61.7) 229 (61.1) 0.36 212 (59.2) 265 (64.5) 0.10 864 (71.2) 864 (71.2) 445 (36.7) 445 (36.7) 195 (16.1) 195 (16.1) 729 (60.0) 729 (60.0) 423 (34.8) 423 (34.8) 186 (15.3) 186 (15.3) 177 (24.7) 177 (24.7) 596 (49.1) 50 (31.2) 210 (47.3) 0.03 0.07 63 (19.6) 165 (46.0) 326 (26.9) 326 (26.9) 81 (18.2) 137 (38.2) 108 (26.3) 279 (23.0) 279 (23.0) 2.8 (± 2.0) 2.0) 2.8 (± 2.0) 2.6 (± 0.07 2.1) 2.9 (± 1.9) 2.7 (± 0.16 5.7 (± 3.0) 3.0) 5.7 (± 3.4) 6.0 (± 0.009 5.2 (±2.3) (±3.0) 5.9 0.001 27.8 (± 5.5) 5.5) (± 27.8 5.6) (± 28.0 0.31 5.6) (± 28.8 5.0) (± 26.8 < 0.001 Initial cohortInitial (N=1214 WRA) WRA) (N=1214 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

Yes Yes Yes Yes Fon Fon Aizo Aizo Toffin Toffin Others Others Vekky Vekky SôAva SôAva Houedo Houedo < 23 cm cm < 23 Illiterate Illiterate Akassato Akassato Polygamy Polygamy ≥ 25kg/m² ≥ 25kg/m² Monogamy Monogamy < 18.5kg/m² 18.5kg/m² < Cohabitation Cohabitation Mean (± SD) (± Mean SD) Mean (± SD) (± Mean SD) Mean (± SD) (± Mean SD) Nulligravidae Nulligravidae 18.5‒24kg/m² 18.5‒24kg/m² For peer review Primigravidae Multigravidae only Characteristics of women included in the initial and final RECIPAL cohorts, Southern Benin, 20142017. cohorts,Benin, final Southern RECIPAL and initial in the of women included Characteristics Maternal length < 150 cm cm 150 < length Maternal 92 (8.0)

Table 4. £

‡

Microscopic malaria, n (%) malaria, Microscopic Body mass index, n (%) n (%) mass index, Body

Schistosomiasis infection, n (%) n (%) infection, Schistosomiasis Subdistrict, n (%) (%) n Subdistrict, (%) group,n Ethnic Pre-conceptionalcharacteristics General characteristics General Characteristics Characteristics Anaemia, n (%) Anaemia, (years) Age n (%) status, Marital

Gravidity, n (%) n (%) Gravidity, Education, n (%) n (%) Education, density Household MidUpperArmCircumference, n (%) n (%) MidUpperArmCircumference, Short stature, n (%) (%) n stature, Short Number of live birth(s) birth(s) live of Number Page 27 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 28 of 43 P P value* 27 (411 vs 359) (411359) vs

5.9) −

(N=411) (N=411) 378 (92.0) 378 (92.0)

5.4 (4.9 Final cohort ). Pregnant women women Pregnant (8.0) 33

118 (57.0) 98 (48.5) 4 (1.9)

(65.7) 136 142 (68.6) 93 (44.9) 51 (25.4) 40 (19.4) 33 (16.1)

1.3) 2.1 (± 1.1) 7.5 (± 12.3 2.5) 6.9 (± 120 (57.9) 7 (3.4) (7.8) 13 continued

follow-up but no no but follow-up pregnancy (N=359) pregnancy WRA with completewith WRA

P value* P value* (1214 vs 444) (1214444) vs

http://bmjopen.bmj.com/ BMJ Open follow-up (N=444) (N=444) follow-up WRA with incomplete incomplete with WRA

on September 25, 2021 by guest. Protected copyright. Initial cohortInitial (N=1214 WRA) (N=1214WRA)

χ χ χ χ χ For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

trimester trimester trimester trimester trimester trimester trimester st rd ≥ 2 2 ≥ st Yes Yes Yes Yes rd nd Median Median Positive Positive Mean (± SD) (± Mean SD) Mean (± SD) (± Mean SD) Mean (± SD) (± Mean SD) Primigravidae Primigravidae Multigravidae Multigravidae In the 1 In In the 3 In During pregnancy During During pregnancy During In the 1 In

For peer review the 3 In only In the 2 In

†

† , n (%) (%) n ,

# †

† µ

†

† Characteristics of women included in the initial and final RECIPAL cohorts, Southern Benin, 20142017 ( 20142017 Benin, cohorts, Southern RECIPAL final and initial in the of women Characteristics included ANC visit (weeks) (weeks) visit ANC st

† § § † Table 4. Table 4.

At least one episode episode one At least Gestational age at 1 at age Gestational Gestational hypertension, n (%) n (%) hypertension, Gestational Inadequate gestational weight gain weight gestational Inadequate

HIV status, n (%) status, (%) n HIV Anaemia, n (%) n (%) Anaemia, respectively. variables, categorical and continuous comparing for used test were chi2 and Tstudent * £ statusinfection Schistosomiasis before conception has been in assessed 718 WRA cohort): (Initial 160 321 followup and preconceptional without with followup, pregnancy, with 237incomplete preconceptional complete in included cohort. the final pregnant). for and 11 g/dL women nonpregnant for (12 g/dL thresholds to WHO § According malaria placental for values 40 missing respectively; pregnancy, of trimester 3rd and 2nd 1st, at values 2 missing 16, and pregnancy, During cohort, respectively. and final initial the in values missing 5 and 78 missing conception, Before ‡ n (%) infection, malaria Placental Fertility incidence rate(95% CI) incidence rate(95% Fertility Median time to pregnancy (months) (months) pregnancy to time Median Gestational Gestational characteristics Characteristics Characteristics Microscopic malaria, n (%) n (%) malaria, Microscopic Gravidity, n (%) n (%) Gravidity, µIncidence rate defined as the number of pregnancies / personsmonth at risk. risk. at personsmonth / pregnancies of number the as defined rate µIncidence women (pre weighted normal ), kg/m² 18.5 < BMI (prepregnancy women underweight in the 5kg and 7kg 11.5kgs, to 12.5kg, below was pregnancy during gain weight total the when inadequate considered was gain weight # gestational A † Results based on the 207 women who have already delivered. delivered. already have 207who women the on based Results † χ massindex Body BMI: care; Antenatal ANC: deviation; Standard SD: age; reproductive of women WRA: Abbreviations: Number of ANC visits (scheduled) (scheduled) visits ANC of Number Number of IPTp doses, n (%) doses, IPTp of Number Number of ANC visits (unscheduled) (unscheduled) visits ANC of Number pregnancy BMI between 18.524.9 kg/m²), overweight women (prepregnancy BMI between 25.029.9 kg/m²) and obese women (prepregnancy BMI ≥ 30 kg/m²)before conception, respectively. respectively. conception, 30 kg/m²)before BMI ≥ (prepregnancy women obese and kg/m²) 25.029.9 between BMI (prepregnancy women overweight kg/m²), 18.524.9 between BMI pregnancy 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 29 of 43 BMJ Open

1 2 3 Table 5. Preliminary data description of newborns at birth (subsample of the cohort). 4 The RECIPAL study (20142017). 5 6 Mean (± SD) or Newborn characteristics† N 7 Proportion (%) 8 9 Gender Male 205 55.2 10 11 Stillbirth Per 1000 live births 206 14.8 12 Preterm birth (< 37 weeks) Yes 207 12.6 13 Smallbirthweightforgestational age ‡§ Yes 191 11.5 14 ¥ § 15 SmallbirthweightforgestationalFor peer age reviewYes only191 18.3 16 Birthweight (g)§ 193 3043.9 (± 432.8) 17 < 2500 193 9.3 18 19 Poor birth outcome £,§ Yes 207 29.5 20 21 † Data based on the first 207 deliveries. §Twins excluded 22 ‡ Smallbirthweightforgestational age: < 10th percentile of birthweight for gestational age using Schmiegelow's charts. 23 ¥ Smallbirthweightforgestational age: < 10th percentile of birthweight for gestational age using Intergrowth 21stcharts; 24 £ Stillbirth, preterm birth, smallbirthweightforgestational age (using Intergrowth 21stcharts) or low birthweight 25 26 27 28 29 30 31 http://bmjopen.bmj.com/ 32 33 34 35 36 37 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 28 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 30 of 43

1 2 3 Figure legends 4 5 6 Figure 1. Geographical location of SôAva and AbomeyCalavi Districts, and of the 35 7 8 villages of the RECIPAL study, Southern Benin, 20142017. 9 10 11 Figure 2. Flowchart diagram of RECIPAL study. 12 13 * Study completion: Followup from enrolment until the end of the study (24month follow 14 15 For peer review only 16 up without pregnancy for women recruited before December 2014 or monthly followup 17 18 without pregnancy for women recruited between December 2014 and December 2016), 19 20 excluding consent withdrawal, migration and lost to followup. 21 22 23 Figure 3. Malaria status (infected vs. non infected) in women of reproductive age at inclusion 24 25 in RECIPAL (initial cohort), Southern Benin, 20142017. For online interactive map, check 26 27 28 this link: Malaria Infection Status of Women at Inclusion of Cohort 29 30 31 Figure 4. KaplanMeier failure estimate of the probability of conceiving; 1214 women of http://bmjopen.bmj.com/ 32 33 reproductiveage included in the initial RECIPAL cohort, Southern Benin, 20142017. 34 35 Probability of conceiving (solid line) and its 95% confidence interval (grey lines).Number of 36 37 pregnancy events are in brackets. Number of censored women at 5, 10, 15, 20 and 24 months 38 39 40 were : 223, 260, 88, 50 and 30, respectively. on September 25, 2021 by guest. Protected copyright. 41 42 43 Supporting information 44 45 46 47 Supplemental file S5. Ultrasound follow-up (RECIPAL study) 48 49 HC: Head circumference; AC: Abdominal circumference; FL: Femur length; BPD: Biparietal 50 51 52 diameter; OFD: occipito frontal diameter; ATD: Abdominal transverse diameter; APAD: 53 54 Anteroposterior abdominal diameter. 55 56 * Uterine, and umbilical blood flows were measured at 2125 wg and from 28 wg, 57 58 respectively 59 60 29 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 31 of 43 BMJ Open

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 http://bmjopen.bmj.com/ 32 Figure 1. Geographical location of Sô-Ava and Abomey-Calavi Districts, and of the 35 villages of the RECIPAL 33 study, Southern Benin, 2014-2017. 34 35 162x123mm (300 x 300 DPI) 36 37 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 32 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 Figure 2. Flowchart diagram of RECIPAL study.

29 * Study completion: Follow-up from enrolment until the end of the study (24-month follow-up without 30 pregnancy for women recruited before December 2014 or monthly follow-up without pregnancy for women 31 recruited between December 2014 and December 2016), excluding consent withdrawal, migration and lost http://bmjopen.bmj.com/ 32 to follow-up. 33 34 133x81mm (300 x 300 DPI) 35 36 37 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 33 of 43 BMJ Open

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 http://bmjopen.bmj.com/ 32 33 34 35 36 37 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 Figure 3. Malaria status (infected vs. non infected) in women of reproductive age at inclusion in RECIPAL 45 (initial cohort), Southern Benin, 2014-2017.

46 215x238mm (300 x 300 DPI) 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from BMJ Open Page 34 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Figure 4. Kaplan-Meier failure estimate of the probability of conceiving; 1214 women of reproductiveage 31 included in the initial RECIPAL cohort, Southern Benin, 2014-2017. http://bmjopen.bmj.com/ 32 Probability of conceiving (solid line) and its 95% confidence interval (grey lines).Number of pregnancy 33 events are in brackets. Number of censored women at 5, 10, 15, 20 and 24 months were : 223, 260, 88, 50 34 and 30, respectively. 35 36 132x89mm (300 x 300 DPI) 37 38 39 40 on September 25, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 43 BMJ Open

1 2 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from 3 Supplemental file S1: Procedures of enrolment of women of reproductive age in the 4 5 RECIPAL study 6 7 All women of reproductive age (WRA), whatever their gravidity, living in the study area 8 were eligible to be included in the study. To be recruited, they should met the following 9 10 criteria: negative urinary pregnancy test, 18 to 45 years old, no current contraception, no 11 previous fertility issues, in willingness to become pregnant, no planned travel for more than 2 12 months within the next 18 months, acceptance of RECIPAL protocol and signed written 13 14 informed consent. Possible fertility issues were: primary or secondary amenorrhea; women 15 unable – but wishing – to become pregnant for 10 years or more; and past history of three or 16 more successive spontaneous abortions. 17 For peer review only 18 19 Three different procedures were used concomitantly to enrol WRAs: 20 21 I First procedure: Sensitization events were organized in each of the 35 selected villages 22 for presenting the study to all inhabitants. If interested, women were invited to register with 23 24 the head of the village and were visited at home the day after for their inclusion. The field 25 investigators were in constant contact with the head of the villages to inquire about women 26 registration and to know whether women or other inhabitants had shown interest in the study. 27 28 Sensitization events were repeated several times throughout the study in villages with a low 29 level of recruitment. Moreover, information sessions were organized in the different certified 30 and accredited heath centres in the area, whether or not participating in the study, in order to 31 32 increase community involvement to and acceptability of the study. 33 34 I Second procedure: Specific sensitization events were organized along with the 35

community leaders (local elected and health authorities, crowned head families), religious http://bmjopen.bmj.com/ 36 37 authorities (traditional, muslim and christian) and mass-media (mainly through radio 38 messages) for supporting the study and mobilizing WRAs during masses, women’s or young's 39 association meetings, church sessions, etc. Women’s population was also informed about the 40 41 project during health facility activities such as antenatal care visits and routine EPI’s 42 (expanded programme on immunization) sessions. Then, WRAs interested in the project 43 could meet the study team (including field investigators and community health workers) to be 44 on September 25, 2021 by guest. Protected copyright. 45 registered. 46 47 I Third procedure: The last way of recruitment consisted of going door-to-door to identify 48 eligible and interested women with the assistance of a network of community health workers. 49 50 Community health workers are members of the community, they are selected by the 51 community and are answerable to the communities for their activities. They must be endorsed 52 53 by the health centre, but do not necessarily belong to it, and have a shorter training than 54 professional workers. "Trackers", who are community members usually solicited by the heath 55 centres for advanced EPI's sessions and national vaccination campaigns, were identified and 56 57 involved in WRAs identification and enrolment after being trained. 58 59 60

(2014-2017) (2014-2017) RECIPAL cohort RECIPAL N N %(n)

5 5 1214 (362) 29.8 0 0 1214 (107) 8.8 1 1 1214 (167) 13.8 35 35 1214 (203) 16.7 19 19 1214 (38) 3.2 ≥ 2-4 2-4 1214 (578) 47.6 BMJ Open ≥ ≤ 25-29 25-29 30-34 1214 1214 (408) 33.6 (271) 22.3 20-24 20-24 1214 (294) 24.2 Poorer Richer 1214 1214 (281) 23.1 (150) 12.4 Higher 1214 (223) 18.4 Middle Middle 1214 (211) 17.4 Poorest Poorest 1214 (427) 35.2 Richest Richest 1214 (145) 11.9 http://bmjopen.bmj.com/ Average 1214 (455) 37.5 No education education No 1214 73.2 on September 25, 2021 by guest. Protected copyright. the 2011/2012 Beninese Demographic Health Survey and Beninese 2011/2012 Demographic the For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

†

: Characteristics of women of reproductive age from RECIPAL cohort in comparison with data extracted from extracted from in comparison with data cohort age from RECIPAL women reproductive of of :Characteristics For peer review only 2 S

Supplemental file Supplemental then categorized according to the quintiles. quintiles. to the according categorized then attainment Education quintiles wealth Household Gravidity exposure media Mass None 1214 (536) 44.1

Age (years) Age (DHS-IV); Survey Health and Demographic Beninese national 2011-2012 the from extracted Data N= 8,701. * Benin 2011/2012 of analysis An Republic: in Benin women age amongst reproductive patterns utilisation services care antenatal with associated Factors AO. Arowojolu AO, J, Adekunle Dansou 2017;24:67–74 J Med Postgrad data. Niger Survey Health and Demographic Republic’s respectively. study and DHS-IV, RECIPAL inthe old to 49 years and 15 45 years to 18 aged § Women was it article; in Dansou’s analysis component principal using study, and by RECIPAL in assets of and ownership occupation combining score synthetic a using approximated was wealth † Household

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 37 of 43 BMJ Open

1 2 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from 3 Supplemental file S3:Follow-up and quality control procedures in RECIPAL study 4 5 6 1. Follow-up procedures 7 8 Women of reproductive age (WRA) were recruited at community-level and followed monthly 9 10 for a maximum period of 24 months; they constituted the RECIPAL initial cohort. The sub- 11 sample of women who became pregnant was then followed-up monthly at the maternity clinic 12 from early pregnancy to delivery; they constituted the RECIPAL final cohort. Data were 13 14 collected by five specifically-trained field investigators, helped by three community health 15 workers (CHW) recruited for the project. 16 17 Pre-conceptionalFor follow-up peer review only 18 19 20 Each field investigator was responsible for the follow-up of a sub-sample of the cohort (~ 243 21 WRAs by investigator). Home visits were scheduled by the field-investigators (~159 22 visits/month/investigator) under site coordinator and deputy coordinator supervision. At each 23 24 visit, the next visit was planned with the woman agreement and according to RECIPAL 25 protocol. Few days before the visit, the investigator contacted the woman by telephone to 26 confirm the date and time of the visit. A list of unreachable women was then established and 27 28 sent to the CHWs, who visited the women at home. 29 30 The following procedures was applied when a woman remained unreachable: 31 32 • In case of absence because of daily activities, the field investigator planned another 33 34 visit within the next few days and asked a family member or neighbor to warn the 35 woman about the visit. In cases of over deadlines, the CHWs revisited the woman on 36 the same day; http://bmjopen.bmj.com/ 37 38 • When a travel was planned for less than 7 days, the investigator called the woman and 39 40 planned a visit when she came back; 41 • 42 When a travel was planned for more than 7 days, a urinary pregnancy test was 43 performed before departure: if positive, the first antenatal care visit was performed 44 before departure; if negative, a new visit was planned when the woman came back on September 25, 2021 by guest. Protected copyright. 45 46 (whether in the two coming months); 47 • 48 A monthly visit was considered as “missing” when not completed within 21 days 49 after the planned date despite several requests from the field investigator. 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 43

1 2 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from 3 Women were excluded from the study in the following cases: 4 5 • 6 Three successive missing visits; 7 • 8 Two successive missing visits with a positive urinary pregnancy test when the woman 9 came back. 10 11 Gestational follow-up 12 13 14 ANC visits were scheduled by midwives after discussion with the RECIPAL team. They were 15 scheduled every month or every six weeks depending on whether an ultrasound scan was 16 17 planned at this visitFor (see Suppl. peer file S5). Inreview case of missing visit, only the procedures were the same 18 than during the pre-conceptional follow-up. 19 20 21 2. Follow-up completion 22 23 A WRA completed the pre-conceptional follow-up when she was followed until the end of 24 the study (from 4 to 24 months of follow-up depending on when she was recruited). A 25 26 pregnant woman completed the pregnancy follow-up when she was followed until delivery. 27 Both follow-ups were considered incomplete in case of (i) lost to follow-up (two or three 28 29 missing visits during the pre-conceptional period, or no information for more than 3 months 30 during pregnancy); (ii) informed consent withdrawal; (iii) migration outside the study area; 31 (iv) no data available at delivery because of adverse pregnancy outcomes such as spontaneous 32 33 abortion or ectopic gestation. 34 35 3. Quality control 36 http://bmjopen.bmj.com/ 37 38 The quality control of the clinical data was carried out in two stages. At the first stage, the two 39 supervisors verified on a weekly basis that data collection was exhaustive (i.e., concordant 40 number of expected and collected forms). A control of missing and abnormal data was carried 41 42 out on all participant forms. The second stage was carried out after data have been entered. 43 The validation of double-entering was realized by the data manager and an audit report of the

44 on September 25, 2021 by guest. Protected copyright. 45 database, including lists of discrepancies, duplications and missing or abnormal data was sent 46 regularly to the supervisors who were responsible for correcting any errors. Regarding the 47 ultrasound data, a quality control of 10% of the pictures was performed monthly by a senior 48 49 sonographer from Oxford University, United Kingdom. 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 43 BMJ Open

1 2 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from 3 Supplemental file S4: Main reasons of study drop out in RECIPAL cohort 4 5 6 Recruitment of WRAs was stopped in September 2016, when only 1,214 WRAs had been 7 8 included. The recruitment period lasted 9 months more than was planned in the protocol (27 9 10 months vs. 18 months), but it could not be extended further because of financial reasons. 11 Besides, because of constraints related to both the study design and the local socio-cultural 12 13 context, a noticeable proportion, 36.5 % (444/1214) of WRAs did not complete the pre- 14 15 conceptional follow-up. At the end, both the number of WRAs and the number of pregnant 16 17 women includedFor in the RECIPAL peer study review were far lower thanonly was planned. Difficulties in 18 recruiting women, and the reasons of drop out before and during pregnancy were the 19 20 following: 21 22 23 I Constraints related to the study design 24 25 To be included in RECIPAL, women should not have planned to travel for more than 2 26 27 months within the next 18 months. However, many women travelled during their follow-up. 28 29 Women with three successive missing visits or with two successive missing visits with a 30 31 positive pregnancy test at their return were excluded from the study (see Supplemental file 32 33 S3). As described in the flow chart diagram, among women with incomplete pre-conceptional 34 follow-up, 31.1% (138/444) moved outside the study area for more than two months and had 35 36 to be excluded. Sô-Ava district is a lakeside city with fishing as principal activity. During the http://bmjopen.bmj.com/ 37 38 period of high fishing activity (~ 3 to 4 months/year), many women travel by ship to Makoko 39 40 city in Nigeria for fish trade. In RECIPAL, 63.2% (86/138) of women who travelled went to 41 Nigeria as final destination. 42 43

44 One of the main constraints related to the study design was the long duration of the pre- on September 25, 2021 by guest. Protected copyright. 45 46 conceptional follow-up. Among the WRAs who withdrew their informed consent, 36% 47 48 (98/272) of them reported study fatigue and disappointment of not being pregnant. No drugs 49 or interventions were administered/implemented by the project to help women to conceive, 50 51 only medical advices and clinical examination in case of symptoms suggestive of genital 52 53 infection were provided. 54 55 56 During pregnancy, the main constraint was related to the number of ANC visits. Indeed, nine 57 ANC visits were scheduled. At each visit, a thick blood smear was performed using capillary 58 59 blood, except at the 2nd and 6th ANC visit when venous blood was collected. The mean 60 number of ANC visits (including both scheduled and unscheduled visits) was 8.1 (SD ±2.8;

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 43

1 2 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from 3 range: 1-15). Although most women were globally satisfied with the RECIPAL follow-up, 4 5 some of them withdrew their informed consent because of too frequent ANC visits. 6 7 8 I African socio-cultural realities 9 10 Sixty-four percent of women (174/272) refused to continue participating in the study because 11 12 of rumours about blood and placental samples collection, as well as suspicion regarding 13 14 medical care provided free of charge for the woman. Indeed, at the beginning of the study 15 16 false rumours circulatedFor about peer the quantity review and number of timesonly blood was collected, and its 17 use for religious purposes. Despite regular sensitization campaigns in the villages, project 18 19 endorsement by the local (health) authorities and religious leaders, as well as organization of 20 21 visits to our laboratory, rumours could not be completely eradicated. The fact that a high 22 23 proportion of women were illiterate (> 70%) and from Toffin ethnicity (fears of foreign 24 25 populations) contributed to the spread of rumours. 26 27 Despite the usefulness of the placenta for biomedical research, personal, socio-economic and 28 29 cultural factors undermine the willingness of mothers to freely donate their placentas for 30 31 research purpose, especially in SAA [1]. In Africa, including Benin, there are strong 32 33 emotional, religious and cultural ties to how placentas are disposed. Indeed, a careless 34 handling of the placenta is associated with a risk of infertility, illness and death in the 35 36 newborn baby [2,3]. In RECIPAL, after a small piece of placenta was drawn for malaria http://bmjopen.bmj.com/ 37 38 histology and 4mL placental blood was collected for malaria screening, the placenta was 39 40 immediately returned to the woman. Information—on why and how placenta will be 41 collected—was provided to the woman and her family throughout the follow-up. Despite 42 43 these precautions, placenta collection remained one of the main reasons for consent

44 on September 25, 2021 by guest. Protected copyright. 45 withdrawal. 46 47 48 The social pressure exerted by the family members was another issue in RECIPAL study. As 49 far as possible, women’s partners and family members (particularly, the women’s mother-in- 50 51 law) were involved at the various stages of follow-up. For many women, family agreement 52 53 was required for their participation in the study and for monthly follow-up at the maternity 54 55 clinic until delivery. Besides, in the RECIPAL study, women benefited from free management 56 of any diseases related to the pregnancy—detected either as part of scheduled follow-up or 57 58 during emergency visits—, free ANC visits and delivery. Eighteen percent (8/45) of consent 59 60 withdrawals were due to the fact that the family-in-law judged the woman’s partner unable to take care of their daughter correctly because of free antenatal care.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 41 of 43 BMJ Open

1 2 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from 3 I Pregnancy cohort attrition 4 5 6 In the final cohort, 32.1% (132/411) of pregnant women did not complete the follow-up, 7 8 mainly because of spontaneous abortion (51.5%, 68/132) and informed consent withdrawal 9 10 (34.1%, 45/132). Reasons for consent withdrawal have been given above. There are few data 11 available on spontaneous abortions, particularly in SSA countries, where they are neither 12 13 detected nor reported. In RECIPAL, the prevalence of spontaneous abortions was 14 15 unexpectedly high (16.5%, 68/411), with more than half (50.7%, 34/68) of cases occurring 16 17 before 6 weeks ofFor gestation. peer review only 18 19 References: 20 21 22 1 Abudu EK, Inyang-Etoh EC, Eziagu UB. Pregnant women perception of placenta 23 donation for biomedical research- experience at a Nigerian Tertiary Health Care 24 Savannah J Med Res Pract 25 Institution. 2015;4:8–14. 26 2 Bazuaye G, Enabudoso E. Willingness of Pregnant Women in Benin City Nigeria to 27 28 Donate Placenta Cord Blood for Stem Cell Transplantation. Ann Biomed Sci 2011;10. 29 30 3 Herlihy JM, Shaikh A, Mazimba A, et al. Local Perceptions, Cultural Beliefs and 31 Practices That Shape Umbilical Cord Care: A Qualitative Study in Southern Province, 32 Zambia. PLoS ONE 2013;8. 33 34 35 36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 25, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 42 of 43 trimester 33–38 weeks 33–38 rd 3 27–32 weeks 27–32 ultrasound scan ultrasound th , 5 th , 4 rd , 3 , BMJ Open trimester nd http://bmjopen.bmj.com/ 21–26 weeks 21–26 2 nd Ultrasound follow-up (RECIPAL study) (RECIPAL follow-up Ultrasound 2 - Fetal viability, amniotic fluid assessment,- placental Identification location of major fetal abnormalities - Fetal biometry measurements (BPD, HC,- AC, Uterine FL, and ATD, umbilical blood APAD, OFD) flow measurements* 15–20 weeks 15–20 on September 25, 2021 by guest. Protected copyright. Unscheduled ultrasound scan in case of emergency visit emergencyof case in scan ultrasound Unscheduled For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Supplemental file S5. S5. file Supplemental Fetal viability assessment, assessment, viabilityplacentalamniotic Fetal location, assessment, fluid fetal position For peer review only diameter. abdominal Antero-posterior APAD: diameter; transverse Abdominal ATD: * Uterine, and umbilical blood flows were measured at 21-25 wg and from 28respectively wg, from 21-25and wg at were measured flows blood umbilical and Uterine, * 9–14 weeks 9–14 trimester st 1 HC: Head circumference; AC: Abdominal circumference; FL: Femur length; BPD: Biparietal diameter; OFD: occipito frontal diameter; frontal occipito OFD: diameter; Biparietal BPD: length; Femur FL: circumference; AC: Abdominal circumference;Head HC: 4–5 weeks 4–5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-019014 on 8 January 2018. Downloaded from Page 43 of 43 BMJ Open