Maternal and child undernutrition contribute to around 50% of global deaths in children under five years of age [1]. Nutrient deficiencies in mothers can worsen during pregnancy due to increased nutritional demands needed to support both the mother and developing fetus [1,2]. The periconceptional and perinatal periods are critical windows for offspring development, and nutritional inadequacies during these times associate with both immediate adverse outcomes for the offspring, such as an increased risk of poor fetal growth, stunting, preterm birth, and mortality [2,3], and long-term health issues, including increased risk of type 2 diabetes, hypertension, and coronary heart disease [4,5]. For mothers, poor nutritional status during these periods can increase the risk of anaemia, hypertension, miscarriage, or mortality [6]. Common nutritional interventions to improve maternal and pregnancy outcomes include daily supplementation with folic acid [7], with and without iron, multiple micronutrients, or calcium [8]. However, these interventions have shown inconsistent success in improving pregnancy outcomes [2,9], in part because the biological mechanisms that determine an intervention’s impact are complex and remain poorly understood.
The placenta plays a crucial, but often overlooked biological role in how nutritional interventions affect maternal, fetal, and infant health outcomes. As a key organ in pregnancy that develops alongside the embryo/fetus, it plays a vital role in fetal growth, facilitating nutrient transport (glucose, amino acids, lipids, and vitamins and minerals [10,11]), waste removal, and protection from harmful substances [10]. Placental nutrient transport is influenced by nutrient availability and fetal nutrient requirements [12], among other factors, and can adapt to buffer against temporary fluctuations in maternal nutrient status, whether deficiencies or oversupply, to maintain a stable nutritional environment for the developing fetus [10]. Importantly, the placenta itself is a developing organ that requires adequate nutritional resources to develop and function properly, and inadequate resources may disrupt optimal placental development and function [13]. Nutritional interventions have the potential to increase nutrient availability for the developing placenta, which can in turn influence placental gene expression, angiogenesis, nutrient transport, and activity in inflammatory and oxidative stress pathways, among other functions [14,15]. A healthy placenta also benefits maternal health in the short-term by releasing placental hormones that can alter maternal physiology in preparation for pregnancy and lactation [16]. At the same time, placental syndromes can have long-term cardiovascular consequences for the mother [17]. Therefore, nutritional interventions aimed at improving maternal and fetal health may be associated with changes in the placenta, and understanding these changes is important for determining why these interventions are successful in some contexts but not others.
To date, reviews on how nutritional interventions influence the placenta have mainly focussed on placental-related pregnancy complications, such as preeclampsia [18–21], rather than on the overall placental phenotype. In this systematic review, we synthesised existing knowledge on how direct maternal nutritional interventions in the periconceptional or pregnancy periods impact placental phenotype. We aimed to determine whether improved maternal and offspring outcomes were more likely to occur with nutritional interventions that reported placental phenotype alterations, and whether associations between nutritional interventions and placental phenotype differed based on placental sex or study location. We hypothesised that direct maternal nutritional interventions in the periconceptional and pregnancy periods would be associated with altered placental phenotype, and that nutritional interventions that improved maternal and offspring health would more likely associate with beneficial placental adaptations compared to interventions without improvements in maternal or offspring outcomes.
METHODS
This study is reported in alignment with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses [22] and Synthesis without meta-analysis (SWiM) [23] reporting guidelines (Table S1 in the Online Supplementary Document).
Inclusion and exclusion criteria
We included randomised and non-randomised controlled trials in human populations that administered a direct nutritional intervention to the mother during the pre-conception, periconceptional, or pregnancy periods. The studies had to be written in English and published between January 2001 and September 2021 (time of search). We set this time restriction to increase the translatability of the review findings to the current day context, as the post-2001 period followed the implementation of folic acid fortification and peri-conceptional supplementation in several countries [24], thereby reducing the potential for including studies where folate deficiency or insufficiency in pregnancy was common and likely to have effects on placental development and function [25].
The population of interest was pregnant people of any age receiving a direct nutritional intervention before or during any stage of pregnancy (Figure S1 in the Online Supplementary Document). Interventions were considered direct if they were nutrition-specific and administered directly to the mother vs. an indirect nutritional intervention, which may seek to address the underlying drivers of undernutrition, and may target household income/food security, health services, water, sanitation, and food production [26]. Studies were required to have included a comparison group that served as a control, and some measure of placental phenotype as an outcome (but not necessarily the primary outcome of that study) (Figure S1 in the Online Supplementary Document). Placental outcomes were categorised into the following five groups: anthropometry (any physical characteristics related to the placenta), molecular (any alterations to placental genotype and/or gene expression, or other molecular changes), pathology (any findings related to placental histopathology), placental abruption (cases of placental abruption), and placenta-related disease (any disease with placental origins, or that significantly impacts the placenta) (Table S2 in the Online Supplementary Document).
Sources and screening process
We searched PubMed, ClinicalTrials.gov, and the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) to identify peer-reviewed publications using predefined search terms. Two authors (VB and MW) conducted the three-level process of title, abstract, and full-text screening (Methods in the Online Supplementary Document).
Data extraction
We extracted the following data from the included studies: study location, maternal clinical characteristics and demographics, type of pregnancy (singleton or twin), and nutritional intervention details (type (micronutrient, macronutrient, and/or diet- and lifestyle-based), composition, timing, dose, and compliance data) (Methods in the Online Supplementary Document). Data on maternal comorbidities reported in the studies under review were noted and used to inform results interpretations. We classified the reported adverse outcomes into two categories: adverse effects (outcomes that were suspected to be in response to the intervention [27]) and adverse events (outcomes that were not suspected to be in response to the intervention [27]). We also summarised data on associations between the nutritional intervention and placental phenotype (primary outcome); placental sex and fetal/infant and maternal outcomes (secondary outcomes); and the proportion of studies that did or did not report placental changes or improvements in maternal and offspring outcomes was calculated for each subtype of nutritional intervention (micronutrient, macronutrient, and diet- and lifestyle-based).
We considered study outcomes to be ‘improved’ if the adverse clinical or physiological condition was prevented or bettered in mothers or offspring. Examples included a reduced preeclampsia or preterm birth risk, or improved maternal nutritional status or fetal growth.
Risk of bias assessments
Two authors (VB and MW) assessed the included studies for risk of bias using the Cochrane Collaboration’s Tool for Assessing Risk of Bias I for randomised studies (n = 50) (Table S3 in the Online Supplementary Document) and the Risk of Bias in Non-randomised Studies of Interventions (ROBINS-I) tool for non-randomised studies (n = 3) [28,29] (Methods in the Online Supplementary Document). Discrepancies were resolved through discussion with a third author (KLC). While RoB in interventions was not a primary focus of this review, assessments of RoB were performed to inform results interpretations.
Statistical analysis
We analysed the data using JMP, version 16.0. (SAS Institute, Cary, NC, USA) Relative risk (RRs) and 95% confidence interval (CIs) were calculated using a 2 x 2 table analysis with P-values from Fisher exact test (two-tail) to determine whether nutritional interventions that improved maternal or offspring outcomes were more likely to be associated with placental changes than interventions that did not improve maternal or offspring outcomes. Due to the high heterogeneity of the interventions administered, placental characteristics measured, and statistics reported across studies, we could not conduct a meta-analysis of intervention effects on placental outcomes, so we summarised the result only descriptively.
RESULTS
Our search retrieved 5299 titles from PubMed (n = 5078), ClinicalTrials.gov (n = 156), and WHO ICTRP (n = 65). Following deduplication and title (level 1) and abstract screening (level 2), 164 records remained for full-text screening (level 3). After exclusion for not including a placental measure (n = 64, including articles reporting on relationships between a nutritional intervention and risk of preeclampsia that did not provide placenta-specific measures), not administering a direct nutritional intervention (n = 9), duplicates (n = 11), or other reasons (i.e. wrong study type, no trial results available or could not be found (n = 21)), 53 articles remained that met our inclusion criteria (Figure 1). Definitions of key technical terms are provided in Table 1.
Figure 1. PRISMA flow diagram for article selection.
Table 1. Definitions of key technical terms
| Term | Definition | Relevance to review objectives |
|---|---|---|
| Amnion-choriodecidua separation | The separation of a protective membrane surrounding the fetus filled with fluid (amnion) and the location of fetal attachment to the placenta (chorion decidua). | Separation that persists after ~ 16 weeks’ gestation can increase risk of preterm birth or stillbirth. |
| Cerebro-placental ratio | A measure of how blood flows to the fetal brain vs. the placenta; derived from the doppler index (a measure of how blood flows) of the middle cerebral artery compared to the umbilical artery. | An abnormal (lower) ratio is caused by changes in vascular dilation and placental resistance, and is associated with preterm birth risk, lower birth weight, and fetal distress. |
| Chorioamnionitis | Infection and inflammation of the amnion and chorion (fetal membranes); characterised by intra-amniotic neutrophil infiltration. | Associated with adverse pregnancy and fetal outcomes, including increased risk of preterm birth. |
| Histomorphometry | The quantitative analysis of measurable traits in tissues (i.e. size, area). | Placental histomorphometry can provide insight into placental development and function. |
| Histopathology | Atypical results confirmed via microscopic analysis. | Positive histopathological finding indicates atypical results (such as the presence of a disease). |
| Hypertension | High blood pressure. | Hypertension is associated with increased risk of maternal and fetal mortality, preterm birth, and preeclampsia. |
| Placental abruption | Separation of the placenta from the uterus during pregnancy. | Can lead to maternal and/or fetal mortality, depending on the stage of pregnancy. |
| Preeclampsia | A placental-mediated hypertensive disorder that can occur after 20 weeks of pregnancy. | A leading cause of maternal and fetal mortality. |
| Premature rupture of membranes | Rupturing of the amniotic sac before labour contractions begin. | Can increase risk of maternal and fetal infection and mortality. |
| Twin-to-twin transfusion syndrome | A condition where there is an imbalance in blood flow and nutrients between twins that share a single placenta. | Can have harmful effects on both fetuses through either lack of or excess blood flow, such as insufficient nutrient intake. |
Study and cohort characteristics
The 53 included studies were randomised controlled clinical trials (n = 43, 81%), interventional clinical trials (n = 7, 13%), controlled feeding studies (n = 2, 4%), and intention-to-treat studies (n = 1, 2%). They administered micronutrient- (n = 33, 62%), lipid- (n = 11, 21%), protein- (n = 3, 6%) and diet and lifestyle-based (n = 8, 15%) interventions, and reported on cohorts from low-income countries (LICs) (n = 4, 7.5%), low-middle income countries (LMICs) (n = 12, 23%), upper-middle-income countries (UMICs) (n = 9, 17%), and high-income countries (HICs) (n = 33, 62%) (Figure 2, Figure 3). Five studies reported findings from multiple cohorts, which included participants from countries with differing socioeconomic statuses.
Figure 2. Geographical distribution of the cohorts included in the 53 articles reviewed.
Figure 3. Income status of study location. Panel A. Placental changes. Panel B. Types of nutritional interventions in the studies included in the review. Diagram created on app.rawgraphs.io website. DLB – diet and lifestyle-based, HIC – high-income country, LIC – low-income country, LMIC – low-middle income country, n – number of studies, UMIC – upper-middle income country, PB – protein-based.
Most studies (n = 46, 87%) included only singleton pregnancies, while the remaining ones (n = 7, 13%) included both singleton and twin pregnancies. Maternal comorbidities were common in the included studies (n = 33, 62%) and were mainly hypertensive or metabolic disorders in pregnancy (Results in the Online Supplementary Document).
Interventions began either peri- (n = 3, 6%) or post-conceptionally (n = 50, 94%), with 14 beginning in the first trimester, 28 in the second trimester, and 11 in the third trimester). The majority continued until birth (n = 52, 98%), while one (2%) finished after 12 weeks of administration.
Risk of bias assessments
All 50 randomised studies under review had low or unclear risk of selection, detection, attrition, and reporting bias (Figure 4). Risk of performance bias was assessed to be high in five trials (10%) where blinding was not possible given the nature of the nutritional intervention. Most studies (n = 38, 76%) had a low risk of bias related to intervention compliance, however, methods for measuring intervention compliance were unclear in over one quarter of the studies under review (n = 14, 26%), and one study was assessed to have a high risk of bias due to limited participant follow up and no reported compliance measures [30].
Figure 4. Quality assessments of randomised studies using the Cochrane tools. Panel A. Randomised studies using Risk of Bias I. Panel B. Non-randomised studies using the ROBINS-I assessment tool.
Of the three non-randomised studies, one had an overall moderate and two had an overall serious risk of bias (Figure 4). Domain-specific contributors to a serious risk of bias were a lack of controlling for confounding variables [30] and methods used for selection of participants into the study [31]. All three non-randomised studies had a moderate risk of bias due to potential deviations from the intended interventions, which was generally due to a lack of recording or reporting on intervention compliance.
We used the results from the risk of bias assessments to inform results synthesis and interpretations if the study reported significant associations between its nutritional intervention and placental phenotype.
Nutritional interventions associated with improved maternal and infant outcomes
Forty studies in total reported maternal and/or infant outcomes, with 20 (50%) and 10 (25%) reporting only maternal or infant outcomes, respectively, and 10 (25%) reporting both (Tables S5 and S6 in the Online Supplementary Document). Maternal and offspring outcomes were reported as being improved following nutritional intervention in 12/30 (40%) and 7/20 (35%) of the studies, respectively, that reported these data (Figure 5, Panels A and B; Results in the Online Supplementary Document).
Figure 5. Nutritional interventions influence maternal, offspring, and placental outcomes. Mosaic plots are proportion (%) of studies reporting improvement, or no improvement, in maternal (Panels A and C) and offspring (Panels B and D) outcomes. Data in Panels C and D are relative risk (95% confidence interval). Panel A. Of the 13 nutritional interventions that associated with improved maternal outcomes, one was diet and lifestyle-based, one was protein-based, four were lipid-based, and seven were micronutrient-based. Mosaic plots are proportion (%) of studies reporting improvement, or no improvement, in maternal outcomes. Panel B. Of the seven nutritional interventions that associated with improved infant outcomes, two were diet-based, one was protein-based, two were lipid-based, and two were micronutrient-based. Panels C and D. Nutritional interventions that improved maternal (C) and offspring (D) outcomes were more likely to associate with placental changes than interventions that did not improve maternal or offspring outcomes. Panel E and F. Visual summary of the nutritional interventions and beneficial maternal (E), offspring (F) and placental (E and F) outcomes reported in the studies under review. *Significance – P < 0.05 from Fisher exact test (two-tail). DHA – docosahexaenoic acid, EPA – eicosapentaenoic acid, FeFol – iron-folic acid, L-Arg – l-arginine, Vit – vitamin, RR – relative risk, SQ-LNS – small-quantity lipid-based nutrient supplements.
Classifications of placental outcomes
Of the 53 included studies, 29 (55%) reported on more than one placental outcome across the five categories, for a total of 88 outcomes across all studies: 19 (22%) of the studies reported on placental anthropometry, 24 (27%) on placental molecular changes, five (6%) on placental pathology, 19 (22%) on placental abruption, and 21 (24%) on a placental-related disease (Table S7 in the Online Supplementary Document).
Direct nutritional interventions associate with placental phenotype
Specific micronutrient-based interventions associate with placental phenotype
Of the 33 studies that used a micronutrient-based intervention, 16 (48%) reported associations between the nutritional intervention and placental phenotype (Figure 6, Figure 7). Micronutrient interventions that were associated with altered placental phenotype began during the first (n = 3), second (n = 9), and third trimesters (n = 4).
Figure 6. Graphical Overview for Evidence Reviews (GOfER) diagram of the impact of micronutrient-based direct maternal nutritional interventions on placental phenotypic alterations (part 1). Income status of study location obtained from World Development Indicators by The World Bank Database. A dash (-) under ‘Details’ indicates no significant placental phenotype alterations. Circles for this review’s secondary outcomes were only placed if the study reported fetal/infant or maternal outcomes as a primary outcome of their study. Anth – anthropometry, GDM – gestational diabetes mellitus, HIC – high-income country, LIC – low-income country, LMIC – low-middle-income country, MMN – multiple micronutrients, Molec – molecular alterations, NSm – non-smoker, PEB – protein-energy ball, Pre – pre-conception, Sm – smoker, UMIC – upper-middle-income country, PA – placental abruption, Patho – pathology, PD – placenta-related disease.
Figure 7. Graphical Overview for Evidence Reviews (GOfER) diagram of the impact of lipid-based direct maternal nutritional interventions on placental phenotypic alterations (part 2). Income status of study location obtained from World Development Indicators by The World Bank Database. A dash (-) under ‘Details’ indicates no significant placental phenotype alterations. Circles for this review’s secondary outcomes were only placed if the study reported fetal/infant or maternal outcomes as a primary outcome of their study. Anth – anthropometry, HIC – high-income country, L-arg – L-arginine, LIC – low-income country, LMIC – low-middle-income country, MTHF – 5-methyl-tetrahydrofolic acid, Molec – molecular alterations, Pre – pre-conception, UMIC – upper-middle-income country, PA – placental abruption, Patho – pathology, PD – placenta-related disease. *Due to a crossover between intervention categories, Vadillo-Ortega et al. [82] appears in Figure 7 and Figure 9, and Larqué et al. [41] appears in Figure 7 and Figure 8.
Two studies reported on associations between micronutrient-based nutritional interventions and placental anthropometry, including increased placental weight at term (following daily vitamin C and zinc supplementation initiated during the first trimester in a population at high risk for malaria and low birthweight [32]), and increased chorionic villi density (following vitamin D supplementation that began in the late first or early second trimester in mothers at high risk of having offspring with asthma [33]).
One study reported on associations between micronutrient-based interventions and placental pathology [34], where daily iron supplementation initiated periconceptionally was associated with a decreased risk of chorioamnionitis (a maternal inflammatory response leading to the infiltration of neutrophils into the placenta) in a population at high risk for malaria and chorioamnionitis [34].
Half (n = 8) of these 16 micronutrient-based intervention studies reported on differences in placental molecular changes between the control and intervention groups. Placental molecular changes included increased expression or activity of the following: placental superoxide dismutase (an antioxidant enzyme [35]; following daily vitamin C supplementation initiated upon gestational diabetes diagnosis [36]), transferrin receptor 1 mRNA (involved in placental iron uptake [37,38]; following a daily multiple micronutrient intervention that began in the second trimester [38]), genes involved in vitamin D metabolism (after daily vitamin D supplementation that started in the third trimester [39]), soluble fms-like tyrosine kinase-1 (SFLT-1; following daily vitamin D plus 5-methyl-tetrahydrofolic acid (MTHF) supplementation that started in the second trimester [40]), and placental phospholipid DHA (after daily 5-MTHF and docosahexaenoic acid (DHA) supplementation that began in the second trimester [41]). Molecular changes also included modified expression of five genes in placental tissue (increased expression of metastasis associated in colon cancer-1 (MACC1), integrator complex subunit 9 (INST9), Von Willebrand factor (vWF), and age-related maculopathy susceptibility 2 (ARMS2); and decreased expression of contactin 5 (CNTN5); following daily vitamin D supplementation initiated in the second trimester [33]). Two studies from a single trial of daily maternal choline supplementation initiated during the third trimester [42] reported on increased methylation of multiple cortisol-regulating genes and dysregulation of 197 different placental biological processes [43,44].
Sixteen studies assessed placental abruption risk following nutritional intervention, where two interventions, which supplemented either daily vitamin C + E [45] (initiated during the first trimester) or magnesium citrate [46] (initiated during the second trimester) were associated with decreased abruption risk. In contrast, two studies from the Diabetes and Preeclampsia Intervention Trial (DAPIT) [47], which supplemented daily vitamin C + E (that began in the first trimester) in mothers with diabetes to lessen preeclampsia risk, reported no differences in placental abruption risk in the full cohort [47] or in follow-up studies with a cohort subset [48]. Two studies from the Combined Antioxidant and Preeclampsia Prediction Studies (CAPPS) trial reported that placental abruption risk was decreased in mothers who smoked [45], but not the intervention group overall [49], following daily vitamin C+E supplementation that started in the first trimester.
Placental-related disease risk was altered in 5 (33%) of the 15 studies that evaluated placental-related disease occurrence following nutritional intervention. These findings included increased risk of premature rupture of membranes (PROM) (following daily vitamin C + E supplementation initiated in the second trimester [50]), and a decreased risk of histopathological-positive placental malaria risk (following daily vitamin A and zinc supplementation that started in the third trimester [32]), and preeclampsia (following daily interventions with either folic acid [51] that began pre-conceptionally, or calcium [52] initiated in the second trimester). In the CAPPS trial [53], no changes in the risk of preeclampsia [45,49], or amnio-choriodecidua [54], were reported following daily vitamin C + E supplementation initiated in the first trimester.
The remaining 17 studies that used a micronutrient-based intervention reported finding no associations between micronutrient interventions and placental anthropometry (n = 5, 29%) [55–58], pathology (n = 1, 6%) [58], molecular changes (n = 5, 29%) [54–56,59,60], abruption risk (n = 8, 47%) [47–49,61–65], or placental-related disease (n = 8, 47%) [47,49,58,61–63,66,67].
Specific macronutrient-based interventions influence placental phenotype
Of the 11 studies that reported on lipid-based interventions, nine (82%) reported placental changes in the intervention group (Figure 8). Lipid-based interventions that associated with altered placental phenotype began during the first (n = 2), second (n = 5), and third trimesters (n = 2).
Figure 8. Graphical Overview for Evidence Reviews (GOfER) diagram of the impact of protein- and diet and lifestyle-based direct maternal nutritional interventions on placental phenotypic alterations. Income status of study location was obtained from World Development Indicators by The World Bank Database. A dash (-) under ‘Details’ indicates no significant placental phenotype alterations. Circles for this review’s secondary outcomes were only placed if the study reported fetal/infant or maternal outcomes as a primary outcome of their study. Anth – anthropometry, EPA – eicosapentaenoic acid, DHA – docosahexaenoic acid, HIC – high-income country, L-arg – L-arginine, LIC – low-income country, LMIC – low-middle-income country, Molec – molecular alterations, MTHF – 5-methyl-tetrahydrofolic acid, N-3 PUFA – omega-3 polyunsaturated fatty acid, UMIC – upper-middle-income country, PA – placental abruption, PD – placenta-related disease, Patho – pathology, Pre – pre-conception, SQLNS – small-quantity lipid-based nutrient supplementation. *Larqué et al. [41] appears in Figure 7 and Figure 8 due to a crossover between intervention categories.
Three lipid-based intervention studies reported on changes in placental anthropometry. Decreased placental weight [68] and fetal/placental weight ratio [69] were reported following daily DHA supplementation that began in the second and third trimester, respectively, while pre-conceptionally initiated daily maternal small-quantity lipid-based nutrient supplementation (SQLNS) was associated with an increase in placental area [70].
Eight lipid-based interventions were associated with molecular changes in the placenta, namely in placental processes related to inflammation and fatty acid transport. The molecular changes included increased expression or concentration of: placental linoleic acid (following daily fish oil supplementation that started in the third trimester [71]), proliferating cell nuclear antigen (PCNA) which plays a key role in nucleic acid metabolism [72]; following daily fish oil supplementation that began in the second trimester [73], and tumour necrosis factor alpha (TNFα) (following daily omega-3 supplementation that started in the second trimester [74]). Decreased expression of placental inflammatory markers (following daily DHA supplementation initiated in the third trimester [69]), and decreased activated protein kinase alpha ratio (AMPKA), a regulator in cell metabolism (following daily pre-conceptionally initiated SQLNS supplementation [70]) were also reported. Two studies from a single clinical trial that supplemented omega-3 fatty acids (beginning in the second trimester) reported a significant reduction in the expression of pro-inflammatory genes in placental tissue and decreased placental lipid storage capacity [75–77]. Two studies that reported findings from a trial of daily DHA and MTHF supplementation that began in the second trimester noted alterations to DHA content in placental phospholipids and expression of PCNA [73] and alterations to placental fatty acid transport proteins [41].
Two of the 11 (18%) lipid-based interventions did not associate with placental changes. In both interventions, a daily fish oil supplement was taken, beginning in the second trimester, and placental weight [78] and preeclampsia risk [79] were not altered.
Three of the 53 studies used amino acid- or protein-based interventions (Figure 9). Two (66%) of these three studies reported associations between the intervention and placental phenotype. Daily supplementation with l-arginine initiated during the third trimester was associated with an increased cerebro-placental ratio (a gestational indicator associated with adverse pregnancy outcomes like low-birth weight [80]) when compared to the control group [81], and twice daily supplementation with l-arginine and antioxidant vitamins that began in the second trimester associated with a decreased risk of preeclampsia [82].
Figure 9. Graphical Overview for Evidence Reviews (GOfER) diagram of the impact of protein- and diet and lifestyle-based direct maternal nutritional interventions on placental phenotypic alterations. Income status of study location was obtained from World Development Indicators by The World Bank Database. A dash (-) under ‘Details’ indicates no significant placental phenotype alterations. Circles for this review’s secondary outcomes were only placed if the study reported fetal/infant or maternal outcomes as a primary outcome of their study. Anth – anthropometry, DiMO – monochorionic diamniotic twin pregnancy, HIC – High-income country, IPTp-SP – intermittent preventative treatment in pregnancy with sulphadoxine-pyrimethamine, L-arg – L-arginine, LIC – low-income country, LMIC – low-middle-income country, Pre – pre-conception, Molec – molecular alterations, PA – placental abruption, Patho – pathology, PD – placenta-related disease. *Vadillo-Ortega et al. [82] appears in Figure 7 and Figure 9 due to a crossover between intervention categories.
The one amino acid- or protein-based intervention that did not associate with placental phenotype was supplementation of l-arginine initiated in the third trimester, which was not associated with changes in placental weight, abruption risk, or thrombosis [83].
Specific diet and lifestyle-based interventions influence placental phenotype
Eight interventions were diet and lifestyle-based and three (38%) were associated with placental phenotype (Figure 9).
An intervention that delivered a daily probiotic supplement and dietary education and support (to promote a diet in line with nutritional guidelines and maintenance of appropriate fat consumption) that began during the second trimester was associated with increased concentration of polyunsaturated fatty acids in the placenta [84].
Two diet and lifestyle-based interventions were associated with a reduced risk of placental-disease. Daily Ensure® liquid nutritional supplementation initiated during the second trimester was associated with reduced twin-to-twin transfusion syndrome (TTTS) diagnosis and incidence at birth [31]. Haematinics and intermittent preventative treatment (IPTp-SP) administered daily during the third trimester until birth was associated with reduced placental malaria incidence in the treatment group when comparing mothers with the same HIV status [85]. Notably, both studies had a moderate risk of bias due to potential deviations from the intended interventions [31,85] and one had a serious risk of bias due to the methods used for selection of participants into the study [31].
The remaining five (63%) diet and lifestyle-based interventions did not associate with placental changes [30,86–89].
Intervention effects on the placenta vary by study location
Studies that took place in LICs and LMICs reported fewer associations between nutritional interventions and placental changes than studies that took place in UMICs and HICs (Figure 3, Panel A). All nutritional interventions in LICs (n = 4) were micronutrient-based. Of these, two (50%) reported associations between the intervention and placental pathology (n = 1) and placental molecular phenotype (n = 1). Nutritional interventions in LMICs were micronutrient-based (n = 9, 75%), lipid-based (n = 1, 8%), or diet and lifestyle-based (n = 2, 16%), of which five (42%) reported placental changes following the nutritional intervention. The five LMIC-based studies reported on associations between the nutritional intervention and placental anthropometry (n = 2, 29%), placental molecular phenotype (n = 2, 29%), and placenta-related disease (n = 3, 43%).
Nutritional interventions in UMICs were micronutrient-based (n = 8, 89%) or lipid-based (n = 1, 11%); of these, six (66%) reported associations between the intervention and placental phenotype. These six UMIC-based studies reported on associations between the nutritional intervention and placental anthropometry (n = 1, 13%), placental molecular phenotype (n = 1, 13%), placental abruption (n = 1, 13%), and placental-related disease (n = 5, 63%).
Nutritional interventions in HICs were micronutrient-based (n = 16; 47%), lipid-based (n = 16, 47%), protein-based (n = 2, 6%), or diet and lifestyle-based (n = 6, 18%) of which 17 (50%) reported associations between the intervention and placental phenotype. These 17 HIC-based studies reported associations between the nutritional intervention and placental anthropometry (n = 4, 20%), placental molecular phenotype (n = 14, 70%), placental abruption (n = 1, 5%), and placental-related disease (n = 1, 5%).
Interventions associated with placental phenotype are overall more likely to improve maternal and offspring outcomes
Maternal outcomes were more likely to be improved in studies that reported placental changes following nutritional intervention than studies that reported no placental changes (n = 11 vs. n = 2 (RR = 3.6; 95% CI = 1.5, 8.7)). Studies that reported placental changes were more likely to report improved offspring outcomes than those that did not report placental changes (n = 6 vs. n = 1 (RR = 1.9; 95% CI = 1.0, 3.6)) (Figure 5, Panels C and D; Results in the Online Supplementary Document).
Data on sex differences in placental response to nutritional interventions are lacking
Only 17 (32%) of the studies under review included data on the number of male and female offspring in the study, and no studies reported data on placental outcomes stratified by sex.
DISCUSSION
Here we reviewed evidence on how direct maternal nutritional interventions influence placental phenotype in humans. Lipid-based interventions were most likely to be associated with altered placental phenotype, often related to changes in inflammatory processes or markers. We also found that interventions associated with placental changes were more likely to also improve maternal and offspring outcomes. Studies conducted in HICs were the most likely to report associations between nutritional interventions and placental phenotype, although LICs were underrepresented in the reviewed studies. There was a lack of reporting on sex-differences or -specific effects of nutritional interventions on placental outcomes. Investigating how direct maternal nutritional interventions during preconception and pregnancy influence placental phenotype strengthens our understanding of the mechanisms underlying intervention effectiveness and could lead to more targeted approaches to improve maternal-fetal health.
Nutritional interventions associated with placental changes were more likely to also improve maternal and offspring outcomes. Specifically, improved maternal iron levels, inflammatory status, and blood glucose control may be accompanied by placental changes such as increased nutrient transporter expression [38] and decreased inflammation and oxidative stress [36,75]. Similarly, decreased placental abruption risk [45], improved cerebro-placental ratio [81], and greater placental area [70] following nutritional intervention corresponded with a decreased risk of preterm birth and intrauterine growth restriction and improved fetal growth. While it is well-established that nutritional interventions can improve maternal and offspring outcomes [90,91], our findings suggest that understanding the placental response to nutritional interventions may reveal insights into the biological mechanisms underlying their effects on maternal or offspring health.
Several studies reported on changes in maternal, placental, or offspring levels of pro-inflammatory biomarkers following micronutrient- and lipid-based nutritional interventions [36,41,69,71,73–75]. For example, supplementation with vitamin C, which has antioxidative properties [92], was found to be associated with reduced levels of oxidative stress markers in placental tissue and maternal and offspring blood [36]. Excess oxidative stress in the placenta can promote inflammation and adversely affect fetoplacental development and future disease risk [93–96]. Lipid-based interventions associated with increased anti-inflammatory DHA and EPA levels in the placenta and maternal circulation [41,69,71,73,74]. During pregnancy, DHA and EPA promote angiogenesis and reduce oxidative stress and inflammation in the placenta and are required for fetal brain development [97]. These findings collectively suggest that common nutritional interventions can impact placental inflammatory processes, which may in turn influence maternal and offspring outcomes [94–96]. Notably, two of these studies had a high risk of bias due to blinding of participants or study personnel [36,69], and an expression of concern was issued regarding the scientific integrity of one of these studies, after the completion of our review [36,113]. Two others had an unclear risk of bias in half of the assessment categories due to inadequate information in the publications [41,73]. High quality studies that adhere to best practice reporting standards are needed to better understand how maternal nutritional interventions affect placental inflammation, especially in pregnancies with maternal comorbidities like obesity and metabolic syndrome, which are associated with chronic inflammation and altered placental nutrient transport [98,99].
None of the reviewed studies reported placental changes by sex. Given the sex-specific differences in placental development, including in nutrient exchange capacities [100], it is likely that placental sex influences the placental response to nutritional interventions [101]. This is an area requiring further research. Additionally, LICs were underrepresented in the reviewed studies, with only four studies taking place in such contexts. As LICs often face higher rates of undernutrition and are where nutritional interventions are most needed [102,103], further research is needed on how nutritional interventions impact the placenta in these regions.
Strengths of this study include it being the first to synthesise knowledge on the relationships between direct maternal nutritional interventions and placental phenotype, extending beyond placental-related pregnancy complications [104]. Further, our focus on only human studies increases the clinical relevancy of our findings [105], and we identified key gaps in the literature, such as the lack of sex-stratified placental data and underrepresentation of LICs and LMICs, which may guide future research. The limited placental data from nutritional interventions in low-resource settings reflects challenges in implementing such interventions in these contexts more broadly, which often stem from issues related to resource availability and mobilisation [106]. A lack of standardised placental biobanking is another barrier to studying the placental impacts of nutritional interventions in low-resource settings. Ongoing biobanking initiatives, such as the Global Pregnancy Collaboration (CoLab), the PREgnancy Care Integrating Translational Science Everywhere (PRECISE) Network, and the Alliance for Maternal and Newborn Health Improvement (AMANHI) study [107–109], aim to address this gap.
Limitations of this review first include the lack of formal meta-analysis due to the high variability in the types of placental outcomes reported in the reviewed studies. In many cases, the scope and detail of placental data available in the reviewed studies were limited and there was high heterogeneity in how placental outcomes were reported. Second, we were unable to assess how individual-level factors like dietary history, nutritional status, or fetoplacental sex influenced the placental response to an intervention, as these data were not available in the reviewed studies. The lack of measurement or adjustment for these key factors is therefore a likely source of bias across the reviewed studies. Third, intervention type, timing, frequency, and duration also varied, which prevented direct comparisons between studies. Fourth, the reviewed studies were largely concentrated in HICs, limiting the generalisability of our findings to other contexts. Fifth, around one quarter of studies did not report or had unclear compliance data, which may have influenced the intervention’s impact on placental phenotype. Lastly, while we have conservatively highlighted consistent themes from the reviewed studies, it is important to recognise that results varied widely even among studies with relatively similar interventions and outcomes. With the available evidence, it is therefore not currently possible to draw conclusions on the strength of associations observed. If future studies collect and report standardised outcomes and the evidence base on how nutrition interventions influence the placenta expands, future reviews can adopt a more focussed approach to better understand the mechanisms that may contribute to success of specific intervention types and their effects on the placenta.
Advancing nutritional interventions and maternal-child health through placental research
Informed by the findings of this review and elaborating on recommendations from other groups [107–109], we propose the expansion of the existing core outcome set, ‘Pregnancy Nutrition Core Outcome Set’ [110], to include the following new variables:
- Fetal/placental sex
- Placental anthropometry (weight, diameter)
- Placental pathology (where applicable, or banking of fixed placental specimens for future histomorphology and pathology assessments)
- Adverse events and effects related to the placenta
- Collection of placental tissue core samples for future analysis of placental phenotype, including function, if not intended for the immediate study
Standardised reporting of placental and maternal variables is essential for understanding placental responses to nutritional interventions, improving comparability across trials, and reducing outcome heterogeneity. These factors limited our ability to make direct comparisons in this review and have been recognised by others as challenges that core outcome sets can help address [111].
We also recommend increased collection and phenotyping of placental samples from nutritional intervention studies in LMICs and LICs. Aligning with efforts to promote standardised placental biobanking like CoLab, the PRECISE Network, and the AMANHI study [107–109], this would improve representation and understanding of placental mechanisms and adaptations to nutritional exposures in diverse populations and geographies.
Finally, future research should focus on determining optimal windows for nutritional interventions (i.e. interventions initiated pre-conceptionally, in early pregnancy, or in later pregnancy) for improving maternal and offspring outcomes, and how the placenta adapts to, and influences pregnancy outcomes, based on timing of exposure. Research focussed on developing human nutritional intervention(s) targeted to the placenta to, in turn, improve pregnancy/offspring outcomes, should also be a priority. Non-nutritional based therapies (such as the use of liposomes to enhance drug delivery to the placenta) have the potential to improve offspring outcomes via placental changes [112], as do nutritional interventions.
CONCLUSIONS
We found that direct maternal preconception and prenatal nutritional interventions that target maternal health and nutritional status also influence placental development and function. These (potentially unintended) placental adaptations are likely beneficial for the pregnancy and may underlie the improved growth and development of offspring observed in these pregnancies. Understanding how the placenta responds to nutritional interventions is key for modifying and improving strategies to address nutrient deficiencies, reducing maternal and neonatal morbidity and mortality and improving offspring health trajectories globally.