Effect of balanced protein–energy supplementation given to pregnant women on birth outcomes: A systematic review and meta-analysis

Abstract

Background:

There is an increased demand for nutrition during pregnancy to improve fetal growth and development. Several dietary interventions have been recommended to pregnant women to meet their maternal needs. Using a larger sample size, we aim to assess the effect of balanced protein–energy supplementation given to pregnant women on birth outcomes.

Objective:

To evaluate the effect of balanced protein–energy supplementation given to pregnant women on birth outcomes.

Design:

Search included PubMed, Cochrane Central, and Embase from initiation till 20 March 2023 to select relevant studies examining perinatal factors associated with balanced protein–energy supplementation. Mean differences (MD) and risk ratios (RRs) with a 95% confidence interval (CI) were calculated using Review Manager.

Data sources and methods:

Randomized controlled trials and observational studies were included. Out of 218,720 studies initially identified, 24 met the inclusion criteria, involving pregnant women who received balanced protein–energy supplementation and reported outcomes related to perinatal death, birthweight, length, or head circumference.

Results:

Meta-analysis of pooled studies revealed that balanced protein supplementation had a significant effect on birthweight (g) (MD: 62.67, CI: 27.92–97.43), low birth weight (LBW) (RR: 0.73, CI: 0.57–0.95), birth length (MD: 0.20, CI: 0.10–0.30), and small-for-gestational-age (SGA) babies (RR: 0.74, CI: 0.59–0.93) and had no significant effect on the birth head circumference (MD: 0.05, CI: −0.09 to 0.20), perinatal death (RR: 0.83, CI: 0.50–1.37), and preterm birth (RR: 0.86, CI: 0.68–1.08).

Conclusion:

Balanced protein–energy supplementation is an effective intervention for birth length, birth weight, risk of LBW, and SGA births, particularly in women who are undernourished.

Plain language summary

Looking at the effects of giving pregnant women a balanced combination of protein energy supplements on the results of their pregnancy outcomes, by reviewing and analyzing data from several previous studies.

Introduction: During pregnancy, there is a greater need for good nutrition to help the baby grow and develop properly. Different dietary supplements have been recommended for pregnant women to meet their increased nutritional requirements. This study aimed to evaluate the effects of providing a balanced combination of protein and energy supplements to pregnant women on birth outcomes.

Methods: The researchers searched various databases to find relevant studies that looked at the relationship between balanced protein–energy supplements during pregnancy and factors like infant birth weight, length, head size, and risk of complications. They evaluated the quality of the studies and used statistical analysis to combine the results.

Results: The analysis of the 24 studies, involving over 11,000 pregnant women, found that balanced protein–energy supplements had these effects: Increased average birth weight Reduced the risk of low birth weight Increased average birth length Reduced the risk of babies being small for their gestational age However, the supplements did not have a significant effect on: Birth head circumference Risk of perinatal death Risk of preterm birth.

Conclusion: Providing a balanced combination of protein and energy supplements to pregnant women, especially those who are undernourished, can be an effective way to improve certain birth outcomes, like increasing birth weight and length.

Introduction

Fetal mortality, defined as the loss of a fetus during gravidity, remains a significant global health concern. ¹ This includes stillbirths (fetal deaths) and neonatal deaths that happen within the first week of life. ² Approximately 2 million stillbirths occur worldwide each year, but a lot of these mishaps can be avoided through antenatal care of satisfactory quality. ³

During pregnancy, sufficient nutritional intake is indispensable for optimal fetal growth and development. ⁴ To address this need, various dietary interventions, particularly balanced protein–energy supplementation, have been recommended for pregnant women.^5,6 Balanced protein–energy supplementation refers to a nutritional intervention where protein contributes to less than 25% of total energy intake, ensuring sufficient caloric and protein intake without excessive protein load. These supplements aim to improve birth outcomes, especially for undernourished mothers. ⁷

We aim to see through analysis if balanced protein–energy supplementation can decrease preterm birth, which means giving birth before completing 37 weeks of gravidity. The exact gestational age at which preterm birth is distinct from spontaneous abortion is differentiated by circumstances and areas. The United States has a preterm birth rate of approximately 12%–13%, while in Europe and other developed regions, it ranges from 5% to 9%.^8,9 Premature birth poses significant health risks for infants, including conditions such as respiratory distress syndrome, growth irregularities, and lifelong impairments in eyesight and hearing.

Each year around 20 million babies have birth weight below 2500 g, and a substantial proportion, approximately 40%, of these cases occur in developing nations.^10,11 Addressing these concerns, a Cochrane review by Ota et al. has indicated that balanced protein–energy supplementation during pregnancy can tremendously enhance both birth weight and birth length. ¹² In contrast, recent meta-analyses led by Kramer and Kakuma have revealed no significant effects of protein supplementation that is unbalanced on pregnancy results, encompassing birth weight and gestational age. Among supplements for undernourished pregnant women, balanced protein–energy supplementation offering less than 25% of total energy is recognized as the most beneficial. ⁶

Prior studies found that giving balanced protein–energy supplementation was associated with birth weight,^{13 –35} low birth weight (LBW),^{14,16,17,19,24,25,31,32} birth length,^{14 –19,21 –25,28,29,31,33,35} birth head circumference,^{14,16,17,19,22 –25,28,29,33,35} perinatal death,^{14,15,17 –19,25 –27} small-for-gestational-age (SGA) baby,^{14,17 –19,24 –28,31,32,35} and preterm birth.^{14,16 –19,24,26,27,33,35} A systematic review by Ota et al. found there is a suggested likelihood of reduced stillbirths with balanced energy/protein supplementation compared to no supplementation. ³⁶ A previous meta-analysis by Stevens et al. found an association between balanced protein supplementation and birthweight, birth length, and head circumference, but the population consisted of pregnant women from only low and middle-income countries Findings demonstrated a significant increase in birth weight with balanced protein–energy supplementation, while no significant effects were observed on birth length or head circumference. ³⁷ Another meta-analysis by Imdad et al. found an association between balanced protein supplementation and birth weight, risk of lower birth weight, SGA babies, stillbirth, and preterm births. ³⁸ Given the emergence of new studies in this field since the most recent meta-analysis, our objective is to provide an updated analysis, encompassing all relevant research published since 1973.We intend to thoroughly assess the influence of administering well-balanced protein–energy supplements to expectant mothers on birth outcomes.

Methods

Data sources and search strategy

The study was reported following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines (Supplemental Table 1). ³⁹ Two separate investigators performed a literature search on databases, including PubMed, Cochrane Central, and Embase, spanning from their initiation to March 20, 2023. The search strategy used: (Protein supplements OR protein supplementation) AND (primigravidae OR pregnant women) AND (fetal loss OR perinatal death OR neonatal death OR fetal death OR stillbirth OR fetal mortality OR birth weight OR low birth weight OR birth length OR birth head circumference OR preterm birth OR small for gestational age baby). We initially applied built-in database find randomized controlled trials (RCTs) and observational studies to refine our search results. However, we acknowledge that built-in filters may introduce limitations due to indexing inconsistencies, potentially excluding relevant studies. To mitigate this, we conducted manual screening of search results and references to ensure a comprehensive selection of studies. The keywords used for the literature search included “Protein supplementation,” “Pregnant women,” “Perinatal death,” “Birth weight,” “Birth length,” “Birth head circumference,” “Small for gestational age baby,” and “Preterm birth.” The following MESH terms were included: (fetal loss OR fetal death) (perinatal OR neonatal death). Detailed search strategy in presented in Supplemental Table 2.

Study selection and inclusion criteria

RCTs and observational studies (e.g., cohort study or case–control study) were included. Quasi-randomized trials (cross-over trials), case reports, reviews, study protocols, comments, or posters as they may introduce bias or lack the necessary controls for causal inference were excluded. The inclusion criteria encompassed pregnant women who received balanced protein–energy supplementation, which means supplementation where protein provided <25% of the total energy content. This criterion ensures consistency in the type of supplementation under consideration. Studies were excluded where the primary intervention involved providing diet health education to gravidas to improve their protein consumption, high protein supplementation (where protein provides greater than or equal to 25% of the sum of energy content), isocaloric protein supplementation (where protein replaced an equal amount of non-protein–energy content), or poor-energy nutrition for gravidas who were either overweight or had experienced significant weight gain early in pregnancy. Eligible control included “routine,” and “no intervention” as this would enable meaningful comparison with the intervention group. The main measure of outcome was perinatal death but in the absence of perinatal death data, we used stillbirth, fetal loss, or fetal death as relevant outcomes. We incorporated studies that involved anthropometric measurements of intrauterine growth, such as birth length, birth weight, and birth head circumference. To identify and eliminate duplicate articles from the systematic search, we utilized Endnote (Clarivate Analytics, Thomson Reuters Corporation, Philidelphia, USA).

Data extraction and outcomes

Two reviewers independently gathered appropriate data using a predefined data collection table. Any disparities between the reviewers were resolved by an impartial third reviewer. For all qualified articles, information including the primary author’s name, study design, publication year, sample size, country of origin, and both raw and adjusted data related to fetal outcomes were extracted and recorded in a prearranged data extraction sheet. In cases where data were missing for certain participants, we assumed that the data were missing at random and performed sensitivity analyses to assess the impact on our results. Fetal outcomes such as birth weight, length, head circumference, SGA babies, perinatal death, and preterm birth were included. For the extraction of data on perinatal death studies providing information on fetal loss, fetal death, and/or stillbirth were also considered.

Statistical analysis

Review Manager, version 5.4 (The Nordic Cochrane Centre, The Cochrane Collaboration), for all our analyses, was employed. The summary estimations for the meta-analyses were depicted as relative risk (RR) for categorical data and mean difference (MD) for continuous data, accompanied by 95% confidence intervals (CIs). Statistical significance was determined using a p-value threshold of 0.05. Substantial heterogeneity was observed among the studies, with an I² value exceeding 50%. To explore potential variations in fetal outcomes based on the study design, we performed a subgroup analysis. This analysis aimed to investigate whether the impact on fetal outcomes differed between undernourished pregnant women and adequately nourished women. Additionally, we performed a leave-1-out sensitivity analysis to examine if any individual study disproportionately influenced the combined results. Furthermore, we performed a sensitivity analysis using adjusted data from studies that provided such data. This analysis allowed us to assess the potential influence of adjusted data on the overall study findings. Funnel plots of key outcomes that showed significance as per analysis were used to explore the publication bias. Two researchers autonomously evaluated the risk of bias in the individual studies with the help of the Cochrane Risk of Bias Tool (RoB 2.0) across several domains: sequence generation, allocation concealment, blinding of personnel, selective reporting, blinding of outcome assessment, handling of incomplete data, and other potential biases. If any discrepancies arose, a third researcher was consulted to reach an understanding. The Risk of Bias assessment was categorized for each domain as either low, moderate, or high. In our analysis, we found that the risk of bias was low. Grade criteria were used to evaluate the quality of evidence. According to this, overall evidence quality was adjudged as “high,” “moderate,” “low,” or “very low.” ⁴⁰

Results

Baseline characteristics

The initial literature search yielded 218,720 records. After a thorough evaluation of 82 potentially eligible articles, we ultimately included 24 studies in our analysis. The PRISMA flowchart visually illustrates the study selection process (Figure 1). The sample sizes of the included studies ranged from 20 to 1708 participants, with a total of 11,305 participants. Of these, 7171 were categorized as adequately nourished, while 4134 were identified as malnourished. Table 1 presents the baseline characteristics, study design, and assessed outcomes of the included studies. Additionally, Table 2 outlines the reasons for the excluded studies from the analysis.

Figure 1.

Flowchart of the systematic review process.

Table 1.

Summary of the included studies in the systematic review (N = 24).

Study details	Study design	Participants	Intervention	Control	Measured outcomes
Blackwell et al. 24	RCT	The study included pregnant women in their third trimester, with at least one son, planning to have another child, and from low-socioeconomic backgrounds. All participants were consuming their typical protein intake at the start of the study.	800 kilocal (around 40 g of protein).	Vitamins and minerals only.	SGA, birth length, birth weight, LBW, birth head circumference, and preterm birth.
Ceesay et al. 25	RCT	Rural population. Chronically undernourished women and this condition exacerbates during the hungry season (June–October).	Total energy: 1017 kcal and 22 g protein.	No supplementation.	Perinatal death, SGA, birth length, birth weight, LBW, birth head circumference.
Elwood et al. 35	RCT	The study was conducted in two South Wales towns with contrasting socioeconomic profiles: one industrialized with a well-off population, and the other predominantly low-income. All participants were considered to have adequate nutrition.	Complimentary tokens of 1/2-pint of milk.	No intervention.	SGA, birth length, birth weight, preterm birth, birth head circumference.
Metcoff et al. 20	Clinical Trial	The study included 900 pregnant women who attended prenatal clinics at the OMH and were eligible for participation based on Oklahoma’s criteria for WIC. These participants were considered to be malnourished.	40–50 g of protein and 900–1000 calories.	Regular diet.	Birth weight.
Mora et al. 26	RCT	Women from lower socioeconomic backgrounds were recruited during their first or second trimester of pregnancy and received supplements in their third trimester.	total energy: 856 kcal.	No supplement.	Perinatal death, SGA, birth weight, preterm birth.
Rush et al. 27	RCT	The study focused on low-income, English-speaking Black women who were identified as being at risk for delivering LBW infants. The criteria for determining this risk included: (1) a pre-pregnancy weight of less than 110 lbs; (2) a pre-pregnancy weight between 110 and 139 lbs combined with inadequate gestational weight gain at the time of enrollment; (3) a pre-pregnancy weight of 110–139 lbs along with a prior history of low-birth-weight delivery; or (4) a pre-pregnancy weight of 110–139 lbs with protein intake below 50 g in the 24 h before registration. Participants in this study were classified as malnourished.	322 kcal (6 g protein).	Only vitamins and minerals.	Perinatal death, SGA, birth weight, preterm birth.
Girija et al. 28	RCT	Lower socio-economic women aged 20–33. Participants of this study were considered malnourished.	417 kcal and 30 g of protein.	Routine diet.	SGA, birth length, weight, and head circumference.
Atton and Watney 29	RCT	Trial has conducted on a subset of Asian women with triceps skinfold increases of 2 mm from 18 to 28 weeks. Participants of this study were considered malnourished.	Energy content—407 and kcal; Protein content—14.6 g.	Routine diet.	Birth weight, birth length.
Huybregts et al. 17	RCT	Pregnant women residing within the study area. All participants were considered to be malnourished.	A spread that meets the Recommended Daily Allowance.	Only multi-micronutrients.	SGA, birth length, birth weight, LBW, birth head circumference, preterm birth, and perinatal death.
Kaseb et al. 30	Clinical Trial	The study included 53 healthy mothers who were not addicted to any substances, did not take any medication, and had no medical conditions. All participants were considered to have adequate nutrition,	400 kcal of energy and 15 g of protein.	No supplementation.	Birth weight.
Mardones-Santander et al. 31	RCT	The study included pregnant women classified as underweight, meaning their weight-for-height ratio at the first prenatal visit (at an average gestational age of 14 weeks) was below the recommended level for those weighing less than 95% of the standard at 12 weeks of pregnancy. Additional inclusion criteria were: age over 18 years, parity between 0 and 5, pregnancy duration of less than 20 weeks based on the last menstrual period, and absence of smoking or alcohol consumption. Women carrying multiple pregnancies were excluded at the time of delivery. Participants in this study were considered malnourished.	A balanced protein supplement with approximately 22% energy content.	Powdered milk.	SGA, birth length, birth weight, LBW.
Prentice et al. 32	Before and after study design	All pregnant women residing in the study area were included. Due to low rainfall, harvests were often insufficient to sustain the population throughout the year in this region, leading to a period of food scarcity between July and September.	Protein content of 17.4 and 2.9 g. The energy content is 468 and 78 kcal.	No supplementation.	SGA, birth weight, LBW.
Ross et al. 13	RCT	Black women <20 weeks of gestation. Participants of this study were considered adequately nourished.	Combined energy: 700–800 kcal. Combined protein: 36–44 g.	Placebo tablets.	Birth weight.
Brown 33	RCT	Aberdeen primiparous women at high risk of LBW delivery starting approximately at 27- week gestation. The risk of LBW was based on low maternal height, weight or weight-for-height at 20 weeks, or weight gain between 20 and 30 weeks. Participants of this study were considered as malnourished.	Supplement: Energy content: 300 kcal. Protein content: 15–20 g.	Routine diet.	Birth length, birth weight, birth head circumference.
Viegas et al. 34	RCT	The study included Asian women residing in Birmingham, UK, who were less than 20 weeks pregnant and appeared to have adequate nutrition based on their weight and height.	273 kcal energy (protein provides 11% of energy).	Ascorbic acid and iron.	Birth weight.
Viegas et al. 34	RCT	Asian women, <20 weeks of gestation (who appeared well-nourished prior to pregnancy but were later considered “nutritionally at risk” based on an inadequate increase in triceps skinfolds between 18 and 28 weeks) stratified at 28 weeks according to increase in triceps skinfold during the second trimester (0.02 versus >0.02 mm/week).	425 kcal energy (protein provides 10% of energy).	Ascorbic acid and iron.	Birth weight.
Janmohamed et al. 19	Cluster-randomized trial	Of Cambodian women who were at least 18 years of age and were in their first trimester. Participants of this study were required to stay in their home village for the duration of their pregnancy.	The everyday CSB Plus ration gave 760 kcal, in which 27 g was 14% of total kcal.	60 mg Iron and 400 mg folic acid tablets.	SGA, birth length, birth weight, LBW, birth head circumference, preterm birth, and perinatal death.
Dwarkanath et al. 15	RCT	Indian women from the same dwelling area and similar socioeconomic background who had no other co-morbidities or chronic illnesses (e.g., diabetes mellitus, hypertension, cardiac disease, thyroid disease, or epilepsy). Women who tested positive for hepatitis B, HIV, or syphilis were excluded from the study. Participants included pregnant women in their first trimester (<13 weeks) with low BMI (below 18.5) and normal levels of plasma folate and vitamin B-12.	Supplement gave 300 kcal/day, in which 20% energy from 15 g/day protein.	Routine diet.	Perinatal death, birth length, birth weight.
Ashorn et al. 18	RCT	Pregnant women who had completed weeks of gestation were residents of the defined catchment area, were available during the period of the study, and signed or thumb-printed informed consent. Exclusion criteria were 15 years, need for frequent medical attention due to a chronic health condition, diagnosed asthma treated with regular medication, severe illness that warranted hospital referral, history of allergy to peanuts, history of anaphylaxis, or serious allergic reaction to any substance that required emergency medical care.	Lipid nutrition supplementation with composition of MMN capsules.	Micronutrient pills.	SGA, birth length, birth weight, LBW, preterm birth, and perinatal death.
Nga et al. 16	RCT	The study included non-pregnant women aged 18–30 who registered to marry. Women were excluded if they were already pregnant, had a history of specific illnesses (HIV, tuberculosis, malaria, diabetes, heart, or kidney disease), or had previously given birth. Participants were withdrawn from the study if they did not conceive within 1 year.	The energy of the supplement was 190 kcal (10% of the everyday energy requirement).	Only prenatal care.	Birth length, birth weight, LBW, birth head circumference, preterm birth.
Lanou et al. 21	RCT	Pregnant women were recruited from Houndé (Tuy province), a rural health district. Women were predominantly young: mean (SD) age = 24.1 (6.3) years, 18.7% were nulliparous, and 89.4% did not attend school. The mean (SD) gestational age was 16.6 (6.7) weeks at recruitment and 39.0 (2.5) weeks at delivery. The nutritional status of the participants was suboptimal: 12.0% had a BMI of 18.5, and 44.9% were anemic (hemoglobin, 110 g/L).	372 kcal and 14.7 g of protein.	MNM cocktail.	Birth weight, birth length.
Huynh et al. 22	RCT	Healthy pregnant women aged between 20 and 35 years, first-time mothers with a singleton pregnancy from 26 to 29 weeks of gestation and pre-pregnancy BMI <25.0 were eligible for the study.	252 kcal (16.8 g of protein)	IFA: Fe (60 mg) and folic acid (400 mcg).	Birth weight, birth length, birth head circumference.
Tran et al. 23	RCT	Healthy and pregnant women, 20–35 years of age, first-time mothers with singleton pregnancies, at 26 to 29 weeks of gestation, and with pre-pregnancy BMI <25.0 (not overweight or obese).	Macronutrients and micronutrients.	Folic acid and Fe supplementation.	Birth weight, birth length, birth head circumference.
de Kok et al. 14	RCT	The study included pregnant women between the ages of 15 and 40 years who were less than 21 weeks pregnant	Supplement 393 kcal (20% energy from protein).	65 mg iron and 400 μg folic acid.	SGA, birth length, birth weight, birth head circumference, preterm birth, and perinatal death.

OMH: Oklahoma Memorial Hospital; WIC: women, infants, and children; SGA: small-for-gestational-age; BMI: body mass index; LBW: low birth weight; RCT: randomized controlled trial; SD: standard deviation; CSB: corn soya blend.

Table 2.

Excluded studies characteristics.

Study details	Reason for exclusion
Kardjati et al. 41	Supplement was provided to treated and control group
Lechtig et al. 42	Energy drink was provided to control group
Chappell et al. 43	Diet health advice only
Hankin 44	Diet health advice only
Hunt et al. 45	Diet health advice only
Kafatos et al. 46	Diet health advice only
Sweeney et al. 47	Diet health advice only
Iyengar 48	Isocaloric supplementation provided
Rasmussen and Habicht 49	More energy than required in the supplement
McDonald et al. 50	Supplementation was started before pregnancy
Meghan et al. 51	Both groups were given nutrition supplement
Okubo et al. 52	Maternal dietary habits were monitored

Quality and risk of bias assessment

According to the quality evaluation of the studies that were included (Supplemental Figure 1), 1 study has a low risk of bias, 1 study has a moderate risk, and 22 studies have a high risk.

Results of meta-analysis

The combined findings from the studies regarding birthweight (g) reveal a statistically significant moderate impact of supplementation (MD: 62.67, CI: 27.92–97.43) (Figure 2). Notably, there was substantial heterogeneity of 81%, prompting the utilization of random-effects models. In one of these studies, despite the fact that the authors classify the supplement as balanced, it had a higher percentage of protein-based energy (>25%). ²⁸ Upon excluding this study, although slightly diminished, the impact of supplementation was still statistically significant (MD: 61.07, CI: 26.12–96.02, I²: 82%) (Supplemental Figure 2). Furthermore, in the intervention group of two studies, balanced protein–energy supplements were administered alongside micronutrient supplements, with micronutrient supplements also administered to the control group. ²¹ Excluding these studies increased the effect of supplementation (MD: 69.02, CI: 33.60–104.44, I²: 78%) (Supplemental Figure 3). According to a stratified analysis based on the nutritional status of mothers, the effect of protein supplementation was significant in both malnourished (MD: 96.17, CI: 53.00–139.34, I²: 70%) and adequately nourished (MD: 36.16, CI: 9.60–62.73, I²: 30%) women (Figure 2).

Figure 2.

Analysis of comparison: balanced protein–energy supplementation versus control, outcome: birth weight.

According to data on LBW incidence (birthweight <2500 g), balanced protein–energy supplementation reduced LBW by 27% (RR: 0.73, CI: 0.57–0.95) (Figure 3). Since there was heterogeneity in the pooled data (I²: 52%), random-effect models were applied.

Figure 3.

Analysis of comparison: balanced protein–energy supplementation versus control, outcome: low birth weight (<2500).

Data from studies on birth length (cm) indicate that supplementation had a significant moderate effect on birth length (MD: 0.20, CI: 0.10–0.30, I²: 12%) (Figure 4). One of these studies was excluded because the intervention supplement contained more energy on protein (>25%) despite the fact that the authors saw that the supplement had a modest but significant effect of supplementation (MD: 0.19, CI: 0.10–0.28, I²: 0%) (Supplemental Figure 4). ²⁸ An analysis of the subgroups depicted that the effect of protein supplementation was significant in only adequately nourished women (MD: 0.23, CI: 0.11–0.35, I²: 16%) and not in malnourished women (MD: 0.09, CI: −0.08 to 0.27, I²: 0%) (Figure 4).

Figure 4.

Analysis of comparison: balanced protein–energy supplementation versus control, outcome: birth length.

Regarding birth head circumference (cm), protein supplementation had no significant effect overall (MD: 0.05, CI: −0.09 to 0.20, I²: 67%) (Figure 5). Subgroup analysis revealed no significant changes in birth head circumference for malnourished (MD: 0.08, CI: −0.16 to 0.33, I²: 35%) and adequately nourished women (MD: 0.05, CI: −0.13 to 0.23, I²: 75%) (Figure 6). However, one study, where the intervention included supplemental foods made from animal sources and locally grown dark green leafy vegetables, showed a significant effect on birth head circumference when excluded (MD: 0.11, CI: 0.01–0.21, I²: 33%) (Supplemental Figure 5). ¹⁶

Figure 5.

Analysis of comparison: balanced protein–energy supplementation versus control, outcome: birth head circumference.

Figure 6.

Subgroup analysis of balanced protein–energy supplementation in adequately nourished women versus malnourished women. Outcome: birth head circumference.

Regarding perinatal death, protein supplementation did not impact the risk (RR : 0.83, CI: 0.50–1.37, I²: 62%) (Figure 7). Even after excluding a study involving corn soya blend (CSB) supplementation, there was no significant difference (RR: 0.73, CI: 0.48–1.08, I²: 34%) (Supplemental Figure 6). ¹⁹ Subgroup analysis revealed a 44% decrease in the risk of perinatal deaths in malnourished women (RR: 0.56, CI: 0.32–0.99, I²: 0%), while no significant effect was noted in adequately nourished women (RR: 1.09, CI: 0.56–2.10, I²: 22%) (Figure 8).

Figure 7.

Analysis of comparison: balanced protein–energy supplementation versus control, outcome: perinatal death.

Figure 8.

Subgroup analysis of balanced protein–energy supplementation in adequately nourished women versus malnourished women. Outcome: perinatal death.

In the pooled results, the intervention group had a 26% lower risk of SGA babies than the control group (RR: 0.74, CI: 0.59–0.93, I²: 87%) (Figure 9). Excluding a study that included fortified formula milk from the analysis reduced the risk reduction to 16% (RR: 0.84, CI: 0.76–0.92, I²: 16%) (Supplemental Figure 7). ³¹ A subgroup analysis revealed a 40% risk reduction in SGA babies (birthweights below the 10th percentile for babies of the same gestational age) for malnourished women (RR: 0.60, CI: 0.41–0.87, I²: 87%), while no significant effect was observed for adequately nourished women (RR: 0.91, CI: 0.83–1.00, I²: 0%) (Figure 10).

Figure 9.

Analysis of comparison: balanced protein–energy supplementation versus control, outcome: small-for-gestational-age baby.

Figure 10.

Subgroup analysis of balanced protein–energy supplementation in adequately nourished women versus malnourished women. Outcome: small-for-gestational-age baby.

Pooled results for preterm birth (<37 weeks) showed that between the groups, there was not a significant difference in the risk of preterm birth (RR: 0.86, CI: 0.68–1.08, I²: 43%) (Figure 11). Subgroup analysis indicated no significant impact on the risk of preterm birth for malnourished women (RR: 0.85, CI: 0.66–1.10, I²: 0%) and adequately nourished women (RR: 0.83, CI: 0.58–1.21, I²: 62%) (Figure 12). Removing a study that involved CSB from the analysis did not change the insignificance of the risk of preterm birth in the intervention group as opposed to the control group (RR: 0.91, CI: 0.75–1.11, I²: 23%) (Supplemental Figure 8). ¹⁹ The table on “summary of findings” using the GRADE criteria is shown in Table 3.

Figure 11.

Analysis of comparison: balanced protein–energy supplementation versus control, outcome: preterm birth.

Figure 12.

Subgroup analysis of balanced protein–energy supplementation in adequately nourished women versus malnourished women. Outcome: preterm birth.

Table 3.

GRADE criteria for quality evidence and pooled analysis.

Quality assessment						Summary of findings
				Directness		Number of occurrences		Effect
Number of studies	Design	Limitations	Consistency	Generalizability to the population of interest	Generalizability to intervention of interest	Intervention	Control	Relative risk [95% CI]
Balanced supplementation versus control: Birthweight (g): Quality—high
24	RCTs/cluster RCT, before and after study design, clinical trial.	Some studies used inadequate processes for allocation and sequence generation.	All studies, except five, depict a positive effect of the intervention.	Developed and developing countries.	The protein in the supplement gave <25% of total energy content.	6043	5262	Mean difference 62.67 [27.92 to 97.43]
Balanced supplementation versus control: Low birth weight (<2500 g): Quality—moderate
08	Same as above	One study used unclear processes for allocation and sequence generation.	All studies, except one, depict a positive effect of the intervention.	Same as above	Same as above	296	367	Risk ratio 0.73 [0.57 to 0.95]
Balanced supplementation versus control: Birth length (cm): Quality—moderate
16	Same as above	Some studies used inadequate processes for allocation and sequence generation.	All studies, except one, depict a positive effect of the intervention.	Same as above	Same as above	4707	4534	Mean difference 0.20 [0.10 to 0.30]
Balanced supplementation versus control: Birth head circumference (cm): Quality—moderate
12	Same as above	Some of the studies had inadequate or unclear methods of sequence generation and allocation concealment.	All studies, except three, depict a positive effect of the intervention.	Same as above	Same as above	3316	3156	Mean difference 0.07 [−0.01to 0.14]
Balanced supplementation versus control: Perinatal death: Quality—moderate
08	Same as above	Unclear methods of sequence generation and allocation concealment in three studies.	All studies, except two, depict a positive effect of the intervention.	Same as above	Same as above	115	109	Risk ratio 0.83 [0.50 to 1.37]
Balanced supplementation versus control: Small-for-gestational-age-baby: Quality—moderate
12	Same as above	Inadequate or unclear methods of sequence generation and allocation in a few of the included studies.	All studies demonstrate a positive impact from the intervention.	Same as above	Same as above	902	946	Risk ratio 0.74 [0.59 to 0.93]
Balanced supplementation versus control: Preterm birth: Quality—moderate
10	Same as above	Sequence generation and allocation concealment was not adequate or unclear in some of the included studies.	All studies, except two, depict a positive effect of the intervention.	Same as above	Same as above	274	306	Risk ratio 0.86 [0.68 to 1.08]

RCT: randomized controlled trial; CI: confidence interval.

Publication bias

For birth weight (g), birth length (cm), SGA (birthweights below the 10th percentile for babies of the same gestational age), and LBW (birthweight <2500 g), the publication bias funnel plots are displayed in Supplemental Figures 9 to 12. No discernible publication bias was found across all of the results as the funnel plot was adjudged to be fairly symmetrical.

Discussion

Our meta-analysis has deduced that the overall effects of protein supplementation on fetal outcomes are positive. The administration of protein supplementation during pregnancy was observed to have beneficial effects on weight and length at birth, as well as a decreased probability of LBW. Moreover, in the subgroup of malnourished women, SGA and perinatal deaths both experienced significant declines.

In our study, the impact of protein supplementation on birth weight and the occurrence of LBW showed positive findings. It adds to the growing body of evidence supporting the benefits of protein supplementation during pregnancy. The observed increase in birth weight aligns with previous findings and underscores the crucial role of adequate protein intake for optimal fetal growth. For instance, a recent systematic review and meta-analysis by Laura Pimpin also reported a positive association between protein supplementation and increased birth weight, although the magnitude of the effect may vary because they used animal protein supplementation. ⁵³ According to the combined data, adding protein–energy to the diet had a moderately notable impact on the birth weight. Neonates whose mothers received protein–energy supplementation had an average birth weight difference of 62.67 g from those whose mothers did not. These findings are comparable with a meta-analysis by Imdad and Bhutta for birth weight which also suggests that prenatal protein–energy supplementation can result in a modest but significant rise in birth weight. ³⁸ We did a sensitivity analysis for this outcome by removing Girija et al., after which the reduction in birth weight caused by protein–energy supplementation was 2.5%. ²⁸ The results imply that a supplement with a greater amount of protein-derived energy might help birth weight when opposed to balanced protein energy supplements. Nevertheless, recent evidence in Ota et al.’s meta-analysis contradicts the previous findings, which showed that a higher risk of having SGA babies has been associated with high-protein supplementation. ¹² About LBW, our study revealed that when LBW prevalence was supplemented with a balanced protein–energy ratio, it decreased by 27% in the intervention group as opposed to the control group. Similar to our findings, a RCT also concluded that providing women with prenatal protein-calorie nutrition helps in decreasing the number of babies with low weight at birth and increasing the average birth weight of their newborn. ²⁷ This aligns with the broader public health focus on improving maternal nutrition to enhance infant health, as discussed in a recent WHO report on maternal nutrition and infant health. ⁵⁴

Our meta-analysis looked at how adding protein to the diet affected head circumference and birth length. The findings imply that prenatal protein–energy supplementation has a moderately noticeable effect on birth length, but not on birth head circumference. Specifically, the infants’ mean birth length differences of intervention and control groups differ by 0.20 cm. In contrast, the MD was only 0.05 cm for birth head circumference which was not statistically significant. While we did not find a significant effect on head circumference, future research should explore other measures of neurodevelopment to understand the full impact of protein supplementation on fetal growth and development. ⁵⁵ A subgroup analysis that took into account the nutritional status of mothers found that in women who received adequate nutrition, protein supplementation had a positive effect on the length of the baby at birth. However, neither the adequately fed nor the undernourished had a discernible impact on the birth head circumference. These results support those of Ota et al.’s Cochrane review which revealed that, while in terms of length at birth or head circumference, there was no significant statistical difference, balanced protein supplementation was linked to a substantial rise in weight at birth. ¹²

Our research also investigated the effect of protein dietary supplements on birth outcomes which include perinatal death, SGA babies, and births occurring before term. When we pooled data from eight studies, the risk of perinatal mortality between the control group and the intervention group did not differ significantly, according to our findings. However, a noteworthy observation emerged when we specifically analyzed malnourished women—protein supplement intake had a link with a remarkable 44% decrease in perinatal mortality risk among this subgroup. This result underscores the urgent need for effective strategies to improve maternal nutrition, especially in low- and middle-income countries. ⁵⁶

Regarding SGA births, our overall analysis indicated a 26% decrease in risk within the intervention group compared to controls. Intriguingly, the subgroup analysis targeting malnourished women revealed an even more substantial 40% risk reduction for SGA infants. This disparity underscores the impact of nutritional status, as the combined results, encompassing both adequately nourished and malnourished women, failed to reveal such a pronounced effect.

In terms of preterm births, our study did not find any significant reduction in risk across the intervention, adequately nourished, or malnourished groups. Our findings are in line with previous research conducted by Kramer and Kakuma, which also demonstrated reduced perinatal mortality and fewer SGA births but no significant alteration in preterm birth rates. ⁶

Limitations

Our study’s strength lies in its diverse inclusion of RCTs from various global income levels. However, limitations include inconsistencies in participant demographics, supplement characteristics, control measures, outcomes, and timing of supplementation. Socioeconomic differences among study populations, coupled with varying dietary habits, lifestyles, supplement types and doses, compliance rates, and genetic factors, contribute to this variability. High heterogeneity in our results is attributed to methodological and sample size differences among the included studies. Additionally, the use of built-in filters during the literature search may have introduced some biases in the selection of studies, limiting the comprehensiveness of the search strategy. Furthermore, although an a priori protocol registration was not conducted, we believe that our systematic approach to the study provides reliable and valuable insights into the topic.

Conclusion

This meta-analysis investigated protein–energy supplementation’s impact on birth outcomes in expectant mothers. Results showed increased length at birth as well as a 27% lower risk of LBW. A 26% decrease in the risk of SGA babies was observed. No notable changes were found in head circumference or preterm births. However, malnourished women experienced more significant benefits in birth weight, perinatal death, and SGA births with balanced protein–energy supplementation. In conclusion, this strategy lowers the possibility of having a bad pregnancy outcome, particularly in undernourished females.