Gut Microbiome-Targeted Nutrition Interventions and Growth among Children in Low- and Middle-Income Countries: A Systematic Review and Meta-Analysis

Background Childhood malnutrition is a public health challenge of much interest and concern globally. However, a perturbed gut microbiome (GM) may limit some nutrition interventions’ effects among healthy children with undernutrition. Objectives This review aimed to evaluate the effects of GM-targeted nutrition interventions on growth outcomes among children (0–59 mo) using published studies in low- and middle-income countries. Methods The methods were guided by the Cochrane methodology. The literature search was conducted to include articles published from inception to July 2023 in PubMed, Google Scholar, and Cochrane Databases. We identified and included 35 studies among 11,047 children. The analysis was conducted considering various growth parameters in the qualitative synthesis and weight gain (kg) in the meta-analysis. Results In the qualitative synthesis, 55.6% of prebiotics, 66.7% of probiotics, 71.4% of synbiotics, and 28.6% of “microbiome complementary feed” studies had significant effects on growth outcomes. Also, prebiotics had more studies with significant effects among healthy children, whereas probiotics, synbiotics, and “microbiome complementary feeds” had more studies with significant effects among children with undernutrition. Nineteen studies were included in the meta-analyses, of which 7 (36.8%) measured GM outcomes. The meta-analysis showed that prebiotics exhibited heterogeneity but had significant effects on weight in the intervention as compared with the control (mean difference [MD]: 0.14 kg; 95% CI: 0.02, 0.25; I2 = 63%, P = 0.02; 4 studies, n = 932). Probiotics had significant effects on weight in the intervention (MD: 0.15 kg; 95% CI: 0.06, 0.25; I2 = 42%, P = 0.05; 8 studies, n = 2437) as compared to the control. However, synbiotics (MD: 0.26 kg; 95% CI: –0.04, 0.56; I2 = 41%, P = 0.17; 4 studies, n = 1896] and “microbiome complementary feed” (MD: –0.03 kg; 95% CI: –0.18, 0.11; I2 = 0%, P = 0.60; 3 studies, n = 733] had no significant effects on weight in the intervention as compared with control. Conclusions Although probiotics and synbiotics may be effective at enhancing growth among children, the selection of interventions should be contingent upon health status. This trial was registered at www.crd.york.ac.uk/prospero/ as CRD42023434109.


Introduction
Globally, childhood malnutrition is a public health challenge of much interest and concern [1].It affects an estimated 149 million children aged <5 y, according to the WHO [1], in low-and middle-income countries (LMICs).These countries account for most child deaths from preventable causes, such as diarrhea, pneumonia, and malnutrition [2].The human gut microbiome (GM) is a complex ecology of microorganisms that coexist with its human host.If unperturbed, such coexistence is usually a mutually beneficial relationship [3].Such a relationship plays a significant role in human physiological processes, including brain function [4] and immunity [5].The GM goes through changes throughout the lifecycle particularly from infancy to preschool age.Notable changes, occasioned by the changes in feeding mode, occur during the transition from exclusive breastfeeding to complementary feeding and later during weaning [6].Other factors, such as the child's mode of delivery, antibiotics intake, environmental exposures, and geographic location all have influence in shaping GM as a child ages [3].Children in LMICs are particularly vulnerable to GM perturbations because of poor water and sanitation, inadequate dietary intake, and infectious diseases coupled with recent rising cesarean section rates [7].
Malnutrition is associated with impaired GM development, reduced immune function, and increased susceptibility to infections and chronic diseases [3].GM dysbiosis which is characteristic of susceptible children or children with undernutrition, may limit the effects of nutrition interventions at promoting growth or enhancing recovery [8].The imbalance between beneficial and harmful bacteria associated with GM dysbiosis could enhance the overgrowth of specific bacterial species, such as Clostridium difficile that could compete for intervention nutrients and hence reduce nutrient availability to the host.Therefore, GM-targeted nutrition interventions have been proposed as a promising strategy to improve child nutritional outcomes.As such, in recent times, nutrition interventions, aside from enhancing growth, have sought to target GM dysbiosis by using prebiotics, probiotics, synbiotics, or specialized complementary feeds [3,9].
Prebiotics are the nondigestible components of carbohydrates such as inulin, galacto-oligosaccharides (GOS), and fructooligosaccharides (FOS) that resist breakdown in the small intestine and reach the large intestines intact, where they serve as a food source for the gut microbiota [10].Among other things, prebiotics stimulate the production of short-chain fatty acids [11].It also helps improve gut barrier function and delay gastric emptying.To the contrary, probiotics are themselves living microorganisms that, when consumed in appropriate quantities, provide health benefits to the host [10].Probiotics are typically beneficial bacteria from the Lactobacillus or Bifidobacterium genus that play a significant role in maintaining a healthy digestive system and overall health [10].They feed on prebiotics and enhance nutrient absorption, improve immune function [5], reduce inflammation, and promote a healthy balance of bacteria in the gut.However, synbiotics are a combination of prebiotics and probiotics that work together to confer health benefits to humans [12].The synergistic effects of combining probiotics and prebiotics may be achieved when both are administered concurrently [13].By combining the 2, synbiotics aim to  enhance the survival and activity of probiotic bacteria in the gut, in addition to being a food source for their growth.
Although GM is an important factor that modifies the effects of nutrition interventions on a child's nutritional status, previous reviews on this subject have been among only healthy infants [14,15].Some other systematic reviews have centered on both LMICs and high-income countries [16] or included interventions among only infants aged <1 y [17].However, the only reviews examining the effects of prebiotics, probiotics, and synbiotics on childhood growth in the context of LMICs are publications by Onubi et al. [18] and Heuven et al. [8].Although the former was published about a decade ago, the latter excluded children aged <6 mo.Heuven et al. [8] also excluded interventions with durations <12 wk.Because children of some undernutrition interventions recover before the twelfth week, it is possible that Heuven et al. [8] may have missed publications of some critical malnutrition interventions.Additionally, Heuven et al. [8] did not include studies that used gut "microbiome complementary feeds" as an approach to enhancing growth.Therefore, this systematic review aimed to evaluate the effects of microbiome-directed nutrition interventions (prebiotics, probiotics, synbiotics, and "microbiome complementary feeds") on growth outcome(s) among children (0-59 mo) using published controlled trials in LMICs.Such a review may inform policy and practice by identifying the most effective and feasible interventions, as well as identifying research gaps and priorities for future studies.

Methods
This review was registered with the PROSPERO and is available at www.crd.york.ac.uk/prospero/ as CRD42023434109.The methods and procedures used in this systematic review followed the PRISMA checklist [19] and the Cochrane Handbook for Systematic Reviews of Intervention Studies [20].

Inclusion and exclusion criteria
The review included all types of controlled trials and studies of GM-targeted dietary interventions directed at enhancing child growth in LMICs.The criteria for classifying a country as an LMIC was based on the World Bank classifications [21].The interventions included any type of "microbiome complementary  This review excluded studies of animal models.Prenatal GM studies among women during pregnancy but with outcomes on infants' GM and studies of child GM with other outcome measures aside from nutrition, such as irritable bowel syndrome, were all excluded.It also excluded GM studies involving preterm babies and studies with part or all participants living in highincome countries.For synthesis, the interventions were grouped into 4, i.e., prebiotics, probiotics, synbiotics, and "microbiome complementary feed" studies.

The search strategy
The literature search was conducted in PubMed, Google Scholar, and Cochrane Library electronic databases.This was done to identify intervention studies published from inception till July 2023.The literature search was carried out from June 2023 to July 2023.Grey literature, such as reference lists of relevant reviews and studies, were also included in the search.EndNote software version X7 (Clarivate Analytics) was used to detect and expunge duplicate studies.Two authors (HYA and CA) independently reviewed and screened the titles and abstracts of the identified studies and, subsequently, the full texts of the potentially relevant studies, using predefined exclusion and inclusion criteria.Identified discrepancies were resolved through discussion and consensus between the 2 authors in consultation with the main supervisor.Details of the search strategy have been included and attached as Supplementary File 1.

Data extraction
The data extraction was performed manually using a prepiloted standardized form that includes study characteristics such as author and year, study country, sample size, age of participants, dietary intervention and duration, method of biospecimen analysis, health status, delivery vehicle(s), growth outcome(s), GM outcome(s) main objective(s), main finding(s) and whether the intervention had a significant positive effect.This data extraction was carried out by 2 authors independently.The standardized forms were then compared after data extraction, and discrepancies were resolved with the inputs of the main supervisor.

Outcome variables
Height or length, weight, height-for-age, weight-for-height, weight-for-age, wasting, stunting, underweight, BMI, or blood parameters such as hemoglobin concentrations, albumin, or serum ferritin were the outcomes of interest in this review.They served as the basis for deciding whether an intervention has had the desired effect on participants.Other outcome variables associated with gut health and microbiota, such as microbiotafor-age z-score and microbiota alpha and beta-diversity, were also noted.However, the weight gain parameter was included in the meta-analysis section.This is because it was the most prevalent nutrition-related outcome indicator among the included studies.

Quality assessment
The Cochrane Risk of Bias (ROB) tool for randomized control trials [22] was used in evaluating the quality of the included trial studies.Areas assessed included bias arising from the randomization process, bias because of deviation from intended interventions, bias in the measurement of the outcome, bias because of missing outcome data, and bias in the selection of the reported results.A composite rating based on the above adjudged each of the interventions as either "low risk," "unclear, " or "high risk" of bias.Details of individual ROB scores for all included studies are attached as Supplementary File 2.

Data analysis
Given that this review had the objective of assessing the effects of various interventions, the end-line, instead of baseline sample sizes, was used in this systematic review.This is because the end-line numbers reflect the participants who completed the study and, therefore, provide a more accurate estimate of an intervention's effect, i.e., mean difference (MD).Weight gain was defined as the differences in preintervention weight and postintervention weight in the control group and the intervention group.MD referred to the arithmetic differences between the mean weight gain in the control group and the mean weight gain in the intervention group.Only the weight gain parameter was included in the meta-analysis as an outcome variable.This was done to minimize heterogeneity.Weight gain was also used because it is the most prevalent growth-related outcome reported in the included studies.Weight is also a part of the other composite anthropometric measures such as BMI, weight-forheight z-score, and weight-for-age z-score.Where end-line sample sizes were different for different outcome measures, sample sizes for nutrition-related outcome variables were used for the qualitative synthesis.For the meta-analysis section, forest plots were used to present results.For studies in which heterogeneity was detected (P < 0.05 and I 2 > 50%), the common effects model analysis was used.For studies with no heterogeneity (P !0.05 and I 2 50%), the random effects model analysis was carried out.The meta-analysis included only articles that reported preintervention and postintervention weight and articles that reported weight gain.All meta-analyses were conducted at a 95% confidence interval (CI) using R software version 4.2.3 (R Core Team).

Results
The PRISMA flow diagram [19] of the included studies is shown in Figure 1.Overall, 35 studies, all published in the last 2 decades approximately, met the inclusion criteria and were included in this systematic review.

Summary of all included studies
The number of trials included was 35, and the total number of children analyzed by all studies was 11,047.Geographically, the review included studies from 16 counties, and most of the studies were conducted in Malawi (5, 14.3%) and Indonesia (5, 14.3%).Eleven (31.4%) of the included studies were conducted in Africa, 16 (45.7%) in Asia, 3 (8.6%) in South America and 4 (11.4%) in the Middle East.Twenty (57.1%) studies had a significant effect on !1 growth outcome.Also, 14 (40%) studies were among children with undernutrition, whereas 20 (57.1%) studies were among healthy children, and 1 study with a not clearly defined health status.The mean intervention duration was 151.3 d, with a range of 28-360.The summary of the location, intervention duration, sample size, health status, and significant positive effect on growth are shown in Table 1.

Prebiotics
Nine prebiotic studies met the inclusion criteria [23][24][25][26][27][28][29][30][31], and their study characteristics are presented in Table 2.The interventions were conducted on 1392 children.Six [25][26][27][28][29]31] of these studies were among healthy children and 3 [23,24,30] FIGURE 2. Forest plot of prebiotic studies that reported weight gain.Growth (weight gain) was measured in kg.Weight gain was defined as the differences in preintervention weight and postintervention weight in the control group and the intervention group.MD referred to the arithmetic differences between the mean weight gain in the control group and the mean weight gain in the intervention group.CI, confidence interval; SE, standard error.among children with undernutrition.Three of the studies used oligosaccharides as the prebiotic, which include GOS [23,24] and FOS [29].However, 2 of the studies used rice bran [25,27] as the prebiotic.The others used a combination of GOS þ iron [26], GOS þ polydextrose [28], Oligofructose þ Zinc [31], and alpha-linolenic acid from flaxseed as prebiotic [30].In general, the vehicles for conveying this prebiotics were RUTF among undernourished children and milk or complementary feeds among healthy children.Five [23][24][25][26][27] of the 9 prebiotic trials (55.6%) had significant positive effects on !1 growth outcome relative to their respective control groups.Of the 5 that reported significant effects, 3 studies [25][26][27] were among healthy children and 2 children with undernutrition [23,24].The mean number of days for intervention duration was 126.2 d, with the least and highest being 48 d and 180 d, respectively.Figure 2 shows the heterogeneity descriptions and the effect sizes MD of selected prebiotic studies and growth outcome (weight gain) with a 95% CI.The included prebiotic studies had a significant overall effect on weight gain (MD ¼ 0.14, 95% CI: 0.02, 0.25).However, significant heterogeneity was also identified among the included studies (I ¼ 63%, P ¼ 0.02).
Figure 3 shows the forest plots with heterogeneity descriptions and the effect sizes MD of selected probiotic studies and growth outcome (weight gain) with 95% CI.Two of the included studies showed a positive effect on the intervention arm.Overall, there was no heterogeneity in the included studies (I 2 ¼ 42%, P ¼ 0.05), and the cumulative effect size was statistically significant (MD ¼ 0.15, 95% CI: 0.06, 0.25).

Synbiotics
Table 4 contains the characteristics of study interventions that used synbiotics (a combination of prebiotics and probiotics) as a means of improving GM and growth.The studies analyzed a total of 2144 children.Seven studies met the inclusion criteria [44][45][46][47][48][49][50].Four [44,46,48,50] of these studies were among children with undernutrition and 3 [45,47,49] among healthy children.The mean intervention duration was 154.4 d, with the least duration being 28 and the highest 365.Five [44,45,47,48,50] of the 7 studies (71.4%) reported a beneficial effect of synbiotics on !1 growth outcome in the intervention group as compared to the control group.Of the 5 synbiotic studies with beneficial effects, 3 [44,48,50] were among children with undernutrition and 2 [45,47] among healthy children.The vehicle(s) of administration were mainly F-100, RUTF, infant formula for children with FIGURE 3. Forest plot of probiotic studies that reported weight gain.Growth (weight gain) was measured in kg.Weight gain was defined as the differences in preintervention weight and postintervention weight in the control group and the intervention group.MD referred to the arithmetic differences between the mean weight gain in the control group and the mean weight gain in the intervention group.CI, confidence interval; SE, standard error.undernutrition, and milk for healthy children, with 1 study using starch powder [48].The specific synbiotics used in these studies include [45,48] and B. infantis þ Lacto-N-neotetraose [50].
Figure 4 shows forest plots with the computed MD and 95% CI for selected synbiotic studies.There is no significant heterogeneity (I 2 ¼ 41%, P ¼ 0.17) among included studies.Although only 2 studies had significant effects on weight gain, the effect sizes are tilted toward the intervention arm.However, there are no overall significant differences in effect sizes (MD ¼ 0.26, 95% CI: -0.04, 0.56) of synbiotic intervention on the growth outcome (weight gain).
Figure 5 shows forest plots with the computed MD and 95% CI for selected "microbiome complementary feed" studies.There was no significant heterogeneity (I 2 ¼ 41%; P ¼ 0.17) among included studies.One study [53] had a significant effect on weight gain, and generally, the effect sizes are tilted toward the intervention arm.However, no overall significant differences in effect sizes (MD ¼ 0.26, 95% CI: -0.04, 0.56) of the "microbiome complementary feed" intervention on the outcome weight gain were detected.

Growth as uniformly positive
This study classified all growth as better because a good number of studies were among children with malnutrition.In these instances, more growth may usually be good.Additionally, for studies among presumably "healthy" children, some interventions were conducted among participants who were at risk of undernutrition.For instance, 2 studies [36,54] described their participants as children "at risk" of undernutrition.The other studies described the location of their interventions as either rural [25][26][27]52,53,56,57], low-resourced [40], urban slum [29,42], or shantytown [31].A significant number of children living in these locations are more likely to be at risk or suffer from undernutrition.For these reasons, this study synthesized growth as uniformly positive and good throughout.

Cochrane ROB tool
Figure 6 shows ratings of the ROB tool 2.0 and the 5 domains for the individual control trials.This is a stacked bar plot, and the horizontal axis depicts the percentage risks in all included studies.Most of the studies were found to have a low ROB.The risks ROB assessments were conducted keeping in mind the tendency of funder influence.Moreover, as such, this study factored this into the rating of studies for ROB.Generally, studies, where funders either own the data or need to approve the manuscript, were termed as at risk of undue funder influence and were rated higher for ROB.It is also relevant to state that the majority of the studies were funded by industry, and all authors declared their conflicts of interest, indicating the funders did not have an influence on the outcome of the research, where necessary.Additional details on the individual scores of all included studies are attached as Supplementary File 2.

Discussion
This systematic review aimed to synthesize existing evidence on dietary nutrition interventions targeting the child GM, assess the effectiveness of interventions on growth outcome, and identify research gaps and priorities for future studies.This present study involving 35 intervention studies with 11,047 children aged 0-5 y is 1 of the largest systematic reviews and meta-analyses on this subject in LMICs.
The qualitative results indicate that 5 out of 9 (55.6%) of prebiotic studies, 8 out of 12 (66.7%) of probiotic studies, 5 out of 7 (71.4%) of synbiotic studies, and 2 out of 7 (28.6%) of "microbiome complementary feed" studies had significant effects on !1 growth outcome in their respective intervention groups as compared with control groups.The effectiveness of synbiotics over other approaches at enhancing growth is consistent with the findings of another systematic review [8].Synbiotics are a combination of probiotics and prebiotics, and this may have synergistic effects, where the probiotic microorganisms derive their food source from the prebiotic substrate, allowing them to grow and multiply more effectively [58].The prebiotic substrate(s) is often specifically selected to stimulate the growth of the probiotic strain(s); this can result in more specificity and more targeted effects on the GM.Together, they provide a range of mutually beneficial effects on the GM, resulting in the reduction of inflammation [59], improving gut barrier function [60], and modulating the immune system [5], which could enhance growth.This kind of synergy may not be plausible in interventions that use only prebiotics or complementary feeds.Contrary to these findings, a systematic review involving 2971 infants with 25 control trials [15] could not establish that prebiotics or probiotics administered separately have lesser effects than when combined into a synbiotic.Their finding accentuates the need for probiotic strains to be well-matched to the specific prebiotic ingredient, or else their synergistic effects would not be harnessed.
This systematic review also revealed that prebiotics had a greater number of studies with significant effects among healthy children than children with undernutrition, whereas probiotics, synbiotics, and "microbiome complementary feeds" had more studies with significant effects among children with undernutrition than healthy children.For instance, in a probiotic study among children with undernutrition [32], L. paracasei had effects on growth outcomes, yet the same probiotic did not have effects among healthy children in another study [37].These differences in effects are anticipated as the physiological states of healthy children and children with undernutrition may be different, as children in each group may respond differently even when exposed to similar nutritional interventions [3,61].Healthy children may have an unperturbed GM, where prebiotics can act as a food source for the already existing beneficial bacteria, promoting their growth and activity.By enhancing the growth of these beneficial bacteria, prebiotics may help to improve overall health outcomes, including growth [62,63].In children with undernutrition, however, GM may be perturbed and may have fewer beneficial bacteria.As such, introducing beneficial bacteria strains in the form of probiotics to restore diversity may be helpful in that context.Furthermore, the concurrent administration of probiotics and prebiotics may likely yield more beneficial outcomes.In sum, the findings of this study provide support for the proposition that the selection of a GM-targeted nutritional intervention aimed at enhancing child growth should be contingent upon the health status of the specific target population [8].However, this should be interpreted with caution because of the very limited number of studies used in arriving at this finding.The finding should be recognized as preliminary and be a basis for future studies.
Generally, although GM-targeted nutrition interventions can have beneficial effects on growth, such effects may be more complex and depend on a variety of factors, including the specific probiotic strains or prebiotic substrates used and other FIGURE 4. Forest plot of synbiotic studies that reported weight gain.Growth (weight gain) was measured in kg.Weight gain was defined as the differences in preintervention weight and postintervention weight in the control group and the intervention group.MD referred to the arithmetic differences between the mean weight gain in the control group and the mean weight gain in the intervention group.CI, confidence interval; SE, standard error.characteristics such as breastfeeding, geographic location, antibiotics intake as well as individual host's GM.
We expected that analysis of specimens of all nutritionrelated GM studies would be conducted using fecal 16s rRNA sequencing.However, this review has revealed that only 4 [26,27,35,50] out of the 28 prebiotic, probiotic, and synbiotic studies used 16s rRNA sequencing.In addition, the low-cost blood hematologic analysis with anthropometric measurements, PCR [37,42,54], and fluorescence in-situ hybridization [47,49] are also acceptable approaches in resource-constrained situations where bacteria rRNA sequencing may not be feasible.Further, analysis of short-chain fatty acids, which include acetate, propionate, and butyrate, that have been associated with improved gut health and a more unperturbed GM [11] could also be used.
The role of infant feeding mode in the included studies can be categorized into 2 main contexts.The first context was mostly associated with breastmilk feeding.Because breastfeeding is associated with inducing higher proportions of B. infantis, a beneficial GM bacterium, supplementation provided to breastfeeding participants may offer no additional clinical benefits [31].For instance, in some included studies that were conducted among breastmilk-fed infants [31,57], the lack of significant growth differences between the intervention and control groups was attributed to the beneficial effects of breastmilk feeding, reiterating the superiority of breastfeeding relative to other modes of infant feeding as a plausible explanation.However, this may require further inquiry as some studies [28,46,54], despite being among breastfed infants, concluded that insufficient dosage, effects of antibiotics, and dietary fiber intake are the reasons for the observed lack of significant differences.In other studies [28,29,40,47,49], the a priori knowledge of the beneficial effects of breastfeeding prompted authors to exclude breastfeeding infants from their studies.This was because of the challenges in measuring the exact quantities of human milk oligosaccharides and other bio-actives supplied through breastfeeding.This could confound with the dietary intervention and its effects on growth.In the second context, where the mode of feeding could limit intervention acceptability and adherence, some studies [25,26,31,39] used locally available complementary feeds as vehicles through which prebiotics and probiotics were administered.Additionally, in some studies [30,51] that used specialized feeds such as RUTF, F75, F100, or microbiome-directed complementary feed, mothers were asked to breastfeed before such therapeutic feeds were given or such feeds were given at half the daily recommended therapeutic dose so infants could still be breastfed.

Limitations
There are several inherent limitations, as captured by included studies, that could help streamline and guide the direction of future studies.First, a number of studies [23,29,39,44], expressed concern about the influence of antibiotics intake and their inability to account for their effects on GM and growth.Some other studies [23,35,44,51,55] stated short intervention duration or short follow-up time as limitations in their interventions.The term short for these studies was an intervention duration of <3 mo and a follow-up time of <2 mo.As such, those studies could not measure long-term outcomes.Additional studies [25,31,35,40] stated that an underpowered sampled size calculation was a limitation.In those studies, although the sample size was powered enough to measure GM characteristics, it was not enough to detect differences in growth outcomes.
Studies [25,35,49] expressed concern about not keeping dietary intake log books meant to measure other feeds.These studies were of the view that aside from the diet used for the intervention, there could be other foods that may have been consumed by participants, which may have gone unnoticed, but such feeds could exert influence on the GM and growth.As such, taking note of such foods would help explain certain unexpected outcomes or control for their confounding effects.Although some studies [26,27,36,39] expressed differences in baseline characteristics as a limitation, others [33,48] had reservations about the safety of probiotics in immune-suppressed children with undernutrition because of probiotic-induced sepsis.Finally, the lack of data on the volume of breastmilk consumed by infants [44] and challenges with sample collection [25], e.g., blood volume and storage process, were additional limitations stated by some authors.Hence, when planning sample collection procedures and storage, it is crucial to account for the half-lives of nutritional outcome measures and ensure adherence to appropriate sample collection procedures.The methodologic limitation of this review is also exogenous in that, because of differences in reported outcomes, not all articles were eligible to be included in the meta-analysis.The final limitation is the age range of included studies (0-5 y).
Although meta-analyzing similar nutrition interventions over such a broad age range has been conducted in the past [16], the plausible confounding effects of the varying feeding modes during infancy and childhood in the context of this study should be acknowledged.This is particularly relevant as feeding mode differences, especially breastfeeding and complementary feeding, and their accompanying behavioral and developmental differences could have profound effects on GM and/or growth.However, because all included interventions are studies with control groups, it is assumed that these limitations would be equally distributed among the intervention and the control groups such that their biased effects may not have significant differences on the outcome(s) of interest.It is presumed that such an assumption of random distribution may isolate only the intervention to be the sole cause of the observed differences between the arms at the end-line.These limitations notwithstanding, promising effects of some synbiotics, probiotic strains, prebiotic substrates, and some microbiome complementary feeds on growth among both undernourished and healthy children were detected in ~60% of the included studies.

Conclusion
Overall, 20 out of the 35 studies demonstrated significant effects on !1 growth outcome.For the qualitative analysis, the synbiotic studies had the highest number of studies significantly influencing growth, whereas the probiotic studies had significant effects on weight in the meta-analysis.However, the observed heterogeneity in prebiotic studies and lack of effectiveness in synbiotic and "microbiome complementary feed" groups could be because of the limited number of studies meta-analyzed.As a result, further intervention research is required to explore the effects of GM-targeted nutrition interventions on growth in LMICs.In future studies, antibiotic and breastmilk exposure should be accounted for.Accounting for antibiotics intake is important because it has the tendency to alter GM composition and possibly influence the effects of the interventions on growth.The sample size should be powered enough to measure not only the GM outcomes but also nutrition-related outcomes.Researchers interested in GM and growth intervention studies should aim for controlled trials with longer durations and follow-up periods of >3 and 2 mo, respectively.This would allow nutrition-related short-to-medium-term outcomes to be adequately assessed.Differences in baseline characteristics and sample collection variability resulting from collection storage processes should also be well addressed.
In conclusion, the relative effectiveness of interventions was found to be dependent on the health status of participating children.Moreover, although probiotics and synbiotics may be effective at enhancing growth among children in LMICs, the selection of a microbiome-targeted nutrition intervention should be contingent upon the health status of the participating children.

FIGURE 5 .
FIGURE 5. Forest plot "microbiome complementary feed" studies that reported weight gain.Growth (weight gain) was measured in kg.Weight gain was defined as the differences in preintervention weight and postintervention weight in the control group and the intervention group.MD referred to the arithmetic differences between the mean weight gain in the control group and the mean weight gain in the intervention group.

FIGURE 6 .
FIGURE 6. Evaluation of the risk of bias on growth outcome for all included studies.

TABLE 1
Summary of location, health status and effects on growth of all included studies

TABLE 2
Characteristics of prebiotic intervention studies

TABLE 3
Characteristics of probiotic interventional studies

TABLE 4
Characteristics of synbiotic interventional studies

TABLE 5
Characteristics of "microbiome complementary feed" studies directed at microbiome and growth