Maternal-focused interventions to improve infant growth and nutritional status in low-middle income countries: A systematic review of reviews

Background Small and nutritionally at-risk infants under 6 months (<6m) are a vulnerable group at increased risk of mortality, morbidity, poor growth and sub-optimal development. Current national and international (World Health Organization) management guidelines focus mainly on infants’ needs, yet growing evidence suggests that maternal factors also influence infant outcomes. We aimed to inform future guidelines by exploring the impacts of maternal-focused interventions on infant feeding and growth. Methods We conducted a systematic review of reviews published since 2008 (PROSPERO, register number CRD 42019141724). We explored five databases and a wide variety of maternal-focused interventions based in low- and middle-income countries. Infant outcomes of interest included anthropometric status, birthweight, infant mortality, breastfeeding and complementary feeding practices. Given heterogenous interventions, we present a narrative synthesis of the extracted data. Results We included a total of 55 systematic reviews. Numerous maternal interventions were effective in improving infant growth or feeding outcomes. These included breastfeeding promotion, education, support and counselling interventions. Maternal mental health, while under-researched, showed potential to positively impact infant growth. There was also some evidence for a positive impact of: women’s empowerment, m-health technologies, conditional cash transfers, water, sanitation and hygiene and agricultural interventions. Effectiveness was increased when implemented as part of a multi-sectoral program. Antenatal supplementation with macronutrient, multiple micronutrients, Vitamin D, zinc, iron folic acid and possibly calcium, iodine and B12 in deficient women, improved birth outcomes. In contrast, evidence for postnatal supplementation was limited as was evidence directly focusing on small and nutritionally at-risk infants; most reviews focused on the prevention of growth faltering. Conclusion Our findings suggest sufficient evidence to justify greater inclusion of mothers in more holistic packages of care for small and nutritionally at-risk infants aged <6m. Context specific approaches are likely needed to support mother-infant dyads and ensure infants survive and thrive.


Introduction
Malnutrition remains a major global health problem, with approximately 47 million children under five years wasted, and 149 million children still stunted in 2020 [1]. Undernutrition is an underlying cause of 45% of childhood mortality [2]. Recently, there has been an increasing awareness that infants aged under six months (<6m) make up a sizable proportion of wasted children [3], with some 3.8 million infants <6m severely wasted and 4.7 million moderately wasted [4]. An even greater number of infants are "small or nutritionally at-risk" [5]. This term includes infants who are at increased risk of morbidity, mortality or suboptimal development due to any of the following: low birth weight (LBW) due to intra-uterine growth retardation and/or prematurity; anthropometric deficit as used in standard case definitions of undernutrition: low weight-for-age, weight-for-length or length-for-age; postnatal growth faltering even if above the usual -2 Z-score criteria for undernutrition.
Short term risks include infection and mortality which are higher among this age group than in older children [6,7]. Poor early-life nutrition has also been linked with adverse longterm outcomes, such as impaired cognitive development and higher risk of non-communicable diseases later in life [8][9][10][11][12].
Despite being so vulnerable, the management of small and nutritionally at-risk infants <6m represents a critical care gap, as available evidence often does not address them directly [13]. In 2013, the World Health Organization Guidelines for the Management of Severe Acute Malnutrition included, for the first time, a chapter focused just on infants <6m, however recommendations were based on limited and low quality evidence [14]. Many of the recommended management interventions focused on the clinical and nutritional needs of the infant, yet growing evidence shows a wide range of underlying causes of nutritional vulnerability [15], many of which could be better addressed through maternal-focused interventions, and management of a mother-infant dyad [16][17][18][19][20].
Maternal factors including poor antenatal and postnatal nutrition are well documented as being associated with adverse neonatal and infant outcomes [2,[21][22][23]. Evidence also suggests that maternal social, environmental, and other physical and mental health factors all also contribute to infant health and growth [16,[18][19][20]24]. What is missing is a collation of this evidence to help understand what might best improve future care for small and nutritionally atrisk infants <6m. In this review we thus aim to collate available evidence from existing literature reviews on the impacts of maternal-focused interventions on infant feeding and growth in low-and middle-income countries (LMICs). This will help to inform future research and policymaking.

Methods
To capture as wide as possible a range of interventions, we conducted a systematic review of reviews. We focused on interventions that impact the immediate and underlying causes of undernutrition as depicted by the UNICEF Conceptual Framework on the Determinants of Maternal and Child Nutrition [25]. We registered our work on the International prospective register of systematic reviews (PROSPERO, register number CRD 42019141724), which we later expanded to also include a search of maternal macro-and micronutrient supplementation (MEDLINE and Cochrane databases only).

Search strategy
We searched five different databases: MEDLINE, EMBASE, Global Health, Cochrane Library and CINALH plus. We considered reviews published since 2008 (the last 10 years since this project was originally started). The most recent search took place on 29 th of October 2020. Reference lists of eligible reviews and personal communications with relevant authors were additional search sources.
Search concepts based around the below PICO were developed for each database. Eligible studies were systematic reviews that included the following: • Population (P): Studies targeting mothers, pregnant women, or women of reproductive age.
• Intervention (I): A wide range of intervention types was considered, broadly categorised as: education and counselling; hydration; water, sanitation and hygiene; women's empowerment; relaxation therapies; multi-sectoral interventions; m-health (i.e. those that use mobile or wireless technologies); conditional cash transfers; agriculture; maternal supplementation during pregnancy; and maternal supplementation during lactation.
• Control (C): Though reviews of randomised controlled trials were ideal, we also included reviews which included observational and uncontrolled original studies.
• Outcomes (O) : We included any review documenting infant growth or nutritional status in the first six months of life, including those aimed at preventing deterioration.
Search results were limited to human studies from LMICs as defined by the World Bank [26], published since 2008, and with full texts available in English language.
Search results were exported to Endnote software X9 (Clarivate Analytics, Philadelphia, USA) where duplicates were removed. Two researchers independently screening titles and abstracts followed by full texts. Where an updated version of an existing review was available, only the most recent version was included.
by the review authors. In order to rank the quality of the review methodology (1), we used the Grades of Recommendation, Assessment, Development and Evaluation (GRADE) Working Group approach according to British Medical Journal Best Practice [27].

Results
The initial search revealed a total of 2320 reviews, with an additional 94 returned by our search specifically on maternal supplementation (Fig 1). We reviewed 224 full texts (214 from the database searches, four recommended by experts via personal communication, and six identified via reference lists) and ultimately identified 55 relevant systematic reviews for inclusion. While all reviews captured studies from LMICs, as per our inclusion criteria, 34 (61%) reviews also incorporated studies from high-income countries (HICs). The studies within the reviews took place across Africa, Asia, The Americas, Europe and Australasia; Asia was the region with the most studies represented. The majority of the evidence related to the prevention of growth faltering, rather than management of existing growth faltering or existing infant undernutrition. Related to this, most studies were aimed at pregnant women rather than during the postnatal period. Many of the reviews also included infants up to 24 months, not just infants <6m.
Interventions took place at household-, community-and medical facility level. Three reviews focused their intervention only on adolescent girls [28][29][30].

Quality of evidence
The overall quality of the reviews was rated high according to the GRADE evaluation for the majority of reviews (S2 and S3 Tables); only eight (15%) reviews were rated moderate or low. Forty-one (75%) reviews included a meta-analysis. The quality of the studies included in the reviews, however, varied greatly. Most, 48(87%) of the reviews incorporated a structured quality assessment of the primary studies. Low quality evidence was mainly attributed to small sample size, high heterogeneity among study participants, inadequate randomisation, and insufficient blinding in the study set up.

Maternal-focused interventions
Education, counselling and peer support. Breastfeeding education, counselling and peer support. Three reviews considered the impact of breastfeeding education and counselling interventions. Given the known benefits of breastfeeding for infant health and survival [31], providing breastfeeding support should be a key management and prevention strategy for small and nutritionally at-risk infants <6m. However, Shakya et al. (2017) did not find a sufficient number of eligible studies to assess the effect of community-based peer support on infant nutritional status [32]. Giugliani et al. (2015) and Lassie et al. (2020) also did not find an impact on infant growth outcomes [33,34]. However, studies largely consisted of healthy infants in HICs, whereas some of the trials that focused on providing breastfeeding education for mothers of LBW infants did find an increase in weight and length [32][33][34]. Besides growth outcomes, several studies found counselling increased exclusive breastfeeding (EBF) and early initiation of breastfeeding [32][33][34].

PLOS ONE
Most education interventions focusing on the appropriate introduction of complementary feeding were associated with an increase in infants' weight and length [34,38] (Table 1). A Cochrane review concluded that complementary feeding education may reduce the risk of growth faltering for term-born infants, but modest effects may not be clinically significant, and long-term effects are uncertain [38].
Community health worker home visits combined with mother/peer support groups showed a positive effect on reducing the risk of wasting, but no effect on stunting or underweight [35,42]. Reviews of antenatal nutrition education programs have found mixed effects, with some suggesting they can improve infant birth weight and reduce risk of preterm birth, but others showed no or limited effect [39][40][41][42][43][44]. Community health education programmes were also shown to reduce neonatal mortality, especially when provided during both antenatal and postnatal periods, and when additional family members are incorporated [37]. However, a Cochrane review of social support programs (home visit, regular antenatal care) for mothers at increased risk of LBW babies concluded that they are unlikely to have a large impact on the proportion of LBW and preterm births [44].
Water, sanitation and hygiene (WASH). The effectiveness of WASH interventions on infant outcomes are mixed [28,41,42,[45][46][47]. A Cochrane review including 14 studies found that an increase in height was more responsive to WASH interventions in children under

Prevention:
No studies reported on infant growth/behaviours; reduction on maternal stress, evidence for increasing milk yield and fat content.

Highmoderate
All studies had limitations regarding design or sample collection procedures-(2/5) not randomised, studies unable to blind participates, many had small sample sizes, did not consider all potential confounders Management: Strongest evidence from increasing milk yields, and possible fat content in mothers of preterm infants.

IPTp/INT-reduction in low birth weight.
Mental health-promising effect of cognitive behaviour theory during pregnancy on depression post-partum, no evidence effect on infant weight or linear growth.

Prevention: iron folic acid:
Positive impact on the prevalence of maternal anaemia.

Low
Not formally assessed, only large scale programs that were adequately evaluated were utilised, high risk of publication bias for grey lit Food supplementation/ nutrition education + counselling: improved caregiver knowledge and behaviour change and improving utilisation of antenatal care, dietary habits, no effect on maternal weight gain, effect on birth weight unclear.

+ conditional cash transfers:
some evidence for positive impact on birth weight.
Preterm birth, birth weight, length-for-age z score

Highmoderate
Overall: moderate quality Inconclusive findings were observed for mHealth effects on maternal knowledge but positive effects on service utilisation.
(Continued ) 24 months of age, whereas increase in weight was more responsive in children 25-60 months of age [45]. Overall, combined WASH interventions were found to be more effective than single interventions in improving infant outcomes [45][46][47]. Only a small selection of possible WASH interventions was applied within the studies, with poor quality assessment and relatively limited duration [45][46][47].
Women's empowerment. Women's empowerment has been hypothesised to improve children's nutritional outcomes by enabling women to use and realign resources and household practices to enhance their children's nutritional well-being. Related interventions were assessed in two different reviews [29,48]. The three most important gender-related levers were identified as being: control over income/assets/resources, decision-making power and education [48]. Higher levels of decision-making power were associated with improved nutritional status in children and a reduced risk of stunting [48]. Interventions targeting both men and women, however, showed mixed results on effectiveness, while adolescent-specific interventions had no effect on infant health outcomes [29].
Relaxation therapy. There are a small but growing number of studies on relaxation therapy, the core idea being that relaxation (via a range of options including music and guided breathing exercises) could be an easy, quick and deliverable intervention with a clear and direct biological pathway to improved infant growth via increased breastmilk supply. One systematic review was identified. Though it found no studies reporting on infant outcomes, it did Height-for-age z score (HAZ); weight-for-age z score (WAZ); mid-upper arm circumference (MUAC); small for gestational age (

PLOS ONE
find that maternal relaxation was associated with reduced maternal stress, increased milk yield and milk fat levels, especially in mothers with preterm infants [49]. m-Health. Three reviews focused on m-health interventions [50][51][52]. One found that LBW was less common among women in the intervention group (short message service, audio voice messages or a combination) compared with the control group [50]. M-Health was also found to be useful in encouraging consumption of micronutrient supplements in pregnancy [50]. Targeted messaging via mobile devices was found to marginally improve breastfeeding rates, but had little or no effect when compared to non-digital strategies [51]. Few studies within the reviews presented growth and mortality outcomes, but one study, which combined educational text messages and two-way voice calls found a significant reduction in perinatal mortality [52].
Conditional cash transfer programs. Three reviews assessed the impact of conditional cash transfers on infant outcomes. However, these reviews were mostly based on programs implemented in Latin and Central America and referred to the same study populations [40, 53,54]. The programs generally combined basic cash transfer with health and nutrition education [53]. An overall positive effect of conditional cash transfers programs on the risk of stunting and underweight was reported [54] alongside a possible impact on birth weight [40]. There was a greater impact of conditional cash transfers on younger children (infants <6m vs children 6-48 months), children from lower socio-economic backgrounds and children with longer exposure to the intervention [53,54]. Studies also found improvements in maternal micronutrient consumption but also consumption of high-fat foods [53,54].
Maternal mental health. While there are strong associations between maternal mental health and infant outcomes in observational work, intervention studies are still lacking. One review included maternal mental health interventions alongside other interventions. They report that cognitive behaviour therapy during pregnancy had no impact on infant weight or growth, but a positive impact on depression post-partum [28]. A Cochrane review of social support programs for mothers at increased risk of LBW babies found a slight positive impact on postnatal depression [44].
Agriculture interventions. Agricultural interventions such as bio-fortification, home gardens, small scale fisheries and aquaculture, dairy development, animal husbandry and poultry development rarely reported the effect on child nutritional outcomes [55]. A limited amount of studies reported a positive impact on Vitamin A status, stunting, underweight and wasting [30,55]; however, studies often lacked high-quality impact evaluations. The effect of agricultural interventions appeared to be bigger when targeting women within empowerment or educational activities to foster their knowledge and skills and increase control over income [30].
Multi-sector interventions. Given that the determinants of infant growth are multi-faceted, eight reviews summarised the evidence base for multiple interventions across sectors [28,30,[39][40][41][42][43][44]. Many of them combined micronutrient supplementation with antenatal education or counselling, however WASH interventions, mental health and social support, agriculture and social safety nets were also captured. Scaling up of nutrition packages that include optimising infant feeding and increasing maternal dietary diversity during pregnancy may successfully reduce the risk for stunting and wasting in children under five years [28]. Implementing a range of multisectoral actions/interventions rather than focusing on one particular intervention appears to increase program effectiveness [28,30]. However, more research is needed as interventions are still often implemented and evaluated in isolation [41]. Furthermore, more comprehensive evaluation tools are needed, to avoid dismissing an effective program due to a lack of effect on a small set of outcomes [40].
Maternal supplementation during pregnancy. Of the 28 reviews focused on maternal supplementation and infant outcomes, most, 23 (82%) review the evidence of supplementation in pregnancy (Table 2). Overall, the reviews found that food fortification and macronutrient interventions had a positive effect on birth weight and length, risk of stillbirth and small for gestational age (SGA) [28,41,43,[56][57][58][59][60] though there was no evidence of longer-term benefits for child growth and development [57]. A recent Cochrane review on lipid-based nutrient supplements in pregnancy found positive impacts on birth weight, length and SGA when compared to iron folic acid supplementation, but no benefit compared to multiple micronutrient (MMN) supplementation [61].

High
Overall very low quality

Zhou et al. (2013) [67]
Iodine Child development and growth Regions severe deficiency: reduced risk of cretinism, but no impact on childhood intelligence, gross development, growth, or pregnancy outcomes, although evidence of improvement in some motor functions.

Highmoderate
None of the studies reported adequate random-sequence generation, high risk of bias due to incomplete data, unclear risk for blinding due to inadequate reporting

Regions mild-moderate deficiency:
No data on child growth and development.

Rumbold et al. (2015) [68]
Vitamin   Uncertain of adverse effect for mothers, those reported were minor complaints.
Infant and child mortality, morbidity, and growth: weight, length, head circumference, WAZ, length-for-age z score Overall: no impact infant and child mortality and morbidity

Moderatelow
Not available for all outcomes: Wide range very low-high Polyunsaturated fatty acid supplementation vs placebo: no impact infant length, weight and head circumference Polyunsaturated fatty acid supplementation during gestation and lactation: no significant growth effects, some evidence for impact on child attention beyond 24 months of age.
An RCT of maternal calorie supplementation + breastfeeding support: no impact WAZ or length for age z score; Increased infant breast milk intake and EBF.

Abe et al. (2016) [81]
MMN Infant mortality, morbidity, adverse effects within three days of supplementation No evidence to quantitatively assess the effectiveness of MMN supplementation in improving health outcomes in mother and baby.

Highmoderate
Poorly reported among original studies; unclear risk of bias Impact on maternal anaemia (1 study) and vitamin B12 and folic acid milk concentration, but no significant dereferences were reported in serum concentrations (1 study).

Zinc, pregnancy:
No impact on LBW, birth weight, one study found a beneficial effect of HAZ at 6months, no impact on other growth parameters. Zinc, lactation: One study (20mg, 40mg and control), no effect on breastmilk, growth or serum zinc concentration in exclusively breastfed infants.

Highmoderate
Majority low-moderate quality Iron, pregnancy: No impact on birth weight, LBW, mental development scores; no data on child growth. Iron, lactation: No impact on birth weight, LBW, mental development scores; no data on child growth.
Standard mean difference (SMD); Height-for-age z score (HAZ); Weight-for-age z score (WAZ); Mid-upper arm circumference (MUAC); Weight-for-height z scores Reviews of MMN supplementation found that compared to no supplementation or supplementation with three or fewer micronutrients, they positively impacted risk of LBW and SGA [41,62]. There was no statistically significant difference between supplements that contained at least 60 mg compared to supplement containing 30mg of iron and there was no difference in the timing of the intervention [62]. There was an increased risk of neonatal death in MMN intervention beginning after the first trimester when compared to iron folic acid supplementation, though this finding was based on low-quality evidence [62].
Vitamin D supplementation is a growing area of research; four reviews found a positive impact on birth weight, particularly if given after 20 weeks' gestation, with potentially continued effect at 3, 6, 9 and 12 months [75][76][77][78]. Deficient women may benefit more [77]. Zinc supplementation has a positive effect on risk of preterm birth but no evidence for other infant outcomes [28,72,82]. Iron folic acid supplementation was associated with increased birth weight [28,41]. Iodine supplementation may improve growth outcomes in severely deficient women [28,[65][66][67]. When stratified by country income there was an association between vitamin B12 and birth weight in LMICs, but not in HICs [70]. There is mixed evidence of the effect of omega-three fatty acid supplementation on birth outcomes [64]. There was also mixed evidence for calcium supplementation on birth outcomes [28,73]. It may be beneficial in low intake populations, or if given after 20 weeks of gestation [28,73]. Cochrane reviews of vitamin A, E and C supplementation showed no clinically significant impact on infant outcomes [63,68,69]. There is no quality evidence on effect of magnesium supplementation on birth outcomes [71].
Maternal supplementation during lactation. We found only five reviews examining the role of supplementation during lactation [7,[79][80][81][82]. There was no conclusive evidence that zinc supplementation, polyunsaturated fatty acid supplementation nor MMN provided to breastfeeding mothers impacted infant health and growth outcomes [82]. A review of galactagogues found some impact of Moringa and mixed botanical galactagogues on infant weight [7,80] (Table 2). One Cochrane review explored extra fluids for breastfeeding mothers, with the rationale that this might increase breastmilk production [79]. Authors concluded that there was "Not enough evidence to increase fluid intake beyond what likely to require for comfort" [79].

Discussion
Our review found a large number of good quality reviews which explore maternal-focused interventions to prevent or address growth faltering in infants <6m. Evidence from these can inform future management strategies for small and nutritionally at-risk infants <6m in LMICs. The options for maternal interventions include a wide range of interventions as reflected in our results categorisation: breastfeeding promotion interventions; education and counselling; WASH; women's empowerment; relaxation therapies; maternal mental health; mhealth; conditional cash transfers; agriculture, maternal supplementation in pregnancy; maternal supplementation during lactation; and multi-sectoral interventions.
Breastfeeding promotion interventions during pregnancy and promotion and counselling during lactation are effective at improving infant feeding practices however evidence is more limited on anthropometric impacts. Outcomes relating to infant feeding practices are generally self-reported and not objectively measured [83]. More research is needed on communitybased breastfeeding education and counselling that effectively improve growth outcomes in small or nutritionally at-risk infants <6m. Most evidence is for prevention of growth faltering, however there is evidence that lactation support can be used to successfully manage growth faltering at birth [84][85][86], and in older infants [87,88]. In the documented experiences of implementing a "MAMI" approach (Management of small and nutritionally at-risk infants under six months) in an inpatient setting, most infants presenting with growth faltering had the possibility to breastfeed and tailored, structured lactation counselling was successfully used to establish EBF and secure weight gain in most infants [87][88][89]. However, there were considerable challenges in achieving this for some higher risk infants and sustaining progress postdischarge in the absence of community follow-up [87]. Educating family members, such as fathers and mothers-in-laws, has also shown a positive effect on neonatal mortality and feeding practices, emphasizing the importance of an inclusive approach and considering the wider family dynamics that influence feeding practices [37,90]. Breastfeeding support packages in the workplace have shown positive impacts on breastfeeding initiation and duration. In high income settings they have also been found to be feasible to implement, presenting an opportunity to further target and support mothers and their infants [91]. However, this is much less evidenced in LMICs that also need to consider the informal work settings that so many mothers worldwide engage in. Education and counselling interventions, such as those focused on complementary feeding or care practices, generally have a positive but modest effect on infant growth outcomes; as with any education intervention, the effectiveness is likely to be proportional to the quality, content and uptake of the education provided [92]. Education interventions in combination with food supplementation or conditional cash transfers appear to be more impactful. Guidance on what constitutes effective education and counselling interventions for improving infant growth would be valuable for programmers faced with managing small and nutritionally at-risk infants <6m.
WASH interventions had some positive effects on linear growth, although less so for infants <6m. Women's empowerment and m-health interventions were also potentially effective when combined with agriculture interventions or nutrition-specific interventions such as micronutrient supplementation. One possible mechanism behind these interventions is that maternal diet is improved, either during pregnancy or lactation, which in turn supports better in utero/infant growth and nutrition. The rationale behind conditional cash transfer programmes is similar and there is evidence of some positive impact on infant growth outcomes, however the evidence notes an important potential side-effect, which is the increased maternal consumption of high-fat or high-sugar foods in middle income settings. While low maternal body mass index and micronutrient deficiencies in pregnancy and during lactation are associated with infant growth faltering, so is high body mass index [2,[93][94][95].
Maternal mental health is perhaps one of the most promising but under-researched areas for post-natal intervention. Evidence for the association between maternal mental health and infant growth is strong; a systematic review found postnatal depression was associated with underweight in infants <6m in two studies and poor weight and height in a third, and is linked to poor breastfeeding practice and increased illnesses in other studies [96]. Alongside maternal depression, increased stress biomarkers are associated with intrauterine growth restrictions and poor growth through the post-natal period [97]. A meta-analysis of cohort studies found maternal anxiety during pregnancy was associated with significant increased risk of pre-term birth and LBW [98]. A meta-analysis estimated that 23% to 29% fewer children would be underweight or stunted if the infant population were unexposed to maternal depressive symptoms [99]. Despite strong associations, we found few reviews on maternal mental health interventions for infant growth and nutrition. One review found that cognitive behavioural therapy during pregnancy had no impact on infant weight or growth, but a positive impact on depression post-partum. However, Fotiou et al (2017) found that cognitive behavioural counselling had a positive impact on breastfeeding initiation and the duration of EBF [100]. In addition, there is promising evidence from relaxation therapy while breastfeeding, which has been found to reduce maternal stress levels, increased milk yield and milk fat levels, especially in mothers of preterm infants. Since that review was published, a randomised controlled trial (RCT) found greater weight gain and body mass index in infants in the relaxation intervention group but no effect on head circumference and length [101]. They also found a significant effect on the duration of sleep at 6-8 weeks and reduced cortisol concentrations in breastmilk [101]. Relaxation therapy is a fast-evolving area of research [102,103]; other context-specific interventions that relieve maternal stress while breastfeeding could have similar effects.
Lastly, interventions to supplement maternal diets with macronutrients, MMN, vitamin D, zinc, iron folic acid, and possibly calcium, iodine and B12 in deficient women, in pregnancy, can improve birth outcomes. However, a recent RCT, not captured by the reviews, found that ready-to-use supplementary food plus daily MMN antenatal supplementation for moderately malnourished pregnant women was only modestly effective at improving length at birth, 6 weeks and 12 weeks, despite the large quantity of calories provided; this suggests that in highrisk settings stunting in utero is unlikely to be reduced by supplementation alone [104]. While there is evidence that low postnatal maternal body mass index is a risk factor for infant growth faltering, there is currently little evidence for the effect of maternal supplementation during lactation to improve infant growth. There were also few intervention studies for improving maternal dietary diversity during pregnancy and lactation, a known risk factor for poor infant growth outcomes [105]. Interventions seeking to address maternal dietary diversity could also influence infant diets during complementary feeding after six months age, another known risk factor for poor later child growth [106].

Limitations
We acknowledge several limitations to our review. First, limiting our search to reviews means that some primary evidence may have been missed. On the other hand, this approach also enabled us to capture a broader range of interventions and set the stage for important future research.
We also acknowledge that we only included reviews published since 2008. This is unlikely however to have altered our overall conclusions since individual reviews still include earlier original studies. Another limitation is that the same primary studies could have been included in several reviews; this may have given unwarranted weighting of the evidence from one study included across multiple reviews. Furthermore, while all reviews included studies from LMIC, several reviews (61%) also included studies from HIC; the inclusion of HIC may affect the applicability of our results to LMIC settings.
We were constrained in our ability to disaggregate data depending on how data was reported. Many reviews did not disaggregate data on infants <6m from older children; recognising infants <6m as an important distinct age category is vital to build evidence for best practice. Similarly, some reviews included studies that combined the exposure of interest with supplementation or included studies with special populations, such as those at increased risk due to underlying conditions. This may have affected the comparability of results but due to the small number of these studies, we do not believe this significantly affected our study's findings. Variability across the infant outcomes measured made comparing and quantifying the effects of different interventions difficult. Whilst the quality of the reviews was generally high, the quality of the respective primary studies was sometimes low. GRADE criteria were used to assess the quality of the included reviews; while it is a standardised and widely recognised method, it does not account for the inclusion of grey literature, which can increase the quality of reviews. There were also several interventions with only very limited evidence, such as a small number of studies from a limited number of contexts, not RCTs, and small sample sizes.
Most evidence we found relates to the prevention of growth faltering, rather than the management of already small and nutritionally compromised infants <6m. This highlights the need for future studies to focus specifically on this group; whilst it is plausible and likely that interventions which work for prevention also work for management, it is not certain and effect sizes may differ. Most studies were also timed around the antenatal rather than the postnatal period. Future programmes should support mothers throughout (and before) pregnancy but better understanding of the relative impact of different interventions at different times is still important to help prioritise and target. Whilst this review focused only on outcomes for the infant, outcomes for the mother are equally important and also warrant attention and action.

Policy implications
The heterogeneity across the zero to six month period, transitioning from the umbilical cord to solid feeding, and the contextual dynamic and needs of the individual mother-infant dyad, will likely warrant an assortment of both maternal and infant-focused interventions, working together in synergy. While a coordinated strategy is required, one approach or set of interventions alone is unlikely to suit all contexts and all infants <6m. Addressing maternal as well as infant needs reflects a recent WHO report on Essential Nutrition Actions and the 2021 Lancet Maternal and Child Undernutrition series, recognising the importance of addressing nutrition throughout the life course [17,107]. The UN 'Global Action Plan' (GAP) on Child Wasting: Framework for Action also stresses the importance of maternal nutrition interventions during the preconception, pregnancy and postnatal period, to improve infant outcomes, specifically for the reduction of LBW [108]. The need for essential nutrition packages to be integrated into national health systems and policies is also highlighted [108]; essential nutrition packages must include multifaced maternal interventions for managing small and nutritionally at-risk infant <6m. This could be achieved though strengthening referral pathways between services that are already available (e.g. maternal mental health, infant and young child feeding) or through the adoption of adaptable community-based interventions that combine maternal and infant nutrition to fill service gaps. One example of such an approach is a recently released integrated care pathway for managing small and nutritionally at-risk infants <6m and their mothers [109].

Conclusion
Our review of reviews identifies a variety of interventions aimed at mothers in the antenatal and postnatal period in LMIC that could be used to also improve infant nutrition outcomes. Several interventions proved to be effective in preventing growth faltering in infants <6m, including: breastfeeding, education and peer support, counselling interventions, antenatal supplementation combined with women's empowerment, m-health technologies, WASH, conditional cash transfers, and/or agriculture interventions. Relaxation therapies and mental health interventions also have potential-but more evidence is needed. Besides galactagogues, evidence on post-natal maternal supplementation for managing infant growth faltering has shown limited impact, although the number of studies is limited. There is limited evidence that directly considers impact of interventions on small and nutritionally at-risk infants <6m, highlighting an important research gap. Based on this collated evidence, more holistic packages of care that include maternal interventions are justified; contextualised support of the mother-infant dyad could help small and nutritionally vulnerable infants aged <6m survive and thrive.
Supporting information S1