The effect of interventions distributing home fortification products on infant and young child feeding (IYCF) practices: A systematic narrative review

Abstract Interventions distributing micronutrient powders (MNPs) and small‐quantity lipid‐based nutrient supplements (SQ‐LNS), or home fortification products (HFPs), have the potential to improve infant and young child feeding (IYCF) practices and children's nutrition. We systematically searched for studies on the effect of interventions distributing HFP on IYCF practices. We identified 12 (8 MNP, 4 SQ‐LNS) studies: seven programmes with IYCF behaviour change communications (BCC) and MNP (IYCF‐MNP) and one provided MNP without IYCF BCC (MNP only). Three SQ‐LNS studies came from randomised trials without an IYCF component (SQ‐LNS only) and one from a programme with both IYCF BCC and SfQ‐LNS (IYCF‐SQ‐LNS). Five IYCF‐MNP programmes reported positive associations with some IYCF practices—four with minimum dietary diversity, two with minimum meal frequency, four with minimum acceptable diet, and three with the initiation of complementary foods at 6 months. Two reported no association between MNP and IYCF indicators, and one reported a decline in IYCF practices during the intervention, although it also reported significant changes to the IYCF programme during the evaluation period. Two studies from interventions that distributed SQ‐LNS (one from a related set of randomised controlled trials and the sole IYCF‐SQ‐LNS programme) reported a positive association with IYCF practices; one trial reported no change in breast milk intake with the provision of SQ‐LNS and one found no association with IYCF practices. SQ‐LNS and MNP can address nutrient gaps for young children in low‐resource settings; our findings indicate that programmes that combine HFP with IYCF interventions may also contribute to improved IYCF practices in some settings.


| INTRODUCTION
Malnutrition during early life increases the risk of mortality and morbidity in children and has lifelong consequences for physical and cognitive development (Black et al., 2013). Improving infant and young child feeding (IYCF) practices is essential to reduce stunting and other forms of malnutrition (Bégin & Aguayo, 2017). The World Health Organization (WHO) recommendations on IYCF include immediate and exclusive breastfeeding for the first 6 months, and the introduction of solid and semisolid complementary foods for children at 6 months of age when breast milk is no longer sufficient to meet children's nutritional needs (World Health Organization [WHO], 2003). In resource-limited settings, nutrient-rich foods are often inaccessible and/or are not provided in sufficient quantities to young children (Dewey et al., 2009) Home fortification (or point-of-use fortification) is an approach to improving the nutrient intake of young children and other nutritionally vulnerable groups which involves the addition of micronutrient powders (MNPs) or other specialised fortified products, such as lipidbased nutrient supplements (LNS), to foods immediately before consumption (Dewey et al., 2009; World Health Organization [WHO], 2011). Common home fortification products include MNP and small-quantity lipid-based nutrient supplements (SQ-LNS). MNPs are single-dose, lightweight, shelf-stable sachets designed to be mixed with a variety of semisolid foods to increase the availability of vitamins and minerals in children's diets (Zlotkin et al., 2005). SQ-LNS are sachets that contain micronutrients as well as protein and essential fatty acids providing energy. SQ-LNS typically contain ≤20 g (≤120 kcal or~4 teaspoons) of an LNS, usually comprised of peanuts, vegetable oil, milk powder, sugar and added vitamins and minerals (Food and Nutrition Technical Assistance III Project [FANTA], 2016). In contrast to medium and large-quantity LNS, which are primarily distributed as part of treatment for acute malnutrition, SQ-LNS are usually provided for home fortification or for direct consumption to prevent undernutrition (Arimond et al., 2015).
Multiple meta-analyses have demonstrated that MNP can reduce the prevalence of anaemia and iron deficiency (De-Regil et al., 2013;Salam et al., 2013;Suchdev et al., 2020;World Health Organization [WHO], 2016) and that SQ-LNS can reduce the risk of stunting, wasting and other forms of undernutrition in children aged 6-23 months (Das, Salam, Hadi, et al., 2019;. The integration of these products into IYCF interventions may also influence IYCF practices (Siekmans et al., 2017). Unlike industrial fortification strategies (when nutrients are added during industrial food processing to household food items such as cooking oil or staple grains), home fortification allows for the targeting of young children with specific doses of several nutrients. Home fortification also, however, requires behaviour change on the part of the caregiver, who must regularly mix the fortification product into the child's food. It has been proposed that incorporating home fortification products into IYCF programmes could contribute to improved IYCF practices by establishing an enabling environment at the governmental and community levels to support IYCF programmes, and through the synergies created by providing a comprehensive package that includes counselling as well as improved access to essential nutrients (Siekmans et al., 2017). In programmatic settings, home fortification distribution is often implemented alongside enhanced IYCF trainings for facility-based and community-based health workers (CHWs) and behaviour change materials that re-enforce key IYCF messages. It is also possible that the introduction of a tangible good (the fortification product) for CHWs to distribute in the community, may improve the demand for and delivery of CHW services (Locks et al., 2018;Reerink et al., 2017;Vossenaar et al., 2017), which could, in turn, improve IYCF practices as well as other child health and nutrition outcomes.
Conversely, there is also concern that home fortification products could negatively impact IYCF by becoming a distraction for caregivers and/or by directly displacing breast milk or other foods, particularly in the case of LNS or other fortification products that include macronutrients (Flax et al., 2015;Kumwenda et al., 2014). To minimise disruption of IYCF practices, it is frequently recommended that SQ-LNS is mixed into local complementary foods (as opposed to direct consumption) (FANTA, 2016).
To determine whether interventions that distribute home fortification products influence IYCF practices, we conducted a systematic search for studies from interventions that provide these products and reported on the effect on IYCF practices.

Key messages
• We identified 12 studies distributing home fortification products (HFP) that assessed infant & young child feeding (IYCF) outcomes; few used experimental or quasi-experimental study designs.
• Interventions associated with improved IYCF practices were predominantly programs that integrated HFP with a new, intensive IYCF behavior change strategy.
• HFP can fill nutrient gaps for infants and children in lowresource settings; programs that combine HFP with IYCF interventions may also contribute to improved IYCF practices in some settings.

| METHODS
A systematic search in accordance with the PRISMA guidelines (Page et al., 2021) was conducted by a CDC staff librarian ( Figure 1). First, they searched PubMed, Embase, Scopus, Ebsco Academic Search Complete, Cochrane Reviews, LLILACS and SCIELO for the following keywords in the title or abstract, or used relevant Mesh terms in PubMed: 'infant or child with micronutrient powders', 'small quantity lipid-based nutrient supplements', 'SQ-LNS', 'fortification or fortificant'.
The search led to 224 independent articles after de-duplification. Two independent readers then screened the abstracts and titles of all of the articles with the following inclusion criteria: (1) English or Spanish language; (2) an intervention that provided home fortification products to children aged under 24 months-intervention was broadly defined to include trials, programmes, policies; home fortification products was broadly defined as any product intended to be added to foods to increase the quantity of nutrients or bioavailability of nutrients to prevent malnutrition; and (3) reported on IYCF outcomes-broadly defined to include any indicators related to breastfeeding and complementary feeding practices (including dietary diversity and frequency). We excluded articles that did not explicitly use a home fortification product for the prevention of malnutrition (i.e., medium or large-quantity LNS as part of treatment for malnutrition) and also excluded articles that only reported on the effect of home fortification products on biological outcomes (i.e., anthropometry and anaemia) without including IYCF indicators. Fourteen articles fulfilled the inclusion criteria and their full texts were retrieved for further review of the full article to further determine eligibility for inclusion. We ultimately excluded one abstract from a conference that did not have a full-text manuscript (Matias & Vargas-Vasquez, 2017) and one trial that provided a larger LNS quantity (46.3-70 g) and did not explicitly counsel caregivers to mix the LNS with food (Flax et al., 2015).

| RESULTS
Our final review included 12 articles-eight on MNP interventions (Ford et al., 2021;Geletu et al., 2019;Lanou et al., 2019;Locks et al., 2017Locks et al., , 2018Mirkovic et al., 2016;Tariqujjaman et al., 2021;Young et al., 2021) and four on SQ-LNS interventions (Arimond et al., 2015;Byrd et al., 2019;Kumwenda et al., 2014;Locks, Nanama, et al., 2019); no other home fortification products were discovered during our search. Table 1 shows all studies from interventions distributing MNP were from integrated IYCF-MNP programmes or pilot programmes that provided MNP along with nutrition education and/or counselling on IYCF except one study from a pilot MNP programme that did not include an IYCF behaviour F I G U R E 1 PRISMA diagram. LNS, lipid-based nutrient supplements. Author ( change component (Geletu et al., 2019). Among the four papers reporting on the effect of an SQ-LNS intervention on IYCF practices, three of the articles report results from randomised controlled trials (RCTs) and only one from a programmatic setting (Table 2). Two RCT studies included nutrition or IYCF messages, one RCT with eight sites included monthly home visits delivering IYCF counselling and free SQ-LNS in the two 'nutrition' arms, and one pilot programme included IYCF counselling. other studies reported a prevalence of MDD at endline that was below 50%. The prevalence of MMF at endline was generally higher than MDD, with wide variability from 27.1% in Uganda (Ford et al., 2021)  T A B L E 2 Studies on the effect of interventions with home fortification with small-quantity lipid-nutrient supplements (SQ-LNS) on infant and young child feeding (IYCF) practices.
Furthermore, the significant difference-in-differences in the quasiexperimental studies were partially driven by declines in IYCF indicators in the control areas (Ford et al., 2021;Lanou et al., 2019).
One article reported a decline in IYCF indicators during the period of an IYCF-MNP programme in Bangladesh (Tariqujjaman et al., 2021).  (Lanou et al., 2019;Locks et al., 2017), one of which found a statistically significant association between the intervention and the timely introduction of solid foods (Lanou et al., 2019). Three studies used maternal recall on the timing of the introduction of solid foods in the full sample (Ford et al., 2021;Locks et al., 2018;Young et al., 2021), two of which found a statistically significant association between the intervention and the introduction of solid foods at 6 months (Ford et al., 2021;Locks et al., 2018).

| Interventions distributing SQ-LNS
There was greater heterogeneity in the study design and assessment  SQ-LNS per day in Malawi and found no significant difference in intake of breast milk or nonbreast milk water.
Of the three papers that reported on the WHO/UNICEF IYCF indicators, two were from clinical trials of home fortification with SQ-LNS. One reported on IYCF practices in the WASH Benefits trial, an eight-arm trial of water, sanitation, handwashing and nutrition interventions in rural Kenya (Byrd et al., 2019). The other study reported on the international lipid-based nutrient supplements (iLiNS) trials-four related RCTs of SQ-LNS in sub-Saharan Africa (Arimond et al., 2017). The WASH Benefits trial found no difference in IYCF indicators at 1 or 2 years of follow-up in children who received the nutrition intervention (SQ-LNS + nutrition counselling) compared to children in the 'active control group' (no SQ-LNS or nutrition counselling, but monthly home visits with assessment of child's arm circumference) in rural Kenya. By contrast, the iLiNS group reported that the prevalence of children receiving the MMF at 18 months was significantly higher among children randomised to receive SQ-LNS (compared to no SQ-LNS) in the two trials that assessed feeding frequency (percentage point differences of 12%-14%, p < 0.0001 and p = 0.005), even though none of the trials were specifically designed to include an IYCF behaviour change communication (BCC) component (the other two trials did not assess feeding frequency).
The iLiNS group did not find significant differences by the treatment group in the prevalence of children consuming the MDD, continued breastfeeding at 18 months or frequent breastfeeding (6 or more times per day) in any of the four trials, but they did report a reduced prevalence of infrequent consumption of animal source foods in the same two trials that found a significant difference in MMF.
Only one study reported on the effect of an integrated IYCF-SQ-LNS intervention from a programmatic setting (Locks, Nanama, et al., 2019). The 2-year pilot intervention was implemented in the Haut-Katanga District in the Democratic Republic of Congo (DRC) targeting 23,000 pregnant women and infants 0-12 months of age in Kasenga health zone with an enhanced IYCF programme that included monthly SQ-LNS distribution, as well as resources to support CHW IYCF counselling (trainings, job aids and bicycles for CHWs to travel to remote areas). The study used a quasi-experimental study design to compare difference-in-differences in the intervention and control zones at baseline and endline and found that mothers in the intervention zone were more likely to report the timely introduction of complementary foods, feeding the MMF the previous day, feeding the child in a separate bowl, awareness of anaemia, owning soap and washing hands after defecating and before cooking and feeding the child in the previous day. There were no significant differences in the prevalence of MDD and MAD in the intervention and control zones, where both remained below 10% at both baseline and endline.

| DISCUSSION
This systematic narrative review found that there is limited research on the effect of interventions with home fortification products on IYCF practices. All but one of the studies on interventions providing MNP are from integrated IYCF-MNP programmes or pilot programmes (there was one pilot programme that did not contain an IYCF component). By contrast, there have been few programmes that have distributed SQ-LNS, and thus all but one of the published studies are from randomised trials.
Most of the studies from integrated IYCF-MNP programmes found either no association with IYCF practices or a small beneficial effect on IYCF practices. One study, from Bangladesh, reported a decline in IYCF practices during the implementation of an integrated IYCF-MNP programme, but important methodological limitationsnamely the removal of CHW incentives for IYCF practices in the middle of the programme period-make it difficult to attribute the decline in IYCF practices to the use of home fortification products.
From the studies available, it is not possible to isolate whether there is an independent effect on IYCF practices of adding home fortificants to existing, intensive IYCF. In most of the studies, the introduction of MNP was part of an integrated IYCF-MNP programme, which included the introduction of a new, intensive IYCF BCC component (Ford et al., 2021;Lanou et al., 2019, Locks et al., 2017Mirkovic et al., 2016). The only study designed to assess the independent effect of MNP (Young et al., 2021)  breastfeeding. In addition, the provision of a home fortificant may improve caregiver and/or health worker motivation and thus improve the quantity, quality and uptake of interpersonal counselling and other IYCF BCC (Siekmans et al., 2017).
Despite the association between integrated IYCF-MNP programmes and improved IYCF practices in several of the studies in this review, the effect sizes were relatively small. MNP fills essential nutrient gaps in the diets of young children; however, improving IYCF, particularly the MDD, will also require interventions that focus on the access and feeding of diverse foods for young children. Furthermore, several of the studies reported a low prevalence of optimal IYCF practices at endline, and several external studies have reported limited MNP coverage and adherence in national and subnational programmes (Locks, Dahal, et al., 2019;Tumilowicz et al., 2019). Taken together, these findings highlight the continued need to strengthen the delivery of communitybased IYCF and MNP programmes (Pelletier & DePee, 2019). Of note, several of the studies in this review reported an association between effective programme coverage (defined by each study, usually as receipt of MNP and/or IYCF counselling) and improved IYCF practices (Locks et al., 2017(Locks et al., , 2018Mirkovic et al., 2016;Tariqujjaman et al., 2021), and all of the programmes that found an association between the intervention and IYCF practices included a strong community-based component (Ford et al., 2021;Lanou et al., 2019;Locks et al., 2017Locks et al., , 2018Mirkovic et al., 2016). In Nepal, Locks et al. (2018) (Menon, Nguyen, Saha, Khaled, Kennedy, et al., 2016;Rawat et al., 2017). Given that optimal IYCF practices and the consumption of home fortificants requires daily engagement of the caregiver, consistent, community-based support for improving both IYCF practices and MNP programme indicators is essential for sustained behaviour change.
In the few studies distributing SQ-LNS included in this review, we found small or no associations between SQ-LNS distribution and IYCF practices. This is consistent with qualitative research showing that SQ-LNS has been well-accepted in many settings (Adu-Afarwuah et al., 2011;Lesorogol et al., 2015;Tripp et al., 2011) and that mothers did not report any interference or changes in breastfeeding or complementary feeding practices (Lesorogol et al., 2015). These findings are also supported by the RCT from Honduras which was excluded from this review because the quantity of LNS distributed was larger than standard SQ-LNS doses (46.3-70 g/day) and was not distributed as a home fortification product (only 30% of mothers reported mixing it with food). Despite the larger quantity of LNS provided to the children in the Honduras study, the 24-h recall data indicated that the provision of LNS did not affect the quantity of food consumed from any of the food groups assessed, with the exception of increased intake of food groups comprised partially or wholly of the supplement (nuts or nut butters, oils and sweets) (Flax et al., 2015).
In contrast to integrated IYCF-MNP programmes, there is limited research on the effect of programmes distributing SQ-LNS on IYCF practices. We found multiple studies from randomised trials assessing whether SQ-LNS displaces breast milk (Kumwenda et al., 2014) or generally impacts IYCF indicators (Arimond et al., 2017;Byrd et al., 2019), but only one from a programmatic setting (Locks, Nanama, et al., 2019). As of 2020, MNP programmes had been implemented in at least 56 countries worldwide (United Nations Children's Fund [UNICEF], 2022). Recent meta-analyses have found that SQ-LNS reduces the risk of stunting and other measures of undernutrition across a range of contexts Keats et al., 2021). Given these promising findings, many IYCF programmes may seek to integrate SQ-LNS; it will be important to continue to assess whether This review has strengths and limitations. We systematically reviewed the scientific literature in accordance with the PRISMA guidelines and identified 12 studies assessing the effect of home fortification products on IYCF practices. Given the substantial heterogeneity in the choice of study design (including some without control groups) and selection of IYCF indicators, we were unable to perform quantitative summaries of the research on this topic.

| CONCLUSION
Our findings indicate that programmes that combine home fortification products and IYCF interventions had positive and/or limited influences on IYCF practices. Given that that home fortification products contain several essential nutrients that are lacking in the diets of infants and young children in low-resource settings, the integration of these products into national and subnational programmes has the potential to improve IYCF practices and nutrition outcomes in vulnerable children (Figure 1).