Impact of nutrient supplementation on maternal nutrition and child growth and development in Sub‐Saharan Africa: the case of small‐quantity lipid‐based nutrient supplements

Abstract Micronutrient deficiencies remain common among women and children in Sub‐Saharan Africa (SSA); in pregnant/lactating women, the intakes of essential fatty acids may also be low. Enriching home‐prepared foods with small‐quantity lipid‐based nutrient supplements (SQ‐LNSs) is a promising new strategy of delivering additional micronutrients, essential fatty acids and good quality protein to women and children. This narrative review aimed to examine the impact of SQ‐LNSs supplementation among women and infants and young children in SSA, and to discuss the differential impact of SQ‐LNS consumption across different settings. Papers reporting randomized trials conducted in SSA in which apparently healthy women and/or ≥6‐mo‐old children received SQ‐LNSs were identified through electronic and manual searches. Prenatal SQ‐LNS consumption reduced the prevalence of low gestational weight gain in Ghana when compared with multiple micronutrients supplementation, and was associated with poorer iron/hemoglobin status when compared with iron‐plus‐folic acid supplementation. SQ‐LNSs received alone or as intervention package improved infant/child growth in two trials in Ghana and one trial each in Burkina Faso, Kenya, Zimbabwe and South Africa, but had no impact on growth in two trials in Malawi. SQ‐LNSs supplementation improved motor development in Ghana, Burkina Faso, Malawi, Kenya, and South Africa, but had no impact on language, socio‐emotional, and executive functions in Ghana and Malawi and on Griffiths’ developmental scores in Malawi. SQ‐LNSs may contribute to improving child growth in SSA. More research is needed to determine the iron level in SQ‐LNSs effective for improving both maternal hemoglobin/iron status and birth outcomes.

of age decreased from about 76% to only 60%, and stunting rates from 49% to 34% .
Besides the high prevalence of micronutrients deficiencies, an analysis of food supply data (Michaelsen et al., 2011) suggests that the essential fatty acid intakes of pregnant and lactating women in some settings in SSA may be low.
It is likely that micronutrient deficiencies may contribute to some of the adverse health and nutrition outcomes among women and children in SSA. In 2015, an estimated 14% of infants born in the region had low birth weight (Blencowe et al., 2019) mostly due to fetal growth restriction linked to micronutrient deficiencies (UNICEF/WHO, 2019). Micronutrient deficiencies may also contribute to the relatively high rates of pregnancy complications (Darnton-Hill & Mkparu, 2015), maternal mortality (Bailey, West, & Black, 2015), and impaired child growth and development in SSA.
Much of the micronutrient deficiencies among women and children in SSA may be attributed to low dietary intakes, likely as a result of over-reliance on poor-quality staple diets low in nutrient density (Bose, Baldi, Kiess, & de Pee, 2019;Deptford et al., 2018). For infants and young children, additional challenges exist, including relatively small gastric capacity, which limits how much can be consumed at a meal, and low meal frequency (Dewey, 2016). Meanwhile, nutritionspecific interventions such dietary diversification (eg. promoting the consumption of locally available nutrient-dense food groups), nutrient supplementation, bio-fortification, and commercial fortification of staple foods may not always yield the desired results in SSA due to a variety of reasons including low affordability, particularly in deprived communities (Lalani, Bechoff, & Bennett, 2019), and low coverage rates of fortified foods (Univ Ghana/GroundWork/Univ Wisconsin-Madison/KEMRI/UNICEF, 2017). In addition, it is difficult to deliver adequate amounts of iron and zinc in complementary foods for infants without any fortification, even when using a combination of several locally-available high quality foods (Bose et al., 2019;Osendarp et al., 2016).
Reports from several studies in SSA (Adu-Afarwuah et al., 2007;Adu-Afarwuah et al., 2015;Adu-Afarwuah et al., 2016;Hess et al., 2015) suggest that the use of small-quantity lipid-based supplements (SQ-LNSs) to enrich home-prepared foods for women and children might hold promise as one of possible interventions for delivering a large number of essential micro-and macro-nutrients, and thereby addressing nutrient deficiencies, in these vulnerable groups. Two recent Cochrane reviews examined the impact of SQ-LNS consumption among women (Das et al., 2018) and infants and young children 6-23 mo of age (Das et al., 2019), but those reviews extended beyond SSA, and the possible reasons for the differential impact of SQ-LNS consumption across different settings were rarely discussed. This paper aimed to examine the impact of SQ-LNSs supplementation among women and children in SSA, and to discuss the possible factors related to the positive impact of SQ-LNS supplementation in some, but not all, study contexts.

| METHODS
In a narrative review, papers reporting randomized trials of SQ-LNSs consumption among women and/or children in SSA were identified by electronically searching PubMed. Papers were selected if they contained "lipid-based nutrient supplements" in the title, abstract or keywords list, and anthropometric, biochemical or child developmental outcomes were measured. The references in identified publications were also searched manually for papers that may not have appeared in the electronic search. Only studies conducted in SSA, in which the

Key messages
• Small-quantity lipid-based nutrient supplements include the family of multiple micronutrient-fortified semi-solid pastes typically given at a dose of about 20 g/d or 110 kcal/day for the purpose of preventing undernutrition.
• Prenatal SQ-LNS consumption reduced the prevalence of low gestational weight gain (GWG) in Ghana. When given alone or as part of a package of interventions, SQ-LNSs consumption has been associated with improved infant or child growth in several settings in Sub-Saharan Africa.
• The impact of SQ-LNS supplementation in promoting fetal and child growth in SSA may be context-specific.
Such interventions might have a greater impact if they address the complex array of underlying factors that contribute to growth faltering.
diets of apparently healthy women and/or children ≥ 6 mo of age were supplemented with SQ-LNSs were reviewed.
This present paper examined the impact of SQ-LNS supplementation on maternal anthropometric and hematologic indices, and on infant and child growth and development, for which considerable data have so far been published. Because the paper was based on a narrative review (rather than a systematic review), it is possible that certain relevant papers published at the time may not have been identified.

| Small-quantity lipid-based nutrient supplements
Small-quantity lipid-based nutrient supplements (SQ-LNSs) include the family of multiple micronutrient (MMN)-fortified semi-solid pastes typically given at a dose of about 20 g/d or 110 kcal/day for the purpose of preventing undernutrition (FANTA, 2016). Typically, they are prepared using vegetable oil, groundnut paste, milk, sugar, and specific amounts of micronutrients depending on the type of product and the target population, and they provide micronutrients as well as good quality protein and essential fatty acids, while leaving enough room for other foods in the diet (Arimond et al., 2015). When mixed with home-prepared foods for women (e.g, during pregnancy and/or lactation) and children, SQ-LNSs increase the nutrient intakes of these vulnerable groups and permit targeting to the individuals consuming them. Because of their low water activity (aw <0.5), SQ-LNSs are not good substrates for the growth of bacteria, yeast and fungi (Arimond et al., 2015), and therefore do not require refrigeration for storage.
SQ-LNSs and multiple micronutrient powders (MNPs) or capsules can be similar in terms of micronutrient contents (De-Regil, Suchdev, Vist, Walleser, & Pena-Rosas, 2013;Suchdev, Pena-Rosas, & De-Regil, 2015), but they differ in the sense that SQ-LNSs come from a food base and often contain energy, protein and essential fatty acids (EFA), as well as being able to include certain minerals (e.g. calcium, magnesium, phosphorus, and potassium) that cannot be incorporated in substantial amounts in MNPs. Other advantages of SQ-LNSs are that they may enhance the taste and acceptability of cereal porridges commonly consumed in developing countries, and they may not adversely affect infant and young child feeding practices . Malawi (Ashorn et al., 2015a;Maleta et al., 2015), and Burkina Faso (Hess et al., 2015); and, more recently, those used in Kenya  Zimbabwe (Humphrey et al., 2019) and South Africa (Smuts et al., 2019  . In the South Africa study (Smuts et al., (Table 1). In most of these studies, the SQ-LNSs were provided post-natally to children at or after 6 mo of age. Among these trials was an earlier trial in Ghana (Adu-Afarwuah et al., 2007), which randomized 6-mo-old infants to consume either to a micronutrient powder, or a crushable micronutrient tablet, or SQ-LNS daily until 12 mo old.
In the Lungwena trial in Malawi (Phuka et al., 2008) infants were randomly assigned to consume either a fortified maize-soy blend (71 g/d), or a medium-quantity (50 g/d) lipid-based nutrient supplement or SQ-LNS (25 g/d) daily from 6 mo to 18 mo of age. The iLiNS-DOSE trial in Malawi (Maleta et al., 2015) included 6-mo-old infants who were randomized into 1 of 6 groups to receive either 10 g/d, 20 g/d, or 40 g/d lipid-based nutrient supplements (LNSs) containing milk powder; 20 g/d or 40 g/d LNS containing no milkpowder; or no supplement until 18 mo of age. In the iLiNS-ZINC trial in Burkina Faso (Hess et al., 2015), infants were randomized into a non-intervention group, or 1 of 4 intervention groups in which they received SQ-LNSs without zinc + placebo tablet, or SQ-LNSs containing 5 mg zinc + placebo tablet, or SQ-LNSs containing 10 mg zinc + placebo tablet, or SQ-LNSs without zinc and T A B L E 1 A summary of the methodology of supplementation trials in Sub-Saharan Africa involving small-quantity lipid-based nutrient supplements (SQ-LNSs) for the prevention of malnutrition 1  (Hess et al., 2015), children in the intervention groups, but not the non-intervention group, received morbidity surveillance and treatment for diarrhea and malaria weekly.
The so-called WASH Benefits cluster-randomized trial in Kenya  involved 8 groups, including an active control group, a passive control group, and 6 intervention groups, namely: water; sanitation; handwashing; combined water, sanitation, and handwashing; nutrition; and combined water, sanitation, handwashing, and nutrition. In the 2 intervention groups that included ii. Impact on maternal hematologic outcomes The hematologic indices reported from the iLiNS-DYAD Ghana trial (Adu-Afarwuah, Lartey, Ashorn, Zeilani et al., 2017b) showed that compared with the women assigned to iron and folic acid (IFA group), those assigned to the SQ-LNSs or multiple micronutrients had significantly lower mean hemoglobin concentration (120 ± 11 vs. 115 ± 12 and 117 ± 12, respectively; P < 0.001), higher mean zinc ii. Impact on post-natal growth All the trials in which SQ-LNS supplementation was provided to children ≥ 6 mo of age in Sub-Saharan Africa evaluated children's growth during the period 11-24 mo of age (Das et al., 2019;FANTA, 2016). In 3 studies in Malawi, including the Lungwena (Phuka et al., 2008), the iLiNS-DOSE (Maleta et al., 2015) and the iLiNS-DYAD (Ashorn et al., 2015a) trials, SQ-LNSs supplementation had no impact on fetal or child growth, except for the Lunguena trial (Phuka et al., 2008), in which children consuming SQ-LNS (as well as those consuming medium-quantity lipid-based nutrient supplement with a dose of 50 g/d) from 6 to 18 mo of age had significantly lower cumulative 12-month incidence of severe stunting (3.5% and 0%, respectively) compared to those who consumed the maize-soy blend (13.3%).
In contrast to the results from Malawi (Ashorn et al., 2015a;Maleta et al., 2015), the provision of SQ-LNSs to infants had a Similarly, in the more recent iLiNS DYAD Ghana trial , children who received SQ-LNSs from 6 to 18 mo of age as did their mothers from pregnancy to 6 mo postpartum had greater attained mean ± SD length (79.7 ± 2.9 cm) and weight (9.9 ± 1.2 kg) and greater mean ± SD of the corresponding LAZ and WAZ values by 18 mo of age than children in the other two groups combined (79.1 ± 2.9 cm and 9.7 ± 1.2 kg, respectively), who received no supplementation and their mothers were assigned to iron and folic acid during pregnancy only, or to multiple micronutrients during pregnancy and the first 6 mo postpartum.
In the iLiNS-ZINC trial in Burkina Faso (Hess et al., 2015), infants who received SQ-LNSs with varied amounts of zinc, along with morbidity surveillance and treatment for malaria and diarrhea from 9 to 18 mo of age had greater mean ± SD length (77.7 ± 3.0 vs. 76.9 ± 3.4 cm; p<0.001) and weight (9.3 ± 1.1 vs 9.0 ± 1.2 kg; p < 0.001) and a lower prevalence of stunting (29.3% vs 39.3%; p < 0.001) than their counterparts who receive no SQ-LNSs, tablets, illness surveillance or treatment.
Results from the WASH Benefits trial in Kenya  showed that compared with children in the active control group who Malawi (Prado, Maleta, Ashorn, Ashorn, Vosti, Sadalaki, & Dewey, 2016) reported no significant group differences in language, socioemotional, or executive function at 18 mo of age. In contrast to these reports (Prado, Adu-Afarwuah et al., 2016, Phuka et al., 2012, Prado, Maleta et al., 2016, the provision of SQ-LNSs to infants from age 9 to 18 mo, along with malaria and diarrhea treatment in Burkina Faso (Prado, Abbeddou, Yakes Jimenez, Some, Ouedraogo, Vosti, Dewey, Brown, Hess, & Ouédraogo, 2016) resulted in greater mean scores language, and personal-social development.
The impact of SQ-LNS consumption on the achievement of motor milestones was evaluated in Ghana (Adu-Afarwuah et al., 2007, Prado, Adu-Afarwuah et al., 2016, Burkina Faso (Prado, Abbeddou et al., 2016) and Malawi (Prado, Maleta et al., 2016). In Ghana, SQ-LNS supplementation (given to infants alone or to both mothers and infants) was associated with a greater percentage of infants being able to walk independently by 12 mo of age in an earlier study (Adu-Afarwuah et al., 2007) when compared with children who received no intervention, as well as in the iLiNS-DYAD trial (Prado, Adu-Afarwuah et al., 2016) when compared with children in the IFA group. Similarly, Burkinabe children who received SQ-LNS supplementation together with morbidity surveillance and treatment for diarrhea and malaria (Prado, Abbeddou et al., 2016) had significantly greater mean motor development scores (indicating higher performance) than children who received no intervention. In Malawi, the supplementation of infants' diet with SQ-LNSs had no impact on locomotor score in the Lungwena study (Phuka et al., 2012), but in the iLiNS-DYAD trial (Prado, Maleta et al., 2016), In the WASH Benefits study in Kenya

| DISCUSSION
This paper examines the results from randomized trials conducted in SSA, in which the diets of apparently healthy women and/or children  and National Research Council, 2007;Ramakrishnan, 2004) showing that a higher energy intake and the consumption of dairy during pregnancy may be associated with a lower risk of inadequate gestational weight gain (as well as a higher risk of excessive gestational gain). Second, it is possible that the 20 mg/d iron dose in the SQ-LNS formulation may have been too low for the population (Mei et al., 2014;Roberfroid et al., 2011), or that, the relatively high dose of zinc (30 mg) in the formulation may have interfered with iron absorption, as suggested by results of a similar supplementation trial in Nepal (Christian et al., 2003).
The positive impact of prenatal SQ-LNSs consumption on mean birthweight of offspring in the Ghanaian sample supports the notion that in that population, the provision of macronutrients (e.g. EFAs) along with multiple micronutrients was necessary to increase fetal growth. As it was shown previously (Michaelsen et al., 2011), the estimated n-3 fatty acid supply in the diet of Ghanaian pregnant women may be low. The stronger impact of SQ-LNSs consumption among primiparous (or younger) women in Ghana suggests that this intervention my offset some of the vulnerabilities associated with primiparity when compared with multiparity (Kramer, 1987). In Ghana, these vulnerabilities appeared to include lower mean hemoglobin concentration, and greater likelihood to be anemic or test positive for malaria at the study enrollment.
Regarding the conflicting results from the Malawi trials (Ashorn et al., 2015a;Ashorn et al., 2015b;Maleta et al., 2015) et al., 2015), Kenya , Zimbabwe (Humphrey et al., 2019) and South Africa (Smuts et al., 2019) with respect to infant and child growth, these findings suggest that the success of SQ-LNS supplementation in promoting fetal and child growth in SSA may be context-specific (Manary, 2015). It appears that the SQ-LNS is more likely to have an impact in settings where constraints on child growth eg. short maternal stature, poor access to health care, high rates of infections, etc. are fewer. In Malawi, children's linear growth may have been restricted by high rates of asymptomatic infections, environmental enteropathy, HIV infection and short maternal stature. (Ashorn et al., 2015b). Comparatively, in a peri-urban setting in Peru (Penny et al., 2005) with a relatively low prevalence of stunting at 18 mo, a nutrition education intervention had a positive impact on stunting, compared to a control group.

| CONCLUSIONS
The provision of SQ-LNSs to pregnant women could be one potential strategy to address inadequate gestational weight gain (GWG) in SSA, without increasing the risk of excessive GWG, or over-

CONFLICTS OF INTEREST
The authors declare that they have no conflicts of interest.

CONTRIBUTIONS
SA-A wrote the manuscript and approved its final version.