Livestock ownership, animal source foods and child nutritional outcomes in seven rural village clusters in Sub-Saharan Africa

This study is one of few to provide a comparative, descriptive analysis of the relationships between livestock ownership, ASF consumption and child nutritional outcomes in Sub-Saharan Africa. We found some evidence that owning livestock was associated with increased household consumption of ASF in some study sites; however, owning livestock and/or consuming ASF was not consistently associated with improved child growth in any of the village clusters. Moreover, the results paint a complex picture with considerable variation across sites, highlighting the need to consider local circumstances in any programme promoting livestock development and/or ASF consumption as a means to alleviate childhood malnutrition.

Farming systems throughout Sub-Saharan Africa are highly variable and dependent on factors such as geography and climate, as well as socio-cultural influences. Across the 15 major farming systems in Sub-Saharan Africa, the contribution of livestock production to rural livelihoods ranges from subsistence or semi-subsistence through to being the primary source of income [23]. Livestock ownership patterns differed considerably across the village clusters, with the proportion of households owning any livestock ranging from less than half in Bonsaaso to nearly all households in Tiby. At least a quarter of households across all seven sites owned poultry, confirming the central role of this livestock species in rural African households [27]. Cattle were important sources of livestock in all but two village clusters: in humid, tropical Bonsaaso, where the presence of tse-tse flies limits the viability of cattle [23], and in Mwandama, where small farm sizes might preclude grazing [28]. Pigs represent an increasingly important sector in Uganda, with 17% of all households keeping pigs at the recent census [29]. In contrast, the lack of pig production in Nigeria, Mali and Senegal can be explained by the fact that these are predominantly Muslim countries [30].

Food consumption patterns were highly variable across the village clusters and likely influenced by many factors beyond whether or not a household owned a particular livestock species (e.g. cultural preferences and different strategies for accessing ASF). A large number of households consumed alternative ASF in addition to those detailed here. For example, 24 and 27% of households in Pampaida and Potou reported eating camel and tripe in the previous 30 days, respectively. These outcomes were not included in our analysis because they did not represent a significant livestock group or ASF across all sites. Food derived from other animal groups undeniably contributes to total food intake, including fish and fish-related products [31, 32], bush meat [33] and insects [34]. We note that nutrition surveys have tended to focus on skeletal meat and food preferences of western cultures. In the future, efforts to truly understand how ASF contribute to food and nutrition security need to consider the whole animal, in particular offal, and other animal groups such as fish and insects.

Although previous studies have demonstrated a positive association between livestock ownership and ASF consumption [21, 35], this is the first study to investigate these relationships across several Sub-Saharan countries. Overall, there were no consistent relationships between livestock ownership and ASF consumption in the preceding 30 days. Our results show that in some village clusters, ASF were significantly more likely to be consumed by households that owned livestock from which the commodity is derived. For instance, we found that cattle, poultry and pig ownership were significantly associated with consumption of milk/milk products, poultry meat and pork, respectively, in Ruhiira, a finding which is consistent with previous studies in Uganda [21]. Similar positive associations were also observed in this study for Mayange, Rwanda. Notably, ASF consumption was amongst the lowest in these village clusters compared to other Millennium Villages. In contrast, poultry, small ruminant and cattle ownership in Tiby were amongst the highest across the 7 village clusters, and the consumption of ASF did not vary significantly between livestock owners and non-owners. This may suggest that there is a threshold effect whereby once a level of livestock ownership is reached in a village, ASF is more readily accessible to all households, including those that do not own a particular livestock species. In other instances, we found consumption of ASF occurred in the absence of a particular livestock species. For example, beef and milk/milk products were consumed with some frequency in Bonsaaso and Mwandama, although no households in these village clusters reported owning cattle. These patterns highlight the complexities of household decision-making regarding food production, trade and consumption [36].

Few studies have investigated the relationship between child nutritional status and livestock ownership and/or ASF consumption in observational studies. Jin and Iannotti [37] found a modest positive association between co-owned/female-owned livestock and WAZ in Kenya, while Mosites et al. [38] found a slight beneficial effect of livestock ownership on stunting prevalence using demographic and health survey data from Ethiopia, Kenya and Uganda. The latter study found no association between ASF consumption (based on 24-h recall) and stunting rates in any country. In this study, we disaggregated ASF into different groups (e.g. poultry meat, pork and eggs), which allowed more detailed analysis of the potential impact of different ASF on child growth. We also captured ASF consumption in the previous 30 days, rather than 24-h recall which does not adequately capture the diversity of a usual diet [39]. Nonetheless, our findings were similar to Mosites et al. [38], in that there was no consistent association between ASF consumption and improved nutritional status in children in any of the village clusters.

There are several limitations of this study. Data analysed in this study are from countries at the very beginning of the nutrition transition [40], with food insecurity and malnutrition still widespread, especially in children [41]. Other research suggests that household diets in Sub-Saharan countries are exposed to large seasonal variety and insecurity around food [42]. It is important to recognise that our study does not account for seasonal variations or extreme weather events (e.g. droughts). Further, it is not representative of national trends as the sample populations were individual villages from the lowest socio-economic bracket and are not representative of rural areas in those countries. Long-term research needs to capture fluctuations in nutritional outcomes, the shifts in livestock ownership and consumption patterns during this transition.

Secondly, interviews were conducted with the household head and we therefore cannot comment on within-household differences in livestock ownership and ASF consumption patterns over the past 30 days. Although 30-day consumption recall has its limitations [43] and is likely to result in bias, we suspect that recall is unlikely to be different between groups (e.g. households that do and do not own livestock), and therefore, any findings would be biased towards the null. Frequency and amount of ASF consumption could plausibly differ, for example, between men, women and children in the household [44, 45]. Given potential intra-household variability of food allocation, particularly between males and females [46], future studies should measure consumption at the individual child level. The importance of women in health and nutritional outcomes of their children is well documented [47–49]. For example, Kariuki et al. [48] demonstrated a positive relationship between female-owned cattle and increased consumption of ASF in Tanzania, Kenya and Mozambique, while Jin and Iannotti [37] found that child HAZ and WAZ scores were higher in households with co-owned/female-owned livestock in Kenya. Future studies are required to investigate how gender roles influence the allocation of food within the household and the impacts on child nutritional outcomes.

Finally, given the high degree of variability across countries and the limited significant findings in the univariate analysis, we elected not to perform further multivariate analysis. Any future study aiming to establish an independent effect of livestock ownership on household food and nutrition security will need to control for factors such as household size, wealth and gender of household head. Additionally, data on breastfeeding and complimentary feeding practices, as well as nutritional status of pregnant or lactating mothers, should be collected and considered in the analysis, given their important influence on child development [50]. Notably, time and labour inputs associated with livestock rearing activities might impact these child feeding practices; more empirical evidence is required to investigate these trade-offs. Access to safe food is also critical for ensuring positive nutritional outcomes in children [6]. Concerns have been raised about possible negative impacts of livestock ownership on child health and development, by increasing infection risks and potentially affecting enteric functions of the gut microbiome [51]. Exposure of infants to livestock faeces has been shown to have a significant impact on bacterial exposure leading to diarrhoea [52, 53]. There is some evidence that housing of poultry in close proximity to the household can have a negative effect on child growth [54]. Given constraints imposed by the data, we were unable to assess the potential effect of livestock housing, zoonoses and child growth outcomes in the current study.

Food and nutritional security in Sub-Saharan Africa is a complex issue with multiple entry points for creating sustainable improvements. It is plausible that livestock production and ASF can play a role in alleviating hunger, although robust evidence for this impact pathway is still widely lacking. Any intervention will need to be multifaceted, multidisciplinary and tailored to the specific context of the country to ensure a sustained impact is realised. A One Health approach that combines assessment of nutritional outcomes, diet, agricultural and social practices is required. In particular, further detailed research is needed to (1) evaluate the quantity and utilisation of ASF in relation to number of livestock owned by a household; (2) assess the time dimension of food security using longitudinal studies to account for seasonal patterns and changes over time; (3) understand how gender roles influence intra-household variation of food consumption and how livestock ownership may impact this, with emphasis on the most vulnerable (i.e. women and children); and (4) investigate how the potential negative impact of zoonotic/food-borne diseases in households that own livestock or consume ASF can be mitigated to ensure improved nutritional outcomes through livestock-based interventions.