Impacts of Agriculture on Nutrition: Nature of the Evidence and Research Gaps

The belief that agriculture contributes to human nutrition and health underpins ongoing efforts to make agricultural policies and programs nutrition-sensitive. However, the search for solid empirical findings of what works in this arena has been stepped up as donors and national governments increasingly call for evidence-informed policymaking. Nine of the 10 reviews considered here set out to answer fundamental questions framed along the lines of "do agricultural interventions improve nutrition?" The types of interventions considered vary considerably across the studies, including home gardens, small animal husbandry, aquaculture, and behavior change communication. Other differences across the review papers include the units of observation used, the metrics of impact, and the threshold of evidence imposed on the studies that they examined. Ruel (2001) reviewed interventions from the 1990s that promoted the production of micronutrient-rich foods through home gardening, small animal husbandry, and aquaculture, as well as behavior change communication efforts aimed at promoting changes in dietary patterns. The author noted that only a few of the home garden and nutrition education studies actually measured the impact of their activities on intended outcomes, and of those that did, few could demonstrate any positive statistical significance. The conclusion was that "although the question of whether home gardens have a positive impact on vitamin A status has been examined in a number of reviews, including some recent studies, evidence is still scant." Berti et al. (2004) broadened the scope of agricultural interventions considered in their systematic review, which pushed back to the 1980s. They only included studies that did measure a range of nutritional outcomes, but this allowed for consideration of irrigation and cash cropping schemes, large ruminant husbandry, and mixed garden-small ruminant interventions. They found "mixed results in terms of improving nutritional status in participating households." That is, some interventions reported significant improvements, while others found no impact or even negative impacts. LeRoy and Frongillo (2007) took a different direction, choosing to narrow down the focus to the role of animal-sourced protein in improving nutrition as generated by interventions promoting animal production. This systematic review generated 14 studies across a range of husbandry, small ruminant, aquaculture, and/or behavior change communication activities. Most of the studies reported positive impacts on production associated with an intervention, but only 4 evaluated nutrition outcomes directly. Those 4 reported improvements in various nutritional parameters, but did not document statistical significance attributable to the intervention or in relation to counterfactuals. The World Bank (2007) compilation of lessons learned included assessment of 52 studies that considered agricultural impacts on food expenditure, caloric intake, and anthropometry. The overriding conclusion was that interventions aimed at increasing production and productivity of staple foods showed impacts on child nutritional status that were "limited and mixed." Similarly, programs focused on promoting animal source foods showed mixed results, while home garden activities failed to achieve significant impacts on nutritional outcomes. The Lancet series on maternal and child nutrition of 2008 (Bhutta et al. 2008) also included a review of interventions designed to improve nutrition under the rubric of dietary diversification strategies. Although most of the 29 studies considered demonstrated various benefits to producer households in terms of increased food production and consumption, especially when combined with a nutrition education component, significant impacts on nutrition were weak. As a result, the Lancet review concluded that "although some promising multidisciplinary nutrition interventions have been implemented, dietary diversification strategies have not been proven to affect nutritional status or micronutrient indicators on a large scale."