Katie Ayling

This thesis aimed to understand the diet and nutritional needs of populations in Côte d’Ivoire and Nigeria, reformulate and fortify food products, and evaluate their effectiveness on nutritional and health outcomes in a real-world setting through a school feeding programme. These aims were addressed through three research themes: (1) understanding population nutritional needs, by identifying nutrient inadequacies and dietary patterns among school-aged children and women of reproductive age in Côte d’Ivoire; (2) Tailoring product (re)formulation via food fortification and assessing its technological and sensory feasibility; and (3) evaluating the effectiveness of a multi-nutrient fortified food product delivered through a school feeding programme in Nigeria. Together, these studies operationalised a practical framework, using the research themes as programme pillars to align public-health nutrition goals with food-industry innovation. The integration of these pillars provides a model for how industry-led nutrition interventions can support national public health priorities within evolving food systems.

Theme 1 Understanding population nutritional needs
Chapter 2 showed that most urban Ivorian women (>50%) were overweight or obese, whereas school-aged children mostly had a healthy weight. Intake of most nutrients was adequate for both school-aged children and women of reproductive age, with notable exceptions for calcium (99% and 96% inadequacy, respectively), riboflavin (53 and 79% inadequacy) and zinc (51 and 53% inadequacy). Older children (11-12 years) experienced higher prevalence of inadequacies than younger children (6-10 years), and for women of reproductive age, the youngest age group (15-17 years) experienced higher prevalence of inadequacies than older age groups (18-49 years). No differences were found between sexes among children. Notably iron was adequate for approximately >85% of children and women, except for children 11-12 years old where slightly higher inadequacies were found (~30%). Most nutrient inadequacies showed a negative linear relationship with socio-economic status for school-aged children, where the highest inadequacies were seen in the lowest socio-economic status groups. Most school-aged children (69%) and women of reproductive age (56%) met their energy requirements. Protein intake was often below the Acceptable Macronutrient Distribution Range (AMDR) (53% children and 45% women). Ivorian nutrition policies and interventions should generally focus on improving calcium, riboflavin and zinc intake, in addition to vitamin A and folate intake specifically in older children and young women.

Chapter 3 evaluated differences in estimated nutrient inadequacy between a modelling approach (nuriR) that was used previously to identify nutrient inadequacies globally, and traditional analytical methods (Estimated Average Requirement (EAR) cut point and probability of adequacy methods). When usual intake data were used, differences between methods were minimal (<1%). In contrast, the use of one-day dietary intake data substantially inflated inadequacy estimates. Similarly, failure to apply age-specific nutrient requirements led to larger discrepancies than the choice of analytical method. By modelling usual intake distributions using log-normal or gamma functions, nutriR produced inadequacy estimates comparable to traditional methods when appropriate intake data were available. These findings support the application of nutriR in global modelling efforts, including those used to characterise nutrient gaps across ages and sexes. However, they also highlight its limitations: nutriR does not reduce the need for high-quality usual dietary intake data, and its accuracy remains sensitive to appropriate age categorisation. Chapter 4, characterised four dietary patterns among school-aged children and women of reproductive age. Two patterns: “Fish, oil and vegetables” and “Dairy and sweet” were highly similar and shared across groups, while a third, labelled “Poultry and citrus fruit” was broadly comparable but in women additionally included red meat and white tubers. In contrast, children additionally displayed a “Refined, red meat and vegetable” pattern whereas women showed a distinct “Low energy intake” pattern, which was least similar to each other. The fish- and dairy-based patterns, more common in Bouaké and Daloa, were associated with lower or moderate socio-demographic indicators but higher diet quality and nutrient adequacy, as well as a lower non-communicable disease risk. In contrast, the poultry pattern among children and low-intake pattern among women were characterised by poorer diet quality. Nutritional status, including overweight and obesity (49% in women and 7% in children), was not clearly explained by dietary patterns. City of residence was the strongest predictor of dietary pattern in both groups. Overall these results suggest that urban Ivorian children and women share dietary patterns consistent with early stages of the nutrition transition. Promoting familiar fish-, vegetable- and dairy based diets, alongside increased dietary diversity, may improve nutrient adequacy although additional work is required to explore causes and to design interventions for the high prevalence of overweight and obesity in women. Theme 2 Commercial and technical feasibility through (re)formulation of two commercial products The bench-scale trials presented in the Appendix assessed the commercial and technical feasibility of improving iron bioavailability by modifying the iron compound used in fortified dairy products. Encapsulated ferric pyrophosphate was the only iron compound that had similar technological and sensory attributes to the reference; however its cost was five times higher, rendering it commercially unviable. As a result, reformulation of the flavoured dairy product available in Côte d’Ivoire was not feasible, and this product was excluded from further evaluation in Theme 3. The bench-scale trials indicated that ferrous fumarate represents a viable alternative to ferric pyrophosphate in non-flavoured dairy applications or cereal-containing formats, given its relatively similar cost price and acceptable colour and taste in short-term testing (4 weeks accelerated). However longer bench-scale trials (>4 months under ambient conditions) are needed to confirm these findings. A commercially available milk-based maize porridge containing ferrous fumarate was instead selected as an appropriate multi-micronutrient fortified product developed for Nigerian families and suitable for inclusion in the National School Feeding Programme.

Theme 3 Effectiveness of a tailored food product on nutritional need
Chapter 5 described the study protocol and statistical analysis plan for a randomised controlled trial, while Chapter 6 presented the trial results. The study evaluated the effectiveness of a multi-nutrient fortified milk- based maize porridge in a school feeding programme in Ibadan, Nigeria. At endpoint, the high-dose group did not differ from control in haemoglobin (−0.5%, 95% CI −3.4 to 2.5) and serum ferritin (1.9%, 95% CI -6 to 10.5) in intention-to-treat analysis (n=945). Secondary outcome vitamin D concentrations were higher in the high dose group compared to control at endpoint (3.8%, 95% CI 0.6 to 7). Vitamin A status improved by approximately 8% (4.6%, 11.5%) across all groups. Overall, the intervention did not improve iron status or anaemia but produced modest improvements in vitamin D status among Nigerian school-aged children.

The limited impact on iron outcomes of the intervention was most likely attributable to the low prevalence of iron-responsive iron deficiency anaemia in the study population. Although anaemia prevalence was high at baseline (74%), inflammation-adjusted iron deficiency affected only 9.5% of participants. The iron deficiency observed was predominately functional, with body iron stores adequate in 99.8% of participants. This profile was likely driven by a high inflammation burden (elevated AGP in 34% and CRP in 15%), high vitamin A deficiency (30%) and malaria infection (12%). In this context, iron was present, but sequestered through inflammation-mediated hepcidin regulation, limiting its availability for erythropoiesis. These findings highlight the importance of understanding the aetiology of anaemia, including the prevalence of iron-responsive deficiency, before implementing iron-based interventions.

Summary of recommendations
This thesis applied a structured framework to guide alignment between commercial food fortification and public health nutrition goals for school-aged children and women of reproductive ages in Côte d’Ivoire and Nigeria. By integrating population-level dietary data, applied product development, and clinical evaluation, the findings demonstrate that:
• Diagnostic-driven fortification is critical, as nutrient inadequacies alone do not reliably reflect population nutritional needs or predict intervention effectiveness.
• Substantial gaps remain in understanding population-specific causes of anaemia, with a large proportion unexplained in this study.
• Further research is needed on the causes and modifiers of inflammation (beyond malaria and parasites) and their role in anaemia, particularly in LMIC contexts.
• Evidence generation and application in resource-limited settings require innovative approaches to dietary assessment, fortificant bioavailability, and evaluation of market-driven fortification strategies.
• Structured coordination between public health and industry needs is needed to enable market-driven fortification to deliver meaningful public health impact under real-world constraints.

The results of this thesis underscore the importance of diagnostic-driven fortification, implementation of feasible solutions in resource limited environments, the need to evaluate programme effectiveness, and collaborative models between public health and industry to improve population nutrition outcomes.