Complementary feeding to nourish and support the developing infant : a thesis presented in partial fulfilment of the requirements of the degree of Doctor of Philosophy in Nutrition Science at Massey University Manawatū, New Zealand

Abstract

During complementary feeding, the gut microbiota undergoes substantial changes in taxonomic composition, functional capacity, and metabolite production. Although diet is recognised as an important modulator of microbiota development, evidence remains limited regarding whether specific complementary foods can influence these trajectories during early infancy. This thesis investigated whether kūmara (New Zealand sweet potato) alone and combined with green banana extract as whole-food prebiotic interventions could influence the developing infant gut microbiota.

Using an integrated experimental approach, controlled in vitro fermentation experiments characterised infant gut microbiota responses to nine substrate mixtures spanning varying combinations of kūmara and green banana extract, which spanned resistant starch contents of 1.2–30.5%. This established intrinsic metabolic capacity under static conditions. The SUN Study, a three-arm randomised controlled trial (n=279), then evaluated kūmara alone, kūmara plus green banana extract (85:15 ratio), and standard complementary feeding from 6 to 10 months of age. Shotgun metagenomics characterised taxonomic composition and functional metabolic pathways, while targeted metabolomics quantified faecal organic acids. Dietary intake assessment confirmed resistant starch delivery and quantified total dietary exposure.

Three principal findings emerged. First, taxonomic composition remained stable across interventions in both in vitro and in vivo settings, despite clear differences in substrate exposure. Second, functional metabolic capacity was responsive to dietary substrates but was dominated by age-driven developmental trajectories in vivo. Functional pathway abundance exhibited a biphasic maturation pattern, with 55% of pathways increasing between 6–8 months and 67% decreasing between 8–10 months, reflecting expansion followed by consolidation characteristic of gut maturation. Intervention effects on pathways were detectable but modest relative to developmental change. Third, faecal organic acid profiles matured predictably with age, characterised by increasing dominance of acetate, propionate, and butyrate (rising from 59% to 92% of total organic acids), but did not differ between intervention groups. Integration of metagenomic and metabolomic data revealed that functional pathway abundance explained three-fold more variation in organic acid profiles than taxonomic composition (12.1% versus 4%), with 94% of taxa-metabolite associations mediated through metabolic pathways.

These findings demonstrate that gut microbiota development during early complementary feeding follows robust, developmentally programmed trajectories that are relatively resistant to dietary variation when introduced alongside continued breastfeeding and other complementary foods. Functional pathway profiling provides substantially greater insight into metabolic outputs than taxonomic composition alone, with important implications for microbiome research design. These findings provide a foundation for designing targeted nutritional strategies that align with developmental timing and dietary context to support healthy microbiota maturation during infancy.