Abstract / Summary
Childhood obesity now affects more than 160 million children and adolescents worldwide, and over the same period ultra-processed foods have come to dominate the early-life diet. This review sets out a single argument: that the effect of ultra-processed foods on childhood obesity is better read developmentally, and that the digestive tract is the organ through which much of that effect is transmitted. Exposure begins before birth and continues through the first 1000 days and complementary feeding, the period in which the gut microbiota assembles and the intestinal barrier matures. We show how ultra-processed foods perturb these two systems in particular: a low-fibre, additive-rich, energy-dense diet reshapes microbial community structure and depletes short-chain fatty acid production, while emulsifiers and other additives thin the mucus layer and raise intestinal permeability, admitting microbial products that sustain low-grade inflammation. We trace how these gut-level changes feed forward into altered appetite regulation, adipocyte biology and, ultimately, excess adiposity, and we situate the other proposed mechanisms (matrix disruption, hyperpalatability, accelerated eating rate, endocrine-disrupting additives and central reward signalling) around this digestive core, marking at each step where the evidence is paediatric and where it rests on adult or animal data. The epidemiological literature is appraised by study design, and the constraints on causal inference are stated plainly: NOVA misclassification, residual confounding, reverse causation, dietary misreporting and the scarcity of paediatric trials. The practical implication is that prevention framed around early-life gut health, rather than around cumulative adult-style exposure, is where intervention is most plausibly effective. We are explicit throughout that most of the mechanistic evidence derives from animal or adult studies and that a causal role for ultra-processed foods in childhood adiposity, though biologically plausible, is not yet established in paediatric populations; the developmental reading we advance is the most coherent interpretation of an associational evidence base, not a demonstrated causal chain.