Interpreting ultraprocessed food trial evidence: evidentiary limits, reporting practices, and policy relevance

Ultraprocessed foods (UPFs) have become a central topic in nutrition science, with extensive observational research linking higher intake to adverse health outcomes. More recent randomized controlled trials have been presented as strengthening causal claims, yet closer examination reveals a pattern of interpretation that extends beyond what the data support. This perspective reviews 4 trials that directly compared ultraprocessed diets with minimally processed diets, examining both their findings and how those findings have been represented in scientific commentary. Across all studies, the comparison group limited the ability to isolate processing as the causal factor, as multiple dietary features differed simultaneously. Reporting consistently emphasized findings aligning with expectations of harm while downplaying neutral or contradictory evidence, including favorable clinical markers and higher adherence in some ultraprocessed conditions. Methodological limitations were also underrepresented in secondary interpretations. The current experimental evidence therefore requires cautious interpretation. Specifically, 4 actionable recommendations follow from this analysis: 1) future trials should match intervention arms on nutrient composition so that processing effects can be disaggregated from nutritional quality effects; 2) adherence and dropout rates should be reported as primary, not secondary, outcomes; 3) terms such as "excess consumption" and "overeating" should be reserved for conditions of positive energy balance and weight gain; and 4) the heterogeneity within NOVA Group 4 warrants subcategory analyses rather than class-wide causal claims. Without these changes, the evidentiary basis for UPF policy risks overstating certainty and misdirecting intervention targets.