{ "$schema_version": "5.4", "review_mode": "fresh", "claim_restatement": { "index_wording": "Ultra-processed food causes excess energy intake and weight gain through properties of processing itself, independent of nutrient composition", "entry_form": "conjunction", "gradeable_restatement": "In adults given ad libitum access to a defined diet, assigning foods classified as ultra-processed (NOVA 4) causes greater energy intake and subsequent weight gain (body weight and fat mass) than a minimally processed comparator, and this effect persists when nutrient composition as consumed (macronutrient and micronutrient amounts, their combinations, and the compositional determinants of energy density) is matched, being carried instead by properties that industrial processing confers on foods (such as texture and eating rate, food-matrix disruption, and additives) that are characteristic of the category.", "restatement_required": true, "split_into": [ "SC1: whole-diet effect on energy intake", "SC2: downstream relative weight or fat gain", "SC3: persistence under selected, as-offered aggregate nutrient matching", "SC4: persistence when nutrient composition as consumed (including compositional determinants of energy density and hyperpalatable nutrient combinations) is matched", "SC5: processing-conferred properties that raise intake are characteristic of the NOVA 4 category" ], "parent_claim": null, "scope_qualifiers": [ { "qualifier": "adults given ad libitum access to a defined diet", "determinable_before_outcome": true, "determined_from": "eligibility, setting, menu assignment, and feeding protocol fixed before outcome measurement", "action": "replaced with measurable antecedent" }, { "qualifier": "independent of nutrient composition", "determinable_before_outcome": true, "determined_from": "prespecified matching and measurement of nutrient composition as consumed: macronutrient and micronutrient amounts, nutrient combinations (including hyperpalatable combinations), and the compositional determinants of energy density", "action": "replaced with measurable antecedent" }, { "qualifier": "properties of processing itself", "determinable_before_outcome": true, "determined_from": "the processing-conferred property proposed as the carrier (for example texture/eating rate, matrix disruption, or an additive class) must be named before outcomes rather than inferred from whichever menu increases intake; such properties are treated as the hypothesis's candidate mediators, not as confounders to be removed", "action": "replaced with measurable antecedent" } ], "redteam_note": "Red-team synthesis (2026-09-30): the original restatement required the effect to survive control of texture, eating rate, matrix, and palatability, which proponents treat as properties of processing [28]; that graded a residual essence no proponent defends. Independence is now required from nutrient composition only (RT-02). The level is Γ under either reading." }, "lay_summary": { "paragraphs": [ "The claim is that industrial processing has an effect of its own: ultra-processed foods lead people to eat more and gain weight because of what processing does to food, such as softer textures that are eaten faster or broken-down food structure, and not simply because of the nutrients the food contains.", "This review assigns Γ — Mechanistically plausible, partial system support, with Moderate confidence. Controlled feeding studies show that some ultra-processed diets make people eat more and gain weight or fat relative to a minimally processed diet over one to eight weeks. That narrower result is causal in the settings tested. The stronger claim, that the effect does not depend on nutrient composition, has not been shown. In a recent inpatient trial whose results are posted in a trial registry but not yet published in a journal, an ultra-processed diet did not clearly increase intake over an unprocessed one when energy density was kept low and foods that combine fat, sugar, or salt in especially appealing ways were limited. Changing the nutrient make-up of ultra-processed diets also changes how much people eat.", "Food texture and eating speed do change how much people eat, but they do so in both processed and minimally processed foods, and some ultra-processed foods are eaten slowly and are filling. The evidence supports particular food properties as contributors: energy density, nutrient combinations, and texture, which processing often but not always creates. It has not identified a property that all ultra-processed foods share and that works regardless of what they contain. The trials are small and short, which is why confidence is moderate rather than high." ], "grade_restated_faithfully": true, "claims_referenced": [ "C1", "C3", "C4", "C19", "C22", "C5", "C6", "C7", "C24", "C25", "C18" ], "jargon_glossed": [ "ad libitum: participants may eat as much or as little as they choose", "NOVA: a classification that groups foods partly by industrial formulation and processing", "mediator: a step through which an exposure produces an outcome", "energy density: calories per gram of food", "hyperpalatable: combining fat, sugar, or salt in especially appealing proportions" ], "no_practical_advice": true, "written_for": "general reader, no specialist nutrition background" }, "hypothesis": { "name": "nutrient-independent ultra-processed-food processing mechanism for intake and weight gain", "central_claim": "Ultra-processed food causes excess energy intake and weight gain through processing-specific properties that act independently of nutrient composition.", "proposed_mechanism": "Industrial processing confers properties on foods (for example softer texture and faster eating rate, food-matrix disruption, or additives) that increase voluntary intake, with weight or fat gain downstream, and the effect persists when nutrient composition as consumed is matched.", "claimed_scope": "broad human food-category causality", "claimed_level": [ "initiation", "progression", "primacy" ], "is_conjunctive": true, "conjunctive_note": "The claim joins a whole-diet intake effect, downstream weight change, independence from nutrient composition, attribution to processing-conferred properties, and category-wide scope; no single study isolates every clause." }, "rubric_routing": { "primary_rubric": "06-metabolism-aging.md", "secondary_rubrics": [], "index_notes_applied": [ "No ultra-processed-food-specific entry was present in the supplied hypothesis index; decomposition and weakest-link aggregation were preregistered from the wording.", "Red-team pass: a stub index entry is recommended in redteam/SYNTHESIS_CHANGE_LOG.md (framework files not modified)." ], "rubric_fit_note": "The supplied packet has no nutrition- or obesity-specific rubric. The metabolism/aging rubric is the closest fit but defines its Δ boundary in lifespan/healthspan terms. This review therefore quotes that rule verbatim, records the mismatch, and treats randomized human changes in measured energy intake and body weight or fat as the functional organism-level analogue. This adaptation cannot promote a processing-specific subclaim when its causal exposure remains unidentified. Red-team routing check (2026-09-30): all seven rubrics were checked; 04 covers microbial effects on obesity and 07 covers psychiatric phenotypes, and neither governs an intake/weight claim; no rubric with a lower binding ceiling applies. Rubric 06's hard rule 'Single-species or single-strain evidence' does not bind because the claim is human-scoped and the evidence is human, so no cross-species inference occurs (cap-precedence rule; recorded in cap_precedence_resolutions)." }, "subclaim_matrix": { "subclaims": [ { "id": "SC1", "statement": "Assigning an ultra-processed-rich whole diet rather than a minimally or unprocessed comparator can increase ad libitum energy intake in humans.", "subclaim_type": "sufficiency", "governing_rubric": "06-metabolism-aging.md", "claimed_scope": "short-term controlled human whole-diet exposure", "scope": "Adults exposed for one to eight weeks to defined whole-diet conditions, with intake measured during ad libitum access and the exact menus retained as the causal package.", "key_evidence_anchor": "Randomized controlled whole-diet feeding with measured ad libitum intake.", "hsef_level": "Ε", "confidence": "Moderate", "governing_cap_or_threshold": "The Ε threshold is crossed because causal effects occur in constrained human menus and settings at three independent sites, and the boundary is mapped: the effect vanishes when energy density and hyperpalatability are low [18] and becomes a relative difference within net weight loss under guideline menus [3].", "rationale": "Hall, Hamano, UPDATE, and the second NIH inpatient trial (registry-posted) show intake increases on UPF-rich diets [1-3,18]. Small samples (9-50), short periods (1-8 weeks), one open-label n=9 trial, self-reported intake in UPDATE, and within-group reproduction limit confidence to Moderate." }, { "id": "SC2", "statement": "The higher intake caused by some ultra-processed-rich whole-diet assignments produces relative short-term body-weight or fat-mass gain.", "subclaim_type": "mechanism", "governing_rubric": "06-metabolism-aging.md", "claimed_scope": "short-term controlled human body-weight and fat-mass response", "scope": "Relative change between randomized diet conditions over one to eight weeks; absolute gain and durable obesity are reported separately.", "key_evidence_anchor": "Randomized controlled feeding with measured intake and repeated body composition or weight.", "hsef_level": "Ε", "confidence": "Moderate", "governing_cap_or_threshold": "The Ε threshold is met for bounded relative weight/fat outcomes; durability and universal absolute gain are not established.", "rationale": "Hall and Hamano produced relative gain under ad libitum exposure; UPDATE produced greater loss on the minimally processed diet while both groups lost weight; a prescribed-intake trial reports UPF-associated weight gain independent of caloric load (abstract-level) [19]. Effects are short-term and small, and a 12-month advice trial achieved only a small, non-clinically-significant weight difference [25]." }, { "id": "SC3", "statement": "The whole-diet intake effect can persist when selected aggregate nutrients and offered overall energy density are matched between menus.", "subclaim_type": "sufficiency", "governing_rubric": "06-metabolism-aging.md", "claimed_scope": "the exact matching variables and menus used in controlled feeding trials", "scope": "Matching of offered total calories, macronutrient percentages, sugar, sodium, fiber, and nominal overall energy density in Hall 2019; no inference to all nutrient forms, food matrix, non-beverage density, or actual consumed composition.", "key_evidence_anchor": "Randomized feeding in which prespecified aggregate nutrients are matched and residual food properties are measured.", "hsef_level": "Ε", "confidence": "Moderate", "governing_cap_or_threshold": "The Ε threshold is crossed only for selected matched variables; the broader phrase 'independent of nutrient composition' is not crossed.", "rationale": "The Hall effect survived matching on several offered aggregate nutrients, which rules out those variables as complete explanations for that menu contrast. Meaningful residual differences in nutrient form, non-beverage energy density, matrix, and eating rate prevent a composition-wide independence claim. Robinson 2026 [15] likewise concludes that trial diets were not matched for nutrient profile." }, { "id": "SC4", "statement": "The intake and weight effect of UPF-rich diets persists when nutrient composition as consumed is matched, including the compositional determinants of energy density and hyperpalatable nutrient combinations.", "subclaim_type": "mechanism", "governing_rubric": "06-metabolism-aging.md", "claimed_scope": "human UPF versus minimally processed diets at matched composition", "scope": "Whole-diet or meal contrasts in which composition as consumed (nutrient amounts, combinations, energy density) is matched and processing category differs; processing-conferred texture, matrix, and additives may differ because they are the proposed carriers.", "key_evidence_anchor": "Adequately powered, composition-matched ad libitum feeding trial comparing UPF with minimally processed diets, with a prespecified processing route and smallest effect of interest.", "hsef_level": "Γ", "confidence": "Moderate", "governing_cap_or_threshold": "The Δ boundary is not crossed: no positive field-valid perturbation shows the UPF effect persisting at matched composition; the most direct contrast is non-significant [18] and matched designs are null [8,11,22].", "rationale": "NCT05290064: UPF-LL vs UNF-LL +128.4 kcal/d (95% CI -8.2 to 265.0; P=0.065; registry-posted). Matched-exposure [8], matched-preload [11], and matched single-meal [22] designs are null; changing macronutrient composition within UPF changes intake [21]. Positive contrasts that exist (soft-texture pairwise contrast [4]; prescribed-intake weight [19]) are not composition-matched. Confidence is Moderate because the key nulls are small, short, and imprecise and cannot exclude a modest residual effect." }, { "id": "SC5", "statement": "Processing-conferred properties that raise intake (such as soft texture and fast eating rate) are characteristic of the NOVA 4 category, so category status predicts excess intake and weight gain beyond composition across foods and populations.", "subclaim_type": "generalization", "governing_rubric": "06-metabolism-aging.md", "claimed_scope": "broad human food-category and population generalization", "scope": "Across product classes, dietary guidance patterns, ages, and durations, with energy density, texture, eating rate, nutrient quality, and social context explicitly assessed.", "key_evidence_anchor": "Independent long-duration randomized trials using diverse menus plus within-category negative controls.", "hsef_level": "Γ", "confidence": "Moderate", "governing_cap_or_threshold": "The Δ boundary is not crossed for category-wide prediction: processing-conferred properties act in both directions within NOVA 4, and the category itself is unreliably assigned.", "rationale": "Within NOVA 4, slow/hard or low-energy-density UPF lowers intake and raises satiety [4,6,23,24]; texture effects operate equally in minimally processed foods [4,5]; energy-intake rate varies widely within category [16]; NOVA assignment agreement is low (Fleiss κ 0.32-0.34) [27]. Observational associations [14] cannot separate processing from composition." } ], "aggregation_rule": "weakest-link", "aggregation_preregistered": true, "aggregation_rationale": "The submitted sentence requires the whole-diet effect, downstream weight effect, independence claim, processing-specific mediator, and category-wide generalization to be jointly true; failure to isolate the independent mediator binds the conjunction.", "primary_subclaim_id": "", "aggregation_derivation": "SC1=Ε; SC2=Ε; SC3=Ε; SC4=Γ; SC5=Γ. Weakest-link therefore returns Γ.", "aggregation_alternatives": { "weakest_link_would_return": "Γ", "primary_subclaim_would_return": "Γ", "scope_matched_would_return": "Γ", "note": "Primary-subclaim with the distinctive independence clause (SC4) designated returns Γ; it would return Ε only if SC1 were designated, which would grade the sentence by its least distinctive clause (corrected in red-team pass, RT-13). Scope-matched scores the claim by SC4/SC5, whose scope matches it, and returns Γ. A claim reduced to SC1 would reach Ε." }, "index_rule_deviation": "" }, "competing_models": { "models_evaluated": [ { "rank_in_tournament": 1, "model_id": "M2", "is_nominal": false, "name": "Proximal food-property model (energy density, hyperpalatable nutrient combinations, texture and eating rate)", "core_claim": "Energy density, hyperpalatable nutrient combinations, and texture-derived eating rate determine the rate and amount of energy consumed; industrial processing often creates these properties but can also produce low-risk versions.", "explains_what_nominal_does_not": "It explains large intake differences within NOVA 4 driven by energy density and hyperpalatability [18,23], texture-derived eating-rate effects within UPF [6], the disappearance of the UPF-versus-unprocessed difference when energy density and hyperpalatability are low [18], and meal-level intake patterns across inpatient diets [20].", "strongest_supporting_evidence": [ "NCT05290064 four-arm inpatient trial [18] (registry-posted)", "Lasschuijt 2025 ED x ER factorial [23]", "Forde 2026 within-UPF eating-rate trial [6]", "Fazzino 2023 meal-level analysis [20]", "Lasschuijt 2023 and Teo 2022 texture trials [4,5]" ], "strongest_contradictory_evidence": [ "Texture manipulations in [4] and [6] were not composition-pure (higher fat in soft/fast diets).", "The complete Hall package effect has not been quantitatively partitioned among these properties [1]." ], "hsef_level_for_this_model": "Ε", "rationale": "Human randomized perturbations establish context-specific effects of energy density, hyperpalatability, and texture/eating rate on intake across and within processing categories. Energy density and hyperpalatability are partly compositional, so this model overlaps M4; durable weight mediation is untested." }, { "rank_in_tournament": 2, "model_id": "M3", "is_nominal": false, "name": "Whole-diet ultra-processed package model", "core_claim": "Some UPF-rich menu packages increase intake and relative weight through the combined effect of formulation, food form, energy density, eating behavior, convenience, and choice, without requiring one residual processing essence.", "explains_what_nominal_does_not": "It preserves the randomized whole-diet effects while allowing heterogeneity among ultra-processed foods and trials.", "strongest_supporting_evidence": [ "Hall inpatient crossover [1]", "Hamano crossover [2]", "UPDATE crossover [3]", "NCT05290064 UPF-HH vs UNF [18]", "Preston prescribed-intake trial [19]" ], "strongest_contradictory_evidence": [ "Null or non-significant results when energy density, hyperpalatability, or composition are controlled [8,11,18,22]." ], "hsef_level_for_this_model": "Ε", "rationale": "Several human randomized interventions establish a causal package effect in bounded menus. The model remains deliberately composite and does not assert a universal category mechanism." }, { "rank_in_tournament": 3, "model_id": "M4", "is_nominal": false, "name": "Nutrient-form, matrix, formulation, and protein-leverage model", "core_claim": "Differences in nutrient form, fibre type, carbohydrate-fat combinations, protein share, food matrix, and formulation account for part of the apparent UPF effect even when aggregate macronutrients are nominally matched as offered.", "explains_what_nominal_does_not": "It explains why aggregate matching does not eliminate biologically meaningful compositional differences and why post hoc meal composition predicts intake.", "strongest_supporting_evidence": [ "Residual Hall diet differences and the authors' protein-leverage estimate (at most ~half) [1]", "Brunstrom post hoc analysis [13]", "Hägele within-UPF protein manipulation [21]" ], "strongest_contradictory_evidence": [ "No single compositional feature has quantitatively mediated the replicated whole-diet effect." ], "hsef_level_for_this_model": "Γ", "rationale": "Human partial-system evidence is plausible and specific, but no one compositional account has been isolated across the main whole-diet trials." }, { "rank_in_tournament": 4, "model_id": "M1", "is_nominal": true, "name": "Nutrient-independent processing model (nominal)", "core_claim": "Properties conferred by industrial processing and characteristic of ultra-processed foods cause excess intake and weight gain when nutrient composition as consumed is matched.", "explains_what_nominal_does_not": "As framed, it would explain a residual category effect after exhaustive matching if such an effect were demonstrated.", "strongest_supporting_evidence": [ "Hall effect under selected as-offered nutrient matching [1]", "prospective association after nutritional-quality adjustment [14]", "soft (ultra-)processed > soft unprocessed at equal energy density, composition unmatched [4]", "prescribed-intake weight gain independent of caloric load (abstract-level) [19]", "positive but non-significant composition-matched contrast [18]" ], "strongest_contradictory_evidence": [ "UPF-LL vs UNF-LL non-significant [18]", "matched exposure, preload, and single-meal nulls [8,11,22]", "composition changes intake within UPF [21]", "bidirectional texture effects within NOVA 4 [4,6,24]", "no detectable post-ingestive striatal dopamine response [26]" ], "hsef_level_for_this_model": "Γ", "rationale": "Processing-conferred properties are real candidate carriers, but no positive composition-matched perturbation exists, and the carriers that do act (texture, energy density) operate in both directions within NOVA 4 and equally in minimally processed foods." } ], "tournament_summary": "The proximal food-property model ranks first because randomized manipulations of energy density, hyperpalatability, and texture/eating rate produce large intake effects within and across NOVA categories, and the UPF-versus-unprocessed difference is small and non-significant when energy density and hyperpalatability are low [18]. The whole-diet package model ranks second, the nutrient-form/protein-leverage model third, and the nominal model fourth because no composition-matched positive perturbation exists.", "study_separating_rank1_from_rank2": "NCT05290064 [18] (registry-posted, not peer reviewed) most directly separates the models: with processing category fixed, energy density and hyperpalatability moved intake by 661.6 and 158.4 kcal/d, and with both low the UPF minus unprocessed difference was 128.4 kcal/d (P=0.065). Lasschuijt 2023 [4] and Forde 2026 [6] are complementary, each with composition differences between arms.", "nominal_wins_only_by_funding_artifact": false, "level_unchanged_by_tournament": true }, "evidence_inventory": [ { "citation": "Hall et al. 2019, Cell Metabolism", "source_pointer": "PMID 31105044; DOI 10.1016/j.cmet.2019.05.008", "relates_to_subclaims": [ "SC1", "SC2", "SC3", "SC4" ], "favors_models": [ "M2", "M3" ], "evidence_class": "Strong", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "A UPF-rich assigned diet package increased intake and produced relative weight and fat gain while several aggregate offered nutrients were matched.", "what_it_does_not_show": "It does not isolate a processing-specific residual from non-beverage energy density, eating rate, nutrient form, matrix, or menu selection.", "confounders_present": [ "two-week periods", "20 participants", "unblinded menu exposure", "residual food-property differences" ], "verification": "verified", "verification_note": "Complete corrected PMC full text read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "A" }, { "citation": "Hamano et al. 2024, Diabetes Obesity and Metabolism", "source_pointer": "PMID 39267249; DOI 10.1111/dom.15922", "relates_to_subclaims": [ "SC1", "SC2" ], "favors_models": [ "M2", "M3" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "A second controlled crossover showed higher intake and relative weight on a UPF menu with less chewing.", "what_it_does_not_show": "Nine men over one week cannot establish broad generalization or a nutrient-independent mechanism.", "confounders_present": [ "n=9", "men only", "one week", "abstract-level audit" ], "verification": "verified", "verification_note": "PubMed abstract and metadata read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "A" }, { "citation": "Dicken et al. 2025, Nature Medicine", "source_pointer": "PMID 40760353; DOI 10.1038/s41591-025-03842-0", "relates_to_subclaims": [ "SC1", "SC2", "SC4", "SC5" ], "favors_models": [ "M2", "M3" ], "evidence_class": "Strong", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Over eight weeks, the minimally processed healthy-guideline diet produced greater relative loss of weight and fat and lower reported intake.", "what_it_does_not_show": "Because both diets produced weight loss and corrected energy density differed, the study does not show inevitable absolute gain or processing-specific independence.", "confounders_present": [ "free-living food delivery", "self-reported intake", "energy-density difference", "unblinded" ], "verification": "verified", "verification_note": "Complete PMC full text and publisher correction read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "A" }, { "citation": "Lasschuijt et al. 2023, European Journal of Nutrition", "source_pointer": "PMID 37452167; DOI 10.1007/s00394-023-03202-z", "relates_to_subclaims": [ "SC4", "SC5" ], "favors_models": [ "M2" ], "evidence_class": "Strong", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "mixed", "what_it_shows": "At equal energy density, hard texture lowered one-day intake by 571 kcal; no processing main effect (P=0.91), but a significant processing x texture interaction (P=0.002): soft (ultra-)processed menus gave the highest intake, significantly above soft unprocessed, while hard (ultra-)processed gave the lowest.", "what_it_does_not_show": "Texture arms differed in fat (40-43% vs 24-27% EN) and fibre, and processed arms carried ~7-14x the sodium; neither texture nor processing is isolated.", "confounders_present": [ "one day per condition", "n=18 pilot", "macronutrient and fibre differences between texture arms", "sodium differences between processing arms", "(ultra-)processed arm mixes NOVA 3 and 4" ], "verification": "verified", "verification_note": "Full text re-read in red-team pass (Tables 3-4).", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "A" }, { "citation": "Teo et al. 2022, American Journal of Clinical Nutrition", "source_pointer": "PMID 35285882; DOI 10.1093/ajcn/nqac068", "relates_to_subclaims": [ "SC4", "SC5" ], "favors_models": [ "M2" ], "evidence_class": "Strong", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "contradicts", "what_it_shows": "Texture changed eating rate and acute intake across minimally processed and UPF lunches.", "what_it_does_not_show": "Residual energy-density differences and an acute lunch prevent isolation of a long-term processing-independent effect.", "confounders_present": [ "single meal", "young healthy adults", "residual energy-density difference" ], "verification": "verified", "verification_note": "Complete PMC full text read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "A" }, { "citation": "Forde et al. 2026, American Journal of Clinical Nutrition", "source_pointer": "PMID 41314613; DOI 10.1016/j.ajcnut.2025.11.012", "relates_to_subclaims": [ "SC2", "SC4", "SC5" ], "favors_models": [ "M2" ], "evidence_class": "Strong", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "contradicts", "what_it_shows": "Texture-derived eating rate causally changed intake by 369 kcal/day within diets that were both approximately 95% UPF.", "what_it_does_not_show": "It has no minimally processed arm, and the served Slow-ER and Fast-ER diets differed in fat (22 vs 34% EN) and protein (22 vs 18% EN); the authors report the intake effect was not attributable to macronutrient intake after adjustment.", "confounders_present": [ "14 days", "41 participants", "served macronutrient differences", "unblinded" ], "verification": "verified", "verification_note": "Full text re-read in red-team pass (Table 1, Results, Figure 6).", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "A" }, { "citation": "van Bruinessen et al. 2026, European Journal of Nutrition", "source_pointer": "PMID 42782361; DOI 10.1007/s00394-026-04111-7", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M2" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": false, "endogenous": true, "direction": "null", "what_it_shows": "Postprandial GLP-1, ghrelin, PP, and PYY responses to a fixed load did not differ between baseline (habitual low-UPF) and after either 14-day high-UPF diet, nor between Fast and Slow diets; fasting PYY was 17% higher after the Fast diet than baseline.", "what_it_does_not_show": "It is a secondary analysis of NCT06113146 and cannot be counted as an independent experiment.", "confounders_present": [ "shared cohort with source 6", "secondary endpoints", "14-day exposure" ], "verification": "verified", "verification_note": "Complete PMC full text read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "A" }, { "citation": "Rego et al. 2026, Obesity", "source_pointer": "PMID 41255123; DOI 10.1002/oby.70086", "relates_to_subclaims": [ "SC4", "SC5" ], "favors_models": [ "M2", "M4" ], "evidence_class": "Strong", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "null", "what_it_shows": "Matched eucaloric UPF exposure did not change overall later buffet intake or eating rate in emerging adults. An exploratory age subgroup found higher intake after the UPF diet in late adolescents (18-21 y) only.", "what_it_does_not_show": "It does not test ad libitum intake during the 14-day exposure diets and the subgroup signal was exploratory.", "confounders_present": [ "n=27", "emerging adults", "post-exposure buffet endpoint", "exploratory subgroup" ], "verification": "verified", "verification_note": "Complete PMC full text read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "A" }, { "citation": "Galdino-Silva et al. 2024, Nutrients", "source_pointer": "PMID 39771019; DOI 10.3390/nu16244398", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M2" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "A matched UPF breakfast was eaten faster with less chewing and less decline in capacity to eat.", "what_it_does_not_show": "It did not demonstrate later excess intake, weight change, or a general metabolic effect.", "confounders_present": [ "acute fixed meal", "n=42", "parallel design" ], "verification": "verified", "verification_note": "Complete PMC full text read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "A" }, { "citation": "Gama et al. 2026 and Barros et al. 2026", "source_pointer": "PMID 42588167; PMID 42633852; trial RBR-10gfr3fb", "relates_to_subclaims": [ "SC4", "SC5" ], "favors_models": [ "M2", "M4" ], "evidence_class": "Strong", "system": "human interventional", "decisive_perturbation": true, "endogenous": true, "direction": "null", "what_it_shows": "The UPF preload altered insulin in a secondary analysis, but the registered primary report found no processing effect on later ad libitum intake.", "what_it_does_not_show": "An acute 19-person preload study cannot determine long-term weight and its hormonal outcome cannot replace the primary intake endpoint.", "confounders_present": [ "n=19", "acute preload", "shared trial reports", "secondary hormonal analysis" ], "verification": "verified", "verification_note": "Complete PMC hormonal report and PubMed primary-report abstract read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "A" }, { "citation": "Sciarrillo et al. 2024 and Brunstrom et al. 2026", "source_pointer": "PMID 39163976; PMID 41475551; NCT03407053", "relates_to_subclaims": [ "SC1", "SC2", "SC3", "SC4" ], "favors_models": [ "M2", "M3" ], "evidence_class": "Moderate", "system": "human interventional", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "Order analysis did not detect UPF diet-order effects, while post hoc meal-property models implicated low-energy-density mass and macronutrient balance.", "what_it_does_not_show": "Both are analyses of Hall's participants and do not add independent causal replication.", "confounders_present": [ "shared NCT03407053 cohort", "post hoc modeling", "partial final-text access for order analysis" ], "verification": "verified", "verification_note": "Final abstract plus full author preprint and complete post hoc PMC full text read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "A" }, { "citation": "Beslay et al. 2020, PLOS Medicine", "source_pointer": "PMID 32853224; DOI 10.1371/journal.pmed.1003256", "relates_to_subclaims": [ "SC5" ], "favors_models": [ "M3", "M4" ], "evidence_class": "Moderate", "system": "human observational", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "Higher UPF share prospectively predicted BMI gain and incident overweight and obesity after extensive adjustment.", "what_it_does_not_show": "Residual confounding, selection, dietary change, measurement error, and category misclassification prevent causal or processing-independent inference.", "confounders_present": [ "residual confounding", "volunteer cohort", "self-reported diet", "NOVA misclassification" ], "verification": "verified", "verification_note": "Complete PMC full text read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "A" }, { "citation": "Forde et al. 2020, Current Developments in Nutrition", "source_pointer": "PMID 32110771; DOI 10.1093/cdn/nzaa019", "relates_to_subclaims": [ "SC4", "SC5" ], "favors_models": [ "M2" ], "evidence_class": "Moderate", "system": "human food-property dataset", "decisive_perturbation": false, "endogenous": false, "direction": "supports", "what_it_shows": "Energy-intake rate was higher on average in more processed categories but varied widely within every category.", "what_it_does_not_show": "A pooled cross-sectional food dataset cannot establish weight gain or causal mediation.", "confounders_present": [ "unbalanced food sampling", "classification ambiguity", "not a feeding trial" ], "verification": "verified", "verification_note": "Complete PMC full text read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "A" }, { "citation": "Robinson et al. 2026, Critical Reviews in Food Science and Nutrition", "source_pointer": "PMID 42746678; DOI 10.1080/10408398.2026.2731838", "relates_to_subclaims": [ "SC5" ], "favors_models": [ "M2", "M3" ], "evidence_class": "Moderate", "system": "systematic review", "decisive_perturbation": false, "endogenous": false, "direction": "supports", "what_it_shows": "Intake effects of UPF arms varied by analytic model; weight gain and eating rate were significantly greater in UPF arms; intake effects appeared confined to trials with higher-energy-density UPF arms; certainty low; trial diets not matched for nutrient profile.", "what_it_does_not_show": "Meta-analysis of heterogeneous trials cannot identify a residual processing mechanism and is not primary causal evidence.", "confounders_present": [ "trial heterogeneity", "model dependence", "low certainty", "abstract-level audit" ], "verification": "verified", "verification_note": "PubMed abstract and metadata read; used for framing only.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable by user report; counts intentionally not used." ] }, "source_tier": "B" }, { "citation": "NIH Clinical Center / NIDDK, NCT05290064 (registry-posted results, 2026-08-12)", "source_pointer": "ClinicalTrials.gov NCT05290064; https://clinicaltrials.gov/study/NCT05290064", "relates_to_subclaims": [ "SC1", "SC4", "SC5" ], "favors_models": [ "M2", "M3" ], "evidence_class": "Strong", "system": "human interventional (inpatient)", "decisive_perturbation": true, "endogenous": true, "direction": "mixed", "what_it_shows": "UPF high in energy density and hyperpalatable foods raised intake 948 kcal/d over an unprocessed diet; within UPF, energy density and hyperpalatability each raised intake; with both low, UPF minus unprocessed was +128 kcal/d, not significant.", "what_it_does_not_show": "Not peer reviewed; one week per diet; unmasked; composition matching beyond energy density and hyperpalatability is not reported in the registry; cannot exclude a residual effect up to ~265 kcal/d.", "confounders_present": [ "registry summary results only", "1-week periods", "same research group as [1]", "n=36" ], "verification": "verified", "verification_note": "Complete posted-results module read via ClinicalTrials.gov v2 API in red-team pass.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable in the red-team pass (account usage limit); counts not used." ] }, "source_tier": "A" }, { "citation": "Preston et al. 2025, Cell Metabolism", "source_pointer": "PMID 40882621; DOI 10.1016/j.cmet.2025.08.004", "relates_to_subclaims": [ "SC2", "SC4" ], "favors_models": [ "M3", "M1" ], "evidence_class": "Moderate", "system": "human interventional (prescribed diets)", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "In a 2 x 2 crossover of prescribed UPF vs unprocessed diets at two caloric loads, body weight increased on UPF independent of caloric load.", "what_it_does_not_show": "Magnitudes and composition matching not verified (abstract-level); prescribed rather than ad libitum intake, so it does not test the intake pathway; free-living adherence.", "confounders_present": [ "abstract-level audit", "composition matching unverified", "men only" ], "verification": "verified", "verification_note": "PubMed abstract and registry NCT05368194 read; publisher full text inaccessible (403).", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable in the red-team pass (account usage limit); counts not used." ] }, "source_tier": "A" }, { "citation": "Fazzino et al. 2023, Nature Food", "source_pointer": "PMID 37117850; DOI 10.1038/s43016-022-00688-4", "relates_to_subclaims": [ "SC4", "SC5" ], "favors_models": [ "M2" ], "evidence_class": "Moderate", "system": "human meal-level analysis within inpatient trials", "decisive_perturbation": false, "endogenous": true, "direction": "supports", "what_it_shows": "Across 2,733 meals from four inpatient dietary patterns, energy density, eating rate, and hyperpalatable foods were consistently positively related to meal energy intake.", "what_it_does_not_show": "Meal-level associations within trials are not randomized manipulations of each property.", "confounders_present": [ "observational within-trial analysis", "same research group as [1]" ], "verification": "verified", "verification_note": "PubMed abstract read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable in the red-team pass (account usage limit); counts not used." ] }, "source_tier": "A" }, { "citation": "Hägele et al. 2025, Nature Metabolism", "source_pointer": "PMID 40082711; DOI 10.1038/s42255-025-01247-4", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M4" ], "evidence_class": "Moderate", "system": "human interventional (calorimeter)", "decisive_perturbation": true, "endogenous": true, "direction": "contradicts", "what_it_shows": "Within two UPF diets matched for calories, fat, and fibre, the high-protein lower-carbohydrate UPF diet lowered ad libitum intake and raised energy expenditure over 54 h.", "what_it_does_not_show": "Very short duration; tests composition within UPF, not processing.", "confounders_present": [ "54-hour exposure", "n=21" ], "verification": "verified", "verification_note": "PubMed abstract read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable in the red-team pass (account usage limit); counts not used." ] }, "source_tier": "A" }, { "citation": "Larcom et al. 2026, Appetite", "source_pointer": "PMID 41115473; DOI 10.1016/j.appet.2025.108352", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M2", "M4" ], "evidence_class": "Moderate", "system": "human interventional (single meal)", "decisive_perturbation": true, "endogenous": true, "direction": "null", "what_it_shows": "Single meals matched for energy, macronutrients, flavour, and texture gave similar intake for non-UPF and UPF of equal nutritional quality; a less palatable low-quality UPF meal gave lower intake.", "what_it_does_not_show": "A single meal cannot test sustained intake or weight.", "confounders_present": [ "single meal", "n=40" ], "verification": "verified", "verification_note": "PubMed abstract read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable in the red-team pass (account usage limit); counts not used." ] }, "source_tier": "A" }, { "citation": "Lasschuijt et al. 2025, Journal of Nutrition", "source_pointer": "PMID 40516651; DOI 10.1016/j.tjnut.2025.06.006", "relates_to_subclaims": [ "SC5" ], "favors_models": [ "M2" ], "evidence_class": "Moderate", "system": "human interventional (single meal)", "decisive_perturbation": true, "endogenous": true, "direction": "contradicts", "what_it_shows": "Within sandwiches made largely of UPF ingredients, energy density and eating rate interacted to produce a twofold intake range.", "what_it_does_not_show": "Single-meal design; no minimally processed comparator.", "confounders_present": [ "single meal", "same research network as [6]" ], "verification": "verified", "verification_note": "PubMed abstract read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable in the red-team pass (account usage limit); counts not used." ] }, "source_tier": "A" }, { "citation": "Teo et al. 2026, Appetite", "source_pointer": "PMID 42176807; DOI 10.1016/j.appet.2026.108604", "relates_to_subclaims": [ "SC5" ], "favors_models": [ "M2" ], "evidence_class": "Moderate", "system": "human interventional (fixed-portion meal)", "decisive_perturbation": true, "endogenous": true, "direction": "contradicts", "what_it_shows": "A hard-textured UPF lunch was eaten about half as fast as a soft UPF lunch, gave the greatest fullness, and the lowest later snack intake.", "what_it_does_not_show": "Fixed portions and a single meal; men only.", "confounders_present": [ "single meal", "men only" ], "verification": "verified", "verification_note": "PubMed abstract read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable in the red-team pass (account usage limit); counts not used." ] }, "source_tier": "A" }, { "citation": "Macena et al. 2026, Nutrition, Metabolism and Cardiovascular Diseases", "source_pointer": "PMID 41381310; DOI 10.1016/j.numecd.2025.104426", "relates_to_subclaims": [ "SC2" ], "favors_models": [ "M3" ], "evidence_class": "Moderate", "system": "human interventional (12-month free-living)", "decisive_perturbation": true, "endogenous": true, "direction": "supports", "what_it_shows": "Adding UPF-restriction advice to energy restriction produced a small, statistically significant but not clinically significant additional weight loss over 12 months.", "what_it_does_not_show": "The achieved UPF-share contrast was small and not significant; advice trial with free-living intake.", "confounders_present": [ "small exposure contrast", "free-living self-report" ], "verification": "verified", "verification_note": "PubMed abstract read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable in the red-team pass (account usage limit); counts not used." ] }, "source_tier": "A" }, { "citation": "Darcey et al. 2025, Cell Metabolism", "source_pointer": "PMID 40043691; DOI 10.1016/j.cmet.2025.02.002", "relates_to_subclaims": [ "SC4" ], "favors_models": [ "M2", "M4" ], "evidence_class": "Moderate", "system": "human mechanistic (PET)", "decisive_perturbation": false, "endogenous": true, "direction": "null", "what_it_shows": "No significant post-ingestive striatal dopamine response to an ultra-processed high-fat, high-sugar milkshake, and responses were unrelated to adiposity.", "what_it_does_not_show": "A single product and one mechanism; does not test intake.", "confounders_present": [ "single product", "detection limits of PET" ], "verification": "verified", "verification_note": "PubMed abstract read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable in the red-team pass (account usage limit); counts not used." ] }, "source_tier": "A" }, { "citation": "Braesco et al. 2022, European Journal of Clinical Nutrition", "source_pointer": "PMID 35314769; DOI 10.1038/s41430-022-01099-1", "relates_to_subclaims": [ "SC5" ], "favors_models": [ "M2" ], "evidence_class": "Moderate", "system": "classification reliability study", "decisive_perturbation": false, "endogenous": false, "direction": "contradicts", "what_it_shows": "Specialists assigning foods to NOVA groups showed low agreement (Fleiss κ 0.32-0.34), even with ingredient lists.", "what_it_does_not_show": "Assessor agreement in one country; not an outcome study.", "confounders_present": [ "French specialist sample", "online survey" ], "verification": "verified", "verification_note": "Full text read (Results).", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable in the red-team pass (account usage limit); counts not used." ] }, "source_tier": "A" }, { "citation": "Monteiro et al. 2025, Lancet", "source_pointer": "PMID 41270766; DOI 10.1016/S0140-6736(25)01565-X", "relates_to_subclaims": [ "SC5" ], "favors_models": [ "M1", "M3" ], "evidence_class": "Moderate", "system": "narrative/systematic synthesis", "decisive_perturbation": false, "endogenous": false, "direction": "supports", "what_it_shows": "Proponents' synthesis concluding that UPF-based dietary patterns are a key driver of diet-related chronic disease, attributing overeating to energy density, hyper-palatability, soft texture, and disrupted matrices.", "what_it_does_not_show": "A synthesis; used only for field framing (Divergence) and to fix the fair reading of 'properties of processing'.", "confounders_present": [ "synthesis by proponents" ], "verification": "verified", "verification_note": "PubMed abstract read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable in the red-team pass (account usage limit); counts not used." ] }, "source_tier": "B" }, { "citation": "Ludwig 2026, Critical Reviews in Food Science and Nutrition", "source_pointer": "PMID 42063358; DOI 10.1080/10408398.2026.2664043", "relates_to_subclaims": [ "SC4", "SC5" ], "favors_models": [ "M2", "M4" ], "evidence_class": "Moderate", "system": "critical review", "decisive_perturbation": false, "endogenous": false, "direction": "contradicts", "what_it_shows": "Critical review concluding that current experimental evidence does not establish a causal role of UPF in obesity beyond energy density and nutrient composition.", "what_it_does_not_show": "A review; used only for field framing (Divergence).", "confounders_present": [ "single-author review" ], "verification": "verified", "verification_note": "PubMed abstract read.", "citation_context": { "retrieved": false, "supporting_count": 0, "mentioning_count": 0, "contrasting_count": 0, "notable_contrasting_statements": [ "Scite unavailable in the red-team pass (account usage limit); counts not used." ] }, "source_tier": "B" } ], "verification_summary": { "load_bearing_claims_checked": 35, "verified": 35, "unverified": 0, "contradicted": 0, "note": "Original edition: 18 claims verified. Red-team pass (2026-09-30): every PMID/DOI/PMCID re-resolved (17/17 correct); full text re-read for [1,3,4,6,7,8] with spot checks of [5,9]; complete posted-results module read for [18]; 17 claims added (C19-C35) and 5 reworded (C5, C7, C8, C15, C18). Scite was unavailable (usage limit), so citation-context counts were not used." }, "factual_issues_found": [ { "quoted_text": "properties of processing itself", "problem": "The phrase does not name a pre-outcome exposure. Red-team correction: processing-conferred properties (texture, eating rate, matrix, additives) are the hypothesis's candidate mediators and must be named before outcomes; they are not confounders to be removed (RT-02).", "source_says": "Proponents attribute overeating to energy density, hyper-palatability, soft texture, and disrupted matrices [28]; direct experiments show texture and energy-density effects operate in both directions within NOVA 4 [4,6,23,24].", "source_pointer": "PMID 41270766; PMID 37452167; PMID 41314613; PMID 40516651; PMID 42176807", "affects_score": true, "action": "scored-on-corrected-understanding" }, { "quoted_text": "independent of nutrient composition", "problem": "Selected aggregate matching is narrower than composition-wide independence, and actual consumed composition necessarily differs when total intake differs.", "source_says": "Hall matched several offered aggregate variables but retained differences in nutrient form, ratios, non-beverage energy density, eating rate, and actual consumed energy density.", "source_pointer": "PMID 31105044; PMID 32110771", "affects_score": true, "action": "scored-on-corrected-understanding" }, { "quoted_text": "causes excess energy intake and weight gain", "problem": "The wording can be read as an invariant absolute outcome, but one major eight-week trial produced weight loss on both diets and only a relative difference.", "source_says": "UPDATE reported -2.06% weight change on the minimally processed diet and -1.05% on the ultra-processed diet, a -1.01 percentage-point relative difference.", "source_pointer": "PMID 40760353", "affects_score": true, "action": "scored-on-corrected-understanding" }, { "quoted_text": "through properties of processing itself", "problem": "Properties caused by processing, such as texture or matrix, are legitimate carriers under the claim; evidence for them supports the claim only if they are characteristic of the category and act at matched composition.", "source_says": "Texture and eating rate alter intake in both processing categories and within NOVA 4; a small factorial trial found no processing main effect but a significant processing x texture interaction.", "source_pointer": "PMID 37452167; PMID 35285882; PMID 41314613", "affects_score": true, "action": "scored-on-corrected-understanding" }, { "quoted_text": "the main effect of processing on total energy intake including snacks was null (P=0.91)", "problem": "Original C5 omitted the significant processing x texture interaction and the composition differences between arms.", "source_says": "Interaction P=0.002 incl. snacks; soft (ultra-)processed 2608 vs soft unprocessed 2205 kcal (significantly different); soft arms 40-43% fat EN vs 24-27%; processed arms ~7-14x sodium.", "source_pointer": "PMID 37452167 (PMC10469122, Tables 3-4)", "affects_score": false, "action": "corrected-in-red-team (RT-04)" }, { "quoted_text": "produced 0.43 kg less fat mass", "problem": "Imprecise; composition differences between diets omitted.", "source_says": "Body fat mass decreased on the UPF Slow-ER diet by 0.43 kg (diet x time P=0.0002); served diets 22 vs 34% fat EN and 22 vs 18% protein EN.", "source_pointer": "PMID 41314613 (PMC13084570, Table 1)", "affects_score": false, "action": "corrected-in-red-team (RT-05)" }, { "quoted_text": "judged the evidence for energy intake and weight effects to be model-dependent", "problem": "Only the intake result was model-dependent; weight gain was significantly greater in UPF arms.", "source_says": "UPF arms tended to have higher energy intake with significance varying by model; weight gain and eating rate significantly greater; certainty low.", "source_pointer": "PMID 42746678", "affects_score": false, "action": "corrected-in-red-team (RT-10)" }, { "quoted_text": "found no diet difference in GLP-1, ghrelin, pancreatic polypeptide, or PYY responses", "problem": "Comparison structure misdescribed.", "source_says": "Responses did not differ between baseline and after 14-day high-UPF exposure, nor between Fast and Slow diets.", "source_pointer": "PMID 42782361 (PMC13612719)", "affects_score": false, "action": "corrected-in-red-team (RT-11)" } ], "cap_rules_applied": [ { "subclaim_id": "SC4", "cap": "Γ", "cap_source": "universal", "triggered": false, "trigger_text_quoted": "Biomarker change without functional/phenotypic change caps at Γ.", "specific_trigger": "Not triggered: the relevant trials measure functional intake and weight/fat outcomes. SC4 remains Γ because the processing-specific exposure is not isolated and therefore fails the Δ boundary.", "resulting_ceiling": "Γ" }, { "subclaim_id": "SC5", "cap": "Β", "cap_source": "universal", "triggered": false, "trigger_text_quoted": "Pure association evidence — however large, however replicated — cannot exceed Β.", "specific_trigger": "Not triggered for the full evidence set because randomized feeding exists. The prospective cohort by itself remains associational and cannot carry category-wide causality.", "resulting_ceiling": "Β" }, { "subclaim_id": "SC3", "cap": "Γ", "cap_source": "rubric", "triggered": false, "trigger_text_quoted": "Biomarker-only improvement without organismal benefit", "specific_trigger": "Not triggered: the matched-menu trial measured intake, body weight, and fat mass. The limitation is the width of the independence claim, not a biomarker-only endpoint.", "resulting_ceiling": "Γ" }, { "subclaim_id": "SC1", "cap": "Γ", "cap_source": "rubric", "triggered": false, "trigger_text_quoted": "Single-species or single-strain evidence", "specific_trigger": "Does not bind: the claim is human-scoped and the evidence is human, so no cross-species inference occurs; the rule governs generalization from one model organism (cap-precedence rule). Applies equally to SC2 and SC3.", "resulting_ceiling": "none" }, { "subclaim_id": "SC4", "cap": "Γ", "cap_source": "universal", "triggered": true, "trigger_text_quoted": "Biomarker change without functional/phenotypic change caps at Γ.", "specific_trigger": "Triggered for the postprandial insulin result [10], which cannot substitute for the null registered intake endpoint [11]; binds that evidence item, not the subclaim's other evidence.", "resulting_ceiling": "Γ" } ], "cap_precedence_resolutions": [ { "collision": "The metabolism/aging Δ wording is lifespan/healthspan-specific, whereas this obesity hypothesis has direct human intake and body-composition interventions.", "subclaim_scope_cap_binds_on": "No universal low-evidence cap binds SC1-SC3 because direct human functional perturbations exist.", "subclaim_scope_threshold_binds_on": "Randomized human feeding changing measured intake and relative weight/fat is treated as the field-valid organismal analogue for the narrow obesity subclaims.", "resolution": "The human dietary perturbations cross Δ and Ε only within the menus and outcomes actually tested; they do not cross the boundary for a residual processing-specific exposure that was never isolated.", "note": "The rubric mismatch is explicit and is not used to promote the broad claim." }, { "collision": "Rubric 06 hard rule 'Single-species or single-strain evidence' (triggers Γ cap regardless of other evidence) versus human-only evidence for a human-scoped claim.", "subclaim_scope_cap_binds_on": "Would bind only an inference from one species to another; no such inference is made.", "subclaim_scope_threshold_binds_on": "SC1-SC3, human-scoped, where randomized human functional outcomes cross the adapted Δ and Ε thresholds.", "resolution": "The cap does not bind SC1-SC3. Recorded explicitly (red-team RT-15) rather than applied or ignored silently.", "note": "If the rubric is revised for nutrition claims, the rule should be restated as 'single-model-system evidence generalized beyond that system'." } ], "boundary_check_BGD": [ { "subclaim_id": "SC1", "field_valid_perturbation_quote": "Genetic or pharmacologic intervention producing reproducible lifespan/healthspan extension; Must separate developmental effects from true aging effects; Biomarker change alone does NOT satisfy this threshold", "qualifying_evidence_present": "yes", "qualifying_evidence_citation": "Randomized controlled human whole-diet perturbations [1-3,18]; boundary-mapping designs [4,8,11,22].", "gamma_hard_rules_triggered": [ "Biomarker-only improvement without organismal benefit — not triggered; energy intake is a functional outcome." ], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "yes", "perturbation_result_note": "The positive result crosses the functional analogue of Δ for bounded human intake, not for a residual processing mediator.", "ceiling_from_boundary_check": "Δ or higher" }, { "subclaim_id": "SC2", "field_valid_perturbation_quote": "Genetic or pharmacologic intervention producing reproducible lifespan/healthspan extension; Must separate developmental effects from true aging effects; Biomarker change alone does NOT satisfy this threshold", "qualifying_evidence_present": "yes", "qualifying_evidence_citation": "Randomized controlled human feeding with repeated weight or fat outcomes [1-3,6,19].", "gamma_hard_rules_triggered": [ "Biomarker-only improvement without organismal benefit — not triggered; weight and fat mass are organism-level phenotypes." ], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "yes", "perturbation_result_note": "Relative short-term weight/fat effects are positive in bounded contexts; durable obesity is not established.", "ceiling_from_boundary_check": "Δ or higher" }, { "subclaim_id": "SC3", "field_valid_perturbation_quote": "Genetic or pharmacologic intervention producing reproducible lifespan/healthspan extension; Must separate developmental effects from true aging effects; Biomarker change alone does NOT satisfy this threshold", "qualifying_evidence_present": "yes", "qualifying_evidence_citation": "Hall 2019 [1] directly randomized menus matched on selected offered aggregate variables.", "gamma_hard_rules_triggered": [ "Biomarker-only improvement without organismal benefit — not triggered." ], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "yes", "perturbation_result_note": "Positive only for the variables actually matched; composition-wide independence is not inferred.", "ceiling_from_boundary_check": "Δ or higher" }, { "subclaim_id": "SC4", "field_valid_perturbation_quote": "Genetic or pharmacologic intervention producing reproducible lifespan/healthspan extension; Must separate developmental effects from true aging effects; Biomarker change alone does NOT satisfy this threshold", "qualifying_evidence_present": "no", "qualifying_evidence_citation": "No positive composition-matched perturbation: UPF-LL vs UNF-LL non-significant [18]; matched exposure, preload, and single-meal designs null [8,11,22]; positive contrasts [4,19] are not composition-matched.", "gamma_hard_rules_triggered": [], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "no", "perturbation_result_note": "The field-valid design exists [18] but its composition-matched contrast is non-significant (pitfall 17: a null of the right design does not cross Δ); unmatched positive contrasts do not qualify either.", "ceiling_from_boundary_check": "Γ" }, { "subclaim_id": "SC5", "field_valid_perturbation_quote": "Genetic or pharmacologic intervention producing reproducible lifespan/healthspan extension; Must separate developmental effects from true aging effects; Biomarker change alone does NOT satisfy this threshold", "qualifying_evidence_present": "yes", "qualifying_evidence_citation": "Multiple human randomized interventions test diverse diets and within-UPF properties [1-11,18,21-24].", "gamma_hard_rules_triggered": [], "disease_phenotype_relevance": "yes", "perturbation_result_positive": "no", "perturbation_result_note": "Processing-conferred properties act in both directions within NOVA 4 [4,6,23,24] and the category is unreliably assigned [27]; the category-wide prediction does not cross Δ.", "ceiling_from_boundary_check": "Γ" } ], "thresholds_tested": [ { "subclaim_id": "SC1", "threshold": "Ε", "requirement_quoted": "Requires causal effect in constrained contexts, backgrounds, or species.", "satisfied": "yes", "supporting_evidence": [ "Whole-diet randomized effects in Hall, Hamano, UPDATE, and NCT05290064 [1-3,18].", "Boundary mapped: effect small and non-significant at low energy density and hyperpalatability [18]; relative difference within net weight loss [3]; null matched designs [8,11,22]." ], "what_would_cross_it": "Already crossed for bounded human menus and exposure periods." }, { "subclaim_id": "SC1", "threshold": "Ζ", "requirement_quoted": "Requires: Reproducible intervention effects; Mechanistic narrowing; Evidence not reducible to general sickness, caloric reduction, or developmental artifacts.", "satisfied": "no", "supporting_evidence": [ "Whole-diet effects replicate directionally, but their size and presence vary with menu properties." ], "what_would_cross_it": "Independent long-duration replications plus prospective quantitative mediation showing which processing-linked properties account for the intake effect." }, { "subclaim_id": "SC2", "threshold": "Ε", "requirement_quoted": "Requires causal effect in constrained contexts, backgrounds, or species.", "satisfied": "yes", "supporting_evidence": [ "Relative weight or fat effects in Hall, Hamano, UPDATE, and Forde [1-3,6]; prescribed-intake weight effect (abstract-level) [19]." ], "what_would_cross_it": "Already crossed for relative short-term human outcomes." }, { "subclaim_id": "SC2", "threshold": "Ζ", "requirement_quoted": "Requires: Reproducible intervention effects; Mechanistic narrowing; Evidence not reducible to general sickness, caloric reduction, or developmental artifacts.", "satisfied": "no", "supporting_evidence": [ "The body-composition effect is consistent with realized energy-intake differences, while no processing-specific mediator has been narrowed." ], "what_would_cross_it": "A longer controlled trial showing durable weight/fat divergence and preregistered mediation by a replicated processing-linked property." }, { "subclaim_id": "SC3", "threshold": "Ε", "requirement_quoted": "Requires causal effect in constrained contexts, backgrounds, or species.", "satisfied": "yes", "supporting_evidence": [ "Hall's effect persisted under prespecified matching of several aggregate offered variables [1]." ], "what_would_cross_it": "Already crossed only for those matched variables and menus." }, { "subclaim_id": "SC3", "threshold": "Ζ", "requirement_quoted": "Requires: Reproducible intervention effects; Mechanistic narrowing; Evidence not reducible to general sickness, caloric reduction, or developmental artifacts.", "satisfied": "no", "supporting_evidence": [ "Residual nutrient-form, matrix, energy-density, and eating-rate differences remain [1,13,16]." ], "what_would_cross_it": "Replication with composition, non-beverage energy density, matrix, texture, eating rate, palatability, and choice prospectively separated." }, { "subclaim_id": "SC4", "threshold": "Δ", "requirement_quoted": "Requires at least one intervention changing lifespan or healthspan in a way consistent with the hypothesis.", "satisfied": "no", "supporting_evidence": [ "Using the explicit obesity analogue, no positive perturbation shows the UPF effect at matched composition; the most direct contrast is non-significant [18] and matched designs are null [8,11,22]." ], "what_would_cross_it": "An adequately powered, composition-matched (as consumed) UPF versus minimally processed feeding trial with a prespecified processing route, showing a replicated positive effect on intake or fat mass." }, { "subclaim_id": "SC5", "threshold": "Δ", "requirement_quoted": "Requires at least one intervention changing lifespan or healthspan in a way consistent with the hypothesis.", "satisfied": "no", "supporting_evidence": [ "The category-wide prediction is contradicted by bidirectional within-UPF effects and null separators [4,6,8,11,18,23,24]; NOVA assignment is unreliable [27]." ], "what_would_cross_it": "Prospectively replicated positive effects across diverse UPF product classes after relevant property matching, with internal UPF negative controls." } ], "generalization_check": { "generalization_tested": "yes", "context_boundary_mapped": "yes", "internal_negative_controls": [ "UPF low in energy density and hyperpalatable foods did not significantly exceed an unprocessed diet [18] (registry-posted).", "Processing x texture interaction with sign reversal; no processing main effect [4].", "Large intake differences between two diets both approximately 95% UPF [6].", "Twofold intake range within UPF-ingredient meals by energy density and eating rate [23]; hard UPF meal most satiating [24].", "No overall later buffet-intake effect after matched eucaloric UPF exposure [8].", "No later-intake effect in the registered 2 x 2 UPF preload trial [11].", "Both UPDATE diets produced weight loss despite a relative minimally processed advantage [3]." ], "note": "The boundary is mapped from both sides: some whole-diet packages are causal, and the effect disappears when energy density and hyperpalatability are low [18]; NOVA status is heterogeneous and unreliably assigned [27]. This earns Ε for the bounded package subclaims, not for independence from composition or category-wide properties." }, "falsification_patterns_met": [ { "subclaim_id": "SC4", "pattern": "Benefits are explained by reduced growth, intake, or sickness", "evidence": "Adapted to this rubric mismatch: weight/fat differences track energy intake, and energy density, hyperpalatability, and texture explain intake without a composition-independent processing effect [18,23]. A prescribed-intake weight effect [19] is the one possible exception and could not be verified beyond its abstract.", "downgrade_applied": false }, { "subclaim_id": "SC4", "pattern": "Pathway manipulation changes biomarkers but not lifespan/healthspan", "evidence": "Adapted to intake: the Gama/Barros trial changed insulin but not the registered later-intake endpoint [10,11].", "downgrade_applied": false }, { "subclaim_id": "SC5", "pattern": "Effects disappear across strains or species", "evidence": "Adapted to human contexts: processing-category effects are null or non-significant in several matched designs [8,11,18,22] and reverse sign with texture [4]; wide variation within category [6,16,23,24].", "downgrade_applied": false } ], "pitfalls_flagged": [ { "subclaim_id": "SC3", "pitfall": "Review-specific caution (no library number): selected as-offered matching treated as exhaustive independence", "where": "Matching selected aggregate nutrients as offered is treated as if composition as consumed were matched." }, { "subclaim_id": "SC4", "pitfall": "Pitfall 8 — Confirmation by biomarker", "where": "Postprandial insulin or appetite hormones cannot substitute for measured later intake or weight [10,11]." }, { "subclaim_id": "SC4", "pitfall": "Pitfall 10 — Mechanism conflation", "where": "NOVA 4 bundles many formulations and processes under one label; evidence for one process-conferred property is not evidence for the category." }, { "subclaim_id": "SC4", "pitfall": "Scope-determination rule (v5.0): qualifier must be determinable before the outcome", "where": "'Properties of processing' must be named before outcomes, not defined as whatever remains after a positive result." }, { "subclaim_id": "SC4", "pitfall": "Pitfall 17 — Negative field-valid perturbation read as a crossed threshold (and its converse)", "where": "The composition-matched NIH contrast [18] is the right design but non-significant; conversely, unmatched positive contrasts [4,19] are not composition-independent." }, { "subclaim_id": "SC5", "pitfall": "Pitfall 7 — Selective generalization", "where": "A bounded positive menu effect is generalized to all ultra-processed foods despite within-category negative controls." }, { "subclaim_id": "SC2", "pitfall": "Review-specific caution (no library number): relative and absolute outcomes conflated", "where": "Greater relative weight loss on a minimally processed diet is described as absolute weight gain on the ultra-processed diet." }, { "subclaim_id": "SC1", "pitfall": "Pitfall 9 — Replication confused with reproduction (shared cohorts)", "where": "Hall, Sciarrillo, and Brunstrom are NCT03407053; Forde and van Bruinessen are NCT06113146; Gama and Barros are RBR-10gfr3fb." }, { "subclaim_id": "SC1", "pitfall": "Pitfall 9 — Replication confused with reproduction (research networks)", "where": "Positive whole-diet evidence and analyses concentrate in one inpatient research group [1,12,13,18,20] (with an author shared with [3]); texture/eating-rate separators in one sensory-science network [5-7,16,23,24] (one member co-authored [1]); matched-meal/preload nulls and the 12-month trial in one Brazilian group [9-11,25]. Independent groups: [2,8,19,21,22]. Informs Moderate confidence; not a criticism of any group." } ], "causal_tally_audit": { "tally_claimed_by_document": 0, "tally_after_audit": 0, "items_rejected": [ "Hall 2019, Sciarrillo 2024, and Brunstrom 2026 are one participant trial, not three independent causal lines.", "Forde 2026 and van Bruinessen 2026 are one trial; the satiety analysis does not add independent replication.", "Gama 2026 and Barros 2026 are one registered trial; the positive insulin secondary endpoint cannot override the null primary intake endpoint.", "Prospective cohort associations cannot establish nutrient-independent processing causality.", "A Tier B meta-analysis summarizes trials but does not add an independent causal perturbation.", "Matching offered macronutrient percentages does not match actual consumed amount, nutrient form, non-beverage energy density, or food matrix.", "NCT05290064 and Hall 2019 come from the same research group: within-group reproduction raises confidence less than independent replication.", "Preston 2025 is abstract-level and uses prescribed intake; it cannot count toward the intake pathway or composition independence.", "Fazzino 2023 is a meal-level association within trials, not a randomized manipulation of each property." ] }, "steelman": { "case_one_rung_higher": "A Δ case for the full claim: randomized whole-diet trials at three independent sites cause large intake and relative weight effects [1-3], a second NIH cohort reproduces the direction [18], Hall matched several nutrients as offered [1], cohort associations persist after nutritional-quality adjustment [14], soft (ultra-)processed menus exceeded soft unprocessed ones by ~400 kcal/day at equal energy density [4], the composition-matched NIH contrast is positive (+128 kcal/d, CI to +265) [18], and a prescribed-intake trial reports UPF-associated weight gain regardless of caloric load [19].", "case_one_rung_lower": "A Β case: 'processing itself' is loosely specified, the category is unreliably assigned (κ ~0.3) [27], and several matched or factorial studies are null for processing category, leaving the independence claim close to a classification-level association.", "which_rejected": "The Δ case is rejected for SC4 and SC5 because no contrast is both composition-matched and significant: the soft-texture contrast carried ~7x the sodium, less fibre, and more saturated fat; the NIH contrast is non-significant and registry-only; the prescribed-intake trial's matching is unverified; composition-matched designs are null [8,11,22]; and composition changes intake within UPF [21]. The Β case is rejected because randomized whole-diet and property perturbations provide genuine causal partial-system evidence.", "blocking_cap_threshold_or_missing_experiment": "The binding blocker is a failed Δ boundary for the independence clause: an adequately powered, peer-reviewed, composition-matched (as consumed) UPF versus minimally processed feeding trial showing a replicated positive effect on intake or fat mass is missing." }, "final_score": { "hsef_level": "Γ", "hsef_descriptor": "Mechanistically plausible, partial system support", "confidence": "Moderate", "confidence_rationale": "Multiple independent controlled human interventions exist, but they are small (9-69 participants), short (one meal to eight weeks), partly composition-confounded, concentrated in a few research networks, and in one key case registry-only [18]; several nulls cannot exclude modest effects. This matches the framework's Moderate definition ('Multiple independent studies; some replication; some assay-level concerns or methodological heterogeneity remains'), not High ('Extensive independent replication ... reproducibility well-established'). Confidence measures evidence density, not certainty that the strong claim fails (red-team RT-03; original: High).", "rests_on_non_peer_reviewed_work": false, "rationale": "Some ultra-processed whole-diet packages causally increase energy intake and relative short-term weight or fat gain, reaching Ε for those bounded subclaims. The submitted proposition additionally requires that the effect persist when nutrient composition is matched and be carried by processing-conferred properties characteristic of the category; those clauses remain Γ because the most direct composition-matched contrast is non-significant [18], matched designs are null [8,11,22], composition changes intake within UPF [21], and processing-conferred properties act in both directions within NOVA 4 [4,6,23,24]. The level does not rest on the registry-only result [18]." }, "role_of_mechanism": { "classification": "reinforcing / stabilizing factor", "derivation_step_matched": "5", "derivation": "Step 1 (downstream consequence?): no — randomized diet assignment precedes the intake change. Step 2 (epiphenomenon: no causal dependency in either direction, only co-occurrence?): no — randomized UPF packages [1-3,18] and processing-modifiable properties such as texture and energy density [4,6,18,23] causally change intake. Step 3 (primary driver: removal prevents, engagement initiates, rate-limiting?): no — excess intake occurs without UPF and UPF low in energy density and hyperpalatability does not clearly raise intake [18]. Step 4 (permissive: required but not initiating?): no — UPF exposure is not required for excess intake. Step 5: processing-linked properties amplify intake in some formulations -> reinforcing / stabilizing factor.", "tournament_used_as_tiebreaker": false, "note": "The classification applies to the processing mechanism as claimed. It does not deny the causal intake effect of particular ultra-processed menus or particular process-created food properties. Red-team pass wrote out steps 1-4 explicitly (RT-14); classification unchanged." }, "extended_scoring_MSD": { "momentum": { "value": "Mechanistic only", "artifact": "NCT05290064 (NIH Clinical Center four-arm inpatient crossover) did prespecify the UPF-LL versus UNF-LL contrast, which is SC4's boundary test, but it completed 2025-08-15 and its posted primary contrast is non-significant (+128.4 kcal/d, P = 0.065) [18]; nothing that could cross the next threshold is being run, and the pending journal report covers a result already in hand. Active mechanistic work on texture, energy density, composition and satiety continues [6,7,21-24].", "justification": "Lead-adjudicated in the 2026-10-14 adversarial review: the framework value requires that someone be running an experiment that could cross the next threshold, and both halves fail - the trial has completed, and its posted contrast would not carry SC4 across \u0394. Returned to the original 2026-09-29 value of Mechanistic only (reversing red-team RT-01/RT-18). Momentum modifies no letter." }, "stability": { "value": "Stable", "trigger": "Named pending trigger: the peer-reviewed report of NCT05290064. Its posted primary contrast (+128.4 kcal/d, 95% CI -8.2 to 265.0, P=0.065) would not carry SC4 across Δ; only a materially different peer-reviewed analysis could, and none is signalled.", "justification": "The one identified pending result is expected to leave the level unchanged." }, "divergence": { "implied_level": "Contested (proponent framing Ζ; critical framing Β-Γ)", "implied_level_basis": "Proponents' Lancet synthesis: 'the totality of the evidence supports' UPF-based dietary patterns as 'a key driver' [28] (implies Ζ or above). Critical syntheses: evidence 'does not establish a causal role of UPF consumption in obesity' [29]; 'evidential certainty is low' [15] (imply Β-Γ).", "assessed_level": "Γ", "gap_in_rungs": "+3 against proponent framing; -1 to 0 against critical framing", "value": "Contested", "justification": "The framework computes divergence from how the field presents the hypothesis; the field is genuinely split, so no single implied level applies. The original value (Severe overstatement, +3) was computed from the wording of the submitted sentence (red-team RT-08)." }, "calibration_note": "M/S/D never modify HSEF level or confidence. Each value is tied to a checkable artifact." }, "change_record": { "applies": true, "prior_level": "Γ", "prior_confidence": "High", "prior_source": "Original HSEF v5.4 edition (OpenAI Codex, 2026-09-29) of this review, audited in the red-team pass.", "derived_before_consulting_prior": false, "new_level": "Γ", "direction": "unchanged (confidence High -> Moderate)", "what_moved_it": "Level unchanged: Γ holds under both the original residual-essence reading and the corrected composition-independence reading. Confidence moved High -> Moderate on the framework's definition (RT-03). Momentum Mechanistic only -> Decisive work underway (RT-01/RT-18); Divergence Severe overstatement -> Contested (RT-08). The 2026-10-14 adversarial review reversed the Momentum change, returning it to Mechanistic only; the Divergence change stands.", "divergence_from_other_reviewers": "Red-team pass by a different model family; see redteam/REDTEAM_REPORT.md and redteam/SYNTHESIS_CHANGE_LOG.md." }, "proposed_decisive_experiments": [ { "name": "Composition-matched equivalence trial with a processing-route factor", "experiment": "Residential randomized crossover at an independent site, at least four weeks per condition, comparing UPF and minimally processed diets matched as consumed on macronutrient and micronutrient amounts, hyperpalatable combinations, and total and non-beverage energy density, crossed with a prespecified processing route (industrial versus home-style preparation of ingredient-identical foods) so that texture and matrix vary by design.", "readout": "Continuous weighed intake, energy-intake rate, nutrient intake, appetite, body weight and fat mass, fecal energy, and metabolizable energy.", "critical_control": "Verify composition as consumed, not only as offered; preregister a smallest effect of interest so that a null result is informative.", "falsifying_result": "Equivalence of UPF and minimally processed diets at matched composition within the preregistered margin.", "supporting_result": "A replicated composition-independent effect on intake and fat mass, carried by the prespecified processing route." }, { "name": "Hall-menu causal mediation reconstruction", "experiment": "Rebuild the Hall menus in sequential versions that equalize non-beverage energy density, hyperpalatable combinations, texture/eating rate, protein share, fibre type, and carbohydrate-fat blend one block at a time, using randomized order and washout.", "readout": "Change in the original approximately 500-kcal/day intake contrast after each block, with meal-level mediation and body-composition follow-up.", "critical_control": "Keep variety, presentation, palatability, meal schedule, availability, and food mass prospectively specified; blind outcome analysis and use the same decision rules in an independent replication site.", "falsifying_result": "The diet effect collapses when named proximal properties are equated, leaving no reproducible residual associated with process history.", "supporting_result": "A substantial reproducible residual remains after every prospectively measured mediator is controlled." }, { "name": "Long-duration within-category negative-control trial", "experiment": "Randomize adults for 12-16 weeks to slow/low-energy-intake-rate UPF, fast/high-energy-intake-rate UPF, and matched minimally processed diets, with menus spanning multiple food families.", "readout": "Ad libitum intake, body fat, weight, adherence, diet acceptability, and durability of effects over time.", "critical_control": "Match macronutrients across texture arms (which earlier texture trials did not), include UPF foods predicted not to increase intake as internal negative controls, and prespecify whether NOVA status adds prediction beyond measured properties.", "falsifying_result": "Measured food properties predict intake and weight while NOVA category adds no out-of-sample information.", "supporting_result": "NOVA category retains a reproducible effect after food-property prediction and internal negative controls." } ], "knowledge_gaps": [ { "name": "Composition-matched test", "statement": "No peer-reviewed whole-diet trial yet matches composition as consumed while varying processing; the most direct contrast [18] is registry-posted and cannot exclude ~250 kcal/d.", "relevance": "This is SC4's boundary test." }, { "name": "Quantitative mediation", "statement": "The fractions of the Hall and UPDATE effects mediated by non-beverage energy density, hyperpalatable combinations, texture/eating rate, protein share, matrix, and food choice are unresolved.", "relevance": "Mediation, not category contrast alone, is required for mechanism-specific attribution." }, { "name": "Composition-pure texture manipulation", "statement": "Texture and eating-rate manipulations to date also changed macronutrient or sodium content [4,6].", "relevance": "Texture effects cannot yet be cleanly attributed." }, { "name": "Durable weight endpoint", "statement": "Most controlled trials last one meal to two weeks; the eight-week trial retained an energy-density difference with weight loss on both diets [3]; the 12-month trial achieved only a small exposure contrast [25].", "relevance": "Short relative changes do not establish durable obesity causation." }, { "name": "Weight at prescribed intake", "statement": "Whether UPF-associated weight gain at prescribed intake [19] reflects composition, metabolizable energy, water/glycogen, or adherence is unknown.", "relevance": "A non-intake route would bear on SC2 and on the claim's stated pathway." }, { "name": "Within-category heterogeneity and classification", "statement": "Within-category negative controls are underused and NOVA assignment is unreliable [27].", "relevance": "A category-wide causal claim requires reliable exposure assignment and mapped failures." }, { "name": "Population boundary", "statement": "Children, adolescents (exploratory signal [8]), older adults, diverse food systems, and people with metabolic disease are underrepresented.", "relevance": "Broad human generalization cannot be inferred from small adult feeding samples." } ], "review_metadata": { "review_date": "2026-09-30", "reviewer": "OpenAI Codex using HSEF v5.4 (original); Claude (Anthropic) cross-model red team and synthesis (2026-09-30)", "evidence_base_note": "Primary human feeding, mechanistic, and prospective evidence through 2026-09-30, including ClinicalTrials.gov posted results. Full text read wherever accessible; abstract-only and registry-only records are labeled. The HSEF v5.4 packet (SHA256 7E097D3D50D2068F68BDB1CD022DF4A045DAC46A9F25E51262EC7568E9DB11EB) was treated as methodology, not scientific evidence.", "literature_tools_available": true, "literature_tools_note": "Original: PubMed, PMC, Europe PMC, author manuscripts. Red-team: PubMed metadata for every PMID, PMC/Europe PMC full text, ClinicalTrials.gov v2 API, web search for currency. Scite unavailable in both passes; no citation-context counts used.", "skill_version": "5.4", "validator_run": "pass (validator 2.2, framework/schema/tools/validate_review.py; see companion validation report)", "validator_warnings_addressed": "See the companion validation report for this edition.", "original_review_date": "2026-09-29", "edition": "Red-Team Synthesis Edition (2026-09-30)", "consensus_consistency_audit": "2026-10-09; supplementary provenance/source reconciliation; original grading retained; prior grades visible." }, "is_calibration_entry": { "value": false, "entry_role": "not_calibration", "blind_graded": false, "completed_blinded_calibration": false, "expected_band_sealed_at": "", "target_sha256": "", "target_authorship": "No sealed calibration target was supplied.", "target_author_saw_review": null, "future_run_status": "not prepared", "unblinded_reason": "Standalone evaluation; no independently authored sealed expected band was supplied." }, "upper_ladder_check": { "final_level_below_zeta": true, "equivalents_considered": [ { "kind": "multiple_human_dietary_perturbations", "available": true, "why_not_sufficient": "Multiple trials establish bounded whole-diet and property effects, but they converge on energy density, nutrient combinations, and texture rather than on a composition-independent processing property; the composition-matched contrasts are null or non-significant [8,11,18,22]." }, { "kind": "multiple_pharmacological_classes", "available": false, "why_not_sufficient": "There is no pharmacological analogue that selectively reverses a category-wide process-history exposure while leaving food composition and behavior intact." }, { "kind": "dose_response_across_independent_populations", "available": true, "why_not_sufficient": "Prospective dose-response associations remain vulnerable to correlated nutritional, behavioral, social, and classification factors (NOVA κ ~0.3 [27]) and cannot separate processing from composition." }, { "kind": "natural_experiment_of_lifelong_exposure", "available": false, "why_not_sufficient": "No natural experiment isolates lifelong exposure to a defined processing operation independently of food system, diet quality, and socioeconomic context." }, { "kind": "withdrawal_rechallenge", "available": true, "why_not_sufficient": "Crossover diet switches show reversibility of the whole-diet package, but the shared causal property remains unresolved and several matched designs are null." } ], "convergence_present": false, "cap_reapplied_to_convergent_set": false }, "provenance_header": { "sources_total": 29, "claims_total": 35, "claims_verified": 35, "claims_unverified": 0, "claims_contradicted": 0, "tier_a_sources": 25, "generated_by": "OpenAI Codex with HSEF v5.4 (original); Claude (Anthropic) red-team synthesis 2026-09-30; validator 2.2", "gate_passed": true }, "claims": [ { "id": "C1", "text": "In Hall et al. 2019, 20 adults consumed 508 ± 106 kcal/day more during two weeks on an ultra-processed diet than during two weeks on an unprocessed diet; body weight changed by +0.9 kg versus -0.9 kg and body fat by +0.4 kg versus -0.3 kg.", "carries_number": true, "load_bearing_for": [ "SC1", "SC2", "SC3" ], "source_ns": [ 1 ], "verification": "verified", "verified_against": "complete corrected PMC full text, abstract, Results, figures, methods, and diet tables", "verified_on": "2026-09-29", "note": "This is the strongest causal whole-diet experiment, but it identifies the assigned diet package rather than a residual processing essence." }, { "id": "C2", "text": "Hall's menus were matched on offered calories, overall energy density, macronutrients, sugar, sodium, and fiber, but they differed in added-sugar fraction, insoluble-fiber fraction, saturated-fat fraction, omega-6:omega-3 ratio, non-beverage energy density, and eating rate; actual consumed energy density was 1.36 versus 1.09 kcal/g.", "carries_number": true, "load_bearing_for": [ "SC3", "SC4" ], "source_ns": [ 1, 16 ], "verification": "verified", "verified_against": "complete Hall full text and complete Forde 2020 full text", "verified_on": "2026-09-29", "note": "Matching selected aggregate nutrients as offered does not establish independence from all nutrient properties, food matrix, or energy-intake rate." }, { "id": "C3", "text": "In a nine-man Japanese randomized crossover, the one-week ultra-processed menu increased ad libitum intake by 813.5 kcal/day (95% CI 342.4 to 1284.7) and body weight by 1.1 kg relative to the less-processed menu, while chewing frequency was lower.", "carries_number": true, "load_bearing_for": [ "SC1", "SC2" ], "source_ns": [ 2 ], "verification": "verified", "verified_against": "PubMed abstract and bibliographic record", "verified_on": "2026-09-29", "note": "The tiny all-male, one-week study replicates the direction of the whole-diet effect but does not isolate nutrient-independent processing." }, { "id": "C4", "text": "In the eight-week UPDATE crossover, both healthy-guideline diets produced weight loss, but the minimally processed diet produced 1.01 percentage point more weight loss (95% CI 0.14 to 1.87) and 0.98 kg more fat-mass reduction; self-reported intake was 327.3 kcal/day lower, while corrected dietary energy density was 1.25 versus 1.60 kcal/g.", "carries_number": true, "load_bearing_for": [ "SC1", "SC2", "SC4", "SC5" ], "source_ns": [ 3 ], "verification": "verified", "verified_against": "complete PMC full text, Results, figures, supplement discussion, and publisher correction notice", "verified_on": "2026-09-29", "note": "The correct outcome is a relative advantage for the minimally processed condition; the ultra-processed condition did not cause absolute weight gain in this trial." }, { "id": "C5", "text": "In a randomized 2 x 2 one-day crossover (n=18) at approximately equal energy density, hard-textured menus reduced intake by 571 ± 135 kcal (33%) relative to soft menus; the main effect of processing on total energy intake including snacks was null (P=0.91), but the processing x texture interaction was significant (P=0.002).", "carries_number": true, "load_bearing_for": [ "SC4", "SC5" ], "source_ns": [ 4 ], "verification": "verified", "verified_against": "PMC full text, Table 4 (re-read in red-team pass)", "verified_on": "2026-09-30", "note": "This is a direct separator between processing category and texture/eating behavior, but it is a small one-day pilot." }, { "id": "C6", "text": "In a 50-person crossover lunch experiment, hard texture reduced food weight by 21% and energy intake by 26%; intake ranged from about 483 kcal for hard minimally processed meals to 789 kcal for soft ultra-processed meals, with residual energy-density differences also contributing.", "carries_number": true, "load_bearing_for": [ "SC4", "SC5" ], "source_ns": [ 5 ], "verification": "verified", "verified_against": "complete PMC full text, Results, tables, and figures", "verified_on": "2026-09-29", "note": "Texture and energy density jointly distinguish conditions; the experiment does not support a category-wide residual processing effect." }, { "id": "C7", "text": "In a 41-person 14-day crossover in which both diets were approximately 95% ultra-processed, a slower texture-derived eating-rate diet reduced intake by 369 kcal/day (95% CI 221 to 517); body fat mass decreased by 0.43 kg on the slower diet (diet x time P=0.0002) and body weight did not differ between diets.", "carries_number": true, "load_bearing_for": [ "SC2", "SC4", "SC5" ], "source_ns": [ 6 ], "verification": "verified", "verified_against": "PMC full text, Results and Figure 6 (re-read in red-team pass)", "verified_on": "2026-09-30", "note": "Large intake differences within the same NOVA category identify eating rate as a causal food property and demonstrate category heterogeneity." }, { "id": "C8", "text": "A secondary analysis of that same 14-day trial found that postprandial GLP-1, ghrelin, pancreatic polypeptide, and PYY responses to a fixed load did not differ between baseline (habitual low-UPF diet) and after either high-UPF diet, nor between the fast and slow diets; subjective satiety AUCs differed slightly from baseline without individual timepoint differences.", "carries_number": false, "load_bearing_for": [ "SC4" ], "source_ns": [ 7 ], "verification": "verified", "verified_against": "complete PMC full text, Results, figures, and Discussion", "verified_on": "2026-09-30", "note": "This is the same NCT06113146 cohort as source 6 and is not an independent causal line." }, { "id": "C9", "text": "After two 14-day eucaloric exposures matched on macronutrients, fiber, added sugar, diet quality, and energy density, 27 emerging adults showed no overall difference in subsequent buffet energy intake, food mass, ultra-processed-food selection, or eating rate.", "carries_number": true, "load_bearing_for": [ "SC4", "SC5" ], "source_ns": [ 8 ], "verification": "verified", "verified_against": "complete PMC full text, Results, tables, and subgroup cautions", "verified_on": "2026-09-29", "note": "The design tests acquired appetite or reward carryover after controlled exposure, not continuous ad libitum access during the assigned diets." }, { "id": "C10", "text": "In an acute matched-breakfast trial in 42 adults with obesity, the ultra-processed meal was eaten in 7 minutes 52 seconds versus 11 minutes 7 seconds, with less chewing and a smaller reduction in capacity to eat, but no overall metabolic-outcome difference.", "carries_number": true, "load_bearing_for": [ "SC4" ], "source_ns": [ 9 ], "verification": "verified", "verified_against": "complete PMC full text, Methods, Results, tables, and figures", "verified_on": "2026-09-29", "note": "The result supports oral-processing behavior as a mediator but is acute and does not establish later intake or weight gain." }, { "id": "C11", "text": "In one 19-person 2 × 2 preload trial, ultra-processed meals increased postprandial insulin (P=0.016), but the registered primary report found no main processing effect on later ad libitum dinner intake (difference -24.5 kcal, 95% CI -87.0 to 38.0) and no processing-by-energy-density interaction.", "carries_number": true, "load_bearing_for": [ "SC4", "SC5" ], "source_ns": [ 10, 11 ], "verification": "verified", "verified_against": "complete PMC full text for the hormonal secondary report and PubMed abstract for the primary intake report", "verified_on": "2026-09-29", "note": "Both reports are RBR-10gfr3fb and count as one experiment; a hormonal response cannot substitute for the null intake endpoint." }, { "id": "C12", "text": "A prespecified-order analysis of the Hall crossover found no significant diet-order effects on energy intake, body weight, or body composition for the ultra-processed versus minimally processed comparison.", "carries_number": false, "load_bearing_for": [ "SC1", "SC2" ], "source_ns": [ 12 ], "verification": "verified", "verified_against": "final PubMed record and abstract plus complete author preprint corresponding to the final analysis", "verified_on": "2026-09-29", "note": "This reduces one design concern but shares NCT03407053 participants with Hall 2019." }, { "id": "C13", "text": "A post hoc analysis of the Hall menus found that low-energy-density food mass and the balance of carbohydrate and fat predicted meal energy intake, reinforcing a food-property explanation without creating a new independent trial.", "carries_number": false, "load_bearing_for": [ "SC3", "SC4" ], "source_ns": [ 13 ], "verification": "verified", "verified_against": "complete PMC full text, Results, models, and limitations", "verified_on": "2026-09-29", "note": "Post hoc explanation is hypothesis-generating and is grouped with NCT03407053." }, { "id": "C14", "text": "In 110,260 NutriNet-Santé participants, each 10-percentage-point higher ultra-processed-food share was associated with incident overweight (HR 1.11, 95% CI 1.08 to 1.14) and obesity (HR 1.09, 95% CI 1.05 to 1.13), with acknowledged residual confounding and possible NOVA misclassification.", "carries_number": true, "load_bearing_for": [ "SC5" ], "source_ns": [ 14 ], "verification": "verified", "verified_against": "complete PMC full text, Results, sensitivity analyses, and limitations", "verified_on": "2026-09-29", "note": "Prospective association supports relevance but cannot isolate processing from correlated diet and social exposures." }, { "id": "C15", "text": "A 2026 systematic review and meta-analysis of randomized trials found that intake effects of UPF arms varied by analytic model, that weight gain and eating rate were significantly greater in UPF arms, that intake effects appeared confined to trials in which UPF arms had higher energy density, and that certainty was low.", "carries_number": false, "load_bearing_for": [ "SC5" ], "source_ns": [ 15 ], "verification": "verified", "verified_against": "PubMed abstract and bibliographic record", "verified_on": "2026-09-30", "note": "This Tier B synthesis checks field framing and heterogeneity; it does not fix the HSEF level." }, { "id": "C16", "text": "Across 327 measured foods, average energy-intake rate rose with NOVA processing category, but there was wide within-category variability, so NOVA status is an imperfect proxy for the food properties most proximal to intake.", "carries_number": true, "load_bearing_for": [ "SC4", "SC5" ], "source_ns": [ 16 ], "verification": "verified", "verified_against": "complete PMC full text, pooled methods, Results, and limitations", "verified_on": "2026-09-29", "note": "This is a pooled food-property analysis, not a randomized weight-gain experiment." }, { "id": "C17", "text": "NOVA defines ultra-processed foods by industrial formulations and processes rather than by a single chemical or physical exposure; the category therefore does not by itself specify the causal property that a mechanistic claim must isolate.", "carries_number": false, "load_bearing_for": [ "SC4", "SC5" ], "source_ns": [ 17 ], "verification": "verified", "verified_against": "PubMed bibliographic record, abstract-level metadata, and the operational definition summarized in opened primary and mechanistic reports", "verified_on": "2026-09-29", "note": "This is conceptual background rather than causal evidence." }, { "id": "C18", "text": "The evidence establishes that some ultra-processed whole-diet packages can causally raise ad libitum intake and relative short-term weight or fat gain, but it does not establish that the effect persists when nutrient composition is matched or that processing-conferred properties characteristic of the category carry it; energy density, hyperpalatable combinations, texture/eating rate, and composition remain the better-supported explanations.", "carries_number": false, "load_bearing_for": [ "SC1", "SC2", "SC3", "SC4", "SC5" ], "source_ns": [ 1, 2, 3, 4, 5, 6, 8, 9, 10, 11, 13, 16, 18, 21, 22, 23, 24 ], "verification": "verified", "verified_against": "cross-study synthesis of the opened randomized and controlled primary records", "verified_on": "2026-09-30", "note": "This is the score-fixing synthesis and distinguishes whole-diet causality from the stronger independence claim." }, { "id": "C19", "text": "In NCT05290064 (36 analysed adults; four 1-week inpatient ad libitum diets), least-squares mean intake was 3424.3 (UPF high energy density/high hyperpalatable), 3265.9 (UPF high/low), 2604.4 (UPF low/low), and 2476.0 kcal/d (unprocessed low/low); UPF-HH minus UNF-LL +948.4 kcal/d (95% CI 811.9 to 1084.8; P<0.0001); UPF-HL minus UPF-LL +661.6 (524.9 to 798.3); UPF-HH minus UPF-HL +158.4 (21.9 to 294.9; P=0.023); UPF-LL minus UNF-LL +128.4 (-8.2 to 265.0; P=0.065); meal eating rate and palatability did not differ.", "carries_number": true, "load_bearing_for": [ "SC1", "SC4", "SC5" ], "source_ns": [ 18 ], "verification": "verified", "verified_against": "ClinicalTrials.gov posted results module (v2 API), outcome measures and statistical analyses", "verified_on": "2026-09-30", "note": "Registry-posted summary results (2026-08-12), not peer reviewed; the level does not rest on this claim." }, { "id": "C20", "text": "In a controlled 2 x 2 crossover trial in young men with prescribed (not ad libitum) diets, UPF compared with unprocessed diets increased body weight independent of caloric load.", "carries_number": false, "load_bearing_for": [ "SC2" ], "source_ns": [ 19 ], "verification": "verified", "verified_against": "PubMed abstract; ClinicalTrials.gov NCT05368194 design", "verified_on": "2026-09-30", "note": "Abstract-level; magnitudes and composition matching not verified and not used." }, { "id": "C21", "text": "Across 2,733 inpatient meals from four dietary patterns, energy density, eating rate, and hyperpalatable foods were consistently positively related to meal energy intake.", "carries_number": true, "load_bearing_for": [ "SC4", "SC5" ], "source_ns": [ 20 ], "verification": "verified", "verified_against": "PubMed abstract", "verified_on": "2026-09-30", "note": "Meal-level association within randomized trials, not a randomized manipulation of each property." }, { "id": "C22", "text": "In 21 adults over 54 hours in a whole-room calorimeter, a high-protein (30%), lower-carbohydrate (29%) UPF diet versus a normal-protein (13%), normal-carbohydrate (46%) UPF diet matched for calories, fat, and fibre lowered ad libitum intake by 196 ± 396 kcal/d and raised energy expenditure by 128 ± 98 kcal/d.", "carries_number": true, "load_bearing_for": [ "SC4" ], "source_ns": [ 21 ], "verification": "verified", "verified_against": "PubMed abstract", "verified_on": "2026-09-30", "note": "Composition changes intake within the UPF category." }, { "id": "C23", "text": "In 40 young adults eating single meals matched for energy, macronutrients, flavour profile, and texture, intake was 585.7 kcal (non-UPF, high nutritional quality) versus 609.7 kcal (UPF, high nutritional quality) and 503.0 kcal for a less palatable UPF meal of low nutritional quality.", "carries_number": true, "load_bearing_for": [ "SC4" ], "source_ns": [ 22 ], "verification": "verified", "verified_against": "PubMed abstract", "verified_on": "2026-09-30", "note": "Single-meal design." }, { "id": "C24", "text": "In 69 adults eating sandwiches made of 67-98% ultra-processed ingredients, energy density and eating rate interacted (P=0.024): the slow/low-energy-density meal (570 kcal) was 573 kcal (50%) lower than the fast/high-energy-density meal (1143 kcal).", "carries_number": true, "load_bearing_for": [ "SC5" ], "source_ns": [ 23 ], "verification": "verified", "verified_against": "PubMed abstract", "verified_on": "2026-09-30", "note": "Within-UPF heterogeneity driven by proximal properties." }, { "id": "C25", "text": "In 43 men eating fixed-portion lunches, the hard UPF meal (25.9 g/min) was eaten about half as fast as the soft UPF meal (61.2 g/min), produced 32% greater fullness iAUC, and was followed by the lowest later snack intake (-118 kcal; texture x processing P=0.004).", "carries_number": true, "load_bearing_for": [ "SC5" ], "source_ns": [ 24 ], "verification": "verified", "verified_against": "PubMed abstract", "verified_on": "2026-09-30", "note": "UPF classification did not by itself imply lower satiety." }, { "id": "C26", "text": "In a 12-month randomized trial of 148 adults with obesity, adding UPF restriction to energy restriction lowered UPF share to 13.86% versus 20.02% (p=0.08) and final weight was 82.9 versus 86.3 kg (p=0.01), a difference the authors judged not clinically significant.", "carries_number": true, "load_bearing_for": [ "SC2" ], "source_ns": [ 25 ], "verification": "verified", "verified_against": "PubMed abstract", "verified_on": "2026-09-30", "note": "Longest randomized UPF-reduction evidence; small exposure contrast." }, { "id": "C27", "text": "In 50 adults, an ultra-processed high-fat, high-sugar milkshake produced no significant post-ingestive striatal dopamine response (p=0.62), and responses were unrelated to BMI.", "carries_number": true, "load_bearing_for": [ "SC4" ], "source_ns": [ 26 ], "verification": "verified", "verified_against": "PubMed abstract", "verified_on": "2026-09-30", "note": "Null for one proposed reward-based mechanism." }, { "id": "C28", "text": "Food and nutrition specialists assigning foods to NOVA groups showed low agreement: Fleiss κ 0.32 for 120 marketed foods (159 evaluators) and 0.34 for 111 generic foods (177 evaluators).", "carries_number": true, "load_bearing_for": [ "SC5" ], "source_ns": [ 27 ], "verification": "verified", "verified_against": "PMC full text, Results", "verified_on": "2026-09-30", "note": "Bears on exposure misclassification." }, { "id": "C29", "text": "A 2025 proponents' synthesis concluded that the totality of the evidence supports displacement of traditional diets by UPF-based dietary patterns as a key driver of diet-related chronic disease, attributing overeating to high energy density, hyper-palatability, soft texture, and disrupted food matrices.", "carries_number": false, "load_bearing_for": [ "SC5" ], "source_ns": [ 28 ], "verification": "verified", "verified_against": "PubMed abstract", "verified_on": "2026-09-30", "note": "Tier B; used for field framing and for the fair reading of 'properties of processing'." }, { "id": "C30", "text": "A 2026 critical review concluded that the available experimental evidence does not establish a causal role of UPF consumption in obesity, citing methodological limitations and imprecision of the Nova classification.", "carries_number": false, "load_bearing_for": [ "SC5" ], "source_ns": [ 29 ], "verification": "verified", "verified_against": "PubMed abstract", "verified_on": "2026-09-30", "note": "Tier B; used for field framing only." }, { "id": "C31", "text": "In the same one-day 2 x 2 trial, including snacks, soft (ultra-)processed menus gave the highest intake (2608 ± 136 kcal), significantly above soft unprocessed (2205 ± 137 kcal), and hard (ultra-)processed the lowest (1706 ± 139 kcal); soft menus provided about 40-43% of energy as fat versus 24-27% for hard menus, and (ultra-)processed menus offered 4681-6184 mg sodium versus 440-640 mg.", "carries_number": true, "load_bearing_for": [ "SC4", "SC5" ], "source_ns": [ 4 ], "verification": "verified", "verified_against": "PMC full text, Tables 3-4", "verified_on": "2026-09-30", "note": "Processing effect changes sign with texture; arms not composition-pure." }, { "id": "C32", "text": "In the 14-day within-UPF trial, the served Slow-ER diet provided 22% of energy as fat and 22% as protein versus 34% and 18% for the Fast-ER diet; the authors report the intake effect was not attributable to macronutrient intake after adjustment.", "carries_number": true, "load_bearing_for": [ "SC5" ], "source_ns": [ 6 ], "verification": "verified", "verified_against": "PMC full text, Table 1 and Results", "verified_on": "2026-09-30", "note": "Eating-rate manipulation not composition-pure." }, { "id": "C33", "text": "In Hall 2019, energy intake declined linearly during the ultra-processed diet (-25.5 ± 6.4 kcal/d per day, p<0.0001) but not during the unprocessed diet (-7.7 ± 6.4, p=0.23); offered meal non-beverage energy density was 1.957 versus 1.057 kcal/g; the authors estimated protein leverage could explain at most ~50% of the intake difference.", "carries_number": true, "load_bearing_for": [ "SC1", "SC3", "SC4" ], "source_ns": [ 1 ], "verification": "verified", "verified_against": "PMC full text, Results, Table 1, Discussion", "verified_on": "2026-09-30", "note": "Bears on durability and compositional explanations." }, { "id": "C34", "text": "UPDATE was free-living with home-delivered foods; 90.9% of randomized participants were female; food diaries were returned by 32 (MPF) and 35 (UPF) of 50 ITT participants; the body-composition-derived energy-imbalance difference was -170.4 kcal/d (s.e. 57.9).", "carries_number": true, "load_bearing_for": [ "SC1", "SC2" ], "source_ns": [ 3 ], "verification": "verified", "verified_against": "PMC full text, Results", "verified_on": "2026-09-30", "note": "Self-reported intake is incomplete; composition-derived estimate is less biased." }, { "id": "C35", "text": "In the Rego trial, an exploratory diet x age interaction (p<0.001) showed higher total intake after the UPF diet in late adolescents aged 18-21 (p=0.03, d=0.79) but not in young adults.", "carries_number": true, "load_bearing_for": [ "SC5" ], "source_ns": [ 8 ], "verification": "verified", "verified_against": "PubMed abstract; PMC full text", "verified_on": "2026-09-30", "note": "Exploratory, hypothesis-generating." } ], "citation_ledger": [ { "n": 1, "citation": "Hall KD, Ayuketah A, Brychta R, et al. Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized Controlled Trial of Ad Libitum Food Intake. Cell Metabolism. 2019;30(1):67-77.e3.", "tier": "A", "kind": "rct", "pmid": "31105044", "pmcid": "PMC7946062", "doi": "10.1016/j.cmet.2019.05.008", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC7946062/", "status_flags": [ "erratum PMID 31269427", "erratum PMID 33027677" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT03407053", "access_depth": "full_text_complete", "access_note": "Complete PMC author manuscript, methods, diet tables, results, figures, discussion, and attached errata read.", "carries_claims": [ "C1", "C2", "C18", "C33" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 2, "citation": "Hamano S, Sawada M, Aihara M, et al. Ultra-processed foods cause weight gain and increased energy intake associated with reduced chewing frequency: A randomized, open-label, crossover study. Diabetes, Obesity and Metabolism. 2024;26(11):5431-5443.", "tier": "A", "kind": "rct", "pmid": "39267249", "doi": "10.1111/dom.15922", "url": "https://pubmed.ncbi.nlm.nih.gov/39267249/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "HAMANO-2024-JAPAN-CROSSOVER", "access_depth": "abstract_metadata", "access_note": "PubMed abstract and complete bibliographic record read; publisher full text was not accessible in this review environment.", "carries_claims": [ "C3", "C18" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 3, "citation": "Dicken SJ, Jassil FC, Brown A, et al. Ultraprocessed or minimally processed diets following healthy dietary guidelines on weight and cardiometabolic health: a randomized, crossover trial. Nature Medicine. 2025;31(10):3297-3308.", "tier": "A", "kind": "rct", "pmid": "40760353", "pmcid": "PMC12532614", "doi": "10.1038/s41591-025-03842-0", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC12532614/", "status_flags": [ "publisher correction to Supplementary Information, 2025-12-15, including dietary energy-density values" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT05627570", "access_depth": "full_text_complete", "access_note": "Complete PMC article, supplement discussion, outcomes, limitations, and publisher correction read.", "carries_claims": [ "C4", "C18", "C34" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 4, "citation": "Lasschuijt M, Camps G, Mars M, et al. Speed limits: the effects of industrial food processing and food texture on daily energy intake and eating behaviour in healthy adults. European Journal of Nutrition. 2023;62(7):2949-2962.", "tier": "A", "kind": "rct", "pmid": "37452167", "pmcid": "PMC10469122", "doi": "10.1007/s00394-023-03202-z", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC10469122/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT04280146", "access_depth": "full_text_complete", "access_note": "Complete PMC article, 2 x 2 design, primary intake results, tables, figures, and limitations read.", "carries_claims": [ "C5", "C18", "C31" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 5, "citation": "Teo PS, Lim AJ, Goh AT, et al. Texture-based differences in eating rate influence energy intake for minimally processed and ultra-processed meals. American Journal of Clinical Nutrition. 2022;116(1):244-254.", "tier": "A", "kind": "rct", "pmid": "35285882", "pmcid": "PMC9257473", "doi": "10.1093/ajcn/nqac068", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC9257473/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT04589221", "access_depth": "full_text_complete", "access_note": "Complete PMC article, meal composition, outcomes, figures, and residual energy-density limitation read.", "carries_claims": [ "C6", "C18" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 6, "citation": "Forde CG, Heuven LA, van Bruinessen M, et al. Eating rate has sustained effects on energy intake from ultraprocessed diets: a 2-week ad libitum dietary randomized controlled crossover trial. American Journal of Clinical Nutrition. 2026;123(4):101122.", "tier": "A", "kind": "rct", "pmid": "41314613", "pmcid": "PMC13084570", "doi": "10.1016/j.ajcnut.2025.11.012", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13084570/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT06113146", "access_depth": "full_text_complete", "access_note": "Complete PMC article, protocol, outcomes, tables, figures, and limitations read.", "carries_claims": [ "C7", "C18", "C32" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 7, "citation": "van Bruinessen M, Liu Z, Heuven LAJ, et al. Satiety responsiveness following dietary exposure to ultra-processed foods differing in texture-derived eating rate: a secondary analysis of a randomized controlled crossover diet intervention. European Journal of Nutrition. 2026;65(7).", "tier": "A", "kind": "secondary analysis", "pmid": "42782361", "pmcid": "PMC13612719", "doi": "10.1007/s00394-026-04111-7", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13612719/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT06113146", "access_depth": "full_text_complete", "access_note": "Complete PMC secondary analysis, appetite curves, hormone results, and limitations read; grouped with source 6.", "carries_claims": [ "C8" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 8, "citation": "Rego MLM, Leslie E, Schmall E, et al. The Influence of Ultraprocessed Food Consumption on Energy Intake in Emerging Adulthood: A Controlled Feeding Trial. Obesity. 2026;34(2):344-356.", "tier": "A", "kind": "rct", "pmid": "41255123", "pmcid": "PMC12850616", "doi": "10.1002/oby.70086", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC12850616/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT05550818", "access_depth": "full_text_complete", "access_note": "Complete PMC article, matched-diet methods, primary null outcomes, exploratory subgroup, and limitations read.", "carries_claims": [ "C9", "C18", "C35" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 9, "citation": "Galdino-Silva MB, Almeida KMM, Oliveira ADS, et al. A Meal with Ultra-Processed Foods Leads to a Faster Rate of Intake and to a Lesser Decrease in the Capacity to Eat When Compared to a Similar, Matched Meal Without Ultra-Processed Foods. Nutrients. 2024;16(24):4398.", "tier": "A", "kind": "rct", "pmid": "39771019", "pmcid": "PMC11676177", "doi": "10.3390/nu16244398", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC11676177/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "GALDINO-SILVA-2024-ACUTE-MEAL", "access_depth": "full_text_complete", "access_note": "Complete PMC article, meal matching, eating-behavior results, metabolic results, and limitations read.", "carries_claims": [ "C10", "C18" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 10, "citation": "Gama IRS, Paula DTDC, Soares DB, et al. Effects of Ultra-Processed Meals with Differing Energy Density on Postprandial Glyco-Insulinemic and Appetite-Related Hormonal Responses: A Randomized Crossover Trial. Nutrients. 2026;18(15):2534.", "tier": "A", "kind": "secondary analysis", "pmid": "42588167", "pmcid": "PMC13468057", "doi": "10.3390/nu18152534", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13468057/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "RBR-10gfr3fb", "access_depth": "full_text_complete", "access_note": "Complete PMC secondary hormonal report, factorial design, results, and limitations read; grouped with source 11.", "carries_claims": [ "C11", "C18" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 11, "citation": "Barros LM, Pereira MR, Peixoto VA, et al. Effect of ultra-processed-rich versus ultra-processed-free preload meals at high and low energy density on ad libitum energy intake: a randomized crossover trial. Appetite. 2026;228:108761.", "tier": "A", "kind": "rct", "pmid": "42633852", "doi": "10.1016/j.appet.2026.108761", "url": "https://pubmed.ncbi.nlm.nih.gov/42633852/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "RBR-10gfr3fb", "access_depth": "abstract_metadata", "access_note": "PubMed abstract and complete bibliographic record read; full publisher text was not accessible in this review environment.", "carries_claims": [ "C11", "C18" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 12, "citation": "Sciarrillo CM, Guo J, Hengist A, et al. Diet order significantly affects energy balance for diets varying in macronutrients but not ultraprocessing in crossover studies without a washout period. American Journal of Clinical Nutrition. 2024;120(4):953-963.", "tier": "A", "kind": "secondary analysis", "pmid": "39163976", "pmcid": "PMC11473439", "doi": "10.1016/j.ajcnut.2024.08.013", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC11473439/", "status_flags": [ "updates medRxiv PMID 37986904" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT03407053", "access_depth": "full_text_partial", "access_note": "Final PubMed record and abstract read together with the complete author preprint that the final article updates; grouped with source 1.", "carries_claims": [ "C12" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 13, "citation": "Brunstrom JM, Schatzker M, Rogers PJ, et al. Consuming an unprocessed diet reduces energy intake: a post-hoc analysis of a randomized controlled trial reveals a role for human nutritional intelligence. American Journal of Clinical Nutrition. 2026;123(3):101183.", "tier": "A", "kind": "secondary analysis", "pmid": "41475551", "pmcid": "PMC12975374", "doi": "10.1016/j.ajcnut.2025.101183", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC12975374/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NCT03407053", "access_depth": "full_text_complete", "access_note": "Complete PMC post hoc analysis, meal-property models, Results, and limitations read; grouped with source 1.", "carries_claims": [ "C13", "C18" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 14, "citation": "Beslay M, Srour B, Méjean C, et al. Ultra-processed food intake in association with BMI change and risk of overweight and obesity: A prospective analysis of the French NutriNet-Santé cohort. PLOS Medicine. 2020;17(8):e1003256.", "tier": "A", "kind": "prospective cohort", "pmid": "32853224", "pmcid": "PMC7451582", "doi": "10.1371/journal.pmed.1003256", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC7451582/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "NUTRINET-SANTE-NCT03335644", "access_depth": "full_text_complete", "access_note": "Complete PMC article, cohort methods, adjusted results, sensitivity analyses, and limitations read.", "carries_claims": [ "C14" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 15, "citation": "Robinson E, Jones A, Evans R, et al. A systematic review and meta-analysis of randomized controlled trials examining the effect of ultra-processed food on energy intake and weight gain. Critical Reviews in Food Science and Nutrition. 2026:1-16.", "tier": "B", "kind": "systematic review", "pmid": "42746678", "doi": "10.1080/10408398.2026.2731838", "url": "https://pubmed.ncbi.nlm.nih.gov/42746678/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "ROBINSON-2026-UPF-RCT-META", "access_depth": "abstract_metadata", "access_note": "PubMed abstract and complete bibliographic record read; used for synthesis and divergence only, not to fix a causal level.", "carries_claims": [ "C15" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 16, "citation": "Forde CG, Mars M, de Graaf K. Ultra-Processing or Oral Processing? A Role for Energy Density and Eating Rate in Moderating Energy Intake from Processed Foods. Current Developments in Nutrition. 2020;4(3):nzaa019.", "tier": "A", "kind": "primary pooled food-property analysis", "pmid": "32110771", "pmcid": "PMC7042610", "doi": "10.1093/cdn/nzaa019", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC7042610/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "FORDE-2020-327-FOODS", "access_depth": "full_text_complete", "access_note": "Complete PMC article, pooled-food methods, category distributions, and limitations read; contextual mechanism only.", "carries_claims": [ "C2", "C16", "C18" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 17, "citation": "Monteiro CA, Cannon G, Levy RB, et al. Ultra-processed foods: what they are and how to identify them. Public Health Nutrition. 2019;22(5):936-941.", "tier": "B", "kind": "concept definition", "pmid": "30744710", "pmcid": "PMC10260459", "doi": "10.1017/S1368980018003762", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC10260459/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-29", "status_checked_on": "2026-09-30", "study_id": "MONTEIRO-2019-NOVA-DEFINITION", "access_depth": "abstract_metadata", "access_note": "PubMed bibliographic record and definition-level metadata read; complete PMC retrieval was blocked by the archive endpoint in this environment. Used only for category definition.", "carries_claims": [ "C17" ], "redteam_check": "Identifier, title, first author, year, journal re-verified 2026-09-30 via PubMed; no retraction or expression of concern found." }, { "n": 18, "citation": "National Institutes of Health Clinical Center / NIDDK. Effect of Ultra-processed Versus Unprocessed Diets on Energy Metabolism (Protocol 22DK0002). ClinicalTrials.gov NCT05290064; summary results first posted 2026-08-12.", "tier": "A", "kind": "registry posted results", "pmid": "", "doi": "", "url": "https://clinicaltrials.gov/study/NCT05290064", "status_flags": [ "not_peer_reviewed: summary results posted to ClinicalTrials.gov 2026-08-12 (QC-reviewed by registry); no journal report found as of 2026-09-30" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "NCT05290064", "access_depth": "registry_results_complete", "access_note": "Complete posted results module (participant flow, baseline, outcome measures, statistical analyses) read via ClinicalTrials.gov v2 API.", "carries_claims": [ "C19", "C18" ], "redteam_check": "Added in red-team pass 2026-09-30; identifier and bibliographic fields verified.", "registry_id": "NCT05290064" }, { "n": 19, "citation": "Preston JM, Iversen J, Hufnagel A, et al. Effect of ultra-processed food consumption on male reproductive and metabolic health. Cell Metabolism. 2025;37(10):1950-1960.e2.", "tier": "A", "kind": "rct", "pmid": "40882621", "doi": "10.1016/j.cmet.2025.08.004", "url": "https://pubmed.ncbi.nlm.nih.gov/40882621/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "NCT05368194", "access_depth": "abstract_metadata", "access_note": "PubMed abstract and registry record NCT05368194 read; publisher full text returned HTTP 403.", "carries_claims": [ "C20" ], "redteam_check": "Added in red-team pass 2026-09-30; identifier and bibliographic fields verified." }, { "n": 20, "citation": "Fazzino TL, Courville AB, Guo J, Hall KD. Ad libitum meal energy intake is positively influenced by energy density, eating rate and hyper-palatable food across four dietary patterns. Nature Food. 2023;4(2):144-147.", "tier": "A", "kind": "secondary analysis", "pmid": "37117850", "doi": "10.1038/s43016-022-00688-4", "url": "https://pubmed.ncbi.nlm.nih.gov/37117850/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "access_depth": "abstract_metadata", "access_note": "PubMed abstract and bibliographic record read.", "carries_claims": [ "C21" ], "redteam_check": "Added in red-team pass 2026-09-30; identifier and bibliographic fields verified." }, { "n": 21, "citation": "Hägele FA, Herpich C, Koop J, et al. Short-term effects of high-protein, lower-carbohydrate ultra-processed foods on human energy balance. Nature Metabolism. 2025;7(4):704-713.", "tier": "A", "kind": "rct", "pmid": "40082711", "pmcid": "PMC12021659", "doi": "10.1038/s42255-025-01247-4", "url": "https://pubmed.ncbi.nlm.nih.gov/40082711/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "NCT05337007", "access_depth": "abstract_metadata", "access_note": "PubMed abstract read; PMC full text exists (PMC12021659) but claims are limited to the abstract.", "carries_claims": [ "C22", "C18" ], "redteam_check": "Added in red-team pass 2026-09-30; identifier and bibliographic fields verified." }, { "n": 22, "citation": "Larcom AB, Lofgren IE, Delmonico MJ, et al. Effects of processing level and nutritional quality on energy intake, ingestive behaviors, palatability, and satiety in a young adult population. Appetite. 2026;217:108352.", "tier": "A", "kind": "rct", "pmid": "41115473", "doi": "10.1016/j.appet.2025.108352", "url": "https://pubmed.ncbi.nlm.nih.gov/41115473/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "access_depth": "abstract_metadata", "access_note": "PubMed abstract read.", "carries_claims": [ "C23", "C18" ], "redteam_check": "Added in red-team pass 2026-09-30; identifier and bibliographic fields verified." }, { "n": 23, "citation": "Lasschuijt MP, Heuven LA, Gonzalez-Estanol K, et al. The Effect of Energy Density and Eating Rate on Ad Libitum Energy Intake in Healthy Adults-A Randomized Controlled Study. Journal of Nutrition. 2025;155(8):2602-2610.", "tier": "A", "kind": "rct", "pmid": "40516651", "doi": "10.1016/j.tjnut.2025.06.006", "url": "https://pubmed.ncbi.nlm.nih.gov/40516651/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "study_id": "NCT05659771", "access_depth": "abstract_metadata", "access_note": "PubMed abstract read.", "carries_claims": [ "C24", "C18" ], "redteam_check": "Added in red-team pass 2026-09-30; identifier and bibliographic fields verified." }, { "n": 24, "citation": "Teo PS, Lim AJ, Goh AT, et al. Impact of food texture and degree of food processing on post-meal satiety and later snack intake. Appetite. 2026;225:108604.", "tier": "A", "kind": "rct", "pmid": "42176807", "doi": "10.1016/j.appet.2026.108604", "url": "https://pubmed.ncbi.nlm.nih.gov/42176807/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "access_depth": "abstract_metadata", "access_note": "PubMed abstract read.", "carries_claims": [ "C25", "C18" ], "redteam_check": "Added in red-team pass 2026-09-30; identifier and bibliographic fields verified." }, { "n": 25, "citation": "Macena ML, Pereira MR, Silva DR, et al. Effectiveness and metabolic impacts of restricting the consumption of ultra-processed foods in individuals with obesity submitted to energy restriction: a randomized clinical trial. Nutrition, Metabolism and Cardiovascular Diseases. 2026;36(5):104426.", "tier": "A", "kind": "rct", "pmid": "41381310", "doi": "10.1016/j.numecd.2025.104426", "url": "https://pubmed.ncbi.nlm.nih.gov/41381310/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "access_depth": "abstract_metadata", "access_note": "PubMed abstract read.", "carries_claims": [ "C26" ], "redteam_check": "Added in red-team pass 2026-09-30; identifier and bibliographic fields verified." }, { "n": 26, "citation": "Darcey VL, Guo J, Chi M, et al. Brain dopamine responses to ultra-processed milkshakes are highly variable and not significantly related to adiposity in humans. Cell Metabolism. 2025;37(3):616-628.e5.", "tier": "A", "kind": "mechanistic human study", "pmid": "40043691", "doi": "10.1016/j.cmet.2025.02.002", "url": "https://pubmed.ncbi.nlm.nih.gov/40043691/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "access_depth": "abstract_metadata", "access_note": "PubMed abstract read.", "carries_claims": [ "C27" ], "redteam_check": "Added in red-team pass 2026-09-30; identifier and bibliographic fields verified." }, { "n": 27, "citation": "Braesco V, Souchon I, Sauvant P, et al. Ultra-processed foods: how functional is the NOVA system? European Journal of Clinical Nutrition. 2022;76(9):1245-1253.", "tier": "A", "kind": "classification reliability study", "pmid": "35314769", "pmcid": "PMC9436773", "doi": "10.1038/s41430-022-01099-1", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC9436773/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "access_depth": "full_text_complete", "access_note": "PMC full text read (Methods, Results).", "carries_claims": [ "C28" ], "redteam_check": "Added in red-team pass 2026-09-30; identifier and bibliographic fields verified." }, { "n": 28, "citation": "Monteiro CA, Louzada MLC, Steele-Martinez E, et al. Ultra-processed foods and human health: the main thesis and the evidence. Lancet. 2025;406(10520):2667-2684.", "tier": "B", "kind": "narrative and systematic synthesis", "pmid": "41270766", "doi": "10.1016/S0140-6736(25)01565-X", "url": "https://pubmed.ncbi.nlm.nih.gov/41270766/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "access_depth": "abstract_metadata", "access_note": "PubMed abstract read; used for field framing only.", "carries_claims": [ "C29" ], "redteam_check": "Added in red-team pass 2026-09-30; identifier and bibliographic fields verified." }, { "n": 29, "citation": "Ludwig DS. Is the experimental evidence on ultra-processed food and obesity reliable? Critical Reviews in Food Science and Nutrition. 2026;66(28):5791-5795.", "tier": "B", "kind": "critical review", "pmid": "42063358", "doi": "10.1080/10408398.2026.2664043", "url": "https://pubmed.ncbi.nlm.nih.gov/42063358/", "status_flags": [ "none_found" ], "opened": true, "opened_on": "2026-09-30", "status_checked_on": "2026-09-30", "access_depth": "abstract_metadata", "access_note": "PubMed abstract read; used for field framing only.", "carries_claims": [ "C30" ], "redteam_check": "Added in red-team pass 2026-09-30; identifier and bibliographic fields verified." } ], "independence_basis": { "required": false, "lines": [], "shared_failure_mode_checked": false, "note": "No subclaim or final level is Ζ or above. Shared trials are grouped by study_id: sources 1/12/13, 6/7, and 10/11 each represent one participant experiment. Red-team: NCT05290064 [18] is a separate cohort from the same research group as [1]; lab-network overlap recorded under Pitfall 9." }, "redteam_synthesis": { "date": "2026-09-30", "red_team_model": "Claude (Anthropic), cross-model red team; authoring model(s) differed", "inputs": [ "ultra-processed-food-processing-independence-hsef-v54-review.md", "ultra-processed-food-processing-independence-hsef-v54-scorecard.json", "ultra-processed-food-processing-independence-hsef-v54-validation.txt", "README-ultra-processed-food-processing-independence-hsef-v54.txt" ], "mode": "single-version red team", "findings_total": 25, "findings_accepted": 20, "findings_rejected": 5, "findings_by_severity": { "Critical": 0, "Major": 7, "Minor": 12, "Editorial": 1 }, "score_before": { "level": "Γ", "confidence": "High", "subclaims": { "SC1": "Ε", "SC2": "Ε", "SC3": "Ε", "SC4": "Γ", "SC5": "Γ" }, "role": "reinforcing / stabilizing factor", "momentum": "Mechanistic only", "stability": "Stable", "divergence": "Severe overstatement" }, "score_after": { "level": "Γ", "confidence": "Moderate", "subclaims": { "SC1": "Ε", "SC2": "Ε", "SC3": "Ε", "SC4": "Γ (re-scoped)", "SC5": "Γ (re-scoped)" }, "role": "reinforcing / stabilizing factor", "momentum": "Decisive work underway", "stability": "Stable", "divergence": "Contested" }, "superseded_after_this_pass": "Momentum was returned to 'Mechanistic only' by the 2026-10-14 adversarial review (lead-adjudicated finding E2); this block records the state as at the 2026-09-30 red-team synthesis, not the current value. See extended_scoring_MSD for the live value.", "grade_changed": false, "grade_change_rationale": "Letter unchanged at Γ. It holds under both the original residual-essence reading and the corrected composition-independence reading: the most direct composition-matched contrast is non-significant and registry-only [18], matched designs are null [8,11,22], composition changes intake within UPF [21], and processing-conferred properties act in both directions within NOVA 4 [4,6,23,24]. Confidence moved High -> Moderate because the framework's High definition ('extensive independent replication ... reproducibility well-established') is not met by small, short, partly composition-confounded trials concentrated in a few research networks.", "citations_checked": 29, "citations_corrected": 0, "citations_removed": 0, "citations_added": 12, "audit_trail": "redteam/REDTEAM_REPORT.md; redteam/SYNTHESIS_CHANGE_LOG.md" }, "consensus_validation": { "completed": true, "status": "documented_historical_discovery_and_reconciliation", "provider": "Consensus", "provider_mode": "Historical browser/connector use as recorded in the source artifacts", "local_audit_date": "2026-10-09", "primary_comparison_count": null, "reported_usage": [], "grade_changed": false, "score_before": { "hsef_level": "Γ", "hsef_descriptor": "Mechanistically plausible, partial system support", "confidence": "Moderate", "confidence_rationale": "Multiple independent controlled human interventions exist, but they are small (9-69 participants), short (one meal to eight weeks), partly composition-confounded, concentrated in a few research networks, and in one key case registry-only [18]; several nulls cannot exclude modest effects. This matches the framework's Moderate definition ('Multiple independent studies; some replication; some assay-level concerns or methodological heterogeneity remains'), not High ('Extensive independent replication ... reproducibility well-established'). Confidence measures evidence density, not certainty that the strong claim fails (red-team RT-03; original: High).", "rests_on_non_peer_reviewed_work": false, "rationale": "Some ultra-processed whole-diet packages causally increase energy intake and relative short-term weight or fat gain, reaching Ε for those bounded subclaims. The submitted proposition additionally requires that the effect persist when nutrient composition is matched and be carried by processing-conferred properties characteristic of the category; those clauses remain Γ because the most direct composition-matched contrast is non-significant [18], matched designs are null [8,11,22], composition changes intake within UPF [21], and processing-conferred properties act in both directions within NOVA 4 [4,6,23,24]. The level does not rest on the registry-only result [18]." }, "score_after": { "hsef_level": "Γ", "hsef_descriptor": "Mechanistically plausible, partial system support", "confidence": "Moderate", "confidence_rationale": "Multiple independent controlled human interventions exist, but they are small (9-69 participants), short (one meal to eight weeks), partly composition-confounded, concentrated in a few research networks, and in one key case registry-only [18]; several nulls cannot exclude modest effects. This matches the framework's Moderate definition ('Multiple independent studies; some replication; some assay-level concerns or methodological heterogeneity remains'), not High ('Extensive independent replication ... reproducibility well-established'). Confidence measures evidence density, not certainty that the strong claim fails (red-team RT-03; original: High).", "rests_on_non_peer_reviewed_work": false, "rationale": "Some ultra-processed whole-diet packages causally increase energy intake and relative short-term weight or fat gain, reaching Ε for those bounded subclaims. The submitted proposition additionally requires that the effect persist when nutrient composition is matched and be carried by processing-conferred properties characteristic of the category; those clauses remain Γ because the most direct composition-matched contrast is non-significant [18], matched designs are null [8,11,22], composition changes intake within UPF [21], and processing-conferred properties act in both directions within NOVA 4 [4,6,23,24]. The level does not rest on the registry-only result [18]." }, "independent_blinded_audit": false, "raw_receipts": [ { "path": "deliverables\\prepublication-enrichment-2026-10-05\\website-series\\ultra-processed-food\\consensus-browser-links.json", "sha256": "5c114b266bbcfeb30c4f7e17f7dd3eabfde35c9ec70fc6f0d6fbad59e9d0df08" }, { "path": "deliverables\\prepublication-enrichment-2026-10-05\\website-series\\ultra-processed-food\\consensus-browser.txt", "sha256": "b546d9f124f0ca9362a5f90f68226d33631453bd6901b263468102b73452cafa" }, { "path": "deliverables\\prepublication-enrichment-2026-10-05\\website-series\\ultra-processed-food\\consensus-targeted-check.txt", "sha256": "3ea601c37be590e8b15d94379154456bc6c1783e6c3effa9d3f95ac4b52ca6d6" } ], "limitations": "This verifies documented historical use and links the record to the final edition. It adds no primary-source, full-text, independence or verification credit. Original access remains limited to the source-specific record.", "existing_primary_reconciliation": { "slug": "ultra-processed-food", "consensus_used": true, "rawproof": [ "deliverables/prepublication-enrichment-2026-10-05/website-series/ultra-processed-food/consensus-browser.txt", "deliverables/prepublication-enrichment-2026-10-05/website-series/ultra-processed-food/consensus-targeted-check.txt", "deliverables/prepublication-enrichment-2026-10-05/website-series/ultra-processed-food/consensus-browser-links.json" ], "reconciliation_existing": "Yes; 9 primary identifiers; selected Results passage reconciled; timeline rejects generated CRISPR caption and marks shared registry/dependent reports.", "primary_sources_matched": [ "31105044", "35285882", "37452167", "39267249", "40043691", "40760353", "41314613", "42588167", "42633852" ], "limitations": "No consumed-density matching inferred; shared registry reports not counted as independent cohorts; service body access is distinct from selected external body check." } } }