Secondary analysis of food-composition databases (food-item level; no human participants)
Geography
United States
Population or data
Liquids, zero-kcal items and infant formula/foods excluded. 3,893 items present in all three databases; models used items present in 2+ databases.
Main result
Using one fixed nutrient-combination definition, the share of US foods classed as hyper-palatable rose from 49% in 1988 to 69% in 2018. The same food items were about 4 times more likely (odds) to meet the definition in 2018 than in 1988, mostly driven by fat+sodium foods.
Supports
Nutrient profiles of commonly reported US foods shifted toward the HPF thresholds between 1988 and 2018, especially fat+sodium combinations. The authors interpret this as probable reformulation: they say it 'likely reflects a reformulation of food products over time to enhance their palatability' and 'may have resulted from the reformulation of existing food products'.
Does not establish
It does not establish who reformulated, why, or that reformulation was deliberately aimed at palatability: no company or brand data are used, and the reformulation reading is the authors' inference. It does not measure consumption, eating behaviour or health outcomes. It does not cover beverages.
Review of primary tobacco-industry documents to code brand ownership, combined with longitudinal analysis of USDA food-composition datasets (PubMed also tags it 'Review'; it is an original empirical analysis)
Geography
United States
Population or data
USDA datasets for 1988-2001 (and 2018); per the abstract, foods that were (n = 105) and were not (n = 587) owned by US tobacco companies. The abstract's wording is ambiguous on whether n counts all items or only HPF items.
Main result
Between 1988 and 2001, foods from tobacco-owned brands were more likely than other foods to meet the fat+sodium and carb+sodium hyper-palatability criteria. By 2018 these food types were common whether or not the brand had ever had a tobacco owner.
Supports
There is an empirical association between tobacco-company ownership and HPF formulation in 1988-2001. By 2018, HPF saturation was not confined to formerly tobacco-owned brands.
Does not establish
It does not establish intent or a transfer of specific techniques. It does not show causation; tobacco firms may have bought already-hyper-palatable brands.
Checked
2026-10-18
Claims
Background source
H09Formulas and reformulation
Consumption of ultra-processed foods in Canada
Polsky JY, Moubarac JC, Garriguet D
Published
Health Reports 31(11), Statistics Canada cat. 82-003-X (released 18 Nov 2020) (2020)
Kilograms per person per year, 1960-2025, not adjusted for losses
Main result
Canadian salad-oil availability rose about 6.6-fold from 1970 to 2025 while refined sugar fell ~40%; shortening and margarine also fell. No dressing/sauce/condiment series exists in the table.
Supports
q5
Does not establish
Availability, not intake; salad oils include all uses; a step increase in 1996-1997 may reflect reporting changes.
Coca-Cola bottlers under 1921 contracts; trial record incl. deposition of Brian Dyson, President of Coca-Cola U.S.A.
Main result
Coca-Cola began replacing sugar with HFCS in January 1980 and had eliminated sucrose from US bottling syrup by November 1984. The court records that HFCS was cheaper and that the Company argued the switch was undetectable.
Supports
A cost-driven sweetener swap in an iconic product, done under the same name and confirmed in court findings.
Does not establish
The district court held that HFCS-55 is not 'sugar' under the 1921 contract. Court records do not show what consumers were told.
consumer-protection agency label comparison + commissioned lab test
Geography
Germany
Population or data
Nutella old vs new ingredient list and nutrition table; lab analysis of cocoa
Main result
The German Nutella recipe gained milk powder and sugar and lost cocoa and fat, and the spread is visibly lighter. Ferrero did not answer vzhh's inquiry.
Supports
Same name, cheaper inputs, and disclosure only through the ingredient list.
Does not establish
Vzhh did not prove cost as the motive; it inferred it. Some reports give the old cocoa figure as 8.4 g/100 g rather than 8.5%.
Reformulation of sugar contents in Canadian prepackaged foods and beverages between 2013 and 2017 and resultant changes in nutritional composition of products with sugar reductions
Bernstein JT, Christoforou AK, Weippert M, L'Abbé MR
Impact of the announcement and implementation of the UK Soft Drinks Industry Levy on sugar content, price, product size and number of available soft drinks in the UK, 2015-19: A controlled interrupted time series analysis
Scarborough P, Adhikari V, Harrington RA, Elhussein A, Briggs A, Rayner M, Adams J, Cummins S, Penney T, White M
controlled interrupted time series (peer-reviewed)
Geography
United Kingdom
Population or data
209,637 observations, 85 time points, Sep 2015–Feb 2019, supermarket websites
Main result
Most reformulation happened between the levy's announcement and its start date. The authors say it was much greater than voluntary initiatives achieved.
Supports
A tax can reformulate a whole category quickly.
Does not establish
Online supermarket data only; exempt categories (juice, milk drinks) were used as controls. The '28.8%' figure circulating online is not in this paper.
Changes in the critical nutrient content of packaged foods and beverages after the full implementation of the Chilean Food Labelling and Advertising Law: a repeated cross-sectional study
Rebolledo N, Ferrer-Rosende P, Reyes M, Smith Taillie L, Corvalán C
Secondary comparison of published USDA nutrient tables (1950 vs 1999), moisture-adjusted
Geography
United States (USDA data)
Population or data
43 garden crops (mostly vegetables); 13 nutrients plus water
Main result
Across 43 garden crops, USDA's 1999 tables show lower median values than its 1950 tables for six nutrients, from 6% lower protein to 38% lower riboflavin. Most individual food-nutrient changes cannot be distinguished from zero.
Supports
There is a group-level signal of lower recorded nutrient concentrations in US vegetables over 50 years. The authors' preferred explanation is cultivar change with yield/nutrient trade-offs.
Does not establish
It does not identify a cause. It does not show soil depletion, and the authors favour cultivar change over soil. It says nothing about health effects. About 28% of the reliably different individual values went up, not down.
Analysis of archived grain and soil samples from a long-term field experiment
Geography
United Kingdom (Broadbalk Wheat Experiment, Rothamsted, est. 1843)
Population or data
Archived Broadbalk wheat grain and soil samples, 1845 onward, across unfertilised, inorganic-fertiliser and manure treatments
Main result
In wheat archived since 1845, zinc, iron, copper and magnesium concentrations held steady until the mid-1960s and then fell, in step with semi-dwarf high-yield varieties. Soil mineral levels did not fall.
Supports
Breeding for yield (a dilution / harvest-index effect) rather than soil depletion is the better-supported driver of lower grain mineral density in this experiment. The decline appeared under organic manure as well as inorganic fertiliser.
Does not establish
It does not establish the size of the decline (no % in the abstract), effects on population nutrient intake or health, or applicability beyond UK wheat. It contradicts, rather than supports, a 'depleted soil' narrative.
A chemical genetic roadmap to improved tomato flavor
Tieman D, Zhu G, Resende MFR Jr, Lin T, Nguyen C, Bies D, Rambla JL, Ortiz Beltran KS, Taylor M, Zhang B, Ikeda H, Liu Z, Fisher J, Zemach I, Monforte A, Zamir D, Granell A, Kirst M, Huang S, Klee H
Accessions grown in Florida (USA); second population in Israel
Population or data
398 modern, heirloom and wild accessions (S. lycopersicum and S. pimpinellifolium); 160 samples / 101 accessions in consumer panels; 48 modern cultivars vs 236 older S. lycopersicum accessions; F2 of 197 plants (Maglia Rosa Cherry x FLA 8059); 352-variety Israel population
Main result
Modern commercial tomatoes carry significantly lower levels of many consumer-liked flavor chemicals than heirlooms (13 volatiles significantly reduced). Sugar loss tracks selection for larger fruit (domestication/improvement sweeps; 47/48 modern cultivars carry low-sugar alleles; strong negative fruit weight-sugar correlation in F2). Visible traits (red color/lycopene-linked MHO) were enriched; volatile alleles not tracked by breeders were lost at random (13/26 loci).
Supports
Selection for production traits (yield, size, firmness, disease resistance) is a different target from flavor and inadvertently reduced flavor chemistry; size and sugar trade off genetically. Also notes Florida 47 and Flora-Dade are not well liked even when harvested fully ripe, so postharvest handling is not the whole story.
Does not establish
Not a nutrient-density study. Does not show that yield selection causes flavor loss in all crops. The volatile loss is described as random loss of untracked alleles, not deliberate selection against flavor. The Lin5 validation used transgenic plants (research tool only).
Uniform ripening encodes a Golden 2-like transcription factor regulating tomato fruit chloroplast development
Powell ALT, Nguyen CV, Hill T, Cheng KL, Figueroa-Balderas R, Aktas H, Ashrafi H, Pons C, Fernandez-Munoz R, Vicente A, Lopez-Baltazar J, Barry CS, Liu Y, Chetelat R, Granell A, Van Deynze A, Giovannoni JJ, Bennett AB
u/u vs U/U tomato lines; transgenic lines expressing Arabidopsis GLK1/GLK2 (AtGLK) or suppressing SlGLK2
Main result
The u mutation, selected for even ripening/uniform color to aid harvest, is a loss-of-function allele of SlGLK2 present in all u/u cultivars examined; it reduces fruit chloroplast development, and ripe u/u fruit have lower soluble solids (U/U had 10% higher soluble solids). Boosting GLK (transgenically) raised sugars and lycopene.
Supports
A trait selected for harvest/appearance convenience inadvertently reduced ripe-fruit sugar and lycopene: production and quality traits can conflict.
Does not establish
The 21%, 40% and 10-60% figures come from transgenic Arabidopsis-GLK lines, not commercial varieties; only the 10% soluble-solids figure compares the natural U and u alleles. Does not measure sensory flavor directly.
Extension presentation of an unpublished 2-year storage trial (slides)
Geography
Michigan, USA
Population or data
Gala, McIntosh, Jonagold, Golden Delicious, Red Delicious, Fuji, Evercrisp; CA or simulated DCA for 3, 6 and 9 or 11 months; half treated with 1 ppm 1-MCP; 5 replicate fruit per treatment; aroma measured up to 5 weeks at 22 C after removal
Main result
Storage duration had the largest effect on apple aroma production; 1-MCP and DCA each inhibited recovery of aroma volatiles after removal from storage, and effects were often additive.
Supports
Technologies that extend apple storage life (CA, DCA, 1-MCP, long storage) suppress aroma: shelf life and aroma are different targets.
Does not establish
No numeric aroma values extracted (charts only, no text values). Not peer reviewed. Does not quantify consumer liking. 'Red Delicious' named in methods but no Red Delicious chart appeared in extracted text.
Experimental (transgenic ethylene-suppressed lines and 1-MCP treatment)
Geography
California, USA
Population or data
Apple fruit from ACC-synthase/ACC-oxidase suppressed trees and 1-MCP-treated apples
Main result
Blocking ethylene action with 1-MCP reduced ester and alcohol aroma production in apples; aldehydes were not much affected; malic acid degradation slowed.
Supports
Primary peer-reviewed evidence that 1-MCP, a commercial storage-life tool, suppresses apple ester (fruity aroma) production.
Does not establish
Magnitude of reduction not in abstract; does not test commercial long-term storage or consumer liking.
Cavendish is the most traded banana and just under half of global production (~50 Mt/yr); world trade ~20 Mt/yr.
Supports
Export trade is concentrated on one cultivar group while global production is not: most bananas grown are not Cavendish.
Does not establish
FAO page gives no explicit percentage of exports that are Cavendish. Figures such as 95-99% of exports circulate in media but were not found on an FAO page.
Checked
2026-10-18
Claims
Background source
P06Crops, grades and storage
The World Banana Economy, 1985-2002 (Chapter 1: Overview of world banana production and trade)
Mature-green Cavendish-subgroup bananas from a retailer distribution centre, held at 5.0, 13.0, 14.0 C with/without 1-MCP
Main result
Mature-green Cavendish bananas are chilling-sensitive; chilling injury appears at 13 C within 10 days and even at the 14 C threshold if storage is extended.
Supports
Export bananas are shipped mature-green under a narrow cold-chain window (~13-14 C) dictated by chilling sensitivity.
Does not establish
Flavor effects; the timed thresholds (1 h at 10 C etc.) are a cited personal communication, not measured here.
Checked
2026-10-18
Claims
Background source
P08Crops, grades and storage
Ethylene: Handling/Processing Technical Report
USDA Agricultural Marketing Service, National Organic Program (technical report)
Literature on ethylene use in postharvest ripening
Main result
Bananas are ripened in sealed ethylene rooms (e.g., 100 ppm, 24 h) to control timing for logistics. Evidence on flavor is mixed: one study found more aroma in naturally ripened fruit; another (Ahmad et al. 2001) found ethylene-treated fruit at 20 C rated better; a review found no effect on final sugar/starch/soluble solids.
Supports
Ripening is scheduled for logistics. Some evidence of aroma differences.
Does not establish
That ethylene ripening as such degrades banana flavor; evidence conflicts.
Checked
2026-10-18
Claims
Background source
P09Crops, grades and storage
Ethylene-induced isocoumarin formation in carrot root tissue
Carrot root discs and roots exposed to ethylene, fungi, auxins
Main result
Ethylene (above 0.3 ppm) induces formation of the isocoumarin MMHD (3-methyl-6-methoxy-8-hydroxy-3,4-dihydroisocoumarin, i.e. 6-methoxymellein) in carrot tissue.
Supports
Storing carrots with ethylene sources (e.g. ethylene-producing fruit, exhaust) can produce the bitter isocoumarin.
Does not establish
Sensory bitterness was not measured; the link between 6-methoxymellein and bitter taste comes from other carrot research.
Checked
2026-10-18
Claims
Background source
P11Crops, grades and storage
Quality of carrots as affected by pre- and postharvest factors and processing
Seljasen R, Kristensen HL, Lauridsen C, Wyss GS, Kretzschmar U, Birlouez-Aragon I, Kahl J
Published
Journal of the Science of Food and Agriculture 93(11):2611-2626 (2013)
Cultivar (genetics) has the largest effect on carrot quality compounds; climate next; retail temperature, storage atmosphere and heat processing cause the largest quality losses.
Supports
Cultivar choice and postharvest handling both strongly change carrot flavor and nutrient compounds.
Does not establish
That commercial cultivars are worse than others; direction of the cultivar effect is not tied to commercial selection in the abstract.
Checked
2026-10-18
Claims
Background source
P12Crops, grades and storage
United States Standards for Grades of Fresh Tomatoes
USDA Agricultural Marketing Service
Published
USDA AMS grade standard (effective 1 Oct 1991, reprinted Jan 1997) (1991)
Grades are defined by maturity (internal seed-cavity development), shape, smoothness, cleanliness, defects (decay, freezing, sunscald, damage), size (diameter) and color stage. Text search found no Brix, sugar, soluble solids or flavor criterion.
Supports
At USDA grading, tomato flavor/sugar is not a measured acceptance criterion.
Does not establish
Private retailer specs, which may add Brix; whether this 1991 standard is the latest (AMS page returned 403).
Checked
2026-10-18
Claims
Background source
P16Crops, grades and storage
Declining Fruit and Vegetable Nutrient Composition: What Is the Evidence?
Three lines of evidence point to declines of some nutrients; side-by-side cultivar trials in broccoli and grains show yield trading off against mineral and protein concentration.
Supports
Yield and nutrient concentration can trade off (dilution), including for a vegetable (broccoli).
Does not establish
That all nutrients declined; historical data are 'inherently limited' per the author.
Kiwifruit Administrative Committee recommendation and research cited
Main result
The floor was lowered from 6.5% to 6.2% in 2000 because 6.2% fruit can 'potentially' ripen to 12.6%, consumer tests found 12.6% acceptable, other exporters used 6.2%, and it would cut reinspection costs and move early-season fruit. A proposal to raise the floor was rejected.
Supports
Explains why the level is set where it is: a floor on ripening potential, tuned to trade and inspection economics, not a target for good eating.
Does not establish
Committee/industry rationale; consumer tests cited are not reproduced in the rule.
Consumer tests (1999, 2008) plus 6-year quality survey
Geography
California, US
Population or data
In-store consumer tests; California vineyard and cold-storage samples
Main result
The 6.2% harvest floor protects storage life; eating acceptance depends on ripe sugar (>=12.5%) and dry matter (>=15.1%), which the legal standard does not measure.
Supports
Quantifies the gap between the legal floor and the eating-quality threshold.
New Zealand (fruit sold into North America and elsewhere)
Population or data
Main result
Zespri refuses fruit below a variety-specific dry-matter floor (15.5% to 17.2%) and pays growers a Taste Zespri payment that rises with dry matter.
Supports
The clearest example of flavour (via dry matter) as a condition of purchase and a basis of grower pay; it is a private brand scheme, not a North American public standard.
Does not establish
Dry matter and soluble solids are different measures; booklet rates may change in season.
Washington sets an early-season sugar floor for Delicious apples; raised from 10% to 11% in 2002 at the Washington State Horticultural Association's request, explicitly to make early apples sweeter.
Supports
Rare explicit eating-quality rationale; floor applies only to early-season shipping of three varieties.
Does not establish
Within-Washington marketing standard; other varieties have firmness/starch rules instead.
Federal grade standard (incorporated by reference)
Geography
Canada
Population or data
Grades for fresh fruit and vegetables in import and interprovincial trade
Main result
No Canadian federal grade, including cantaloupes and grapes, contains a sugar, Brix, acid or dry-matter test; 'mature' is defined only as able to complete ripening. Since 2023 CFIA retains only apple, onion and potato grades in the volume and points to DRC-maintained requirements for other produce.
Supports
Canada sets no sugar floor at all in its grade system.
Does not establish
DRC-maintained Fresh Fruit and Vegetable Grade Requirements and Volume 9 import grades not fully reviewed;
The grade standards for strawberries, tomatoes and peaches contain no sugar or flavour criterion; tomato maturity is anatomical, and colour stages are descriptive.
Supports
Supports that general grades are cosmetic/handling based.
Most Florida tomatoes are picked mature-green and ethylene-ripened; a fifth or more of a green lot may be too immature ever to ripen properly, and M2 fruit is fit only for food service.
Supports
Documents mature-green harvest and its eating-quality consequences.
Does not establish
No exact share given; 'majority' only.
Checked
2026-10-18
Claims
Background source
S29Crops, grades and storage
Taking Tomatoes Back To Their Tasty Roots (transcript)
Palca J (reporter); quoting Harry Klee, University of Florida
When both diets followed the UK Eatwell Guide and were delivered free to participants' homes to eat at will for 8 weeks, people lost weight on both. They lost about 1 percentage point more (~1 kg) on the minimally processed diet, which also produced greater fat loss, lower triglycerides and fewer cravings.
Supports
Within guideline-compliant diets, a minimally processed version produced modestly more weight and fat loss than an ultra-processed version over 8 weeks, in adults with high habitual UPF intake.
Does not establish
It does not identify which property of UPF caused the difference: energy density was higher on the UPF diet and taste ratings were lower on MPF, and either could explain it. It shows no clinically meaningful cardiometabolic difference (LDL-C favoured UPF). It says nothing about long-term effects; the 9-13% vs 4-5% one-year figures are the authors' extrapolation, not data. It does not generalise to men (91% female) or to people with low UPF intake.
Cross-sectional, two studies; self-report questionnaires (Yale Food Addiction Scale) plus forced-choice (Study 1) and 7-point Likert ratings (Study 2) of 35 foods
Geography
United States
Population or data
Study 1: 120 University of Michigan undergraduates. Study 2: 398 recruited via Amazon MTurk, 384 analysed (paid $0.40); not nationally representative
Main result
Highly processed foods with added fat and/or refined carbohydrates (high glycemic load) were most often linked to self-reported addictive-like eating; minimally processed foods ranked lowest.
Supports
Perceived food-level differences: people associate processed, high-fat, high-GL foods with loss-of-control eating far more than whole foods.
Does not establish
That any food is clinically addictive, causes addiction, or causes weight gain; no consumption, biological or behavioural measurement.
Checked
2026-10-18
Claims
Background source
H02Eating, palate and health
Premature Mortality Attributable to Ultraprocessed Food Consumption in 8 Countries
Nilson EAF, Delpino FM, Batis C, Machado PP, Moubarac JC, Cediel G, Corvalan C, Ferrari G, Rauber F, Martinez-Steele E, Louzada MLC, Levy RB, Monteiro CA, Rezende LFM
Published
American Journal of Preventive Medicine 68(6):1091-1099 (2025)
Dose-response meta-analysis of observational cohort studies plus population attributable fraction (comparative risk) modelling
Geography
Colombia, Brazil (low UPF); Chile, Mexico (intermediate); Australia, Canada, United Kingdom, United States (high)
Population or data
Pooled cohort studies for the risk estimate; national dietary surveys for each country's UPF energy share
Main result
Applying a 3% mortality increase per 10-point rise in UPF share to each country's intake yields 4% to 14% of premature deaths modelled as attributable to UPF.
Supports
A population-level modelled burden estimate across 8 countries, including Canada and the US.
Does not establish
Observed deaths, individual risk, or causation. The figure is a modelled attributable fraction built on an observational association.
Inpatient randomised crossover controlled trial (2 weeks per diet, ad libitum)
Geography
United States (NIH Clinical Center, Bethesda)
Population or data
Main result
Under controlled inpatient conditions, an ultra-processed diet caused people to eat about 500 kcal/day more and gain weight versus an unprocessed diet matched for presented nutrients.
Supports
Causal evidence that ultra-processed diets, as tested, drive excess calorie intake and short-term weight gain.
Does not establish
Long-term disease or mortality effects; which property of UPFs (speed of eating, energy density of consumed food, palatability) is responsible; generalisation beyond 20 adults over 4 weeks.
Official UN report; true cost accounting modelling
Geography
Global, 156 countries
Population or data
GBD 2021 dietary risk data and other inputs
Main result
Hidden costs of agrifood systems exceed 10 trillion 2020 PPP dollars (11.6 trillion); 70% are health costs from dietary risks for NCDs.
Supports
The scale of health costs that market prices exclude, broken down by GBD dietary risk factor.
Does not establish
Any cost attributable to ultraprocessed foods. The report attributes costs to GBD dietary risks (low whole grains, high sodium, low fruit, processed meat, SSBs, etc.). It uses UPF retail sales only as a descriptive indicator (Food Systems Dashboard) of trends across agrifood-system types.
Higher UPF share associated with higher prevalence of obesity, diabetes and hypertension in Canadian adults; no significant association reported for heart disease in the abstract.
Cites a pooled analysis of 281 studies from 36 countries
Main result
Argues that UPFs high in refined carbohydrates and added fats may be addictive and that this framing has social justice, clinical and policy implications.
Supports
Reporting what the addiction-framing researchers claim.
Does not establish
That UPFs are addictive in a clinical consensus sense; prevalence is YFAS self-report.
Checked
2026-10-18
Claims
Background source
H13Eating, palate and health
Quantification of the effect of energy imbalance on bodyweight
Hall KD, Sacks G, Chandramohan D, Chow CC, Wang YC, Gortmaker SL, Swinburn BA
Mathematical model of adult human metabolism, validated against feeding studies
Geography
United States (NIH)
Population or data
Dynamic energy balance model; US adult population averages
Main result
Every permanent change of energy intake of 100 kJ (about 24 kcal) a day leads to an eventual bodyweight change of about 1 kg, equivalently 10 kcal a day per pound, with half the change reached in about 1 year and 95% in about 3 years.
Supports
The time course of weight change from a sustained change in daily intake: 500 extra kcal a day gives about 25 lb in the first year and about 50 lb at the plateau.
Does not establish
A population-average rule of thumb for adults; individual responses differ with body composition and activity.
Low to moderate sugar-sweetened beverage consumption impairs glucose and lipid metabolism and promotes inflammation in healthy young men: a randomized controlled trial
Effect of Important Food Sources of Fructose-Containing Sugars on Inflammatory Biomarkers: A Systematic Review and Meta-Analysis of Controlled Feeding Trials
Food source matters: sugars delivered with SSBs increased CRP in substitution trials; total fructose-containing sugars did not raise CRP overall; fruit, juice, soy beverage lowered it in some designs.
Supports
q1
Does not establish
Moderate-to-low certainty; short-to-medium term; authors from a group with past industry funding (see paper disclosures).
WHO guideline; free sugars defined as added sugars plus sugars in honey, syrups, juices
Main result
Strong recommendation: free sugars <10% of total energy for adults and children; conditional: <5% (~25 g/6 tsp/day). Does not cover sugars in fresh fruit, vegetables or milk.
Supports
q1, q3
Does not establish
Verbatim fragments are the WHO definition of free sugars; the 10%/5% figures are paraphrased from the release. Guideline rests mainly on weight and dental caries evidence, not inflammation.
Checked
2026-10-18
Claims
Background source
N07Eating, palate and health
Dietary sugars and body weight: systematic review and meta-analyses of randomised controlled trials and cohort studies
63 Black 2-year-olds previously tested at birth and 6 months
Main result
Early sugar-water feeding predicted higher sugar-water intake at age 2, but not higher intake of sweetened fruit drink: the effect was specific to the medium.
Supports
q3
Does not establish
Small, observational; specific to sugar water.
Checked
2026-10-18
Claims
Background source
N28Eating, palate and health
The Sweet Tooth Trial: A Parallel Randomized Controlled Trial Investigating the Effects of A 6-Month Low, Regular, or High Dietary Sweet Taste Exposure on Sweet Taste Liking, and Various Outcomes Related to Food Intake and Weight Status
Cad EM, Mars M, Pretorius L, van der Kruijssen M, Tang CS, de Jong HB, Balvers M, Appleton KM, de Graaf K
180 healthy adults; 7%, 35% or 80% of provided foods sweet (sugars, low-calorie sweeteners, fruit, dairy)
Main result
Raising or lowering dietary sweet-taste exposure for 6 months did not change sweet liking, energy intake or weight.
Supports
q4, q3
Does not establish
Adults only (says nothing about infants); authors disclose funding/speaker links to sugar, sweetener and beverage firms; published issue year 2026 (online 2025).
Short-term trials: small weight reduction (-0.71 kg) mostly when NSS replace sugar; long-term cohorts: higher obesity, T2D, CVD and mortality risk at low/very low certainty.
Supports
q4
Does not establish
Cohort associations vulnerable to reverse causation and residual confounding (critiqued e.g. in PMC10630128); RCTs short.
Ultra-processed foods cause weight gain and increased energy intake associated with reduced chewing frequency: A randomized, open-label, crossover study
Hamano S, Sawada M, Aihara M, Sakurai Y, Sekine R, Usami S, Kubota N, Yamauchi T
Laboratory experiments (pilot + 4 studies), hypothetical menu choices
Geography
USA
Population or data
Undergraduate students: Study 1 N=104, Study 2A N=70, Study 2B N=79, Study 3 N=183, Study 4 N=81
Main result
Adding a healthy item to a hypothetical menu roughly doubled the share choosing the most indulgent item in small student samples; in Study 1 the effect held only for high self-control participants and reversed for low self-control participants.
Supports
Yes, for stated choice in the lab; no field or consumption data.
Does not establish
Small student samples, hypothetical choices with no real food, all p-values near .05, interaction depends on a measured self-control scale. No independent direct replication was found in searches on 2026-10-18 (absence of evidence, not a failed replication).
Checked
2026-10-18
Claims
Background source
Y09Eating, palate and health
A Meta-Analytic Review of Moral Licensing
Blanken I, van de Ven N, Zeelenberg M
Published
Personality and Social Psychology Bulletin 41(4):540-558 (2015)
Altering product placement to create a healthier layout in supermarkets: Outcomes on store sales, customer purchasing, and diet in a prospective matched controlled cluster study
Vogel C, Crozier S, Penn-Newman D, Ball K, Moon G, Lord J, Cooper C, Baird J
The share of Canadians eating fruit/vegetables five or more times a day fell by about a third from 2015 to 2023.
Supports
B: Canadian intake declining.
Does not establish
Frequency, not quantity; self-report; 2020+ collected during pandemic with mode changes: check StatCan comparability notes before stating a trend as definitive.
Adult tasters, sucrose solutions and sweetened foods
Main result
Moskowitz's own primary work establishing an optimum ('breakpoint') sweetness level: the scientific basis of the later 'bliss point' term.
Supports
C: optimisation of palatability is documented primary science.
Does not establish
The phrase 'bliss point' does not appear in this abstract; its popular industrial application (e.g., Prego) rests on Michael Moss, Salt Sugar Fat (2013): journalism. No Moskowitz patent verified.
Checked
2026-10-18
Claims
Background source
Z20Eating, palate and health
Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized Controlled Trial of Ad Libitum Food Intake
Historical analysis of internal industry documents
Geography
UK (research site, Univ. of Birmingham) / USA (funder)
Population or data
SRF/ISRF internal reports on Project 259 (W.F.R. Pover), 1967-1971
Main result
ISRF ended funding for an animal study shortly before completion after preliminary results linked sucrose (vs starch) to triglycerides via gut bacteria and to higher urinary beta-glucuronidase; results were never published.
Supports
Industry termination of and non-disclosure of unfavourable animal research.
Does not establish
Sugar causes bladder cancer in humans; authors state 'It is not clear why the results of Project 259 were never published.' Rodent data, preliminary.
20 members of the 2020 Dietary Guidelines Advisory Committee
Main result
Most members of the 2020 Dietary Guidelines Advisory Committee had documented industry ties, 95% at least one; public disclosure of conflicts recommended in 2017 did not appear to occur.
Supports
Does not establish
Any guideline recommendation was changed because of COI.
Secondary company history citing contemporaneous news
Geography
USA
Population or data
Philip Morris-Kraft deal
Main result
Kraft accepted $13.1bn offer at end of Oct 1988.
Supports
Deal value and timing.
Does not establish
The final terms are as reported by the Wall Street Journal; initial bid amounts vary across sources (about $11.0 to 11.8 billion).
Checked
2026-10-18
Claims
Background source
C15Ownership, money and rules
Regulations Amending the Food and Drug Regulations (Nutrition Symbols, Other Labelling Provisions, Vitamin D and Hydrogenated Fats or Oils): SOR/2022-168
Towards unified global action on ultra-processed foods: understanding commercial determinants, countering corporate power, and mobilising a public health response
Baker P, Slater S, White M, Wood B, Contreras A, Corvalán C, Gupta A, Hofman K, Kruger P, Laar A, Lawrence M, Mafuyeka M, Mialon M, Monteiro CA, Nanema S, Phulkerd S, Popkin BM, Serodio P, Shats K, van Tulleken C, Nestle M, Barquera S
Published
The Lancet 406(10520):2703-2726 (Series paper 3 of 3, Ultra-processed foods and human health) (2025)
Review / Series paper (narrative synthesis and policy analysis)
Geography
Global
Population or data
Literature and industry analysis
Main result
The UPF industry and co-dependent actors are a key driver of the global shift to ultra-processed diets; their political activity is the principal obstacle to regulation; authors propose reducing industry power and building a global advocacy response.
Supports
An authoritative, peer-reviewed framing of corporate power, profitability and political activity in the UPF system (Nov 2025).
Does not establish
Quantified profit margins or specific corporate conduct findings in the abstract; it is an advocacy-oriented review, not an empirical study.
Sysco is the largest foodservice distributor in the US and Canada, holding about 18% of a $377B US market by its own estimate; its product mix is dominated by meats, canned/dry and frozen goods, with fresh produce only 8% of sales; about a third of US Broadline cases are Sysco's own brand.
Market share is Sysco's estimate on a Technomic denominator, not an audited market measure; product categories are storage/commodity classes, not processing levels (frozen 'other' and canned/dry include both raw ingredients and prepared items); Canada figure in US dollars and includes all Sysco Canada sales, no Canadian share disclosed.
The court found Sysco and US Foods were the only two national broadliners and that the merger (even after divesting 11 facilities to PFG) would give them 59-71% of sales to national customers such as GPOs, contract caterers and chains; it enjoined the deal.
Supports
Q2; Q5 (national customers include healthcare GPOs and foodservice management companies)
Does not establish
Shares are for a 'national customer' submarket (~$20-30B), not all foodservice; 2013 data; preliminary-injunction standard; some figures redacted in public version.
Produce is a high-gross-margin department (38%), above typical total-store gross margin (~28% US independents, ~22% Kroger), even after shrink (~5.5%) and department labor (~7.4%).
Supports
Contradicts the claim that produce is sold at low or negative margin; supports produce as traffic driver (framing only).
Does not establish
Small self-selected sample (15 companies); produce-trade association with interest in showing produce is profitable; excludes non-department overhead, DC/handling costs.
NAICS 445 Food and beverage retailers (incl. liquor stores), Q3 2025-Q2 2026, sum of 4 quarters (vectors v1456718040, v1456718042, v1456718046, v1456718053)
Main result
In Canada, fresh fruit and vegetables are about 12.6% of sales at food and beverage retailers (14% of food and beverage), about the same as packaged dry goods alone.
Supports
Context: produce share in Canada similar to US (11-15%).
Does not establish
NAICS group includes liquor stores (alcohol ~18% of sales) and specialty food stores; fresh meat values suppressed in this cut; researcher-computed shares.
Checked
2026-10-18
Claims
Background source
G04Stores, shelves and kitchens
Retail Grocery Market Study Report: Canada needs more grocery competition
Financial data from Loblaw, Sobeys/Empire, Metro and others, 2017-2022
Main result
Food gross margins rose modestly; private label flagged as possible driver; supplier payments acknowledged but not quantified; Bureau did not examine supplier dealings.
Supports
Supports that supplier payments and private label are part of grocer economics; does not provide department or produce margins.
Does not establish
Grocers did not supply food-only or granular data; no department breakdown; Bureau lacked compulsion powers.
Checked
2026-10-18
Claims
Background source
G06Stores, shelves and kitchens
Question Period Note: Grocery Retailers Charging New Fees to Food Suppliers (AAFC-2021-QP-00010)
Vendor allowances (~6.9% of sales) were about 4x Kroger's operating profit in 2017.
Supports
Strongly supports that supplier payments are a large share of retailer gross margin relative to profit.
Does not establish
Much of the allowance funds promotional price cuts and is passed to shoppers; not all is retained profit; allowances relate mostly to branded (centre-store) goods but category split not disclosed.
FAO's own wording: export trade is 'almost exclusively' Cavendish; race 1 destroyed Gros Michel in the 1950s; TR4 first seen on Cavendish in Taiwan late 1960s.
Supports
Q1: export monoculture; Gros Michel precedent
Does not establish
2014 document; title inferred from file name/session (verify exact title). Dating of Gros Michel collapse differs from V03 (1960s).
New premium apples are increasingly released as closed 'club' varieties restricted to licensed growers, packers and marketers, with trademark control outliving patents.
Supports
Q2: closed licensing
Does not establish
Original 10-year Cosmic Crisp exclusivity has been extended to 2032 (single press source for extension).
Checked
2026-10-18
Claims
Background source
V15Stores, shelves and kitchens
Garden variety / More vegetables, please! (Heald & Chapman findings); Veggie Tales: Pernicious Myths About Patents, Innovation, and Crop Diversity in the Twentieth Century
Heald PJ, Chapman S
Published
UGA Today (2009); University of Illinois Law Review (2012) (2012)
1903 USDA seed list vs 2004 commercial catalogues, 48 vegetable crops
Main result
Most 1903 named varieties are no longer sold, but total commercial vegetable variety counts were nearly unchanged by 2004; RAFI's arithmetic understated survival.
Supports
Q3 (criticism of RAFI)
Does not establish
Catalogue names are not genetic measures; synonymy inflates 1903 counts. The commonly cited '93%' (National Geographic 2011) is a different rendering of RAFI across 66 crops.
Checked
2026-10-18
Claims
Background source
V19Stores, shelves and kitchens
Increasing homogeneity in global food supplies and the implications for food security
Khoury CK, Bjorkman AD, Dempewolf H, Ramirez-Villegas J, Guarino L, Jarvis A, Rieseberg LH, Struik PC
global time-series analysis (FAOSTAT food balance sheets 1961-2009)
Geography
Global
Population or data
52 crop commodities, national food supplies
Main result
National diets gained crops locally but converged globally on wheat, rice, soy, palm and sunflower oil while millets, sorghum, cassava and yams declined.
Supports
Q3
Does not establish
Commodity-level (species), not variety-level; FAOSTAT aggregation.
Checked
2026-10-18
Claims
Background source
V20Stores, shelves and kitchens
Supermarket Facts
FMI – The Food Industry Association
Published
FMI website (The Food Retailing Industry Speaks) (2025)
Loblaw is majority-owned by George Weston Ltd (52.6%), itself ~59.5% controlled by Galen G. Weston; control brands are ~1/3 of grocery sales and are made by third-party vendors.
Fresh Del Monte Produce Inc. Form 10-K for fiscal year ended December 26, 2025; and 8-K Exhibit 99.1 'Fresh Del Monte Completes Acquisition of Select Del Monte Foods Assets' (March 19, 2026)
In March 2026 a fresh-produce company bought the bankrupt US packaged Del Monte business, re-uniting fresh and canned Del Monte for the first time in ~40 years.
Supports
Strongest documented current case of one firm spanning fresh produce and branded packaged produce (licensing arrangements in other regions remain).
Does not establish
Scope as described.
Checked
2026-10-18
Claims
Background source
W11Stores, shelves and kitchens
Dole Food Company, Inc. Form 8-K (Sept 21 2012: Acquisition Agreement with ITOCHU) and Form 8-K (April 2 2013: completion and Trademark Rights Agreement)
A major branded packaged-food maker bought into fresh produce in 2012 and sold out in 2019 at about a third of the price.
Supports
Contrary evidence: packaged-food majors have tried and exited fresh produce.
Does not establish
Scope as described.
Checked
2026-10-18
Claims
Background source
W15Stores, shelves and kitchens
Backgrounder: Competition Bureau seeks to block Nortera's proposed acquisition of B&G's Green Giant and Le Sieur Canadian vegetable business; and Statement by the Interim Commissioner on the abandoned Nortera-B&G Foods Canada merger
One Quebec-based processor makes its own vegetable brands, retailers' private labels and most of its main rival's Green Giant product in Canada; the Bureau blocked it from buying that rival.
Supports
Documented Canadian case of a single manufacturer behind several competing labels on the same shelf.