How we score

Free sugar is the sugar that counts: added sugars and syrups, plus the sugar in juice, purée and pressed fruit, which the body treats like added sugar.

• WHO, Guideline: Sugars intake for adults and children, 2015. Free sugars under 10 % of energy, ideally under 5 %; free sugars include those in juice, purée and concentrate. guideline
• SACN, Carbohydrates and Health, 2015. UK free-sugar definition and the 5 % of energy target. guideline

Ingredients are listed by weight. Sugar first or second means the product is built on it, whatever the front of the pack says.

• FSANZ Food Standards Code, Standard 1.2.4, 2016. Ingredients must be listed in descending order of ingoing weight. regulation

Additives with a cosmetic job, isolates, extracts and unnamed refined oils are the fingerprints of an industrial formulation.

• Monteiro et al., Public Health Nutrition 22(5), 2019. NOVA: the markers of ultra-processing are cosmetic additives, isolates, extracts and industrial ingredients. review

Carboxymethylcellulose and polysorbate 80 are the two emulsifiers taken beyond animal work.

• Chassaing et al., Gastroenterology 162(3), 2022. Controlled feeding trial in humans: carboxymethylcellulose altered the gut microbiota and metabolome within two weeks. trial
• Chassaing et al., Nature 519, 2015. Dietary emulsifiers (carboxymethylcellulose, polysorbate 80) thinned the gut mucus layer and promoted colitis and metabolic syndrome in mice. trial

Synthetic azo and coal-tar dyes. Six of them were linked to increased hyperactivity in children in the Southampton study, which led to warning labels in the EU.

• McCann et al., The Lancet 370, 2007. Southampton trial: mixes of six synthetic colours with sodium benzoate increased hyperactivity in 3- and 8/9-year-olds. trial
• Regulation (EC) 1333/2008, Annex V, 2008. EU warning label on the six Southampton colours: 'may have an adverse effect on activity and attention in children'. regulation

Sweeteners are ranked, worst first. Aspartame: IARC called it possibly carcinogenic in 2023, it carries a phenylalanine warning, and its record is the weakest, so it costs the most.

• WHO, Use of non-sugar sweeteners: guideline, 2023. Advises against non-sugar sweeteners for weight control; no long-term benefit, possible long-term harms. guideline
• IARC Monographs vol. 134 (summary in Lancet Oncology 24), 2023. Aspartame classified Group 2B, possibly carcinogenic to humans; JECFA left the ADI at 40 mg/kg. monograph
• Suez et al., Cell 185(18), 2022. Randomised trial: sucralose and saccharin altered the gut microbiome and impaired glucose tolerance in healthy adults. trial
• Witkowski et al., Nature Medicine 29, 2023. Higher blood erythritol associated with cardiovascular events and enhanced platelet reactivity; causality contested. trial
• FSANZ Food Standards Code, Standard 1.3.1, 2016. Permitted sweeteners and additives in Australia and New Zealand. regulation

Two or more sweeteners, or one high in the list, means the product is engineered to taste sweet without sugar.

• WHO, Use of non-sugar sweeteners: guideline, 2023. Advises against non-sugar sweeteners for weight control; no long-term benefit, possible long-term harms. guideline
• Monteiro et al., Public Health Nutrition 22(5), 2019. NOVA: the markers of ultra-processing are cosmetic additives, isolates, extracts and industrial ingredients. review

Over 600 mg of sodium per 100 g is the UK's 'high salt' threshold; children need far less than adults.

• UK Department of Health, Front of Pack nutrition labelling guidance, 2016. High salt is more than 1.5 g salt (600 mg sodium) per 100 g. guideline
• NHMRC, Nutrient Reference Values: sodium, 2017. Adequate intake for children is far below adult levels: 200 to 400 mg a day for 1 to 8 year olds. guideline

A short ingredient list can still be a hard-processed food, and the evidence line names which kind this pack is.

• Stickler scoring policy, 2026. Where the literature is contested (seed oils, homogenisation, heavy processing of clean-label foods) the app takes a stated position and shows it as such. stance
• Deeth & Lewis, High Temperature Processing of Milk and Milk Products (Wiley), 2017. UHT at 135 to 150 °C denatures whey proteins, drives Maillard reactions and lowers heat-labile vitamins relative to pasteurisation. book
• Claeys et al., Food Control 31(1), 2013. Review of raw versus heated milk: heat treatment reduces some vitamins and changes protein structure; pasteurisation removes pathogen risk. review
• EFSA CONTAM Panel, EFSA Journal 13(6):4104, 2015. Acrylamide forms in fried and baked starchy foods above 120 °C and is a probable carcinogen at dietary levels of concern. review
• Monteiro et al., Public Health Nutrition 22(5), 2019. NOVA: the markers of ultra-processing are cosmetic additives, isolates, extracts and industrial ingredients. review

Homogenisation forces milk through a fine valve at high pressure so the fat globules are broken small enough that the cream never rises.

• Stickler scoring policy, 2026. Where the literature is contested (seed oils, homogenisation, heavy processing of clean-label foods) the app takes a stated position and shows it as such. stance
• Michalski & Januel, Trends in Food Science & Technology 17(8), 2006. Homogenisation changes fat globule structure; health effects remain unproven either way. review

Supermarket milk is collected from many farms into one tanker, then standardised: the cream is spun out and blended back to hit a fixed fat percentage, so every bottle of the same brand tastes the same all year.

• Stickler scoring policy, 2026. Where the literature is contested (seed oils, homogenisation, heavy processing of clean-label foods) the app takes a stated position and shows it as such. stance

Supermarket mince is trim from many carcasses and whatever cuts were left, ground together and sold by fat grade: the stars are a fat percentage, not a quality mark.

• Stickler scoring policy, 2026. Where the literature is contested (seed oils, homogenisation, heavy processing of clean-label foods) the app takes a stated position and shows it as such. stance

Oils are ranked. Hydrogenated and interesterified fats cost the most: industrial hardening, and trans fats are the one fat every authority agrees to eliminate.

• Stickler scoring policy, 2026. Where the literature is contested (seed oils, homogenisation, heavy processing of clean-label foods) the app takes a stated position and shows it as such. stance
• WHO, REPLACE trans fat action package, 2018. Industrially produced trans fat should be eliminated from the global food supply. guideline
• EFSA CONTAM Panel, EFSA Journal 14(5):4426, 2016. 3-MCPD and glycidyl esters form when vegetable oils are refined at high temperature; highest in palm oil. review
• Estruch et al., NEJM 378:e34 (PREDIMED), 2018. Mediterranean diet with extra virgin olive oil or nuts cut major cardiovascular events by about 30 %. trial
• Ramsden et al., BMJ 346:e8707, 2013. Sydney Diet Heart re-analysis: replacing saturated fat with linoleic acid (safflower oil) raised mortality. trial
• Hooper et al., Cochrane Review CD011094, 2018. Omega-6 fats: little or no effect on cardiovascular events; the counter-view to the linoleic acid case. review
• Simopoulos, Biomedicine & Pharmacotherapy 56(8), 2002. Modern diets carry an omega-6 to omega-3 ratio near 15:1 against an evolutionary 1:1. review

Ingredients are listed by weight. A refined oil in the first three means the food is fried in it or built on it: chips, crackers, dressings, most snack bars.

• FSANZ Food Standards Code, Standard 1.2.4, 2016. Ingredients must be listed in descending order of ingoing weight. regulation

When refined flour, starch, sugar or oil is the bulk of the recipe, the named vegetable or fruit is a costume.

• Monteiro et al., Public Health Nutrition 22(5), 2019. NOVA: the markers of ultra-processing are cosmetic additives, isolates, extracts and industrial ingredients. review
• NHMRC, Australian Dietary Guidelines, 2013. Prefer wholegrain and high-fibre cereal foods; limit foods with added sugar, salt and saturated fat. guideline

Refined flour, added fat and added sugar together make a pastry, a biscuit or a cake.

• Stickler scoring policy, 2026. Where the literature is contested (seed oils, homogenisation, heavy processing of clean-label foods) the app takes a stated position and shows it as such. stance
• NHMRC, Australian Dietary Guidelines, 2013. Prefer wholegrain and high-fibre cereal foods; limit foods with added sugar, salt and saturated fat. guideline

Cacao concentrates cadmium and lead from soil and drying.

• Consumer Reports, Lead and cadmium could be in your dark chocolate, 2022. 23 of 28 dark chocolate bars tested exceeded California's maximum allowable dose level for lead or cadmium per ounce. review
• Commission Regulation (EU) 488/2014 (cadmium in cocoa and chocolate), 2014. Sets maximum cadmium levels for chocolate by cocoa content and for cocoa powder, in force from 2019. regulation
• EFSA CONTAM Panel, Risk assessment of ochratoxin A in food, 2020. Cocoa and cocoa products are among the contributors to ochratoxin A exposure; the toxin is nephrotoxic and genotoxic. review

Rice takes up inorganic arsenic from paddy water more than any other grain, and rice cakes, puffed rice and rice crackers concentrate it.

• EFSA CONTAM Panel, Update of the risk assessment of inorganic arsenic in food, 2024. Rice and rice-based products are the main dietary source of inorganic arsenic; margins of exposure are of concern, especially for children. review
• Karagas et al., JAMA Pediatrics 170(6), 2016. Infants eating rice cereal and rice snacks had roughly double the urinary arsenic of infants who ate none. trial
• Swedish Food Agency (Livsmedelsverket), Arsenic in rice, 2015. Children under six should not eat rice cakes; everyone should vary grains and rinse rice. guideline

The bag is grease-proofed with PFAS-class chemicals that migrate into the corn when heated; the US FDA only completed the phase-out of those coatings in 2024.

• US FDA, Grease-proofing substances containing PFAS no longer sold for US food packaging, 2024. PFAS-based grease-proofing agents used in microwave popcorn bags and fast-food wrappers were phased out of the US market by 2024. regulation
• NIOSH, Flavorings-related lung disease (diacetyl and 2,3-pentanedione), 2016. Inhaled butter flavourings cause bronchiolitis obliterans; NIOSH set exposure limits for diacetyl. guideline

Three things about commodity milk that the carton cannot show.

• Harrison et al., Cancer Causes & Control 28:497–528, 2017. Systematic review found moderate evidence linking milk intake to higher circulating IGF-I, and higher IGF-I to prostate cancer risk; it does not isolate homogenisation or pooling. review
• Roddam et al., Annals of Internal Medicine 149(7):461–471, 2008. Pooled data from twelve prospective studies: highest versus lowest IGF-I quintile had prostate cancer odds ratio 1.38 (95 % CI 1.19–1.60). This measures circulating IGF-I, not a milk treatment. review
• Lu et al., Nutrition Journal 15:91, 2016. Eleven cohorts overall. The whole-milk prostate cancer mortality subgroup showed RR 1.50 for high versus low intake and 1.43 per additional daily serving; total dairy was not associated with overall cancer mortality. review
• Gao et al., Frontiers in Nutrition 11:1390379, 2024. Reviews residual sex steroid hormones in milk, variation with reproductive state and possible breast and prostate cancer pathways; further research is needed to establish the health effects. review
• FSANZ, Compendium of Microbiological Criteria for Food, Table 5.1, 2022. Recommends bulk milk cell counts below 400,000 cells/mL when verifying dairy primary-production controls; higher counts may indicate mastitis. This is process-control guidance, not evidence that a compliant carton causes harm. guideline

Peanuts are a major dietary route to aflatoxin, a mould toxin the IARC classes as a Group 1 human carcinogen and the clearest dietary cause of liver cancer worldwide.

• IARC Monographs vol. 100F, Aflatoxins, 2012. Aflatoxins are Group 1 human carcinogens causing liver cancer; maize and peanuts are major dietary routes in affected regions. monograph
• FSANZ Food Standards Code, Standard 1.4.1 and Schedule 19, 2024. Schedule 19 sets an aflatoxin maximum of 0.015 mg/kg for peanuts and tree nuts. regulation
• EFSA CONTAM Panel, EFSA Journal 18(3):6040, 2020. Aflatoxins are genotoxic and aflatoxin B1 causes liver cancer. Dietary exposure raises a health concern; grains supplied the largest mean B1 exposure in the European data assessed. review