Editorial Standards

Every claim on Ultra Processed Food List is backed by data from the USDA FoodData Central database, which contains nutrition and ingredient information for over 2 million branded food records sold in the United States.

We do not accept sponsored content, paid product placements, or advertiser influence over our analysis. Our processing scores are computed algorithmically from ingredient lists, so no product receives a better or worse score based on brand relationships.

Our content creation follows a structured process:

1. Data extraction. We query the USDA FoodData Central database for the relevant products, nutrients, or categories.
2. Processing Score calculation. Each product receives a Processing Score (uncapped, starting at 1.0) based on ingredient analysis. Products are classified into four processing levels: Minimally Processed, Processed, Highly Processed, and Ultra-Processed.
3. Analysis and writing. Our editorial team interprets the data and writes guides that explain what the numbers mean for consumers.
4. Fact-checking. All nutrient values, product names, and statistical claims are verified against source data before publication.
5. Review. Content is reviewed for accuracy, clarity, and adherence to these editorial standards.

We cite the following types of sources:

• USDA FoodData Central. Primary source for all product nutrition data, ingredient lists, and brand information.
• Peer-reviewed research. Published studies from journals including BMJ, JAMA, Cell Metabolism, The Lancet, and Nature for health-related claims. All cited studies link directly to their DOI (Digital Object Identifier) on PubMed or the publisher's site.
• NOVA classification system. The internationally recognized framework developed by researchers at the University of São Paulo for categorizing foods by degree of processing.
• Government dietary guidelines. USDA Dietary Guidelines for Americans and FDA Daily Reference Values for nutrient recommendations.

We do not cite social media posts, anonymous blogs, or manufacturer marketing materials as authoritative sources.

Our health-related guides include numbered inline citations that link to a references section at the bottom of each page. Every referenced study links directly to its DOI on PubMed or the publisher's website, allowing readers to verify claims against the original research.

We prioritize the following types of evidence:

• Randomized controlled trials (RCTs). The gold standard for establishing causal relationships, such as the NIH's 2019 controlled feeding trial on ultra-processed food consumption.
• Systematic reviews and meta-analyses. Studies that aggregate findings from multiple trials or cohorts to provide a broader evidence base.
• Prospective cohort studies. Large-scale observational studies that follow populations over time, such as NutriNet-Santé, the UK Biobank, and the Framingham Offspring Study.

When we describe research findings, we include study authors, publication year, journal name, and specific outcomes reported. We distinguish between association and causation, and we note when evidence is preliminary or limited to animal models.

Our processing scores are derived from ingredient list analysis. The scoring algorithm considers:

• Number of ingredients
• Presence of additives (artificial colors, flavors, preservatives, emulsifiers)
• High-fructose corn syrup and other ultra-processed sweeteners
• Hydrogenated and modified oils
• Protein isolates and other industrially refined components

For complete methodology details, see our Processing Score Guide.

We are building an Expert Review Board of registered dietitians and food scientists to review health-related content. Pages that have been reviewed by a credentialed expert display a "Reviewed by" byline with the reviewer's name and credentials.

Until expert review is in place, all content is published under our organizational byline and clearly labeled as informational, not medical or nutritional advice.

We update content when the underlying data changes, such as when the USDA updates its database or when new research is published that affects our methodology.

If you find an error in our data or content, please contact us. We take corrections seriously and will update or correct published content promptly.