Where Our Data Comes From
OxalateGuard is built on transparency. Every value in our database is traceable to peer-reviewed research, and every calculation is documented here.
15+ Peer-Reviewed Sources
We don't rely on a single lab or a single study. Our database aggregates measurements from independent research teams across the world.
Harvard T.H. Chan School of Public Health
Comprehensive fresh-weight oxalate database
Wake Forest University
Clinical oxalate measurements for common foods
Oxalosis & Hyperoxaluria Foundation
Patient-focused food oxalate reference data
Siener et al. (2006-2021)
German clinical studies across multiple food categories
MDPI 2023
Updated measurements from recent peer-reviewed publication
Savage & Vanhanen
Indian spices, mushrooms, and specialty foods
Judprasong 2006
Thai foods and Southeast Asian ingredients
Avila-Nava 2021
Mexican ethnic foods including moles, nopales, and chilies
Honow 2010
Comprehensive tea oxalate measurements
Abdel-Moemin 2014
Egyptian herbs and traditional ingredients
Plus additional data from KidneyCop, Dr. Duke's Phytochemical Database, Sally Norton, Oxalate.org aggregations, and community contributions via Scout Mode.
How We Calculate: The Consensus Algorithm
When multiple studies measure the same food, values can vary. Instead of trusting a single measurement or taking the maximum, we use a consensus approach designed to give you the most representative value.
Collect all measurements
For each food, we gather every published measurement from our source library. A single food might have values from 2 to 8 independent studies.
Identify dry-weight outliers
Some older studies (notably USDA 1984 and Dr. Duke's Phytochemical Database) reported values on a dry-weight basis. Because water makes up 70-95% of most fresh foods, dry-weight values can be 10-40x higher than what you actually consume. We flag these and exclude any that exceed 5x the median of other measurements.
Compute the median
After filtering, we take the median of the remaining values. The median is more robust than the average because it resists the pull of any single extreme measurement. This gives you a value that represents the typical oxalate content of the food as you eat it.
Example: Spinach, cooked
Why Numbers Differ From Other Sources
If you have compared oxalate values across different websites, you have probably noticed they don't agree. There are real reasons for this, and understanding them helps you make better dietary decisions.
Dry-weight vs. fresh-weight reporting
The single biggest source of discrepancy. A dried spinach measurement might report 970 mg/100g, while the same spinach measured fresh (as you eat it) might be 755 mg/100g. Many popular lists unknowingly mix these values.
Measurement methods
Different labs use different analytical methods (HPLC, enzymatic assays, colorimetric methods). Each has slightly different accuracy profiles, which contributes to natural variation between studies.
Cultivar and growing conditions
Oxalate content varies based on the specific variety of a plant, the soil it grew in, the climate, and the time of harvest. A beet grown in Germany may have different oxalate levels than one grown in Thailand, even though both are reported as “beet.”
Single-source lists
Many websites simply copy from one source without cross-referencing. This means their values are only as good as that one study. Our multi-source consensus approach accounts for natural variation and catches outliers that a single source cannot.
Full Transparency on Every Food
We don't ask you to trust us blindly. Every food detail page in our database of 2,500+ foods includes a Source Breakdown section where you can see each individual measurement, the study it came from, and whether it was a fresh-weight or dry-weight value. You can trace any number back to its origin.
Risk levels are clearly labeled using established thresholds: Low (<25 mg/100g), Moderate (25-99 mg), High (100-299 mg), and Very High (300+ mg). We also show per-serving values so you know the actual impact of a realistic portion.
Ongoing Updates
Our database is not static. We actively integrate new research as it is published and process community contributions through our Scout Mode feature, where users photograph product ingredient labels that we then analyze and verify.
New research sources are added as peer-reviewed studies are published. Recent additions include Avila-Nava 2021 (Mexican ethnic foods) and updated Siener data.
Community contributions via Scout Mode let users photograph ingredient labels on packaged products. Submissions are verified before being added to the database.
Deduplication and quality control processes run regularly to merge duplicate entries, reconcile cross-aggregator source overlaps, and remove entries that relied solely on dry-weight data.
Important Limitations
- Oxalate content varies by growing conditions, season, and preparation method
- Values are estimates based on published research averages, not guarantees for any specific item you purchase
- Cooking reductions depend on water volume, cook time, and method
- Individual absorption varies based on gut health, calcium intake, and other dietary factors
- This tool is not a substitute for medical advice from your urologist or dietitian
Frequently Asked Questions
Where does OxalateGuard get its oxalate data?
We aggregate measurements from 15+ peer-reviewed sources including Harvard T.H. Chan School of Public Health, Wake Forest University, the Oxalosis and Hyperoxaluria Foundation (OHF), Siener et al. (2006-2021), MDPI 2023, Savage & Vanhanen, Judprasong (2006, Thai foods), Avila-Nava (2021, Mexican ethnic foods), and several others. Each food may have data from multiple independent studies.
Why do your oxalate values differ from other websites?
Most free oxalate lists rely on a single source, often the USDA 1984 dataset which reported dry-weight values. Dry-weight measurements can overstate oxalate content by 10-40x compared to the food as you actually eat it. OxalateGuard cross-references multiple modern sources and filters out dry-weight outliers so you get values that reflect real-world portions.
How does the consensus algorithm work?
For foods with a single measurement, we use that value as-is. For two measurements, we compare and exclude any dry-weight value that exceeds 5x the other. For three or more, we compute the median, exclude dry-weight sources whose values exceed 5x the median, and return the median of the remaining cluster. This prevents any single extreme measurement from distorting the result.
Can I see which studies measured each food?
Yes. Every food detail page in our database includes a Source Breakdown section showing each measurement source, its reported value, and whether it is a fresh-weight or dry-weight measurement. You can verify exactly how we arrived at the consensus value for any food.
Explore the Database
Browse 2,500+ foods with multi-source consensus values, per-serving data, and full source breakdowns.