GlucholGluchol methodology and prediction limits
Separate food recognition, nutrition estimates and dietary-impact categories. Understand assumptions, uncertainty and what Gluchol does not measure.
Identification is not measurement
Gluchol's public methodology distinguishes describing a food from estimating its dietary impact. Photos, barcodes, menus and food entries provide different levels of information. A product label can offer nutrient details; an image usually requires more assumptions about recipe and portion.
The public website does not establish the precise formula, thresholds, model weights or validation performance of every app version. These must not be inferred or fabricated. Ask Gluchol via the contact page for clarification before using the model in research or clinical decisions.
Three separate educational outputs
Blood sugar/glucose impact, cholesterol-related dietary risk and blood-pressure-related dietary factors address different questions. Low, Moderate and High are educational categories, not diagnostic criteria. A category does not prove a future clinical outcome.
Personalised estimates and sensitivity-related insights may incorporate recorded history, but do not demonstrate measured physiological sensitivity. Food records are incomplete and may contain errors.
What evidence would strengthen the model?
Transparent validation would require a defined population, an appropriate reference measurement, a specified app version and reported uncertainty. This website does not claim clinical validation or publish invented accuracy statistics.
Authoritative nutrition references explain general relationships between diet and health. They do not by themselves validate the Gluchol scoring system. Keep that distinction when quoting this website.
Continue reading
Related pages
Build a clearer picture of your meals
Explore food choices with Gluchol’s educational estimates, with context for what they can and cannot tell you.