Back to Trends

How Your Two Scores Work

This app uses two separate scores from 0 to 100. Higher is better. Both scores help fast decisions, but they measure different things and different time horizons.

70-100: Good
40-69: Mixed
0-39: Bad

Overview

Use these scores together, but do not treat them as identical. One score is mainly about near-term glucose response, while the other is mainly about longer-term heart/lipid pattern.

Blood Sugar Score (Detailed)

This score focuses on glucose spike risk. It is mostly driven by glycemic load, added sugar, fiber, protein, and refined/liquid sugar penalties.

Practical meaning: this score helps answer, "How likely is this food to push my glucose up quickly?" It is not a lab glucose reading and not a diagnosis tool. It is a consistent decision aid.

available_carbs = max(carbs_g - fiber_g, 0) GL = (GI * available_carbs) / 100 BloodSugarScore = clamp( 100 - 1.8 * GL - 0.7 * added_sugar_g + 0.8 * fiber_g + 0.25 * protein_g - 12 * liquid_sugar_flag - 8 * refined_carb_flag, 0, 100 )

ApoB Score (Detailed)

This score focuses on likely ApoB/LDL-friendly profile. It penalizes saturated/trans fat and processed red meat, and now also penalizes sugary refined foods so candy-like foods do not look falsely healthy.

Practical meaning: this is a long-run pattern score, not a short-term glucose spike score. It helps separate foods that may be okay for glucose today but less ideal for heart-lipid goals over time.

ApoBScore = clamp( 100 - 5 * saturated_fat_g - 20 * trans_fat_flag - 8 * processed_red_meat_flag + 1.2 * fiber_g + 6 * unsaturated_fat_flag - 1.0 * added_sugar_g - 10 * refined_carb_flag - 8 * liquid_sugar_flag, 0, 100 )

Confidence & Score Versions

The confidence badge describes how clearly the app could identify the food, portion, or package label. It does not describe medical certainty. High, medium, and low are qualitative evidence categories, not calibrated probabilities.

Quick Interpretation