How the Health Freak Score Works

A transparent, evidence-based nutritional quality score for restaurant food and everyday meals — built on peer-reviewed science, not proprietary black-box algorithms.

Download Scoring Methodology (PDF)

What is the Health Freak Score?

The Health Freak Score is a numerical rating from 0 to 10 that tells you, at a glance, how a dish or meal performs across evidence-based nutritional quality markers. Unlike calorie counts alone — which tell you how much energy something contains but nothing about whether it is actually nourishing — our score reflects the full picture: protein quality, fibre, micronutrient breadth, plant diversity, food processing level, and the balance of saturated fat, salt, and sugar.

Higher is better. A score of 7 or above means a dish performs strongly across all major quality dimensions. A score below 3 usually indicates limited nutritional value — high in free sugars, saturated fat, or ultra-processed ingredients with little protein or fibre in return.

The score is not a verdict on you or your food choices. It is information — the same kind a registered dietitian would consider when advising someone on the nutritional quality of their diet.

Why We Built It

UK adults now eat approximately one in four meals outside the home. Mandatory calorie labelling has helped, but calories alone are a deeply insufficient guide to nutritional quality. A 500 kcal salmon bowl and a 500 kcal croissant carry the same calorie label. They are not nutritionally equivalent.

Health Freak exists to close that gap — for restaurants, for diners, and for anyone who wants to understand their food better. Our scoring system is published in full. Every coefficient, every scientific reference, every calibration decision is documented. If you disagree with a weighting, you can tell us why.

What the Score Measures

The Health Freak Score is built from four components:

1. Macronutrient Quality Score (MQS)

Evaluates the balance of beneficial and harmful macronutrients, normalised per 100 kilocalories so portion size doesn't distort the comparison.

Positive factors

  • Protein density — rewarded above 3g per 100 kcal (Leidy et al. 2015)
  • Dietary fibre — rewarded above 2g per 100 kcal (Reynolds et al. 2019, Lancet)

Penalty factors

  • Saturated fat — dose-response for LDL cholesterol (Hooper et al. 2020)
  • Free sugars — blended formula (WHO 2015; Te Morenga et al. 2012)
  • Salt — UK's leading modifiable CV risk factor (Aburto et al. 2013)

2. Micronutrient Richness Score (MRS)

Measures how broadly a meal covers your daily micronutrient needs across 12 essential nutrients:

Vitamin A · Vitamin C · Vitamin D · Vitamin B12 · Folate · Iron · Calcium · Magnesium · Zinc · Potassium · Selenium · Omega-3

Each nutrient is measured as a fraction of the UK Dietary Reference Value (DRV), capped at 100% so no single nutrient can inflate the score unfairly.

3. Plant Diversity Bonus

A bonus of up to +0.5 points for the number of distinct plant food groups present: wholegrains, legumes, vegetables, fruits, and nuts/seeds. Evidence from the American Gut Project (McDonald et al. 2018) and Wastyk et al. (2021, Cell).

4. Food Processing Penalty

Based on the NOVA classification. Ultra-processed foods carry a 0.8 point deduction — proportionate to the 14% higher all-cause mortality found in the NutriNet-Santé cohort (Srour et al. 2019, BMJ). The only RCT of ultra-processed vs unprocessed diets (Hall et al. 2019) found participants consumed 508 more kcal/day on the ultra-processed diet.

Final blend

Base Score = (MQS × 75%) + (MRS × 25%) + plant bonus − processing penalty

Five Scoring Profiles

Because nutritional goals differ, the Health Freak Score adapts to five evidence-based profiles. The same dish, scored through five different lenses:

Base

General healthy eating

Balanced macronutrients + micronutrient breadth

Protein Prime

Active individuals, training goals

Protein density above all else

Fibre First

Gut health, portfolio diet

Fibre density + plant diversity

Lean Cut

Fat loss, body composition

High protein + low total energy density

Bio Hacker

Longevity, healthy ageing

Longevity-weighted micronutrients — omega-3, vitamin D, selenium, B12

The Bio Hacker profile uses a longevity-weighted MRS where nutrients are re-weighted according to the strength of evidence for healthy ageing. Based on the REDUCE-IT trial (Bhatt et al. 2019, NEJM) and the VITAL trial (Manson et al. 2019, NEJM).

Daily Meal Score

When you log multiple foods throughout the day, Health Freak aggregates your nutritional intake and calculates a rolling daily score. This is more meaningful than scoring items in isolation — a banana scores modestly on its own, but as part of a day that includes oily fish, legumes, and leafy greens, it contributes to strong potassium and vitamin C coverage. The formula is designed at meal and day level, not ingredient level.

Development Sample Results

Our scoring formula was calibrated against 318 dishes from four UK restaurant chains:

Restaurant Dishes scored Average base score
itsu785.37 / 10
Farmer J445.29 / 10
Pret a Manger1694.97 / 10
The Salad Project274.94 / 10

6% of dishes scored 7.0+, 57% scored 5.0–7.0, 16% scored below 3.0. Highest-scoring dishes were oily fish preparations — salmon sashimi, miso salmon, shawarma chicken.

What the Score Is Not

Frequently Asked Questions

Why does a banana score lower than a chicken dish?

The formula scores complete meals, not individual ingredients. A banana has excellent potassium and vitamin C but limited protein as a standalone item. In the context of a full day's eating it contributes positively. Use the daily score for single whole foods.

Why is the Bio Hacker score generally lower?

It weights omega-3, vitamin D, selenium, and B12 very heavily — nutrients that are genuinely sparse in most restaurant food. Only oily fish dishes score highly. A Bio Hacker score of 6 represents an excellent eat-out choice for longevity nutrition.

Does a high score mean I should eat this every day?

No. Nutritional diversity matters. A varied diet — even with some lower-scoring items — may be more beneficial long-term than identical high-scoring meals.

Can restaurants improve their scores?

Yes — by providing analytically verified nutritional data, by reformulating dishes to reduce salt or ultra-processed ingredients, or by having their existing profile accurately reflected. We work directly with restaurant partners on all of the above.

References

Health Freak scoring methodology v0.3 · Last updated March 2026 · methodology@healthfreak.ai

Download the full white paper

Complete methodology, every formula coefficient, all 35+ scientific references, and our full development dataset analysis.

Download Scoring Methodology (PDF)