CalorieSciencePopulation guidance misses individual response
Standard guidelines matter for public health. They do not explain why two clients on the same plan get different results.
Grounded in evidenceExplore the science
800
participants · 46,898 meals
wide variation in blood glucose response to identical foods
Zeevi et al., Cell, 2015 7 RCTs873 participants · prediabetes and metabolic syndrome
precision nutrition interventions consistently improved HbA1c and waist circumference, with mixed results on lipids and blood pressure
Robertson et al., Nutrients, 2024 3,500+BANT-registered practitioners in the UK
a profession built on individualised care
BANT Practitioner RegisterThe gap in generic plans
Why generic diet plans fail your clients
Two people can follow the same “healthy” plan and get very different results. The issue is rarely effort. Biology, medications, routine and preferences rarely fit a static template.
A plan that ignores individual context quickly becomes hard to follow, hard to explain and hard to adapt.
Personalised advice changes behaviour
The Food4Me trial showed personalised online nutrition advice led to more appropriate dietary changes than generic population guidance.
The same meal can land differently
Postprandial glucose studies show substantial variation in blood sugar responses to identical meals, even in people without diabetes.
Personalisation in practice
What personalised nutrition means in practice
Personalised nutrition puts the individual's biology, history and lived experience at the centre of care, instead of fitting them into a generic protocol.
UK bodies such as BANT describe personalised nutrition as tailoring recommendations to symptom patterns, clinical markers, genetics, microbiome, lifestyle and preferences, within a structured, evidence-based framework.
On CalorieScience, that means:
One client view
Bring dietary intake, blood biomarkers, client-reported medications, lifestyle patterns and goals into a single workspace.
Structured practitioner workflow
Turn assessment data into food and lifestyle strategies shaped around clinical priorities, risk factors and readiness for change.
Clinical control stays with you
Use AI-assisted tools to move faster, then review, edit and approve the plan before it reaches the client.
Evidence base
The science behind personalised nutrition
The evidence for personalised nutrition has grown rapidly over the last decade, especially for digital models that combine assessment data with practitioner-led advice.
The literature points toward more precise, behaviour-change-oriented interventions when individual context is actually used in the plan.
Trial-backed dietary improvement
Randomised controlled trials such as Food4Me showed personalised advice delivered digitally can improve diet quality beyond generalised advice.
Multiple data streams matter
Recent reviews show that combining diet, phenotype and clinical markers with decision support enables more precise interventions.
Designed for applied practice
CalorieScience is built around evidence-based personalisation, data integration and practitioner-guided behaviour change, not quick-fix meal plans.
From information to action
How CalorieScience supports personalised meal planning
Unified client nutrition data
Bring together the information you already collect in one secure workspace:
- Food diaries and 24-hour recalls
- Symptom timelines, health history and medications, as reported by the client
- Anthropometrics and blood report results (for example HbA1c, lipids, vitamin and mineral panels)
- Activity, sleep and stress patterns, as reported by the client
This mirrors the whole-person assessment registered practitioners in the UK and Europe are encouraged to use, rather than isolated metrics.
Biomarker and medication intelligence
Traditional nutrition software stops at macronutrient and micronutrient targets. CalorieScience builds on that foundation:
- Upload a client's blood report and see markers flagged against reference ranges, with trends across reports
- Get medication-linked nutrient alerts, for example metformin and B12, PPIs and magnesium, statins and CoQ10
- Analyse diet quality against your chosen guidelines and reference values
Personalised meal planning, not just calorie counting
Using your assessment, generate and refine meal plans that:
- Respect clinical constraints (for example carbohydrate distribution, sodium limits) where appropriate to the client's care plan
- Reflect cultural background, food access, budget and preferences, key drivers of long-term adherence
- Include practical swaps that support incremental change and reduce discretionary foods
You stay in full control of the plan. The software makes it faster and more consistent to turn complex data into real-world menus.
Key inputs: biomarkers, medications, lifestyle and preferences
A personalised plan is only as strong as the data behind it. The most clinically useful inputs fall into a few categories, and the combination matters as much as any single one.
Supported today
Biomarkers and phenotype
Objective measures anchor planning around risk, response and progress:
What this includes
- Glycaemic markers (for example HbA1c, fasting glucose)
- Lipids and cardiometabolic risk markers
- Vitamin and mineral status
- Weight, body composition and waist measures
Supported today
Medications
Client-reported medications are checked against known drug–nutrient interactions, so depletion risks are flagged when you plan.
Supported today
Lifestyle, environment and preferences
The most clinically elegant plan still fails if it doesn't fit how the client lives:
What this includes
- Sleep, stress, activity and work patterns, which shape appetite, food choices and metabolic response
- Dietary preferences, cultural context, ethical choices and practical constraints
On our roadmap
Genetics, microbiome and continuous glucose
These inputs can add nuance when interpreted within the wider clinical picture. Nutrigenetic results can indicate variation in nutrient metabolism, but should be interpreted cautiously. Microbiome-based suggestions complement, not replace, clinical judgement.
Availability
- Genetic, microbiome and CGM inputs are on the CalorieScience roadmap. Blood biomarkers and medications are supported today.
Each input interacts with the others. A client's ideal macronutrient split may need adjusting to fit shift work or family cooking. A medication may change which nutrient gaps matter most. Holding all of this together for every client, at caseload scale, is where software helps.
Data governance
Any platform processing health data must comply with the EU GDPR for clients in the EU, and with UK GDPR and the Data Protection Act 2018 for clients in the UK. Health data is special category data and needs appropriate safeguards.
CalorieScience uses encryption and access controls. Our privacy notices explain how personal data is handled. Client data is hosted in the EU (Frankfurt) and is never used to train AI models.
More time for care
How AI enables personalisation at scale
The science behind personalised nutrition has been understood for years. The barrier has been operational. A practitioner with 50 or 100 clients cannot manually cross-reference blood results, medications, intake, lifestyle constraints and preferences to build and update a tailored plan for each person every few weeks.
That takes software. The Weizmann Institute study showed a machine-learning model combining individual data could predict personalised glucose responses to real meals, and that diets based on those predictions outperformed standard advice.
Personalised meal planning without the admin
Meal planning is one of the most time-intensive parts of practice. You balance clinical requirements, nutrient targets, preferences, health conditions, cultural needs and adherence, and doing that by hand for every client doesn't scale.
CalorieScience generates a personalised meal plan from structured client information, biomarker results and your parameters in minutes, compared with the 30–60 minutes many practitioners spend building one by hand. You get more time for consultations.
The missing piece: adherence
A personalised plan only creates value when it's followed. Many practitioners say adherence, not plan creation, is their biggest challenge: forgetfulness, complexity, lifestyle constraints, gaps between appointments and poor visibility into progress.
That's why CalorieScience covers both planning and follow-through. From 21 October, clients can log and rate meals in the companion app, so you can see what's working before the next session.
Why practitioners choose CalorieScience
- AI-assisted meal planning you review and approve
- Blood biomarker analysis with medication-linked alerts
- Meal plans in minutes, not 30–60
- Structured client onboarding and assessment forms
- Practice management: calendar, appointments, invoicing
- Client companion app
- EU-hosted data, never used to train AI
A connected clinical workflow
How it works: from data to delivered plan
Bring client data together
Lab results, intake questionnaires, food diaries and client-reported medications and lifestyle combine into a single client profile.
Generate a personalised meal plan
Using that profile and your clinical priorities, the AI drafts a plan balanced across nutrient targets, preferences, cultural context and practical constraints, in minutes.
Review and approve
Every recommendation comes to you for review, annotation and override before the client sees it. The platform handles the computation; you keep the clinical judgement.
Track adherence and adapt
Coming 21 October: clients log and rate meals so you can see what's being followed.
Deliver to the client
Clients receive their plan, recipes and progress tracking in the CalorieScience companion app.
Your expertise. Always at the centre.
How AI supports, not replaces, your expertise
- 01Flags blood markers outside reference ranges and medication-linked nutrient risks
- 02Drafts meal plans that match a client's targets and preferences, for you to review and adjust
- 03Automates repetitive work, such as turning diaries into analysable data, so you spend more time on consultation and behaviour change
You decide which suggestions to accept. Every recommendation is transparent and editable.
Personalisation in the real world
Example practitioner journey (illustrative)
This is a composite scenario based on common use cases in UK and European practice, not an individual case or a promised outcome.
The starting point
A registered dietitian in Manchester sees a client with fluctuating energy, weight concerns and borderline glycaemic markers from their GP. The client takes metformin.
Blood biomarkersMedication contextIndividual care
- 01
She uploads the client's blood report and food diary, and records reported sleep and stress patterns.
- 02
CalorieScience flags the borderline HbA1c, a low-normal B12 alongside metformin use, and frequent late-night high-glycaemic snacks.
- 03
She builds a phased plan: better carbohydrate quality and timing, more protein- and fibre-rich options, B12-rich foods, and fewer discretionary snacks. She edits the AI draft before approving it.
- 04
At follow-up she reviews meal logs and repeat bloods, and refines the plan.