Recommendations

Recommendations are personalized, actionable lifestyle guidance generated from a User's health data and profile context. Your application can use recommendations to help Users understand which actions may support their biomarkers, goals, and overall health experience.

InsideTracker recommendations are non-prescription health optimization guidance. They are not medical diagnoses or treatments.

The Recommendations API returns recommendations grouped by marker abbreviation or marker group, then by recommendation type. Use this page when your application needs to display recommendation content outside of the Action Plan workflow.

Group by marker

Retrieve recommendation groups keyed by biomarker abbreviation or marker group.

Show action content

Display recommendation title, text, frequency, imagery, and supporting references when present.

Connect to plans

Use Action Plan APIs when Users need to select, schedule, or check in on recommendations.

Where This Fits

Use this page after the User has enough health data and profile context for InsideTracker to generate recommendations. This page focuses on retrieving recommendation content; Action Plan covers selection, scheduling, notifications, goals, and check-ins. ProTips covers timely insights that reinforce recommended actions in context.

What Recommendations Are

InsideTracker recommendations are selected from a curated, evidence-based Knowledge Base, not generated as open-ended chatbot advice. A recommendation is delivered only when a User's biomarker data, stated goals, and physiological context indicate a likely benefit.

This gives your application personalized guidance that is consistent, explainable, and grounded in curated human clinical research. The goal is not to surface novel advice for its own sake, but to identify which actions are relevant for a specific User, why they apply, and how they should be implemented.

InsideTracker's science team reviews peer-reviewed human clinical research for each recommendation, including the intervention, biomarker effect, study population, dose, frequency, evidence consistency, and safety considerations. Recommendations are added to the InsideTracker Knowledge Base only when the evidence supports clear, actionable guidance.

Impact Score And Scientific Grade

Two recommendation fields help you prioritize and explain recommendation content:

Field	Meaning	How to use it
`impactScore`	Ranks recommendations from `0` to `10` by evaluating relevant blood biomarkers, the recommendation's Scientific Grade, and association with the User's selected goal when goal context is available. Higher scores indicate greater potential impact on health and wellness.	Use Impact Score to prioritize recommendations in experiences that rank or select actions for a User. Recommendations are sorted from greatest to least impact; ties are resolved alphabetically.
`gradeScore`	The recommendation's Scientific Grade on a `0` to `5` scale. It reflects the source material that contributed to the recommendation and is also used as part of the impact score.	Use Scientific Grade as evidence-strength context when your experience explains why a recommendation is well-supported.

Data Needed For This Output

Data	Required?	How InsideTracker receives it	Why it matters	If it is missing
Linked InsideTracker customer record and customer-level token	Required	Create the linked customer record for the User and store the returned credentials.	Recommendation endpoints are customer-level APIs.	Your application cannot retrieve User-specific recommendations.
Blood result data	Required for biomarker-based recommendations	Add Blood Result, Blood Result Upload, OCR, or Lab Connect.	Recommendations are grouped by marker abbreviation or marker group.	Marker-specific recommendation groups may be empty or unavailable.
Profile and onboarding context	Recommended	Adding Profile Data and onboarding questionnaire value sets.	Profile context can improve personalization and helps outputs interpret the User's answers.	Recommendations may be less personalized, or some recommendation logic may not apply.
Inner Age result	Conditional	Inner Age calculation from eligible blood result data.	Needed only when retrieving recommendations tied to the latest Inner Age result.	Inner Age-specific recommendation content is not available.

Workflow

Use the general Recommendations endpoint when you need recommendation content grouped by marker. Use Action Plan when the User is selecting recommendations, setting a schedule, receiving notifications, or recording adherence.

GET /api/customer/recommendations/distributed-by-marker-abbrs
Authorization: Bearer {customer_access_token}

The Draft Action Plan API can also return recommendations for plan creation. That workflow is intentionally narrower than the general Recommendations endpoint: it returns the most impactful recommendations for the User, limited to items with impactScore greater than 6.0.

Output Availability

State	What it means	Application handling
Recommendation groups returned	The response includes an `items` object with one or more marker or marker-group keys.	Render recommendations by marker or map them into your own display model.
Marker has no recommendations	A marker may have no recommendation entries even when marker data exists.	Avoid implying that every marker always produces recommendations.
Empty or missing recommendation groups	The User may not have enough applicable data or recommendation logic for the requested context.	Show an empty state or route the User to the next data step when that is appropriate in your product.

Response Shape

The response includes an items object. Keys are marker abbreviations or marker groups. Each key contains recommendation type groups, and each type group contains recommendation entries.

{
  "items": {
    "RDW": {
      "LIFESTYLE": [
        {
          "type": "LIFESTYLE",
          "title": "Daily Walking",
          "text": "Walking for 30 minutes every day can help improve overall health.",
          "frequencyDescription": "Daily",
          "impactScore": 8.5,
          "gradeScore": 4.2,
          "basedOn": ["Exercise studies", "Health data"],
          "imageUrl": "https://example.com/images/walking.jpg",
          "references": [
            {
              "title": "The Health Benefits of Walking",
              "date": "2023-03-01",
              "sourceName": "Health Journal",
              "url": "https://example.com/articles/walking-benefits"
            }
          ]
        }
      ]
    }
  }
}

Recommendation Fields

Recommendation entries can include:

Field	Description
`type`	Recommendation type or category returned by the API, such as lifestyle or nutrition.
`title`	Short recommendation title.
`text`	Recommendation body text.
`frequencyDescription`	Human-readable frequency guidance, when present.
`impactScore`	Impact score returned by the API on a `0` to `10` scale. See Impact Score And Scientific Grade.
`gradeScore`	Scientific Grade returned by the API on a `0` to `5` scale, when present. See Impact Score And Scientific Grade.
`basedOn`	Inputs or rationale categories returned by the API.
`imageUrl`	Recommendation image URL, when present.
`references`	Scientific or supporting references returned with the recommendation.
`extra`	Additional recommendation or User context returned by the API.
`checkIn`	Check-in context returned with the recommendation, when present.
`blogLinks`	Related InsideTracker blog links, when present.

Inner Age Recommendations

The API Reference also includes an InnerAge-specific endpoint for recommendations tied to the latest Inner Age result:

GET /inner-age/customer/v1/result/latest/recommendations
Authorization: Bearer {customer_access_token}

Use that endpoint only when your experience is specifically explaining recommendations in the context of Inner Age contributors. For general recommendation browsing or marker-grouped display, use the Recommendations endpoint above.

Best Practices

Implementation guidance

Treat the marker and recommendation-type keys as API data, not a fixed taxonomy hard-coded in your application.
Do not assume every biomarker has recommendations.
Use Impact Score and Scientific Grade consistently in your UI, and avoid inventing score ranges or labels that are not supported by your product experience.
Preserve scientific references and blog links when your UI exposes supporting context.
Keep recommendation retrieval separate from Action Plan selection, scheduling, notifications, and check-ins.

Implementation Checklist

You have a customer-level access token for the User.
The User has blood result data or other source data that can support recommendations.
Profile and onboarding data is submitted when your experience depends on personalized recommendation context.
Your UI handles empty recommendation groups.
Your data model supports marker keys and recommendation type groups returned by the API.
Your display handles optional fields such as imageUrl, references, extra, checkIn, and blogLinks.
Action tracking flows use Action Plan APIs rather than the Recommendations retrieval endpoint.

Related API Reference

GET /api/customer/recommendations/distributed-by-marker-abbrs
Inner Age-specific: GET /inner-age/customer/v1/result/latest/recommendations
Related topic: Action Plan
Related topic: ProTips