TRANSPARENCY

Metabolic Syndrome

This document follows the CHAI Applied Model Card format (v0.1). It is not CHAI-certified or CHAI-endorsed.

Header

Model name
Metabolic Syndrome
Developer
Zenlo LLC
Release stage
Research Tool (not FDA-cleared)
Version
1.0
Availability
United States
Regulatory status
Not applicable — academic and transparency positioning
Pattern slug
metabolic_syndrome
Biomarkers
GLU, INSULIN, TG, HDL, HBA1C

This document follows the CHAI Applied Model Card format (v0.1).

Summary

The Metabolic Syndrome pattern evaluates a cluster of metabolic laboratory abnormalities aligned with ATP-III–style criteria available from standard fasting panels: HOMA-IR–derived insulin resistance, elevated triglycerides, low HDL cholesterol, and dysglycemia (fasting glucose or HbA1c). A deterministic detector counts qualifying criteria; sufficient abnormalities fire the pattern. Claude Haiku 4.5 provides narrative synthesis for physician review. For licensed functional-medicine physicians, this is supportive clinical decision support — it does not measure waist circumference or blood pressure, which are required components of full ATP-III metabolic syndrome diagnosis. Outputs highlight metabolic clustering in laboratory data for clinical correlation, not standalone syndrome diagnosis.

Uses & Directions

Intended use

Clinical decision support for licensed physicians evaluating metabolic clustering in adult fasting laboratory panels.

Primary users

Licensed functional medicine physicians and similarly qualified clinicians.

How to use

Review flagged metabolic criteria with blood pressure, waist circumference, weight, and lifestyle context.

Target population

Adults aged 18 and older in the United States.

Out of scope

  • Direct patient use without physician oversight
  • Pediatric populations
  • Standalone ATP-III metabolic syndrome diagnosis without clinical measures
  • Type 1 diabetes management

Warnings

Clinical risk level

Low — supportive tool; the treating physician retains full clinical judgment and responsibility.

Known limitations

  • NHANES audit used a 4-lab subset only — blood pressure and waist circumference were unavailable, so detected prevalence is lower than full ATP-III.
  • HOMA-IR requires fasting state; non-fasting samples reduce reliability.
  • HDL thresholds are sex-specific; missing sex defaults conservatively.

Trust Ingredients

AI system facts

  • Deterministic pattern detector (metabolic_syndrome) plus Claude Haiku 4.5 for narrative synthesis
  • Primary inputs: Fasting glucose, fasting insulin, triglycerides, HDL cholesterol, and HbA1c from structured extraction.
  • Output: Pattern flag plus narrative on metabolic criteria met.

Security & compliance

  • Anthropic Business Associate Agreement with zero-data-retention configuration
  • HIPAA-aligned design; no patient data used for model training

Ongoing maintenance

Versioned, transparent, and reproducible via a public independent audit harness (see Resources).

Transparency

Self-funded development; no third-party sponsor for this pattern card.

Key Metrics

Usefulness / Efficacy

Zenlo's detection approach was benchmarked across five models in a separate study; see the medRxiv preprint in Resources. No per-pattern efficacy metric is published for this pattern.

Source: 5-model benchmark, medRxiv MEDRXIV/2026/346284

Fairness / Equity

Overall prevalence detected: 10.3% (615 / 5,992)

By sex

SubgroupnPrevalence detected
Female3,1059.6%
Male2,88711.0%

By age group

SubgroupnPrevalence detected
18–392,2267.1%
40–591,86512.4%
60+1,90111.8%

By race / ethnicity

SubgroupnPrevalence detected
Mexican-American1,06412.7%
Other Hispanic79812.3%
Non-Hispanic White1,91410.9%
Non-Hispanic Black1,2656.7%
Non-Hispanic Asian7268.3%
Other / Multiracial22512.4%

This audit applied a 4-laboratory subset (fasting glucose, triglycerides, HDL, HbA1c) without blood pressure or waist circumference — two of five ATP-III criteria were unavailable in NHANES 2015–2016 for this run. Detected prevalence is therefore lower than full ATP-III metabolic syndrome rates in the literature.

These are descriptive population rates from the audit harness — not model error rates or benchmark accuracy metrics.

Source: Independent NHANES 2015–2016 audit harness (N=5,992, commit c10afe8)

Safety / Reliability

  • Supportive-only; not intended as a standalone diagnostic
  • Physician authorization required before clinical use
  • Deterministic detector is reproducible for inputs: GLU, INSULIN, TG, HDL, HBA1C

Resources

Footer

Note: The mention or sharing of any examples, products, organizations, or individuals does not indicate any endorsement of those examples, products, organizations, or individuals by the Coalition for Health AI (CHAI).

HIPAA-aligned design · ZDR active · Anthropic BAA · Pending legal review

Zenlo Labs is a clinical decision support tool intended exclusively for use by licensed healthcare providers. Not a substitute for professional medical judgment. Not intended for direct patient use.