A Case Study in Diagnostic Failure

The data existed, but there was no system to connect the dots.

I had various seemingly unrelated symptoms for years and brushed them off until they became unbearable. I saw 5-10 doctors over the past 5 years—different specialists—all unable to figure out what was going on. I finally got bloodwork through Function Health and discovered I had Hashimoto's thyroiditis, which opened my eyes and changed how I approach my health—viewing it in a more holistic, systemic way. But it took 5 years of being in the dark because no one had the tools to integrate the evidence that was already there. Now I'm finally making progress.

SCROLL

The Timeline

The diagnostic journey

A fragmented healthcare system that couldn't connect the dots.

2020

Symptoms Begin

Severe fatigue, exercise intolerance. Testing shows impairment but no diagnosis.

2021

POTS Diagnosed

TSH spikes to 8.6, then normalizes. Thyroid antibodies never tested despite family history.

Missed for 3.5 years
2024

Hashimoto's Found

Function Health's comprehensive panel catches elevated thyroid antibodies (176 IU/mL).

Diagnosis #1
2025

Infections Discovered

Specialized testing reveals Lyme, Babesia, Bartonella—never tested in 5 years of conventional care.

Diagnosis #2 & #3

The Evidence

What the data revealed

Key biomarkers that told a story—once someone finally looked at them together.

My Function Health Results

Real data showing markers indicating Hashimoto's thyroiditis

TSH Results from Function Health - 5.72 mIU/L (Above Range)

TSH: 5.72 mIU/L - Above optimal range

Thyroglobulin Antibodies Results from Function Health - 176 IU/mL (Above Range)

Thyroglobulin Antibodies: 176 IU/mL - Elevated

Pattern Recognition

How Function Health connected the dots

After five years of fragmented care across a dozen specialists, Function Health's comprehensive approach finally revealed what conventional medicine missed. Their panel doesn't just test for symptoms—it systematically screens for patterns across interconnected systems.

When Function Health tested my thyroid antibodies—something never ordered despite family history and clear warning signs—the diagnosis was immediate. Elevated TSH combined with high thyroglobulin antibodies told a clear story: Hashimoto's thyroiditis, quietly causing inflammation for years.

This wasn't a breakthrough in testing technology. These are standard labs. The breakthrough was in integration—running a complete panel proactively, not waiting for symptoms to worsen. Function Health's model proves that comprehensive, data-driven diagnostics can catch what reactive, siloed care misses. That realization changed everything about how I understood both my health journey and the massive opportunity in preventive diagnostics.

Data-Driven Diagnostics

The data was there—the integration wasn't

Two datasets that could have accelerated diagnosis by years, if only they'd been analyzed systematically.

7 Years of Physiological Data

Pattern recognition before symptoms become debilitating

HRV Analysis 2018-2026

In 2018, my Apple Watch began recording heart rate variability data. By 2021, I was experiencing severe autonomic dysfunction—but the warning signs had been in the data all along.

My HRV was chronically low—consistently around 30-40ms. For comparison, healthy men in their mid-20s to early 30s typically have HRV values of 60-100ms or higher. My values were in the range typically seen in people decades older with cardiovascular disease. The data showed clear autonomic dysfunction years before I developed debilitating symptoms.

The challenge wasn't data availability. It was interpretation. Standard medical care doesn't analyze longitudinal consumer health data. There's no system to flag concerning patterns before they become clinical emergencies. I had to export 7+ years of Apple Health data and analyze it myself to see what had been there the entire time.

Data source Apple Health export (XML)
Processing Python analysis with custom visualization
Key insight Chronically low HRV (consistently 30-40ms for a healthy 25-30 year old) indicated autonomic dysfunction years before POTS diagnosis—a pattern that could have triggered earlier intervention

Genetic Susceptibilities

Your genome has answers—if you know what to ask

Gene
Variant Status
POTS Relevance
NOS3
Double homozygous
(T-786C + rs1800779)
Severely impaired nitric oxide production → vascular dysfunction
ADRB2
GG homozygous
(R16G)
Beta-2 receptor downregulation → orthostatic intolerance

After years of unexplained POTS symptoms, I uploaded my Ancestry DNA raw data to an AI-powered genetic interpretation tool. The analysis revealed something striking: I carry double homozygous NOS3 variants (both T-786C and rs1800779) plus ADRB2 R16G—a genetic combination that creates a "perfect storm" for autonomic dysfunction.

The NOS3 variants severely reduce nitric oxide production, impairing my blood vessels' ability to dilate properly. The ADRB2 variant causes beta-2 receptor downregulation, meaning my cardiovascular system can't respond appropriately to positional changes. Together, these explain the core pathophysiology of my POTS: impaired vascular tone combined with inadequate compensatory mechanisms.

Here's what matters: this wasn't exotic genetic testing. This was standard consumer DNA data ($99 from Ancestry) that I'd had for years. What was missing wasn't the data—it was the infrastructure to interpret it in clinical context. No doctor I saw had the tools to say "Your symptoms + these variants = high probability autonomic dysfunction, let's test for POTS."

That integration gap—between having data and making it clinically actionable—is exactly what Function Health is solving at scale.

Data source Ancestry DNA raw data (677K+ SNPs)
Analysis AI-assisted interpretation using Claude + genetic databases
Key insight Multiple compound genetic vulnerabilities explain clinical phenotype—but only when analyzed systematically

The opportunity: Building the connective tissue

These two datasets—physiological tracking and genetic variants—existed in isolation for years. One lived in my Apple Watch. One lived in an Ancestry DNA file. Neither healthcare provider I saw had access to both, let alone the tools to integrate them.

Function Health is building the infrastructure to close this gap. Comprehensive lab testing plus systematic analysis plus longitudinal tracking. Not reactive medicine, but proactive pattern recognition.

I want to help build that future.

AI-Powered Optimization

Function Health's AI: From diagnosis to optimization

How AI-powered analysis transforms test results into actionable lifestyle interventions

Real Examples of Function Health's Chat

Function Health comprehensive health summary and optimization strategies

Comprehensive Action Plan

Function Health synthesizes your complete biomarker profile into a prioritized action plan. Rather than isolated test results, you get a holistic view of your health status with specific, evidence-based strategies for optimization. This includes detailed treatment protocols to discuss with your doctor—from targeted antimicrobial therapy for infections to thyroid support, metabolic optimization, and supplement recommendations tailored to your specific genetic and clinical picture.

Function Health AI explaining Vitamin D and Th1/Th2 immune balance

Mechanistic Understanding: The Vitamin D Example

Conventional Medicine: "Your Vitamin D is 30 ng/mL. That's within the normal range of 20-50 ng/mL. No treatment needed."
Function Health's AI: "While technically 'normal,' optimal Vitamin D for immune function is 50-70 ng/mL. Low Vitamin D shifts your immune system toward Th2 dominance—weakening your ability to fight the chronic Lyme and Bartonella infections you're battling, while increasing autoimmune susceptibility that's driving your Hashimoto's."

Function Health's AI doesn't just flag a low vitamin—it explains why it matters for you specifically. The analysis connects a "technically normal" Vitamin D level to immune system dysregulation, explaining how Th1/Th2 imbalance affects my ability to fight chronic infections while increasing autoimmune risk. This mechanistic understanding transforms a simple vitamin deficiency into a strategic intervention point for infection control and autoimmune regulation.

Function Health doesn't just tell you what biomarkers are out of range—it helps you understand why they matter and how to optimize them.

Traditional medicine often settles for "acceptable" ranges. Function Health's AI targets optimal ranges and provides mechanistic understanding of how to get there through lifestyle, diet, and supplementation.

Why This Matters:

Traditional medicine would have said "your Vitamin D is acceptable" and moved on. Function Health's AI:

  • Connects biomarkers to mechanisms: Explains how low Vitamin D weakens Th1 immunity
  • Personalizes recommendations: Links deficiency to my specific conditions (chronic infections, autoimmunity)
  • Targets optimization, not just normalization: 50-70 ng/mL for immune function vs. "20+ is fine"
  • Provides actionable protocols: Specific supplementation targets and monitoring plans to implement with your physician

The Bottom Line

I have POTS, Hashimoto's thyroiditis, and chronic tick-borne infections (Lyme, Babesia, Bartonella) that may be exacerbating both conditions. These diagnoses were delayed by years because healthcare operated reactively—testing only what symptoms suggested, never looking at the complete picture.

Better use of analytical tools and preventative diagnostics rather than reactionary methods improves healthcare outcomes. Function Health proved this with my case: comprehensive testing plus AI-powered interpretation caught what five years of conventional care missed.

This isn't just about me. It's about building systems that prevent diagnostic delays for millions of patients navigating the same fragmented healthcare maze. The tools exist. The data exists. What's missing is integration—and that's exactly what I want to help build.

Why This Mission

What this journey taught me about healthcare

Systemic failures that better tools—and better data—can solve.

Snapshot medicine fails patterns

My TSH fluctuated from 8.6 to 2.27 within days. Conventional medicine sees isolated data points. AI can see trajectories, correlations, and early warning signals across time.

Silos miss the bigger picture

Cardiologist sees POTS. ENT sees sinusitis. Endocrinologist sees "normal" TSH. Nobody connects the dots because nobody sees all the data. The opportunity: unified biomarker analysis.

Prevention beats treatment

By the time symptoms forced action, I had years of untreated autoimmune inflammation. Comprehensive, proactive testing could have caught this before the damage accumulated.

The tools exist—the integration doesn't

The tests that finally diagnosed me weren't cutting-edge. They were standard. The innovation was simply running them together, looking at the full picture. That's a solvable problem.

What I Bring

Skills shaped by engineering and experience

A unique combination of quantitative training, personal insight, and technical capability.

📊

Data Analysis

7+ years of personal health data analysis. Built quantitative risk frameworks processing real-time market data. Expertise in pattern recognition across complex datasets.

Python Data Viz Statistics
🤖

AI-Assisted Development

Experienced "vibe coder" leveraging AI tools to build production applications. Created dashboards, trading systems, and analytical tools through AI collaboration.

Claude Cursor React
⚙️

Systems Thinking

Developed the ORION portfolio risk management framework combining multiple signal inputs into actionable decisions. Industrial Engineering training focused on optimizing complex systems.

Process Design Optimization
🔬

Biomedical Knowledge

Minor in Biomedical Engineering. Deep personal research into autonomic dysfunction, autoimmune conditions, and chronic infections. First-hand experience navigating the healthcare system.

Research Medical Lit
📈

Product Building

Built flowtrackdata.xyz—a market intelligence dashboard. Experience taking ideas from concept to deployed product. Comfortable with full-stack development and user-facing design.

Product Full-Stack UX
💡

Patient Perspective

Lived the problem I want to solve. Understand the frustration, the gaps, and the opportunities from both a patient's view and an engineer's analytical lens.

Empathy Advocacy
Justin Strubel

About Me

I'm Justin—an Industrial Engineering graduate from Penn State with a Biomedical Engineering minor. I was trained to see systems, identify inefficiencies, and optimize processes. I never expected to apply that training to my own healthcare journey.

After years of navigating a fragmented system, I've become passionate about how data and AI can transform diagnostics. I've built data products, developed risk frameworks that process complex real-time information, and experienced firsthand the gap between what's possible and what patients actually receive.

The tools to prevent what happened to me already exist. They just aren't being applied at scale—yet.

B.S. Industrial Engineering, Penn State
Biomedical Engineering Minor
Former Cost Analyst, NAVAIR
Research Analyst, Flowtrack Capital
7+ Years Personal Health Data
Function Health Member

Let's build better diagnostics together

I'm looking to join a mission-driven team using technology and AI to transform how we detect and prevent disease.

Connect on LinkedIn →