Continuous Glucose Monitoring (CGM) technology and investment research

Technical bottleneck

Physics & chemistry: why it's hard

The glucose oxidase electrochemical sensor:

A CGM sensor is a miniature amperometric biosensor. The core reaction:

`

Glucose + O₂ → gluconic acid + H₂O₂ (catalyzed by glucose oxidase, GOx)

H₂O₂ → O₂ + 2H⁺ + 2e⁻ (electrochemical oxidation at electrode surface)

`

The resulting current is proportional to glucose concentration. The engineering challenges cascade:

1. Biocompatibility — the foreign body response:

The sensor filament is a foreign object. Within minutes of insertion, proteins adsorb to the surface. Within hours, inflammatory cells (macrophages, neutrophils) arrive. Within days, a fibrous capsule begins forming around the sensor, creating a diffusion barrier that slows glucose transport. The sensor must maintain accurate readings for 10–15 days despite this progressive encapsulation. Abbott and Dexcom achieve this through proprietary polymer coatings (biocompatible hydrogels, anti-biofouling surface treatments) — the exact formulations are trade secrets.

2. Enzyme stability — GOx degradation over time:

Glucose oxidase denatures at body temperature (37°C). The enzyme must remain active for the full sensor lifetime. Cross-linking GOx to the electrode surface (via glutaraldehyde or other crosslinkers) slows degradation, but sensor sensitivity inevitably drifts over 10–15 days. Calibration algorithms compensate for drift, but the fundamental stability of the enzyme-electrode interface limits sensor lifetime to ~15 days — a deliberate engineering choice, not a biological accident. Longer-life sensors (Senseonics Eversense, 180-day implantable) use optical fluorescence rather than electrochemical detection, sidestepping the enzyme stability problem entirely.

3. Interference rejection — the selectivity problem:

Blood and interstitial fluid contain dozens of electroactive species that can oxidize at the electrode and produce false signals: acetaminophen (paracetamol), ascorbic acid (vitamin C), uric acid, and many drugs. Early CGM sensors had dangerous acetaminophen interference. Modern sensors use permselective membranes (e.g., Nafion, polyurethane, cellulose acetate) that block interferents based on size and charge while permitting H₂O₂ diffusion. The membrane must be thin enough for fast response (~5-minute lag from blood glucose) but thick enough to reject interferents — a fundamental diffusion trade-off.

4. Calibration and the factory-calibration breakthrough:

First-generation CGMs required 2–4 fingerstick calibrations per day. Dexcom G6/G7 and Abbott Libre 3+ are factory-calibrated — no fingersticks needed. This required developing manufacturing processes with sensor-to-sensor reproducibility tight enough that a single calibration curve works for every sensor. The physics: each sensor's sensitivity (nA/mM glucose) is determined by GOx loading, electrode surface area, and membrane permeability — all of which vary at nanometer scale. Factory calibration means these tolerances are controlled at wafer-scale manufacturing precision.

5. The adhesive problem — the mundane bottleneck:

A CGM must stay on the skin for 10–14 days through showers, exercise, swimming, and sleep. Sensor failures due to adhesive loss are the #1 complaint in user forums. Dexcom G7 addressed this with a new adhesive formulation, but it remains a persistent issue — and a competitive differentiator. The 15-day sensor window (vs. 10-day for earlier generations) means the adhesive must survive longer under harsher conditions.

Economic constraints

• Duopoly: Abbott and Dexcom control ~87% of global CGM shipments. This is not by accident — FDA 510(k) clearance for a new CGM requires clinical accuracy studies (MARD <10% vs. reference glucose), biocompatibility testing, and manufacturing quality systems that take 3–5 years and $50–100M+ to build. Medtronic has been #3 for a decade and cannot close the gap.
• Recurring revenue: A Libre 3 user replaces the sensor every 14 days — 26 sensors/year. A Dexcom G7 user replaces every 10 days (+12-hour grace period) — 36 sensors/year. At $30–70/sensor wholesale, this is $780–2,500/year per patient, recurring for life. CGM is the purest razor-blade model in medtech.
• Reimbursement expansion: CMS coverage for Type 2 non-insulin users (potential ADA 2026 data readout) would add millions of patients to the reimbursed market. Prime Therapeutics (pharmacy benefit manager) expanded commercial coverage in Q1 2026.

Adoption

Why it matters now

Dexcom Q1 2026:

• Worldwide revenue: $1.19B (+15% reported, +12% organic)
• U.S.: $832M (+11%); International: $360M (+26% reported, +17% organic)
• G7 15-Day sensor launched — ~50% existing base conversion expected by year-end; fewer sensor changes = better user experience = stickier users
• Stelo OTC: crossed $100M revenue in first ~12 months, >140,000 users — the non-diabetic wellness market is real and growing
• Gross margin 63.5% (expanding); operating margin 23–23.5% (raised)
• FY2026 guidance: $5.16–5.25B revenue (11–13% growth); cash $2.4B post-Q1
• [SEC] DXCM 10-K (search sec.gov for Dexcom latest)

Abbott (ABT) — FreeStyle Libre:

• Libre grew +7.5% YoY in Q1 2026 — notable deceleration from double-digit growth
• Still the global leader (~53% share) with unmatched retail presence: Lingo OTC in 3,500+ Walmart stores
• Libre 3+: smaller sensor, factory-calibrated, 15-day wear
• [SEC] ABT 10-K (search sec.gov for Abbott latest)

OTC/wellness market: Projected ~$650M in 2026. Dexcom Stelo and Abbott Lingo are building the consumer CGM category from scratch — targeting non-diabetics for metabolic health, athletic performance, and general wellness. This is a new TAM layer on top of the ~$15B medical CGM market.

Key trends

• Type 2 non-insulin expansion: The largest untapped market. CMS coverage would be a step-change event — ADA 2026 data expected to support. Dexcom gaining share in this segment.
• GLP-1 crossover: CGM paired with GLP-1s for metabolic monitoring — does the drug actually stabilize glucose? Early clinical data showing GLP-1 + CGM improves outcomes vs. GLP-1 alone.
• AI insights: Dexcom + Google Vertex/Gemini for glucose pattern analysis and predictive alerts. Abbott partnering with WeightWatchers and other wellness platforms.
• Automated insulin delivery (AID): Dexcom G7 + Tandem t:slim X2 / Insulet Omnipod 5 = closed-loop system. CGM is the sensor input; insulin pump is the actuator. The tighter the CGM accuracy, the more aggressive the algorithm can be.
• Implantable sensors: Senseonics Eversense (180-day implantable, Ascensia-distributed) — niche (<1% share) but a technology proof-of-concept. The future: a 1-year implantable CGM with no external transmitter?

Key players

Horizon

• Horizon 1 (0–2yr): G7 15-day conversion; Stelo/Lingo OTC ramp; Type 2 non-insulin CMS coverage (ADA 2026 catalyst); GLP-1 + CGM bundled care models
• Horizon 2 (3–5yr): 30-day wear sensors; AI-driven predictive alerts (warn of hypoglycemia 30 minutes before it happens); implantable sensors at scale
• Horizon 3: Multi-analyte sensors — glucose + ketones + lactate on one filament; CGM as the healthspan operating system (everyone wears one, like a fitness tracker); non-invasive optical glucose sensing (Apple Watch rumors — game-changing if real, likely 5+ years away from FDA accuracy requirements)

Related Technologies

• Sterile Fill-Finish — insulin and GLP-1 drugs require the fill-finish bottleneck; CGM + AID closed-loop increases injectable demand
• Proteomics — multi-analyte sensors will measure protein biomarkers alongside glucose; proteomics provides the biomarker targets
• Bioprocessing Consumables — insulin manufacturing is a bioprocessing application; GLP-1 peptides are synthetic but share supply chain bottlenecks