CGM wellness products (Lingo, Stelo, Levels, ZOE): what they actually offer

The consumer CGM market has expanded rapidly. Abbott's Lingo, Dexcom's Stelo, Levels, ZOE, and others now offer CGMs to healthy adults for self-experimentation. The products differ in price, app sophistication, and accompanying coaching.

The underlying sensors#

Three sensors underpin the consumer market:

• Abbott Freestyle Libre 2/3: the sensor underlying Lingo and many consumer programmes
• Dexcom G7 — the sensor underlying Stelo
• Abbott Lingo sensor: newer, designed specifically for wellness use; slightly different specs from Libre

All measure interstitial fluid glucose every few minutes. MARD vs lab plasma glucose is similar across modern devices (7–9%). See what-a-cgm-measures.

The consumer products#

Abbott Lingo (£89 / 14 days): UK direct-to-consumer. Subscription model. App focuses on "glucose spikes" and food impact. No coaching.

Dexcom Stelo (US-only currently; not yet UK) — direct-to-consumer in the US for non-diabetic adults.

Levels (£199 for 2 sensors + 1 month app): US-focused with limited UK availability. App-heavy with detailed analytics, food logging integration, and notifications.

ZOE (£300+ multi-month programme): UK; built around personalised nutrition. Bundles CGM with microbiome stool test, blood test, and structured 4-week programme.

Other brands are emerging quickly; the market is moving.

Where the value is#

The trial evidence on CGM in healthy adults is limited but the practical value sits in a few places:

Personalised meal experiments. Comparing your own glucose curves to the same meal made differently is genuinely informative. You learn that your response to specific foods may differ from population averages; Zeevi et al. 2015 found postprandial response varies widely between people for identical meals.

Behaviour modification. A real-time signal of glucose impact reinforces behaviour change in a way population averages can't. Many users find the visibility itself moves them toward more whole foods and post-meal movement.

Detection of unexpected patterns. Some people discover their glucose curves are more variable than expected, prompting clinical investigation.

Alcohol, sleep, and stress impact. The 24–48 hour next-day patterns from a poor night's sleep, alcohol, or stress are often more impactful than meal choices.

Where the value is overstated#

"Optimising metabolic health" in healthy adults. No good trial evidence that reducing post-meal spikes in healthy non-diabetic adults produces meaningful long-term health benefits. HbA1c is the validated long-term measure of glucose-related disease risk.

"Personalised nutrition" based on CGM alone. Day-to-day variation means same-meal repeats produce different curves; a single CGM session isn't precision nutrition.

"Avoiding metabolic disease" claims. Diabetes prevention is well-evidenced for lifestyle interventions (sleep, exercise, weight loss, dietary patterns). CGM doesn't independently demonstrate prevention benefits.

"Diagnosing your metabolism." Healthy non-diabetic ranges aren't well-validated; what looks abnormal on CGM is often within normal physiology.

UK regulatory framing#

• CGM as medical device for diabetes management: regulated as a medical device; widely available on the NHS for type 1 and certain type 2 diabetics
• CGM for non-diabetic wellness use: regulated as a medical device but used in a wellness context; sold direct-to-consumer without prescription

The MHRA doesn't restrict non-diabetic use of CGM; it's a personal-choice product within the medical device framework.

The cost reality#

For a typical user:

• Curiosity / 1–2 sensor experiment: £89 (Lingo) — reasonable
• 2–3 month experiment: £200–£400; significant
• Ongoing year-round use: £1,000+/year; substantial

For most users, 2–4 weeks of CGM data captures most of the personal learning. Continued use shows diminishing returns unless there's a specific clinical question.

Wellness coaching layers#

Levels and ZOE add coaching, analytics, and structured programmes. The trial evidence for these specifically is small but suggestive of behaviour-change effectiveness. Whether this justifies the substantial price premium is personal.

For someone capable of interpreting CGM data themselves, the bare sensor (Lingo) is probably sufficient. For someone who wants structured guidance, the bundled products add value at cost.

Common patterns users notice#

After a few weeks of CGM data, most users notice:

• Their own response to specific foods varies more than population averages suggest
• The 24–48 hour effect of alcohol or poor sleep on next-day glucose
• Post-meal walks measurably blunt curves
• Stress events produce glucose rises
• Some "healthy" foods produce surprisingly large responses (and vice versa)

These observations have practical value even if "optimising glucose" isn't formally evidenced in healthy adults.

When CGM data should trigger clinical action#

If a CGM consistently shows:

• Fasting glucose >7 mmol/L
• Post-meal peaks >11 mmol/L
• HbA1c estimate >48 mmol/mol on app
• Very low overnight glucose

these warrant HbA1c testing with a GP. Wellness-CGM data isn't a substitute for clinical diagnosis but can flag patterns worth investigating.

What we'd actually say#

For curious adults who want to understand their own glucose responses, a 2–4 week CGM trial via Lingo (or equivalent) is a reasonable, evidence-supported self-experiment. The personal learning is real.

For long-term ongoing CGM use in healthy adults, the cost is substantial and the marginal learning declines. Most of the value is captured in the first few weeks.

For wellness coaching add-ons (Levels, ZOE), the value depends on whether you want structured guidance vs. just data.

For someone with prediabetic markers or family history of type 2 diabetes, CGM is more justifiable; for healthy adults with no risk factors, it's an interesting but optional experiment.

When to talk to a clinician#

• CGM showing persistent fasting >7 mmol/L or post-meal >11 mmol/L
• Family history of type 2 diabetes + concerning CGM patterns
• Unexpected low glucose readings
• Suspected reactive hypoglycaemia
• Using CGM while on insulin or sulfonylureas (you should already have clinical CGM access)