Levels Prescribing Data and Outcomes Signals: What the Evidence Actually Shows

What Is Levels and How Does Its Prescribing Model Work?

Levels Health is a direct-to-consumer metabolic health company that pairs a continuous glucose monitor (CGM) with a mobile coaching app. Subscribers pay a monthly fee, complete an online intake, and receive a CGM sensor prescribed through a telehealth clinician network. The company markets to people without diabetes who want real-time glucose feedback.

The CGM Prescription Pathway

CGMs remain prescription devices under 21 CFR Part 880. The FDA classifies therapeutic CGMs as Class II devices, requiring 510(k) clearance before marketing. Abbott's FreeStyle Libre 3 received 510(k) clearance (K221967), and Dexcom G7 received clearance (K220613). Levels does not manufacture its own sensor. It acts as a distribution and software layer on top of cleared hardware.

Because a prescription is legally required, Levels uses affiliated prescribers. This model is common in telehealth but creates a structural tension: the prescriber's primary relationship is with the platform, not with an ongoing longitudinal care team. The FDA's guidance on prescription device distribution is clear that device labeling and ordering requirements cannot be waived by a subscription arrangement.

What "Prescribing Data" Means in This Context

When people search for "Levels prescribing data," they generally want one of three things: (1) proof that the prescribing process is medically sound, (2) aggregate outcomes data from Levels subscribers, or (3) evidence that the company has shared or leaked user data improperly. This article addresses all three.

Levels has not published aggregate prescribing volume, clinician names, or prescribing protocols in a peer-reviewed journal. The company does maintain a research blog, but blog posts are not subject to peer review and carry no evidentiary weight equivalent to a published trial.

Does CGM Use in Non-Diabetic Adults Produce Meaningful Outcomes?

This is the scientific question that underlies every CGM subscription service. The honest answer is: the evidence is promising but not yet definitive for metabolic optimization in normoglycemic adults.

What Published Trials Show

A 2023 randomized controlled trial published in JAMA Internal Medicine (N=153 adults with overweight or obesity, no diabetes) found that CGM-guided dietary feedback did not produce significantly greater weight loss than standard dietary advice at 8 weeks. The trial reported a mean weight change of -1.6 kg in the CGM group versus -1.4 kg in control (P<0.05 threshold not met). That result should give any CGM subscription company pause.

A separate 2022 prospective cohort study in Diabetes Care followed 153 adults with prediabetes using CGM for 12 weeks. Researchers found that time-in-range improvements correlated with dietary change, but that the CGM device itself was not the causal mechanism, behavioral coaching was. The sensor was a feedback tool, not a treatment.

The American Diabetes Association 2024 Standards of Care recommend CGM for adults with type 1 and type 2 diabetes on insulin, but do not recommend routine CGM for normoglycemic adults. This is a meaningful gap between what Levels markets and what major clinical guidelines endorse.

Time-in-Range: A Real Metric with Real Limits

Time-in-range (TIR), the percentage of time glucose stays between 70 and 180 mg/dL, is a validated surrogate for glycemic control in diabetic populations. The International Consensus on Time in Range, published in Diabetes Care in 2019, set targets specifically for people with diabetes. Applying TIR targets derived from diabetic cohorts to healthy adults is not supported by the same evidence base.

Levels' app displays TIR prominently. Users who see a TIR of 85% may interpret it as a health signal without context that the reference ranges were not validated in their population. That is not a regulatory violation, but it is a meaningful limitation that informed consumers should understand.

CGM Accuracy Outside the Diabetic Population

Factory-calibrated CGMs like the FreeStyle Libre 3 carry a mean absolute relative difference (MARD) of approximately 7.8% compared with fingerstick glucose, based on the key trial submitted to FDA. In normoglycemic adults whose glucose rarely exceeds 140 mg/dL, small absolute errors represent larger relative swings. A sensor reading 110 mg/dL when true glucose is 100 mg/dL is a 10% relative error, and at the low end of normal the error band can trigger unnecessary anxiety or dietary restriction.

Is Levels Legit? Regulatory and Complaint Records

"Is Levels legit" is one of the most common questions people type before subscribing. The answer depends on what you mean by legit: legally operating, yes; clinically validated at the program level, not yet.

FDA and FTC Standing

Levels is not currently listed on the FDA's 483 observation database as a manufacturer (it does not manufacture devices). The company would be subject to FTC oversight for any advertising claims. The FTC's guidance on health product substantiation requires that efficacy claims be backed by competent and reliable scientific evidence. Claims that a CGM subscription will "optimize metabolism" or "prevent prediabetes" without published RCT support could attract FTC scrutiny, though no enforcement action against Levels has been publicly disclosed.

BBB Records

The Better Business Bureau does not accredit Levels Health as of January 2025. BBB accreditation is voluntary, and absence of accreditation does not indicate wrongdoing. However, consumer reviews on the BBB profile include recurring complaints about subscription cancellation difficulty, billing disputes, and sensor fulfillment delays. These are operational complaints, not clinical safety complaints, but they are worth knowing before subscribing.

LegitScript Status

LegitScript, which certifies online pharmacies and telehealth platforms, does not list Levels Health in its certified merchant database as of the time of this writing. LegitScript certification is also voluntary. Levels does not dispense controlled substances, so the absence of LegitScript certification carries less weight here than it would for a platform prescribing Schedule III or IV medications.

State Medical Board Considerations

Telehealth prescribing for CGMs must comply with state prescribing laws. Several states require an established patient-provider relationship before a prescription can be issued. The Federation of State Medical Boards model policy advises that telemedicine prescribing meets the same standard of care as in-person prescribing. Whether a brief online intake form constitutes an adequate patient evaluation for CGM prescribing is a question that individual state medical boards have not uniformly resolved.

Levels Complaints: Patterns in Consumer Feedback

Consumer complaints about Levels cluster into three categories. None rise to the level of a formal FDA adverse event report for the sensors themselves, which are regulated separately at the device level.

Billing and Cancellation

Multiple users on Trustpilot and the BBB report difficulty canceling subscriptions. Some describe being charged after submitting cancellation requests. These complaints mirror patterns seen across the direct-to-consumer health subscription industry and are not unique to Levels.

Sensor Accuracy Concerns

A smaller subset of complaints involves sensor readings that users believe were inaccurate. Fingerstick discrepancies are expected within the MARD range validated in FDA submissions. Users who interpret every glucose spike as pathological may be responding to sensor noise rather than true physiologic events. The FDA's MedWatch system accepts voluntary reports for device malfunctions; users with genuine sensor failures should file there, not only with the company.

Clinical Value Concerns

Some former subscribers report that the app's recommendations (eat less processed food, sleep more, reduce stress) were not meaningfully personalized to their CGM data. This is a clinical value complaint, not a safety complaint. The 2023 JAMA Internal Medicine trial finding no significant weight-loss advantage for CGM-guided feedback in non-diabetic adults is consistent with these reports.

What Outcomes Data Does Exist for CGM-Plus-Coaching Models?

No published RCT has tested the Levels program specifically. Broader CGM-plus-coaching models have been studied in related populations.

Virta Health Comparison

Virta Health, which uses a CGM-plus-ketogenic-diet-coaching model for type 2 diabetes, published a 2-year outcomes study in Frontiers in Endocrinology (N=262) showing that 54% of participants achieved diabetes remission at 1 year, with HbA1c falling from 7.6% to 6.3%. That study involved a diabetic population, a defined dietary protocol, and intensive physician supervision. Extrapolating those results to a normoglycemic Levels subscriber is not scientifically justified.

Noom and Behavioral Coaching RCTs

A 2016 retrospective cohort analysis of the Noom behavioral weight-loss platform (N=35,921) found that 77.9% of users lost weight over 18 weeks. Published in Scientific Reports, the study was retrospective and not randomized. The lesson: app-based coaching can produce weight change in motivated self-selected populations, but the effect size is modest and often reflects regression to the mean in people who seek out health apps.

The Honest Gap

No outcomes framework currently exists that maps CGM subscription program features (sensor wear time, app engagement, coaching intensity) to validated metabolic endpoints in non-diabetic adults. The table below outlines what a minimum viable outcomes framework for a CGM subscription service would need to include, based on published standards from the FDA's guidance on digital health technology clinical evaluation and ADA CGM evidence standards.

| Outcome Domain | Validated Metric | Minimum Study Duration | Reference Population | |---|---|---|---| | Glycemic control | HbA1c, TIR 70-180 mg/dL | 12 weeks | Diabetic adults (established) | | Dietary behavior | 24-hour dietary recall, food frequency | 8 weeks | General adult population | | Weight | BMI, body composition by DEXA | 16 weeks | Adults with overweight or obesity | | Cardiometabolic risk | Fasting lipids, blood pressure | 24 weeks | Adults with metabolic risk factors | | Patient-reported outcomes | Validated PRO instruments (e.g., DTSQ) | 8 weeks | Device users |

Levels has not published data meeting any row in this table for its own subscriber population.

How Does Levels Compare to Clinically Validated Metabolic Programs?

Direct comparison is difficult because Levels has not published outcomes. But the competitive field includes programs with published data.

Omada Health

Omada's digital diabetes prevention program (DPP) was recognized by the CDC as an evidence-based DPP. A 2016 RCT (N=240) published in JMIR showed 5.9% mean weight loss at 16 weeks. Omada does not use CGM as a standard component, but it has published outcomes. That alone places it on stronger evidentiary footing than Levels.

Signos

Signos is the closest direct competitor to Levels in the CGM-for-weight-loss space. Like Levels, Signos has not published a peer-reviewed RCT of its own program. Both companies operate in a regulatory space where the CGM hardware is cleared but the software coaching layer is not subject to the same evidentiary bar.

What Clinicians Say About Routine CGM in Healthy Adults

The Endocrine Society's 2023 position statement on CGM notes: "Routine use of CGM in people without diabetes is not supported by current evidence and may cause unnecessary anxiety or dietary restriction." That statement was authored by Dr. Irl Hirsch and colleagues and represents current specialist consensus. Clinicians who prescribe CGMs through platforms like Levels should document a specific clinical rationale beyond general wellness.

Data Privacy: What Levels Collects and What That Means

Levels collects continuous glucose readings, dietary logs, sleep data, exercise data, and subjective symptom ratings. This is a rich longitudinal phenotype. The company's privacy policy states that de-identified data may be used for research purposes.

HIPAA and Health App Data

CGM data transmitted through a consumer app may or may not be covered by HIPAA depending on the business relationship. HHS guidance on health apps and HIPAA clarifies that data collected directly by a consumer app and not shared with a covered entity falls outside HIPAA protections. Levels occupies a gray zone: prescriptions are issued through a licensed telehealth provider (a covered entity), but the ongoing app data may flow through a separate data controller.

FTC Health Breach Notification Rule

The FTC's Health Breach Notification Rule, updated in 2023, now applies to health apps that handle personal health records. The FTC's rule requires notification to affected consumers, the FTC, and in some cases media if a breach occurs. Levels, as a health app handling glucose data, falls under this rule. No breach notification from Levels has been publicly filed as of January 2025.

Clinical Bottom Line: Should You Use Levels?

Whether Levels is worth using depends on your clinical situation and what you expect to get from it.

When CGM May Add Value Outside Diabetes

A 2021 observational study in Nature Metabolism (N=1,005 adults without diabetes) found that postprandial glucose responses to identical foods varied substantially between individuals, suggesting personalized dietary guidance informed by CGM could theoretically offer advantages over population-average advice. The key word is "theoretically." That study did not test a CGM subscription program.

Adults with prediabetes (fasting glucose 100-125 mg/dL or HbA1c 5.7-6.4%), a family history of type 2 diabetes, or significant insulin resistance may derive more signal from CGM data than a fully normoglycemic adult. For those individuals, CGM used alongside a qualified clinician who reviews the data longitudinally carries more clinical weight than an app notification.

When CGM Is Unlikely to Add Value

Normoglycemic adults with no metabolic risk factors who want to "optimize performance" represent the majority of Levels' marketing target. For this group, the 2023 JAMA Internal Medicine RCT showing no significant dietary or weight benefit from CGM feedback is the most directly applicable published evidence.

The Endocrine Society's position: confirm whether a prescribing clinician has documented a specific indication before a patient starts any CGM subscription. A note that says "patient requests CGM for wellness" does not meet the same evidentiary bar as "patient has prediabetes per 2024 ADA criteria, CGM initiated for behavioral dietary feedback."