The daily rhythm of life transforms when glucose levels are under constant surveillance. A little over a year ago, the author embarked on a routine that began with the meticulous application of a continuous glucose monitor (CGM) to each arm—a Dexcom Stelo on the right, an Abbott Lingo on the left. This ritual, involving an isopropyl alcohol wipe, a small applicator, and a decisive “ka-thunk,” became a prelude to travel or any significant day. The immediate check of accompanying smartphone apps confirmed data transmission, while mental notes were made about potential influences like high altitudes on readings. Surprisingly, the insertion process was consistently painless.
The Expanding Horizon of Glucose Tracking
Initially, there was no pressing medical imperative for this continuous glucose tracking. The author was not diabetic, and A1C levels—the long-term measure of blood sugar—had always been within healthy limits. However, the landscape of glucose monitoring has dramatically shifted. What was once a tool exclusively for diabetics is now a burgeoning trend in wellness, embraced by doctors, social media influencers, biohackers, and athletes alike. As a health tech tester, the author’s curiosity was piqued, prompting a decision to explore the benefits for a non-diabetic. This initial curiosity unexpectedly led to a year-long deep dive, marked by extensive device testing, academic research, consultations with experts, and moments of profound personal confusion and anxiety.
From Medical Necessity to Wellness Trend
The journey of CGMs began in 1999 with the Food and Drug Administration’s clearance of the first “professional use” device. These devices, often mistakenly thought to measure blood sugar directly, actually provide real-time glucose measurements from the interstitial fluid, located just beneath the skin between cells. This capability allows for tracking glucose trends over an extended duration, offering a more dynamic picture than traditional finger-stick tests. Until 2024, CGMs primarily required a prescription and were indispensable for Type 1 diabetics, who produce little to no insulin.
The recent shift has seen Dexcom and Abbott introduce over-the-counter CGMs, explicitly targeting non-diabetics, prediabetics, and Type 2 diabetics who manage their condition without insulin. To delineate these consumer-grade devices, companies often market them as “glucose biosensors.”
The Promise vs. The Reality for Non-Diabetics
For prediabetics and Type 2 diabetics, the utility of CGMs is unequivocal. These conditions typically progress over time as the body develops insulin resistance. Early detection and intervention through lifestyle modifications, such as dietary adjustments and exercise, can potentially “reverse” their progression. Given that Type 2 diabetics constitute approximately 95 percent of the 38.4 million Americans with diabetes (as of 2021, per the American Diabetes Association), and an estimated 98 million Americans have prediabetes, CGMs offer a substantial opportunity for a large population to gain insights into how their dietary choices affect a genuine metabolic condition.
However, the benefit for non-diabetics remains ambiguous. Despite this, the adoption of CGM technology is aggressively promoted by both manufacturers and political figures. Notably, figures like Health Secretary RFK Jr. envision widespread CGM use, and controversial surgeon general nominee Casey Means, cofounder of the CGM startup Levels, champions the technology in her book Good Energy as a fundamental tool for rectifying metabolic dysfunction—a claim that several medical experts publicly dispute due to a lack of supporting evidence for non-diabetics.
The Allure of Biohacking
The current era of wearable technology holds a compelling allure: not just to detect potential illnesses, but to “optimize” one’s biometric data for a longer, healthier life. In the context of a often-flawed healthcare system, this technology is frequently presented as a means for individuals to regain control over their health—wear a device, track your health, understand your body better, and consequently make superior choices.
Metabolic health optimization has been a growing trend for years, initially manifesting as food logging. CGM-driven metabolism tracking has amplified this trend by adding a layer of real-time data. Companies like January AI evolved from CGM startups to offer glucose spike predictions within meal-logging apps, leveraging collected data. Similarly, Oura has partnered with Dexcom to integrate glucose readings, meal logging, and AI interpretations into its smart ring app, while Abbott collaborated with Withings for a comparable integration. With the increasing use of GLP-1 medications and a resurgence of “ultra-skinny diet culture” reminiscent of the 2000s, CGM use is increasingly positioned as the next frontier in fitness tracking.
For the author, this proposition was particularly appealing. With a family history of Type 2 diabetes and high cholesterol, a diagnosis of polycystic ovary syndrome (PCOS)—a condition often linked to insulin resistance or chronic inflammation—and struggles with energy levels during long-distance running despite standard fueling practices, the idea of biohacking nutrition to manage diabetes risk, understand PCOS, and enhance athletic performance seemed logical.
A Personal Deep Dive: Data, Anxiety, and Diagnosis
The Daily Scrutiny
Initially, the author meticulously reviewed CGM data every morning, after each workout, and a few hours post-meal. Most readings appeared normal: pasta caused a glucose spike, salmon and salad resulted in a minimal increase, and carb-loading before a long run led to a significant spike followed by a sharp drop. (The author found CGMs impractical for mid-run fueling due to a five-minute data delay).
The Dawn of Doubt and Medical Consultations
However, this constant data review soon triggered anxiety. Both Dexcom and Abbott apps provided educational materials outlining the ideal glucose range for healthy non-diabetics (70 to 140mg/dL) and fasting glucose levels (70 to 99mg/dL). To the author’s alarm, morning readings consistently exceeded 100mg/dL, even without late-night snacks. Sometimes, Dexcom would even issue spike alerts during sleep. This was in stark contrast to a previous two-week trial of Nutrisense (another CGM-based service) a year prior, where morning glucose levels were never elevated. The conclusion drawn was unsettling: diabetes had finally arrived.
This realization spiraled into heightened anxiety, a flurry of doctor’s appointments, and an intense fixation on diet and exercise. During medical consultations, doctors often responded with shrugs and skepticism. One nurse even questioned the necessity of non-diabetics wearing such devices. Despite the author’s concerns, A1C tests consistently returned normal results, indicating no diabetes. However, blood work revealed elevated liver enzymes and cholesterol.
Unpacking the Fatty Liver Diagnosis
Out of caution, the doctor ordered an ultrasound, leading to a diagnosis of non-alcoholic fatty liver, described as “nothing to worry about right now.” This condition is commonly associated with PCOS. The doctor’s advice was familiar: reduce alcohol intake (though the author rarely drank) and lose body fat (a long-standing, difficult goal). A follow-up was scheduled for a year later.
From one perspective, this could be seen as a CGM success story: the devices flagged a change, leading to a new official diagnosis. Control over one’s health, it seemed, was regained. However, a more critical interpretation suggests that elevated liver enzymes and cholesterol had been present in previous blood work, with a former doctor having already suspected fatty liver but deemed an ultrasound unnecessary unless levels worsened. The CGM-inspired tests didn’t show significantly worse levels, and the medical advice remained unchanged. In essence, no truly new information was gleaned.
It was much later, after direct consultations with Dexcom and Abbott, that the author discovered side sleeping could compress the CGM, leading to inaccurate overnight glucose readings—either lower or higher than actual levels. Despite attempts to mitigate this by swapping arms, consistent high overnight and morning readings persisted, leaving the author unable to distinguish between inaccurate data and an undiagnosed underlying issue.
Navigating the Complexities and Inaccuracies of CGM Data
Practical Challenges and Costs
The first month of continuous CGM wear revealed practical challenges. While the author received test units, for an average user, the cost could be around $100 per month. Each sensor is designed to last about 15 days, but malfunctions or accidental removal (snagging on clothing, bumping into objects) were common. The adhesive also left persistent residue. To manage supply and give the skin a break, the author reduced usage to two weeks a month, then quarterly.
The Problem of Inaccurate Readings
Beyond the practicalities, the fundamental issue of data accuracy proved deeply unsettling. The inability to definitively know if the high readings were real or a product of sensor compression or other factors fueled significant anxiety.
Lack of Clinical Consensus
Compounding this problem is the disparate approaches of CGM manufacturers in presenting data. Dexcom Stelo, for instance, provides spike alerts after an hour or two, while Abbott’s Lingo CGM opts for a “Lingo Score” that simplifies raw data into a digestible number, indicating adherence to a healthy glucose range of 70 to 140mg/dL.
However, Dr. Nicole Spartano, an assistant professor at Boston University, highlights the critical lack of research on follow-up outcomes for non-diabetics using CGMs. While one small study linked more time above the healthy range to a higher diabetes risk, it involved a population likely already predisposed. Spartano emphasizes the absence of context for how CGM numbers relate to individual health, particularly the distinction between blood glucose (measured by doctors) and interstitial glucose (measured by CGMs), which can differ. More research is needed to understand these discrepancies.
Dr. David Klonoff, medical director at the Diabetes Research Institute at Mills-Peninsula Medical Center and editor-in-chief of the Journal of Diabetes Science and Technology, further explains that CGMs are not exactly accurate, despite meeting FDA accuracy requirements, leaving “wiggle room.” This means that simultaneously wearing two CGMs could yield different results, and even sequential sensors on the same person could vary.
In a clinical study, Spartano found no consensus among 18 endocrinologists asked to interpret non-diabetic CGM data. Some viewed glucose peaks as normal physiological responses, while others saw them as indicators for further testing. Klonoff concurs, stating, “Essentially, there was no consensus. Even clinicians who read CGM data all day, every day don’t know what to do with this data.” There is currently no vetted and agreed-upon reference data set for non-diabetics, making accurate interpretation challenging and years away.
Influencer Standards vs. Medical Reality
This lack of expert consensus means that a non-diabetic seeking help interpreting CGM data might receive widely varying advice from different doctors. While medical experts reassured the author that her optimal A1C and generally healthy CGM range meant she was fine and that spikes were normal, this advice conflicted with the “optimization” narrative. Influencers like Casey Means advocate for stricter criteria: post-meal glucose below 115mg/dL, spikes no more than 30mg/dL, and optimal morning fasting glucose between 75 and 80mg/dL. Adhering to these unscientific standards can lead to a sense of “suboptimal” metabolism even when medically healthy, particularly for perfectionists.
The Hidden Toll: Mental Health and Disordered Habits
The Spiral into Obsession
The prolonged use of CGMs led the author into an unhealthy obsession with food. A single slice of pizza at a social gathering could induce a cold sweat, driven by the fear of a spike alert or a poor score. This fear sometimes led to skipping meals or snacks entirely, even when hungry. Over-exercising also became a pattern, with self-worth tied to achieving specific, often arbitrarily strict, glucose numbers. The stress of constant monitoring created a vicious cycle of being stressed about stress.
Social Isolation and Intervention
Social events became unbearable, prompting avoidance. A particularly stark moment occurred during a family Thanksgiving dinner, where the author found herself negotiating with herself about food choices based on CGM readings. After six months, the situation deteriorated to the point where loved ones felt compelled to intervene, highlighting the author’s unnoticed fixation on “optimizing her metabolism.”
The Link to Disordered Eating
This experience underscores a crucial, often overlooked aspect of the metabolism optimization narrative. While studies indicate that wearables, diet, or fitness apps don’t have a definitive causal link to disordered eating or eating disorders, they have been consistently associated with exacerbating symptoms in individuals already prone to them. For some, CGM use might be benign; for the author, it undeniably triggered a dark side.
The author now limits CGM use solely for testing new features, having recognized the detrimental impact on her mental well-being.
Finding True Metabolic Improvement
The Turning Point
After approximately a year, a final self-test revealed troubling results: elevated morning glucose, prolonged post-meal spikes, and a higher daily average glucose than ever before. The author was consistently exhausted, gaining weight despite vigilant nutrition and exercise. This prompted a search for a new doctor and further blood work.
A Comprehensive Diagnosis and Effective Treatment
The new tests still showed no diabetes or prediabetes, but bad cholesterol had worsened, and two liver enzymes had more than tripled. An ultrasound confirmed the fatty liver had progressed from mild to moderate. Crucially, insulin resistance was tested for the first time and found to be on the “high side of normal.”
The new doctor concluded that the author’s non-medicated efforts to manage her “chaotic metabolism,” while commendable, were insufficient. Prescriptions were issued, and a treatment plan was initiated. Four months into this new regimen, the author’s CGM data and blood work have dramatically improved. For the first time in a decade, bad cholesterol is normal. Fifteen of the 25 pounds gained have been lost, and liver enzymes have dropped by approximately 65 percent. Morning glucose levels are no longer elevated.
Beyond Biohacking: The Role of Medical Intervention
Proponents of non-diabetic CGM use might hail this as an unqualified success, and in many ways, it is. However, the author remains cautious. While grateful for improved health, she emphasizes that medication, not solely CGM use or “taking control of her health” through lifestyle changes, was the ultimate catalyst for improvement. The journey involved 13 months of CGM testing to identify metabolic issues and 17 months to see significant improvement, alongside considerable frustration and tears, and ongoing adjustment to medication side effects.
Conclusion
Continuous Glucose Monitors can be valuable tools, but their application for non-diabetics demands a nuanced understanding. While they can illuminate individual responses to food and activity, the lack of clinical consensus on interpreting non-diabetic glucose data, coupled with potential inaccuracies and the pervasive “optimization” culture, can easily lead to anxiety, disordered eating tendencies, and a false sense of control. The author’s experience highlights that while CGMs played a role in flagging underlying metabolic changes, true and sustainable health improvement ultimately came through professional medical diagnosis and prescribed treatment, not through self-directed biohacking alone. Therefore, while access to these sensors for non-diabetics is positive, it must be accompanied by a clear understanding of their limitations and the potential for mental health risks, rather than being presented as a simplistic “silver bullet” for metabolic optimization.
