Is a Glucose Sensor Really Necessary? For Whom Does Wearing One Make Sense?
In this article
For the first time, glucose and ketones can be monitored continuously with the same wearable sensor
The FDA granted De Novo marketing authorization to Abbott's Libre Duo 10 Day system for people with diabetes aged 2 years and older. The ten-day sensor monitors glucose and ketone levels together in fluid beneath the skin every minute. This development shows sensors moving from devices that merely “display a glucose curve” toward safety systems that may help identify a life-threatening complication such as diabetic ketoacidosis earlier.
A small sensor attached to your arm produces a new number every few minutes. You watch the curve rising after breakfast, glucose falling with a walk and a sudden alarm appearing during sleep. For some people, this information is a treatment tool that safely adjusts insulin doses, detects nighttime hypoglycemia and can save lives. For others, it can be an expensive source of unease that turns every bite into a test and presents normal biological fluctuations as “failure.”
Continuous glucose monitoring—commonly called a glucose sensor or CGM—is now marketed not only to people with diabetes, but also to those with prediabetes or excess weight, and even to completely healthy people. As technology advances rapidly, the fundamental question remains the same: Does more data really mean better health?
Short answer: In type 1 diabetes, in people using insulin and in those at high risk of hypoglycemia, often yes. In type 2 diabetes without insulin use, it may make sense to address a specific clinical problem. In prediabetes, obesity alone or healthy individuals, however, the evidence is not yet sufficient for everyone to use it continuously.
Why is Libre Duo more than an ordinary sensor update?
Seeing the direction of ketones alongside glucose may provide earlier and more actionable information than a single measurement when diabetic ketoacidosis is a risk.
Ketones form when the body begins to use fats intensively for energy instead of glucose. When insulin is insufficient, ketone production can increase uncontrollably and progress to diabetic ketoacidosis (DKA) , which acidifies the blood. DKA can present with nausea, vomiting, abdominal pain, rapid or deep breathing, dehydration, weakness and altered consciousness; it can quickly become a life-threatening emergency.
Until now, home ketone monitoring mostly relied on single point-in-time measurements using a drop of blood or a urine strip. Libre Duo measures glucose and ketones in interstitial fluid every minute, day and night ; it sends the values and their rising or falling direction to a compatible smartphone. It can issue an automatic alarm when ketones reach a specified warning threshold. This may give the user or caregiver an opportunity to notice rising ketones before DKA develops. According to the FDA, Libre Duo is the first wearable device in the US to monitor ketones continuously and the world's first system to monitor glucose and ketones together continuously in a single device.
Evidence evaluated by the FDA
The clinical program, involving people aged 2 years and older, included more than 600 participants across six studies. According to the FDA, the device accurately tracked clinically meaningful ketone differences throughout ten days of use and was able to detect elevated ketone levels before DKA began.
The device had previously received Breakthrough Device status. The final decision is not a drug “approval”: the FDA granted marketing authorization through the De Novo pathway , which is used for novel devices with low-to-moderate risk, and established special controls for this new device class, including labeling and performance testing.
A ketone alarm alone is not a diagnosis of DKA
The FDA particularly emphasizes that ketone information should be interpreted alongside glucose levels and symptoms. Normal or not very high glucose does not rule out DKA in every situation. Especially if vomiting, abdominal pain, rapid breathing, marked weakness or altered consciousness is present, merely watching the sensor alarm at home is insufficient; urgent medical assessment should be sought.
The strongest potential uses
Type 1 diabetes; interruption of insulin pump delivery or injections; infection and other sick days; vomiting or inability to take enough fluids; a history of DKA and other situations in which ketone monitoring is clinically necessary.
What does it not mean?
This authorization does not show that healthy people should continuously monitor their ketogenic diet, exercise or “fat burning.” The device was authorized for people with diabetes ; its clinical purpose is to improve diabetes safety, not satisfy metabolic curiosity.
How this development answers our article's main question: A sensor's value comes not from the amount of data it produces, but from the importance of the clinical problem it solves. While scoring every meal in a healthy person may be an unproven use, noticing rising ketones at night in a child at risk of DKA may directly provide safety.
What does a glucose sensor actually measure?
It shows not a snapshot of blood glucose, but a moving film of glucose beneath the skin.
A CGM sensor measures glucose in interstitial fluid beneath the skin and sends regular values to a phone or reader. Unlike a conventional fingerstick measurement, it therefore answers not only “What is the level now?” but also “In which direction and how quickly is it changing?” Nighttime drops, rises after meals, exercise responses and variability throughout the day become visible.
However, the sensor does not measure blood in a vessel directly. Interstitial glucose may lag several minutes behind blood glucose, especially during rapid rises and falls. Older physiological studies generally found a delay of approximately 4–10 minutes ; the device's algorithm and measurement conditions may affect the total difference. Confirmation with a fingerstick measurement is therefore required if symptoms and the sensor disagree, glucose is changing rapidly or the reading is unexpectedly low/high.
Strength: it shows patterns
Rather than a single fasting glucose value, it reveals patterns associated with nighttime, meals, exercise and medication. It may show at what time treatment falls short.
Limitation: it is not a laboratory test
According to the ADA 2026, CGM data are insufficient to diagnose prediabetes or diabetes. Diagnosis is made using validated tests such as HbA1c, fasting plasma glucose or an OGTT.
For whom does wearing one really make sense?
The best question is not “Is the sensor good?” but “What problem will it solve for this person?”
Often recommended
Type 1 diabetes; insulin through multiple doses or a pump; type 2 diabetes using basal insulin; risk of severe or unrecognized hypoglycemia.
May be considered depending on the problem
In type 2 diabetes without insulin: HbA1c outside target, medications that cause hypoglycemia, a treatment change, unexplained fluctuations or a need for short-term education.
Routine use cannot be recommended
Indefinite personal monitoring in prediabetes without medication, excess weight/obesity alone or healthy individuals without diabetes.
Type 1 diabetes
CGM is one of the core diabetes technologies that should be offered from diagnosis to children and adults able to use it appropriately. It plays a central role in insulin dosing, hypoglycemia alerts and the operation of automated insulin systems.
Type 2 diabetes using insulin
In people using multiple doses, a pump or basal insulin alone, it can improve HbA1c and time in range; it reveals nighttime drops and excessive basal insulin use.
People at high risk of hypoglycemia
People with hypoglycemia unawareness, a history of severe hypoglycemia, insulin or sulfonylurea/meglitinide use, older people with irregular eating patterns and those with reduced kidney function may benefit more.
Pregnancy and special situations
There is strong evidence in pregnant women with type 1 diabetes; in other types of diabetes, individual assessment is made throughout pregnancy. It may also reduce burden in cognitive or physical limitations where a caregiver monitors the readings.
Type 2 diabetes without insulin
Short-term or intermittent use may be useful if HbA1c is above target, meal responses are unclear, treatment has recently changed or data-based education is planned. The sensor is effective with structured feedback, not on its own.
Short-term professional CGM
For someone who does not want to wear a device continuously, a 10–14-day “metabolic Holter” approach may answer specific questions such as nighttime hypoglycemia, rises after meals or disagreement between HbA1c and home measurements.
How much benefit do studies show?
The benefit is real; however, its size depends on the person, treatment and how sensor data are used.
In 2026, a comprehensive review published in JAMA Internal Medicine reported that in type 2 diabetes, CGM reduced HbA1c by an average of approximately 0,3 percentage points compared with fingerstick monitoring or usual care. This is not a “30% reduction”: it is an absolute change, such as HbA1c falling from 8,5% to approximately 8,2%. The average is modest; benefit may be greater in some people with high baseline values who use insulin and make actual treatment changes based on the data.
A current meta-analysis involving people with type 2 diabetes who were not using insulin evaluated 8 randomized studies and 541 participants . CGM was associated with HbA1c differences of −0,37 percentage points, differences in time in range of +8,84 percentage points and differences in time above 180 mg/dL of −8,14 percentage points . These data make it incorrect to say “no one who does not use insulin benefits”; however, they alone do not determine who should use it permanently and who should use it for a short period.
The sensor, or the education given alongside the sensor? In a meta-analysis of 21 randomized studies focused on diet and lifestyle, HbA1c improved by −0,46 percentage points, time in range by +7,18 percentage points and body weight by approximately −2,06 kg. In another behavioral-change meta-analysis of 25 studies, however, HbA1c and time in range improved while differences in weight and BMI were not significant. The most likely explanation is that the sensor is not a weight-loss device on its own; its effect emerges together with education, goals, follow-up and treatment adjustment.
What is the situation in people with prediabetes, obesity or who are healthy?
Metabolic curiosity and clinical necessity are not the same thing.
Prediabetes and insulin resistance are important conditions that increase the risk of future type 2 diabetes. Nevertheless, the ADA 2026 explicitly states that evidence is still insufficient to support CGM use for screening and diagnosis of prediabetes or diabetes. A post-meal peak in a sensor curve does not, on its own, make someone “prediabetic.”
| Group | Current evidence | The most reasonable approach |
|---|---|---|
| Prediabetes | Glucose variability may be higher on average than in normoglycemia; however, CGM screening or continuous use has not been shown to prevent progression to diabetes. | First HbA1c, fasting glucose and, when needed, an OGTT; afterward, a limited-duration trial may be considered as part of a structured lifestyle program. |
| Obesity without diabetes | Short-term behavioral feedback is possible; there is no strong evidence that the sensor alone produces lasting weight loss. | A plan that treats obesity as a chronic disease, based on nutrition, movement, sleep and medication when needed. |
| Healthy, with low metabolic risk | There is no good evidence that CGM improves health, prevents diabetes or extends lifespan. | Evidence-based lifestyle habits and standard screening appropriate to age/risk, rather than routine continuous monitoring. |
| Healthy, but highly curious | A brief personal experiment may reveal some behaviors; its clinical meaning is uncertain and there is a risk of misinterpretation. | One predetermined question, a short duration, interpretation without panic and avoidance of unnecessary food restrictions based on the results. |
In a multicenter study monitoring 153 people without diabetes or obesity using blinded CGM, mean glucose was approximately 98–99 mg/dL and median time in the 70–140 mg/dL range was %96 . Nevertheless, participants were observed above or below these levels for a median of approximately 30 minutes above 140 mg/dL and 15 minutes below 70 mg/dL per day. Brief rises or falls, device error and physiological variability can therefore also occur in healthy people; a single “spike” is not a disease.
Why might the conclusion “Fruit spikes my glucose, so I should stop eating it” be wrong? A food's health value is not determined solely by its highest glucose peak. Fiber, vitamins, satiety, total energy, portion size, meal composition and the long-term eating pattern also matter. Unnecessarily removing beneficial foods such as fruit, legumes or whole grains to flatten the sensor curve may improve the number while worsening nutrition. A Mediterranean-style eating pattern offers a broader health framework than individual glucose peaks.
When does more data begin to cause harm?
The sensor's risks are not limited to the skin; misinterpreted data are also an adverse effect.
Alarm fatigue and interrupted sleep
Frequent high/low alerts can create stress, disrupt sleep at night and eventually lead to genuine alerts being ignored. Alarm thresholds should be individualized.
Fixation on data and fear of food
Treating every rise as a mistake may contribute to banning healthy foods, constant checking behavior or worsening symptoms in people predisposed to eating disorders.
False lows and false highs
Lying on the sensor may cause a “compression low,” while rapid changes may produce differences between blood and interstitial fluid. Some drugs and substances may affect measurements depending on the device type.
Skin problems and device burden
Irritant or allergic dermatitis related to the adhesive, pain, local infection, premature sensor detachment and discomfort from wearing the device continuously may occur.
Cost and opportunity cost
Money, time and clinician effort devoted to a sensor may be diverted from more effective medication, nutrition support, an exercise program or access for patients who truly need a sensor.
Privacy and the algorithm problem
Who stores health data, which target a commercial application considers “good” and whether advice is scientific or marketing-based are not always transparent.
In a systematic review of 54 qualitative studies combining the experiences of 1.845 people with diabetes, CGM provided control, confidence, freedom and motivation for self-management. In the same review, users also reported technological complexity, distrust of algorithms, data overwhelm, inaccurate measurements, device malfunction, alarm stress and cost barriers. The psychological effect is therefore not one-directional: a device that reduces burden for the right person can create a new burden for the wrong person.
If a sensor is used, what should correct use look like?
A small clinical protocol is needed instead of “Let's attach a sensor and see what happens.”
| Metric | General target for most adults with diabetes | How should it be interpreted? |
|---|---|---|
| TIR: 70–180 mg/dL | >%70 | Shows how much of the day is spent in range. A 5% increase corresponds to approximately 72 minutes per day. |
| TBR: <70 mg/dL | <%4 | Shows the burden of hypoglycemia. In older or frail people, the target is generally stricter: <1%. |
| TBR: <54 mg/dL | <%1 | Clinically significant hypoglycemia; the treatment plan and measurement accuracy should be reviewed promptly. |
| TAR: >180 mg/dL | <%25 | Shows the total time above target; it should not be reduced to a single post-meal peak. |
| Coefficient of variation | Generally ≤36% | Summarizes fluctuations; lower variability becomes important in people at risk of hypoglycemia. |
These targets were developed for diabetes management; they should not be used to give healthy people a “perfect glucose score.” Age, pregnancy, frailty and hypoglycemia risk alter the targets.
A special use in oncology practice: Dexamethasone and other glucocorticoids may raise afternoon and evening glucose more than fasting glucose. During repeated steroid courses, short-term CGM may capture the pattern in selected patients with pre-existing diabetes or marked hyperglycemia. However, the ADA particularly emphasizes postprandial or random glucose monitoring in people taking steroids; wearing a sensor is not mandatory for every patient receiving steroids. If reduced food intake, vomiting and insulin use coexist, hypoglycemia risk should also be assessed.
The aim of dietary adjustment is not to make the sensor curve mathematically perfect, but to achieve sustainable metabolic health. Sugary drinks and whole fruit should not be assessed under the same heading of a “rise”; the sugar load of drinks and healthier alternatives make this distinction concrete.
Will everyone monitor their glucose continuously in the future?
More people will probably be monitored; however, the real advance will be better-filtered meaning, not more numbers.
A sensor beneath the skin and a smart algorithm
Real-time or intermittently scanned CGM is becoming central to diabetes care alongside insulin pumps and automated dosing systems.
Multiple biomarkers
On 25 August 2026, the FDA authorized the Libre Duo 10 Day system through the De Novo pathway as the first device to monitor glucose and ketones continuously in the same sensor. Adding lactate and other metabolites is being investigated.
Context-sensitive, quiet monitoring
Systems that combine meal, exercise, sleep, medication and illness data and issue alerts only when action is needed may be more useful than a flood of raw data.
The idea of measurement without needles is appealing: technologies attempting to measure glucose through watches, rings, optical spectroscopy, sweat or tears are being developed. However, the FDA's safety communication states that no smartwatch or ring claiming to measure blood glucose on its own without piercing the skin has been approved. An unvalidated device sold today as a “noninvasive glucose meter” may cause serious harm, particularly if it leads to incorrect insulin dosing.
The best system of the future will probably not display a number to everyone every five minutes. A system that gives early warnings to people at risk, avoids unnecessary alarms in healthy people, presents summaries to clinicians and measures the user's psychological burden is more valuable. Maturity in digital health will not mean continuously monitoring everything we can monitor; rather, knowing which data should be visible and when will define it.
Conclusion: A sensor is a tool, not a goal
A glucose sensor is powerful for the right person: it reveals hypoglycemia, makes insulin treatment safer, increases time in range and reduces the burden of fingersticks. However, when the purpose of use changes while the technology stays the same, the benefit–harm balance changes too.
In type 1 diabetes, insulin treatment or high hypoglycemia risk, a sensor is often a clinical necessity. In type 2 diabetes without insulin, use that answers a specific problem and is linked to education and treatment changes makes sense. In prediabetes, obesity and healthy people, however, evidence is not yet sufficient for routine, indefinite use.
The best prescription is this: define the question first, then measure; link the data to an action; do not hesitate to remove the device if it is not helping.
Scientific sources
- Dower JA, Johansson M, Camp AW, Montori VM, Lipska KJ. Continuous Glucose Monitoring in Type 2 Diabetes and Beyond: A Review. JAMA Internal Medicine. 2026. doi:10.1001/jamainternmed.2026.2772.
- Johansson M, Dower JA, Lipska KJ. Continuous Glucose Monitoring for Persons With and Without Diabetes. JAMA Internal Medicine. 2026.
- American Diabetes Association. Diabetes Technology: Standards of Care in Diabetes—2026. Diabetes Care. 2026;49(Suppl 1):S150–S182.
- American Diabetes Association. Diagnosis and Classification of Diabetes: Standards of Care in Diabetes—2026. Diabetes Care. 2026.
- American Diabetes Association. Glycemic Goals, Hypoglycemia, and Hyperglycemic Crises: Standards of Care in Diabetes—2026. Diabetes Care. 2026.
- Martens T, Beck RW, Bailey R, et al. Effect of Continuous Glucose Monitoring on Glycemic Control in Patients With Type 2 Diabetes Treated With Basal Insulin. JAMA. 2021;325:2262–2272.
- Aronson R, et al. Continuous glucose monitoring in noninsulin-treated type 2 diabetes: updated systematic review and meta-analysis. 2025.
- Foo V, et al. Comparative effectiveness of glucose monitoring methods and glycaemic control in type 2 diabetes: an umbrella review. Diabetes Obesity and Metabolism. 2026.
- Continuous Glucose Monitoring to Guide Lifestyle Choices With a Focus on Nutrition in Type 2 Diabetes. Systematic review and meta-analysis. 2025.
- The efficacy of CGM as a behaviour change tool in populations with and without diabetes. Systematic review and meta-analysis. 2024.
- Shah VN, et al. Continuous Glucose Monitoring Profiles in Healthy Nondiabetic Participants. Journal of Clinical Endocrinology & Metabolism. 2019.
- Natale P, et al. Patient experiences of continuous glucose monitoring: a systematic review of qualitative studies. Journal of Diabetes. 2023.
- U.S. Food and Drug Administration. Do Not Use Smartwatches or Smart Rings to Measure Blood Glucose Levels. Safety Communication. 2024.
- U.S. Food and Drug Administration. FDA Authorizes First Wearable Device That Continuously Monitors Both Ketone Levels and Blood Sugar. 25 August 2026.
This content is intended for general information; it is not a recommendation for personal diagnosis or treatment. People who use insulin or medications that can cause hypoglycemia should not change their treatment based on sensor data without a physician's recommendation.
Was this article helpful?
Your feedback helps us improve our content.
