Story
CGM for the Metabolically Healthy
Last updated
In one pass A continuous glucose monitor (CGM) measures sugar in the interstitial fluid under the skin — the fluid between cells — not in the blood inside your vessels.
Educational content, not medical advice — consult a clinician.
Story path
Chapter 1
What a glucose monitor measures
A (CGM) measures sugar in the interstitial fluid under the skin — the fluid between cells — not in the blood inside your vessels. A fine sensor stays under the skin and gives a reading every few minutes, and those readings join up into the line on your phone. Sugar has to seep out of the vessels into that fluid first, so the monitor runs several to fifteen-odd minutes behind fingertip blood, and its number is not exactly blood glucose.
Many people panic when they see that line shoot up after a meal. In a metabolically healthy person, a curve that rises and then falls back is what this system does every day: sugar enters the blood, and the body moves it into cells. It is not diabetes, and it is not a quiet loss of control.
There is a ready reference for what normal looks like. In a multicenter observational study, 153 healthy people without diabetes and without obesity wore a CGM, and they spent about 96% of their time inside the normal range (Shah 2019). So the question is not whether your glucose moves, but how much movement counts as normal.
Many people panic when they see that line shoot up after a meal. In a metabolically healthy person, a curve that rises and then falls back is what this system does every day: sugar enters the blood, and the body moves it into cells. It is not diabetes, and it is not a quiet loss of control.
There is a ready reference for what normal looks like. In a multicenter observational study, 153 healthy people without diabetes and without obesity wore a CGM, and they spent about 96% of their time inside the normal range (Shah 2019). So the question is not whether your glucose moves, but how much movement counts as normal.
Numbers · Where a healthy person's glucose sits
Shah 2019 is a multicenter prospective observational study, not a . It had one aim: to set a reference for what a healthy person's continuous glucose curve looks like. 153 healthy children and adults without diabetes and without obesity wore a (*JCEM* 2019). Devices of this kind give a reading every 1–15 minutes; the interval depends on the model.The results: the median participant spent about 96% of the time between 70 and 140 mg/dL (3.9–7.8 mmol/L). Mean glucose was about 98–99 mg/dL, a little higher in people over 60. Across the whole day, the median time above 140 was about 2%, roughly 30 minutes a day; time above 160 was smaller still, a median of only about 0.3%.
By this reference, a healthy person whose reading climbs to 120–140 after a meal and then comes back down is still inside the usual range of this group. It is something the system does every day, and does steadily.
Know its limits: this is a descriptive study. It answers what a healthy curve usually looks like; it does not answer whether straying a little outside that range does any harm. The later chapters ask exactly that: how much movement is normal, how the body reins it in, and which flatten-the-curve claims have substance.
Chapter 2
How the body brings glucose back
What pulls post-meal glucose back down is a pair of hormones that push in opposite directions: one works the brake, the other the accelerator.
Insulin is the brake. When sugar enters the blood, beta cells in the pancreas release it, and muscle, liver and fat move the sugar into their cells. At the same time the liver stops releasing sugar into the blood and stores the extra as glycogen. Glucagon is the accelerator. When glucose runs low during fasting or exercise, alpha cells in the pancreas release it, and the liver breaks glycogen apart and puts sugar back into the blood, so glucose does not fall too far.
When you eat, the gut also releases a class of hormones called incretins ( is one): they prompt insulin, hold back glucagon, and slow the emptying of the stomach, so the wave of sugar arrives more smoothly.
The small hill you see on the curve is this system at work, not a sign it is broken. In a healthy person the peak usually comes about 1 hour after the first bite, and glucose is back at baseline within 2–3 hours; in diabetes the curve climbs and does not come back.
Insulin is the brake. When sugar enters the blood, beta cells in the pancreas release it, and muscle, liver and fat move the sugar into their cells. At the same time the liver stops releasing sugar into the blood and stores the extra as glycogen. Glucagon is the accelerator. When glucose runs low during fasting or exercise, alpha cells in the pancreas release it, and the liver breaks glycogen apart and puts sugar back into the blood, so glucose does not fall too far.
When you eat, the gut also releases a class of hormones called incretins ( is one): they prompt insulin, hold back glucagon, and slow the emptying of the stomach, so the wave of sugar arrives more smoothly.
The small hill you see on the curve is this system at work, not a sign it is broken. In a healthy person the peak usually comes about 1 hour after the first bite, and glucose is back at baseline within 2–3 hours; in diabetes the curve climbs and does not come back.
Mechanism · When the post-meal peak comes and goes
The default result of this brake-and-accelerator system is a regular timetable: in a healthy person, post-meal glucose peaks about 1 hour after the first bite and returns to baseline within 2–3 hours (a review by Dimitriadis 2021).Why does the peak not keep climbing? Because after a meal several gates work at once, not insulin alone:
What you eat and in what order: composition and sequence decide how fast sugar reaches the bloodHow fast the stomach empties: when it releases food slowly, sugar enters the small intestine a little at a timeGut hormones: incretins prompt insulin and hold back glucagonInsulin and glucagon themselves: one sends sugar into cells, the other stops topping up the bloodThe liver: the review puts it at the center, because a good share of the sugar from a meal is intercepted and stored by the liver first instead of rushing into the general circulation
So the rise-and-fall hill on a is what these gates look like when they work together in real time. The review's authors also spell out what this teamwork is for: it keeps the acute rise in post-meal glucose and insulin within bounds, protecting blood vessels and tissues from marked high blood sugar.
Mechanism · Same climb, healthy or diabetic
Diabetes is this system failing. In type 1, there is not enough insulin, because the body has almost stopped making its own. In type 2, cells respond poorly to insulin and the pancreas gradually cannot make up the difference. Either way the result is the same: glucose climbs and does not come back, and stays high for the long term.A healthy person's curve climbs too, but it comes back down cleanly. The same rising line can mean a broken system or a system doing its job. So how high the line climbs tells you nothing on its own; what matters is whether it comes back, and where it sits the rest of the time.
That is the dividing line for this whole story. A healthy person spending most of the day in the normal range is what this brake and accelerator achieve, steadily, every day. The next question is whether the ups and downs on a healthy person's curve can themselves be read as a sign of disease.
Chapter 3
What variability predicts
The ups and downs on a healthy person's curve are the brake and accelerator dealing with sugar that was just eaten. They are not a disease tracing.
Marketing's favorite line is that a can catch you quietly losing control even when your lab results are normal. The line has a real origin, but it has been inflated. Some people with normal standard lab tests do, now and then, drift into the higher ranges on the curve. But the classification behind that claim came from an exploratory study of 57 people, and its reproducibility has been questioned by other researchers in a formal comment. The device also reads high: in a randomized crossover trial, it ran systematically about 0.9 mmol/L above fingertip blood. A good share of the scary spikes are the device being off, not the body swinging that much.
In people without diabetes, a systematic review found no clear link between variability and insulin sensitivity, fatty liver, obesity, blood lipids or blood pressure. And whether deliberately flattening a healthy person's curve buys fewer diseases or a longer life has never been tested in any trial.
Marketing's favorite line is that a can catch you quietly losing control even when your lab results are normal. The line has a real origin, but it has been inflated. Some people with normal standard lab tests do, now and then, drift into the higher ranges on the curve. But the classification behind that claim came from an exploratory study of 57 people, and its reproducibility has been questioned by other researchers in a formal comment. The device also reads high: in a randomized crossover trial, it ran systematically about 0.9 mmol/L above fingertip blood. A good share of the scary spikes are the device being off, not the body swinging that much.
In people without diabetes, a systematic review found no clear link between variability and insulin sensitivity, fatty liver, obesity, blood lipids or blood pressure. And whether deliberately flattening a healthy person's curve buys fewer diseases or a longer life has never been tested in any trial.
Evidence · Does the glucotype idea hold up?
The real origin of this claim is Hall 2018, an exploratory study of 57 people that proposed the idea of a glucotype: grouping people by the shape of their curve. It reported that some people judged normal by standard lab tests still, on a CGM, drifted into the prediabetic range now and then (about 15% of the time) and even into the diabetic range (about 2% of the time). That result was later used to argue that everyone is quietly losing glucose control.The classification itself drew serious criticism from other researchers. Hulman 2021 published a formal comment in the same journal that examined, point by point, whether glucotypes carry over to other populations, whether the results can be interpreted, and how they relate to the traditional CGM measures. Their conclusion was that none of these stands on firm ground.
So it reads more like an exploratory finding worth further study than a mature tool for labeling healthy people. Keep one more thing in mind: those percentages were themselves read off a CGM, and the next question is exactly how accurate that device is.
Evidence · How far a CGM reads high
Hutchins 2025 is a randomized crossover trial. On separate lab visits, 15 healthy adults took different test drinks or foods; each time their glucose was measured with a and with fingertip blood at the same time, and the two were compared directly, with fingertip blood as the reference answer.The CGM read both fasting and post-meal glucose about 0.9 mmol/L higher than fingertip blood. It overestimated the time spent above 7.8 mmol/L (140 mg/dL) roughly 2–4-fold: about 4-fold using the raw readings, and still about 2-fold after subtracting each person's fasting baseline difference. For the same commercial fruit smoothie, the CGM gave a glycemic index of 69 (medium), while the standard method gave 53 (low). That difference sat right at the edge of statistical significance, but it was enough to move a food from low to medium. The size of the bias also differed from person to person.
In other words, a good share of the scary spikes a healthy person sees on the device are the device reading high, not the body swinging that much. That is device bias, not secret diabetes. The authors' own advice is direct: to know accurately how much a food raises blood sugar, use fingertip blood first.
Evidence · What variability predicts without diabetes
So what does glycemic variability predict in people without diabetes? The most honest answer today: the evidence is weak and inconsistent.Hjort 2024 is a systematic review and that pooled studies measuring variability with a in people without diabetes. Most of those studies were cross-sectional: they compare several measures at a single point in time, so they cannot show which came first. Its findings come in three layers:
Variability is higher in people with prediabetes, and it runs opposite to beta-cell function: the weaker the pancreas's ability to make insulin, the bumpier the curveVariability shows no clear link with insulin sensitivity, fatty liver, obesity, blood lipids, blood pressure or oxidative stressIn people who already have coronary heart disease, variability may be linked to how advanced their atherosclerosis is and to cardiovascular events
The authors did not close the door. They suggest variability may still be a risk measure worth tracking in people without diabetes, and they call for prospective studies to test whether it predicts future disease. That describes where things stand: it is a research question, not a conclusion.
The more important point is that no trial has ever tested whether deliberately flattening a healthy person's variability buys fewer diseases or a longer life. Not studied and studied and found no effect are two different things, and this question has not even reached the first stage. Seeing variability is not the same as finding disease.
Myth · Do glucose spikes cause every disease?
"Glucose spikes are the root of all disease — they inflame you, age you, make you store fat and give you diabetes" is the core line of this business. It takes something true in diabetes and moves it onto healthy people, swapping the premise along the way.In diabetes, high blood sugar that stays high over the long term, again and again, genuinely damages blood vessels, nerves, kidneys and eyes, and that evidence is solid. But the small peak a healthy person has after a meal — up to 120–140 and back down within two hours — is a completely different thing: brief, controlled, happening every day, the system working normally.
The premise is swapped in three steps:
Long-term uncontrolled high blood sugar is harmful becomes any rise after a meal is harmfulVariability is linked to complications in people with diabetes becomes variability is quietly causing disease in healthy people too — when Hjort 2024 shows precisely that this link is weak in people without diabetesCorrelation is presented as causation, and then turned around to sell you a promise that flattening it prevents disease — a promise with no outcome evidence at all in healthy people
The real, usable stance: if you already have prediabetes or diabetes, controlling post-meal glucose is a meaningful clinical goal. But if you are metabolically healthy, "my glucose hit 140 after a mango" does not mean you are getting sick; it means the mango contains sugar and your body is handling it normally. What actually shapes long-term metabolic health are the big levers — weight, activity, dietary fiber, sleep, not smoking — not the height of any single meal's curve.
Chapter 4
Which curve-flattening tips work
Eating vegetables and protein before the starchy part of a meal, and getting up for a walk afterward, both slow how fast that meal's sugar reaches the blood and let the muscles take up more of it on the spot. The same meal can also draw noticeably different curves in different people.
These effects are real, but they play out over the few hours after a meal. Protein and dietary fiber slow the emptying of the stomach and draw on the incretin hormones; when you move, the working muscles pull sugar out of the blood; and differences between people come from what is in the meal, from the person, and partly from the gut microbiome.
The parts that matter for long-term metabolism — more vegetables and protein, more walking — are good advice anyway, and they do not need a to approve them. At most, the device lets you watch these effects with your own eyes; being able to see them does not mean you need to. The studies behind these tips are all small, and they measured the curve over a few hours after a meal, not whether people get sick later.
These effects are real, but they play out over the few hours after a meal. Protein and dietary fiber slow the emptying of the stomach and draw on the incretin hormones; when you move, the working muscles pull sugar out of the blood; and differences between people come from what is in the meal, from the person, and partly from the gut microbiome.
The parts that matter for long-term metabolism — more vegetables and protein, more walking — are good advice anyway, and they do not need a to approve them. At most, the device lets you watch these effects with your own eyes; being able to see them does not mean you need to. The studies behind these tips are all small, and they measured the curve over a few hours after a meal, not whether people get sick later.
Evidence · How much eating vegetables first helps
If you eat the vegetables and protein of a meal before its starchy part, the post-meal glucose curve comes out clearly flatter. The mechanism suggests why: protein and dietary fiber slow the emptying of the stomach and draw on the incretin hormones.Shukla 2015 is a very small crossover trial: 11 people with type 2 diabetes ate the same meal on two different days, changing only the order, each person serving as their own control. On the vegetables-and-protein-first day, the area under the glucose curve over the 120 minutes after the meal — the total glucose rise — was about 73% lower. Small trials have since seen the same direction in prediabetes and in healthy young adults.
But this is a number over a few hours after one meal. Whether reordering food makes healthy people less likely to get sick has never been tested. It is a good habit, not a medicine, and not a way for healthy people to prevent disease.
Evidence · What a few minutes of walking does
Buffey 2022 is a systematic review and of several one-day randomized crossover trials. The same adults spent one day sitting continuously and another day getting up every so often to stand or to take an easy walk, each break lasting 2–5 minutes, and that day's measurements were compared.The results: breaking up sitting with standing lowered post-meal glucose a little, with no clear change in insulin. Breaking it up with light walking clearly lowered both post-meal glucose and insulin, and did so more than standing. Neither changed systolic blood pressure.
Know its limits: these were one-day trials that looked at glucose and insulin on that day, not health outcomes months later.
Walking's benefits go far beyond blood sugar anyway, so this advice holds even if you never look at a . It is one of the big levers, and it does not need a curve to approve it.
Evidence · How much responses differ between people
Even on identical meals, people's post-meal glucose can differ by a wide margin. PREDICT (Berry 2020) was an observational study of 1,002 twins and unrelated healthy adults in the UK who ate standardized meals in a clinic and at home. For the same meals, the between-person coefficient of variation in post-meal glucose response (the standard deviation as a share of the mean) was about 68%. For glucose, the composition of the meal itself explained more than personal factors such as the gut microbiome. That is the scientific foundation of the personalized-nutrition business.Individual differences are real, but that does not mean every healthy person needs a to eat well. The ZOE company's METHOD (Bermingham 2024) put 347 US adults on an 18-week diet program that scored foods based on their personal test results, against general advice from the US Dietary Guidelines. Of the two primary outcomes, fell a little and low-density lipoprotein () cholesterol did not change. Among secondary outcomes, weight, waist size and (HbA1c, a two-to-three-month average of blood sugar) improved, and more people in the personalized group reported better energy and mood (participants knew which group they were in, so this kind of self-report is easily biased).
Note three things: this was the effect of a whole lifestyle program, not a CGM alone; the outcomes were intermediate markers, not rates of disease; and most of the authors were employees of or advisers to the company. It shows personalization has some signal, but it is nowhere near proving that a healthy person must wear a CGM.
Chapter 5
Who really needs one
A earns its keep when its number changes a decision, and that decision concerns a system that has already failed or is being treated: how much insulin to take, whether to change medicines, whether to go to the hospital.
For people with type 1 diabetes and anyone taking insulin, real-time readings and high and low glucose alarms bear directly on safety, which is why the American Diabetes Association (ADA) guideline clearly recommends CGM for them; for people with type 2 diabetes on other glucose-lowering medicines, the guideline says it can be considered. For a metabolically healthy person, however pretty the curve, it rarely changes what they should do: eat more vegetables, walk more — the same as before. Information is worth something only when it can change an action.
If you have already been diagnosed with diabetes or prediabetes, this story is about healthy people and is not aimed at you. Follow your doctor and your guidelines, and do not stop monitoring or medication you need because of one popular-science article.
For people with type 1 diabetes and anyone taking insulin, real-time readings and high and low glucose alarms bear directly on safety, which is why the American Diabetes Association (ADA) guideline clearly recommends CGM for them; for people with type 2 diabetes on other glucose-lowering medicines, the guideline says it can be considered. For a metabolically healthy person, however pretty the curve, it rarely changes what they should do: eat more vegetables, walk more — the same as before. Information is worth something only when it can change an action.
If you have already been diagnosed with diabetes or prediabetes, this story is about healthy people and is not aimed at you. Follow your doctor and your guidelines, and do not stop monitoring or medication you need because of one popular-science article.
Clinical · Who the ADA guideline names
Under the American Diabetes Association's Standards of Care in Diabetes—2025 (the Diabetes Technology chapter), is recommended for these groups:Type 1 diabetes, starting early after diagnosis. These people's bodies make almost no insulin of their own, so injected insulin has to make the body's decisions for it, and real-time readings with high and low alarms bear directly on safety.Anyone with diabetes who takes insulin (type 1 or type 2): it helps adjust doses, catches low blood sugar at night, and shows which meals push glucose highest.Type 2 diabetes treated with glucose-lowering medicines other than insulin: from the 2025 edition this is also in the guideline, worded as "consider", to help reach each person's own glucose target.
What these people share is a glucose-control system that has already gone wrong or is being treated, so the information a CGM gives actually changes a decision. The guideline does not recommend CGM for people without diabetes.
Clinical · Uses outside the guideline list
Outside the guideline list, a has a few other uses, but they carry different weight:Prediabetes: some people use one to see which foods and activities best pull their own glucose back into the normal range, as a form of feedback. The guideline does not list this group as recommended, and no trial has yet answered whether using it this way brings long-term benefit; decide with your doctor whether to wear one.Research: it is a fine instrument for studying post-meal metabolism, differences between people and drug effects, and many of the observations cited in this story were made with it.When a diagnosis is unclear: doctors sometimes have a patient wear one for a short period to catch high or low glucose patterns that lab tests miss.
See the difference? In these settings, the number has a chance to change an action. For a metabolically healthy person, however pretty the curve, it rarely changes anything they should do. A reference for where a healthy person's glucose sits across the day is in the chapter What a glucose monitor measures.
Chapter 6
The cost of over-monitoring
When a healthy person treats normal ups and downs as an alarm, the monitoring itself becomes a new source of stress.
A review of use in people without diabetes found that consistent, high-quality evidence that it helps this group is lacking: detecting abnormal glucose, changing behavior, improving metabolism — none of the three has held up. The harms it might bring have also barely been studied: misreading normal fluctuation as a defect, starting to fear healthy foods such as fruit and oats, checking the reading dozens of times a day, even turning the pursuit of a pure diet into an obsession.
The default stance: you do not need a device to prove you are healthy. Rising after a meal and falling back is your body at work. If you really want to wear one, treat it as a short experiment, not a verdict. This is popular science and no substitute for a doctor.
A review of use in people without diabetes found that consistent, high-quality evidence that it helps this group is lacking: detecting abnormal glucose, changing behavior, improving metabolism — none of the three has held up. The harms it might bring have also barely been studied: misreading normal fluctuation as a defect, starting to fear healthy foods such as fruit and oats, checking the reading dozens of times a day, even turning the pursuit of a pure diet into an obsession.
The default stance: you do not need a device to prove you are healthy. Rising after a meal and falling back is your body at work. If you really want to wear one, treat it as a short experiment, not a verdict. This is popular science and no substitute for a doctor.
Safety · Possible harms when sold to healthy people
Oganesova 2024 is a narrative review that looks specifically at use in people without diabetes. It first breaks the market's pitch into three steps: detecting abnormal glucose, prompting behavior change, and improving metabolic health. Then it looks for evidence on each, and concludes that all three lack consistent, high-quality evidence. When the review was written, clinical guidelines recommended CGM only for people with type 1 diabetes and for people with type 2 diabetes treated with insulin.The review also points to several gaps that have barely been studied, one of them being the possible harm to healthy people's eating habits. Reasoning from how the device is used, the risks look roughly like this, but how often they happen and how serious they are has not been measured:
Misreading normal fluctuation as a defect: many people believe blood sugar should not rise at all after eating, so a normal post-meal bump makes them anxious — when rising and falling back is exactly what healthy looks like.Starting to fear food: to avoid spikes, some people cut out healthy foods such as fruit, oats and carrots and eat an ever-narrower diet. That is backwards.Getting tied to the numbers: checking the curve dozens of times a day and letting a reading set your mood.Feeding disordered eating and an obsession with healthy food: in people already inclined that way, it may push the pursuit of a pure diet into a pathological fixation.
The review's authors therefore argue that commercial claims that the device helps healthy people should be labeled misleading, and that oversight of these devices after they reach the market should be strengthened.
Disclaimer: this is popular science, not medical advice, and no substitute for a doctor. It discusses metabolically healthy people; if you have, or suspect you have, diabetes, prediabetes or another metabolic condition, follow your doctor's advice, use monitoring tools as your guidelines direct, and do not change a plan your doctor gave you because of one popular-science article.
In practice · Three kinds of reader, three answers
Here is should you wear one in three cases:Diagnosed diabetes, or on insulin: if you take insulin, the guideline recommends it; if you have type 2 diabetes without insulin, the guideline says it can be considered. Follow your doctor and the ADA guideline; the debunking in this story is not aimed at you. If you have been diagnosed with prediabetes, the guideline does not recommend it, so decide with your doctor.Metabolically healthy and simply curious: optional. If you want to, treat it as a two-week experiment: see for yourself that walking and eating vegetables first really do flatten the curve, and stop once your curiosity is satisfied. Do not wear it long-term, and do not make it the judge of every meal.Metabolically healthy but prone to anxiety, or with a history of disordered eating: better not to wear one. For you, the curve is more likely to become a new source of anxiety, while the information it gives can barely change anything you should already be doing.
One overall test: information is worth collecting only when it can change an action and does not harm you. For people with diabetes under treatment, a usually meets both conditions; for healthy people, it usually meets neither.
References · 12
- Shah, V. N., DuBose, S. N., Li, Z., et al. (2019). Continuous glucose monitoring profiles in healthy nondiabetic participants: a multicenter prospective study. Journal of Clinical Endocrinology & Metabolism, 104(10), 4356-4364. Multicenter prospective observational study: 153 healthy, nonobese, nondiabetic participants aged 7-80 wore a blinded Dexcom G6 for up to 10 days. Median time at 70-140 mg/dL 96% (IQR 93-98); mean glucose 98-99 mg/dL in all age groups except over 60 years (104 mg/dL); within-person CV 17 ± 3%. Median time > 140 mg/dL 2.1% (~30 min/day) and < 70 mg/dL 1.1% (~15 min/day). Full-text Table 2 (24 h, all participants): > 160 mg/dL median 0.3% (IQR 0.1-0.9), > 180 mg/dL 0.0%. In the >= 60 group: > 140 mg/dL 4.1%, > 160 mg/dL 0.6%, 70-140 mg/dL 93% (abstract, PMID 31127824; full text, PMC7296129). 10.1210/jc.2018-02763
- Dimitriadis, G. D., Maratou, E., Kountouri, A., Board, M., & Lambadiari, V. (2021). Regulation of postabsorptive and postprandial glucose metabolism by insulin-dependent and insulin-independent mechanisms: an integrative approach. Nutrients, 13(1), 159. 10.3390/nu13010159
- Hall, H., Perelman, D., Breschi, A., Limcaoco, P., Kellogg, R., McLaughlin, T., & Snyder, M. (2018). Glucotypes reveal new patterns of glucose dysregulation. PLoS Biology, 16(7), e2005143. 10.1371/journal.pbio.2005143
- Hulman, A., Foreman, Y. D., Brouwers, M. C. G. J., et al. (2021). Towards precision medicine in diabetes? A critical review of glucotypes. PLoS Biology, 19(3), e3000890. 10.1371/journal.pbio.3000890
- Hutchins, K. M., Betts, J. A., Thompson, D., Hengist, A., & Gonzalez, J. T. (2025). Continuous glucose monitor overestimates glycemia, with the magnitude of bias varying by postprandial test and individual: a randomized crossover trial. American Journal of Clinical Nutrition, 121(5), 1025-1034. 15 healthy adults (9 female), 7 laboratory visits, randomized crossover; CGM compared with capillary sampling (the criterion) every 15 min for 120 min after 50 g glucose or fruit and smoothie carbohydrate loads. CGM read fasting and postprandial glucose 0.9 ± 0.6 and 0.9 ± 0.5 mmol/L higher than capillary samples, overestimated time > 7.8 mmol/L about 4-fold (about 2-fold after baseline adjustment), and gave a smoothie glycemic index of 69 vs 53 by capillary (P = 0.05); the fasting bias varied between participants (abstract, PMID 40021059). 10.1016/j.ajcnut.2025.02.024
- Hjort, A., et al. (2024). Glycemic variability assessed using continuous glucose monitoring in individuals without diabetes and associations with cardiometabolic risk markers: a systematic review and meta-analysis. Clinical Nutrition, 43(4), 915-925. 71 studies, the majority cross-sectional, searched to April 2022. Glycemic variability is higher in prediabetes and inversely associated with beta-cell function, but not clearly associated with insulin sensitivity, fatty liver disease, adiposity, blood lipids, blood pressure or oxidative stress; it may be associated with atherosclerosis and cardiovascular events in people with coronary disease. The authors call for prospective studies (abstract, PMID 38401227). 10.1016/j.clnu.2024.02.014
- Shukla, A. P., Iliescu, R. G., Thomas, C. E., & Aronne, L. J. (2015). Food order has a significant impact on postprandial glucose and insulin levels. Diabetes Care, 38(7), e98-e99. Pilot letter: 11 adults with metformin-treated type 2 diabetes, within-subject crossover (the letter does not say the order was randomised), the same 628 kcal meal on 2 days a week apart. Eating vegetables and protein before carbohydrate lowered glucose by 28.6%, 36.7% and 16.8% at 30, 60 and 120 min and glucose iAUC by 73% vs the reverse order (full text, PMC4876745). 10.2337/dc15-0429
- Buffey, A. J., Herring, M. P., Langley, C. K., Donnelly, A. E., & Carson, B. P. (2022). The acute effects of interrupting prolonged sitting time in adults with standing and light-intensity walking on biomarkers of cardiometabolic health: a systematic review and meta-analysis. Sports Medicine, 52(8), 1765-1787. 10.1007/s40279-022-01649-4
- Berry, S. E., Valdes, A. M., Drew, D. A., et al. (2020). Human postprandial responses to food and potential for precision nutrition. Nature Medicine, 26(6), 964-973. PREDICT 1: 1,002 twins and unrelated healthy UK adults eating standardized meals in clinic and at home (observational; US validation cohort n = 100). Between-person variability (population coefficient of variation) in responses to identical meals: triglyceride 103%, glucose 68%, insulin 59%. For postprandial glycemia, meal macronutrients explained more of the variance (15.4%) than person-specific factors such as the gut microbiome (6.0%); for lipemia it was the reverse. Genetic variants explained 9.5% for glucose. Many authors are consultants to, or current or former employees of, Zoe Global Ltd (abstract and conflict-of-interest statement, PMID 32528151). 10.1038/s41591-020-0934-0
- Bermingham, K. M., et al. (2024). Effects of a personalized nutrition program on cardiometabolic health: a randomized controlled trial. Nature Medicine, 30(7), 1888-1897. METHOD trial: 347 US adults aged 41-70 randomized to ZOE's app-based personalized dietary program (n = 177) or standard USDA-guideline advice (n = 170) for 18 weeks. Primary outcomes: triglycerides fell more with the program (mean difference -0.13 mmol/L, P = 0.016); LDL cholesterol did not differ. Weight, waist, HbA1c, diet quality and microbiome beta-diversity improved (P < 0.05), particularly in highly adherent participants; blood pressure, insulin, glucose, C-peptide, apoA1, apoB and postprandial triglycerides did not differ. Self-reported improvement: energy 43% vs 11%, sleep 35% vs 9%, mood 33% vs 15%, hunger 22% vs 14%. Participants wore CGMs for up to 14 days and were unblinded to the results; the between-group analysis was done by a blinded researcher. ZOE co-founders, consultants and current or former employees are among the authors (abstract, PMID 38714898; full text, PMC11271409). 10.1038/s41591-024-02951-6
- American Diabetes Association Professional Practice Committee. (2025). 7. Diabetes technology: Standards of Care in Diabetes-2025. Diabetes Care, 48(Suppl. 1), S146-S166. 10.2337/dc25-S007
- Oganesova, Z., Pemberton, J., & Brown, A. (2024). Innovative solution or cause for concern? The use of continuous glucose monitors in people not living with diabetes: a narrative review. Diabetic Medicine, 41, e15369. 10.1111/dme.15369