Creating an optimal diet by tracking glucose with a continuous glucose monitor (CGM)

Continuous glucose monitoring (CGM) can offer real-time, actionable data to create an personally optimized diet.

Casey Means MD; Josh Clemente; and Arlo Crawford


Article highlights

  • Viewing diet through the lens of glucose levels can be a powerful way to personalize food choices and optimize metabolic health.
  • Post-meal spikes in glucose can be uniquely damaging to health and minimizing them should be a dietary priority. Keeping glucose levels steady over time leads to better health outcomes.
  • Having a fasting glucose level in the "normal" range may not be good enough: research shows that a higher fasting glucose level (even in the "normal" range) increases risk of chronic disease. Dietary choices have a strong impact on fasting glucose levels.
  • There is simply no one-size-fits-all metabolic eating plan that works for everyone. Different people can have very different glycemic reactions to the exact same foods, and the metabolic impact of lifestyle choices and other individual factors are difficult to measure with traditional tools.
  • Continuous glucose monitoring (CGM) can help. With real-time, objective, actionable information about how food and lifestyle choices affect glucose levels, it’s possible to create an individualized diet to support optimal health.

Humans are complex organisms, each with a unique genetic, biochemical, and microbial blueprint, and the reality is that every bite we consume has an impact on whether our bodies are moving towards a state of optimal health or a state of dysfunction. While a standard guidebook for living would be nice, the truth is that each individual must determine the personalized diet that supports their body’s highest level of functioning. Fortunately, continuous glucose monitoring (CGM) can be a powerful tool in this pursuit. CGM can provide data and feedback to help determine how daily choices affect glucose levels in real-time. We can finally write our own guidebooks, linking the meals we eat with metabolic health, and quickly access quantitative, actionable information that can optimize diet.

Make no mistake, it’s tough out there. We’re surrounded by what feels like unlimited sources of food, plenty of which are marketed as “healthy.” But despite this abundance, our country is sicker than ever. Recent research shows that in 2018, 80% of consumers found conflicting information about food and nutrition, and 59% say that the conflicting information makes them doubt their choices. The beauty of objective data is that it cuts through the noise.

What’s An Optimal Diet? Glucose Levels Can Help Identify It.

Glucose levels, and how they change over both the long and short term, have a great deal to do with health and well-being. These glucose levels are largely determined by diet, and chronically elevated levels and post-meal spikes can lead to serious dysfunction, including increased risk for type 2 diabetes, cardiovascular disease, stroke, liver cirrhosis, obesity, death from cancer, and more.  It follows then that supporting the healthy regulation of glucose is a fundamental building block of an optimal diet.

In the past, there have been a few tools to help understand how food affects glucose levels; a glycemic index chart could offer rough estimates of predicted glucose impact of foods for the general population, and more personalized information was available from daily finger sticks to spot check glucose levels. CGM has changed all this. This powerful tool can track glucose trends 24-hours a day, not only in response to diet, but for lifestyle behaviors including exercise, sleep, and many others that are known to affect glucose.

A diet that optimizes glucose levels should have three main goals:

  1. Minimizing spikes in glucose levels after meals
  2. Maintaining glucose levels in a fairly narrow and healthy range
  3. Keeping fasting glucose (glucose levels measured after consuming no calories for at least 8 hours) in a range that carries the lowest risk

Minimizing Glucose Spikes After A Meal

The term postprandial hyperglycemia refers to larger than normal elevations, or spikes, in glucose levels after eating a meal. Excessive spikes are dangerous for many reasons; among other issues, they’re a risk factor for development of type 2 diabetes, cardiovascular disease, thickened carotid artery walls, liver disease, obesity, stroke, retinopathy, renal failure, cognitive dysfunction, cancer, and mortality.

The mechanisms that link glucose spikes with chronic disease are thought to include oxidative stress and inflammation, both of which can contribute to insulin resistance.  Insulin resistance refers to the state when the body’s cells become less responsive — or “numb” — to signals from insulin, a hormone that allows cells to take up glucose. Insulin resistance can be the first step toward full-fledged diabetes.

It’s thought to proceed like this: Insulin resistance leads to loss of post-meal blood glucose control, which is followed by the development of elevated fasting glucose levels, which leads to high glucose levels sustained over time. As a point of reference from the International Diabetes Federation, healthy people should rarely ever exceed glucose levels of 140 mg/dL after a meal, and glucose should revert to pre-meal levels within two to three hours.

However, studies of non-diabetic populations wearing CGMs suggest that we may benefit from keeping an even tighter range after meals. One study showed that young healthy adults spent about 80% of the time at glucose levels <100 mg/dL and less than 1% of the time >140 mg/dL. Another study showed that healthy individuals spend 94.4% of the time at glucose levels <120 mg/dL, and again, less than 1% of the time >140 mg/dL. As such, shooting to just stay below a level of 140 mg/dL after meals is likely too lenient of a benchmark for optimal health; more likely, a healthy individual should be maintaining a glucose level of less than 100 mg/dL for the vast majority of the day, and rarely ever spend time at glucose levels above 120 mg/dL.

The implications are clear. An optimal diet, when seen through the lens of glucose levels, should focus on foods that minimize post-meal spikes. Diets that reduce these spikes can potentially reduce the risk of developing diabetes, heart disease, and some cancers.  Additionally, these diets can improve insulin sensitivity and cholesterol profiles.

The concept of glycemic index was developed with relatively simple parameters; it describes the rise in glucose levels observed after the intake of 50g of carbohydrates of a specific food. Under this paradigm, a food with a higher glycemic index raises glucose levels more than a food with a lower index.  But this measure has limited application in the real world because very few people eat 50g of a particular carbohydrate at a given time. Despite this, some studies have found that a diet’s glycemic index is a good predictor of blood sugar fluctuations and that low glycemic index diets can reduce post-meal glucose elevations (Figure 1).

Figure 1: Low glycemic diet for 7 days reduces mean area under the curve for glucose. Byrnes, et al.

While these studies suggest that understanding glycemic index can help, more recent research has shown that two individuals’ glucose response can vary significantly, even after eating the exact same food (Figure 2). This introduces some uncertainty into the idea of standardized scales to predict glycemic response to foods.

Figure 2: Different levels of glucose elevation in different individuals in response to the same foods. Zeevi, et al.

Traditional metrics have shown trends in favor of low-glycemic diets in the general population, but on an individual level, they may not be an accurate predictor of glycemic response. Rather than the notion that a food may have an intrinsic quality that affects glucose levels, it’s more likely that personal genetics, lifestyle, physical activity, body type, and microbiome also play an important role in reducing glucose spikes after a meal.‍

Limit The Variability In Glucose Levels

Glycemic variability refers to large swings in glucose levels. Also known as glycemic excursions, these are potentially more harmful than sustained high glucose levels alone. It’s thought that excessive peaks and dips in glucose can lead to tissue-damaging metabolic byproducts such as free radicalsdamage to blood vessels, damage to the nervous system, triggering of inflammation, and activation of the stress hormone cascade (sympathetic nervous system activation).

Glycemic variability increases as people move along the continuum from normal glucose regulation toward diabetes. As a person becomes more insulin resistant, they’ll also tend to show more variability in glucose levels, including in their fasting and post-meal levels. Simply put, poor glycemic regulation isn’t just about having higher glucose levels, it’s also about having increased variability in those levels.

It’s been shown that the average height of glucose spikes and dips–known as “mean amplitude of glycemic excursions” or MAGE–is lowest in those without diabetes or obesity. A normal range of average glycemic excursions for non-diabetic, non-obese individuals has been shown to be between 26-28 mg/dL. In contrast, a non-diabetic morbidly obese individual displays a MAGE value of 48.6 mg/dLMAGE increases in those who are prediabetic or obese, is higher in those with stable diabetes, and is highest in those with uncontrolled diabetes. Clearly, for best health, we want to eat in a way that produces minimal glycemic excursions. Unfortunately, a hemoglobin a1c test (a standard test which measures approximate glucose levels averaged over 3 months) doesn’t take these spikes and dips into account, so it may miss this important independent risk factor for diabetes and many chronic diseases (see Figure 3).

Figure 3: People with the same mean blood glucose concentration can have very different glycemic variability; Knowing the glycemic variability is an independent risk factor for health problems,  such measures of averages are limited in utility. Suh, et al.

Fasting Glucose And “Normal” Ranges

“Fasting glucose” is a measurement of glucose levels after consuming no calories for at least 8 hours. A high number is a strong predictor of developing diabetes down the road, and higher values may even be a risk factor for heart disease and stroke, among other health problems, even when this level is in the non-diabetic “normal” range (Figure 4).

The American Diabetes Association places fasting plasma glucose levels into three categories: “normal” (<100 mg/dL), “prediabetes” (100-125 mg/dL), and “diabetes” ( >126 mg/dL) (Figure 5). It’s thought that up to 70% of individuals at the prediabetes level will eventually develop diabetes, and approximately 90% of those at prediabetic levels aren’t aware of their condition!

Additional research indicates that even people in the high “normal” range are at an increased risk of developing diabetes. Research in children has shown that having a fasting glucose measurement of 86-99 mg/dL, while still in the “normal” range, increases the risk of developing prediabetes and type 2 diabetes more than two-fold when compared with children with a fasting glucose less than 86 mg/dL. A large study in adults showed that as fasting glucose increases from <81 up to 99 mg/dL, there is an increase in the risk of developing diabetes by as much as three times, despite all of these individuals meeting criteria for “normal” fasting glucose (Figure 6). What’s more, there is a sharp increase in the risk of cardiovascular disease, heart attacks, and thrombotic stroke as fasting glucose increases (Figure 4). This increase in risk begins at ~90 mg/dL fasting glucose, well below the “normal” fasting glucose level cut-off of 100 mg/dL.

The idea of a fasting glucose of <100 mg/dL being a standard threshold for complete safety and risk-avoidance is flawed. Physiology is a spectrum, and it is up to every individual to stay vigilant and aware of where they fall on this spectrum. Unfortunately, traditional tools used for assessing glucose don’t give us this level of granularity. Fortunately, tools like CGM are now accessible, software like Levels can make sense of the data, and glucose levels are highly modifiable with dietary and lifestyle modifications.

Figure 4: Risk (as indicated by increased hazard ratio) of cardiac outcomes and thrombotic stroke are lowest at fasting glucose levels of ∼90 mg/dL, and increased sharply above these levels. Park, et al.

Figure 5: Fasting glucose levels diagnostic criteria from American Diabetes Association.

Figure 6: Even in normal range, as fasting glucose goes from <81 up to 99,  diabetes risk increases from 1x risk to 3x risk. Tirosh, et al.

In addition to sustained, elevated fasting glucose levels, variability in these levels between fasting glucose readings may represent a risk factor. Fasting glucose levels that “bounce around” from test to test appear to be correlated with increased risk of heart disease and mortality. Research shows that those with the least variability between tests have a lower risk of future health problems than those with the most variability (Figure 7).

Figure 7: Increased mortality and cardiovascular disease in general population stratified by variability of fasting glucose readings (quartile 1 having lowest variability, quartile 4 having highest variability). Wang, et al.

Considering these findings, it’s clear that an optimal diet should take into account the diet’s effect on keeping fasting glucose on the lower end of the normal range, as well as minimizing variability between test-to-test fasting levels. The good news is that CGM can be a powerful tool to monitor glucose and determine which foods lead to better or worse glycemic function. Consistent use of these devices can support lifestyle changes that lower fasting blood glucose. And these lifestyle changes can work dramatically–even for those who meet the criteria for prediabetes, lifestyle modification can reduce the risk of progressing to diabetes by up to 70%.

Optimal Diet: There’s No “One Size Fits All”

The data makes it clear: there’s no one-size-fits-all diet that will optimize glucose levels. Neither a food’s carbohydrate content nor its glycemic load/index can predict an individual’s exact glucose response to a real-life meal. Standardized scoring systems like the glycemic index simply aren’t personalized enough to be as useful as we’d like (Figures 2). What’s more, metrics like carbohydrate content or glycemic load/index fail to take into account genetics, weight, sleep quality, stress levels, gut microbiome, insulin sensitivity, mixed meals and food combinations, and other factors which all affect glycemic response.

So what’s the best way to determine an optimal diet? It’s simple, but it requires data and analytics. We know that an optimal diet is one that minimizes post-meal glucose spikes, reduces glycemic variability, and maintains fasting blood glucose in an optimal range. CGM can aid in achieving these aims.

Using CGM, coupled with Levels software to interpret the data and parse out the individual dietary and lifestyle drivers of glycemic trends, individuals can track post-meal spikes, glycemic variability, and fasting glucose in a way that’s accurate, individualized, applicable to real-life, and actionable. Reading nutrition labels simply isn’t enough; Levels can reveal an individual’s glycemic reaction to any food or meal, facilitating enhanced control and power in pursuit of optimal health and well-being.  Now is the time to utilize this technology to craft personalized diets that optimize biologic wellness and take the guesswork out of this aspect of eating.