Published: October 14, 2024
Edited on October 15, 2024.
In this blog post, we are looking at how different forms of time-restricted eating, or TRE, affect blood glucose homeostasis, that is, our blood sugar levels as well as related factors, including glucose tolerance, insulin sensitivity, and insulin secretion by the pancreatic beta-cells.
We will be looking at the effects of TRE that are mediated by weight loss, and separately at weight loss-independent effects.
We will consider the effects on healthy people and those with insulin resistance or glucose intolerance, such as prediabetes or diabetes.
And we will consider whether the impact of TRE on blood sugar regulation depends on what form of TRE we use, whether it’s early or late TRE, and how long the actual eating window is.
And for each of these questions, we’ll look at the entire cumulative evidence, consisting of several dozen randomized controlled trials.
How to Measure the Impact of An Intervention on the Regulation of Blood Sugar
The randomized controlled trials we will discuss that have studied the impact of TRE on blood glucose regulation have used quite a few different variables we can look at. Let’s first review these and make sure we understand clearly what they measure. For more detail, I recommend you read my prior blog post about The Regulation of Blood Sugar.
When it comes to blood sugar, one of the key measures we are interested in is our glycemic control, that is, the degree to which our blood glucose levels are within the normal range. The most common measures of glycemic control are the fasting glucose concentration and the glycated hemoglobin, or HbA1c, test. The HbA1c test provides information about our average blood glucose concentration roughly over the last three months. Another way to assess whether our glucose levels are in the normal range is by wearing a continuous glucose monitor (CGM). The interpretation of CGM data is not as standardized as that of fasting glucose and HbA1c, but CGM data can give us a pretty good idea of whether our body is able to keep our blood glucose levels within the normal, homeostatically regulated range. I discussed this in some detail in a recent blog post.
Another measure we are interested in is our glucose tolerance. This is most commonly measured with an oral glucose tolerance test (OGTT) where blood glucose levels are regularly measured over two or three hours after consuming a beverage containing 75g of pure glucose. Clinically, we mostly look just at the time point exactly two hours after starting to drink the glucose-sweetened beverage, but in research, we often collect blood samples at multiple time points over two or even three hours and measure, for example, the area-under-the-glucose curve as a measure of glucose tolerance, which is this entire area under the glucose curve over 2 or 3 hours in the figure below.
Whether we are glucose tolerant is largely determined by our insulin sensitivity and the amount of insulin our pancreatic beta-cells can produce. We often call this beta-cell function.
For insulin sensitivity, we have various measures. HOMA-IR is a simple one that is based on fasting glucose and insulin concentrations, and that provides an estimate of insulin resistance, which is the opposite of insulin sensitivity. So, if someone is insulin resistant (high HOMA-IR), it means they have low insulin sensitivity. In research, we also often look at the total insulin response in an OGTT by measuring insulin levels repeatedly over two or three hours after drinking 75g of glucose in water. What can the insulin concentrations during an OGTT tell us about insulin sensitivity? I have explained this previously using Ben and Jack (see figure below). Let’s assume both complete an OGTT, and both are glucose tolerant, with identical blood glucose curves. Now, while their glucose tolerance is identical, their insulin concentrations during the OGTT look very different. Ben has higher baseline insulin than Jack, and his insulin levels take off like a rocket after he drinks the OGTT beverage sweetened with glucose. Jack’s insulin levels, in contrast, remain pretty low throughout the test. This means that Jack is very insulin-sensitive because he doesn’t need a lot of insulin to keep his blood glucose concentrations in the normal range. Ben is very insulin resistant; he needs much higher insulin levels throughout the test to keep his glucose levels in the normal range. And fortunately, Ben’s pancreatic beta-cells are able to produce enough insulin to fully compensate for his high level of insulin resistance. So, by looking at both glucose and insulin, we can get a good overall picture of glucose tolerance, insulin sensitivity, and even the ability of the pancreatic beta-cells to produce insulin.
To measure insulin secretion or beta-cell function there is also a simpler measure based just on fasting glucose and insulin called HOMA-B. From the OGTT or similar test, we can calculate several measures of beta-cell function. They generally consider the increase in blood insulin relative to the increase in blood glucose. Importantly, measures of insulin secretion need to be interpreted relative to a measure of insulin sensitivity. That is because if we looked at Jack’s insulin curve in isolation, without also considering his level of insulin sensitivity or at least his blood glucose levels, we would not be able to tell whether he CANNOT make more insulin, or whether he just doesn’t HAVE to make more insulin because he is very insulin sensitive.
For several of these measures, we have ranges that define normal. For example, for fasting glucose, normal means 70-100 mg/dL or 3.9-5.6 mmol/L. For HbA1c, the normal range is between 4% and 5.6%. And for the 2-hour time point in an OGTT, we want to be below 140 mg/dL or 7.8 mmol/L.
For others, we don’t always have such clearly defined ranges, but one thing is very important to be clear about: lower is not always better. As long as our values are in the normal range, we may not be able to lower them further, or it may even be harmful to lower them further, because we don’t want our blood sugar as low as possible. Because then we’d drop dead. We want our blood sugar levels to be appropriately regulated within the normal ranges.
The Impact of TRE on Blood Sugar Regulation in Healthy People
Let’s start by looking at a few RCTs in which healthy, mostly insulin-sensitive people with normal or close to normal blood sugar levels were randomized to either TRE or a non-TRE control. There are seven such studies. Let’s start with the two studies by Moro et al. (2016) and Brady et al. (2021) shown in the table below. Both of these were done in athletes, and in both trials, the interventions lasted 8 weeks. Participants who followed TRE lost some weight, about two to four pounds more than participants in the control groups. And what do we think happened to their fasting glucose and fasting insulin levels or HOMA-IR, the measure of insulin resistance? Well, nothing. None of these measures changed, so TRE, even with modest weight loss, did not affect these measures.
Now, does that mean that TRE doesn’t benefit glycemic control or insulin sensitivity?
No, of course, it doesn’t. It probably just means that if someone is an athlete and already perfectly glucose tolerant and insulin sensitive at baseline, he or she is not likely to become even more perfect.
The table above shows two more studies in people who were overweight or obese, but who still had normal glucose levels and normal or close to normal insulin sensitivity, and in those individuals, again, TRE didn’t have an impact.
Again, in my interpretation, the lack of response here is mainly because we are studying participants with mostly healthy blood glucose regulation. Also consider that these studies were fairly small, so the minor changes that could be triggered by TRE may not be detectable. What we need to consider is that fasting glucose and insulin levels, for example, are fairly variable within the normal range in healthy people. Now, add an intervention, and that variability will increase. What I am arguing is that changes in healthy people are unlikely to be large enough to be consistently detectable on this natural background variability.
However, there is another potential explanation for why these studies were all null, meaning why they didn’t detect an effect of TRE: three of these studies used late TRE, meaning the eating window was later in the day. Chow and colleagues didn’t specify the exact time of the eating window, and participants could choose whichever window was most convenient for them. And what participants chose was mostly late TRE. In other words, the form of TRE studied in all four of these trials shown in the table above was late TRE.
This is in contrast to these two studies shown in the table below by Zhang et al. (2022) and Xie et al. (2022) that directly compared early TRE to late TRE to a non-TRE control group in healthy men and women. In both cases, improvements in insulin sensitivity and glycemic control were seen with early TRE, and not with late TRE. Note also that these studies were substantially larger, leading to greater statistical power to detect small differential changes.
This study by Moro and colleagues was a bit of an outlier in that they used late TRE, and also saw reductions in fasting glucose and insulin resistance, as measured by HOMA-IR. Notably, however, this was a very long-term study of 12 months that led to a major difference in body weight of more than 5 kg, or 12 pounds. And in this context, the substantial reduction in body weight and fat mass was probably the primary driver of the reductions in glucose levels and insulin resistance.
Our first conclusion is, therefore, that in healthy people, late TRE does not affect glycemic control or insulin sensitivity in a significant way, unless it leads to major weight loss over time, which certainly is a possibility. This is in contrast to early TRE, where the eating window is early in the da, and no food intake usually occurs after 4 or 5 PM. With early TRE, we see modest improvements in insulin sensitivity and glycemic control, even in healthy people who start out with close to normal blood sugar regulation. And we even see these improvements in shorter studies and with fairly modest weight loss.
So that’s pretty encouraging.
The Impact of TRE on Blood Sugar Regulation in People with Impaired Glucose Tolerance and/or Insulin Resistance
If you have read some of my other blog posts on this topic, you know that one of the most common causes of elevated blood sugar levels is excess body fat mass. The way we think about this is that if we carry excess body fat, we often also store excess fat in visceral fat depots surrounding the inner organs, and also in the form of ectopic fat, that is, in tissues not designed for fat storage, such as the liver, the pancreas, and muscle.
Carrying excess visceral fat and having excess liver and muscle fat is one of the most common causes of insulin resistance, along with a low-grade chronic inflammation that tends to be associated with excess body fat (see figure below). On the other hand, having fat in the pancreas contributes to a reduced ability of the pancreas to secrete insulin. In other words, pancreatic fat deposition is a common cause of beta-cell dysfunction. Now, insulin resistance and a reduced ability of the pancreas to secrete insulin causes us to become glucose intolerant, which then leads to elevated blood glucose levels and, eventually, diabetes.
This means that often, one of the best ways to improve glucose tolerance and lower those elevated blood glucose levels is to lose some of that excess body fat. In fact, we see this in any study in which glucose-intolerant people, such as those with type 2 diabetes or prediabetes, lose weight. Whether they lose weight due to bariatric surgery, medication, or calorie restriction, their insulin sensitivity tends to improve. If there is enough weight loss, we often also see improvements in beta-cell function. As a result, glucose tolerance often improves, and blood glucose levels drop. So, we understand that if we are glucose intolerant due to excess body fat mass and lose weight, the condition improves.
The question is, does this also work with weight loss triggered by TRE?
As we discussed in the last blog post, restricting the eating window in TRE leads to a spontaneous reduction in calorie intake, body weight, and fat mass, unless we experimentally try to prevent that weight loss. In other words, TRE is one way through which people can reduce their excess body weight and fat mass. And as we also saw in the last blog post, any such reduction in body weight and fat mass is associated with a reduction in the types of body fat that are strongly linked to insulin resistance, visceral fat, and ectopic fat.
So what we’ll do first is review studies in which 4-10-hour TRE caused participants to lose weight to see whether this led to the expected improvements in insulin sensitivity, possibly insulin secretion, glucose tolerance, and glycemic control.
After that, we will look at studies investigating weight loss-independent effects of 4-10-hour TRE on blood sugar regulation.
We’ll focus on 4-10-hour TRE first, because more extreme forms of TRE with eating windows shorter than 4 hours, such as one-meal-a-day or OMAD, seem to have very different effects on glucose tolerance. We will cover this specific case of OMAD later in the blog post.
Weight Loss-Dependent Effects of TRE on Blood Sugar Regulation
Let’s start with the weight loss-dependent effects of 4-10-hour TRE. We’ll do this by first looking at trials in which participants adopted TRE and where their weight and fat mass spontaneously dropped.
The table below shows four randomized controlled trials in people with type 2 diabetes or insulin resistance at baseline. Participants in all studies were overweight or obese, which suggests that many likely had excess visceral and ectopic fat. In all studies, they were randomized to TRE or a control group that ate in the normal eating window. Two of these used early TRE, the other two late TRE.
And even though these studies differed a lot by type of TRE, using early or late TRE with a very short or rather long eating window, participants lost more weight in the TRE group in all studies. And that more significant weight loss was associated with the expected changes in measures of insulin sensitivity, insulin secretion, and glycemic control.
Let’s take a closer look at the study by Che and colleagues, for example. Over 12 weeks, adults with type 2 diabetes on TRE lost an average of 2.2 kg or almost 5 pounds more than those in the control group. And that combination of following TRE and losing a modest amount of weight reduced both HbA1c and fasting glucose quite substantially, and also improved both their insulin sensitivity and insulin secretion, that’s HOMA-IR and HOMA-B in the table, while also reducing the number of diabetes medications they required. That is quite a success in just 12 weeks.
In other studies, such as the one here by Cienfuegos and colleagues, only those parameters were improved by TRE that were abnormal at baseline: people were insulin resistant at baseline, and fasting insulin and HOMA-IR were both reduced by following either 6-hour or 4-hour TRE, which caused them to lose on average 3.2% of their baseline weight. However, fasting glucose and HbA1c, mostly normal at baseline, did not change significantly. This goes back to what we discussed earlier: if a measure is in the normal range, we should not expect improvements. It may not be possible to improve further in this regard, or only very small improvements may be possible that can be difficult to detect in a trial.
One additional study was not considered in the YouTube video, because it was published on October 1, 2024, after our literature search had been completed. This study describes a three-month RCT by Manoogian and colleagues who compared the impact of 10-hour TRE combined with nutritional counseling vs. nutritional counseling only (control) on blood glucose regulation in participants with the metabolic syndrome. They found statistically significantly greater improvements in HbA1c, but the effect size of that difference was only 0.1 percentage points and therewith considerably smaller than in the other RCTs discussed above. They also did not detect any significant differences in fasting glucose, fasting insulin, or HOMA-IR, even though participants were fairly insulin resistant at baseline with a HOMA-IR around 4.5. And this lack of a response was particularly staggering given that participants lost significantly more weight and fat mass in the TRE intervention compared to the control intervention. The results from this RCT are an outlier in that they suggest that TRE plus the weight loss it tends to trigger do not meaningfully benefit blood sugar regulation even in insulin resistant people. One potential explanation may be that participants used 10-hour TRE relatively late in the day, whereas other studies shown in the table above either used early TRE or late TRE with a shorter eating window.
So, looking at our chart above, we can conclude that TRE combined with modest weight loss improves glycemic control in people with high blood glucose levels, and it also improves insulin sensitivity in people who are insulin-resistant at baseline. We also have some modest evidence that following TRE and losing weight may improve insulin secretion.
What we don’t have good data on is whether TRE, with some modest weight loss, also improves glucose tolerance. Only in one of the four trials did researchers conduct an OGTT, and they only measured the 2-hour glucose value, not the entire glucose response throughout the test. In that study, no differential effect on the 2-hour glucose value was detected. That may have been an issue of statistical power, though, because the 2-hour glucose value was clearly lower after the TRE phase in that study, and by a margin that I would consider clinically highly relevant. However, the variation in the 2-hour glucose concentration was very large in both groups, and the differential change was too small relative to the large variability in this measure. With a 50% larger sample size, this would have likely been statistically significant, however. Because we don’t have any other data on the impact of TRE plus weight loss on glucose tolerance, I’d say we don’t know yet whether and to which degree glucose tolerance is improved. That’s why I put a cautious little plus in the chart above, with a large question mark to indicate that there MAY be an improvement in glucose tolerance with TRE and weight loss.
Weight Loss-Independent Effects of TRE on Blood Sugar Regulation
So, TRE plus the modest weight loss it tends to trigger have the anticipated benefits for blood glucose regulation that are also seen in response to weight loss resulting from other interventions. An open question is whether TRE also has weight loss-independent effects here?
Several studies have investigated the impact of TRE on glucose tolerance, insulin sensitivity, or insulin secretion while keeping calorie intake and body weight artificially stable. Let’s take a look at these.
One trial that is often cited because it showed weight-loss independent improvements in glucose tolerance and insulin sensitivity is the one by Jones and colleagues shown in the table above. This is indeed a nice 2-week RCT comparing 8-hour early TRE vs. a calorie intake and body weight-matched control group eating in their non-restricted habitual eating window. The study showed indeed statistically significant differences in the changes in the AUC glucose and AUC insulin; however, in my interpretation, these differences were driven by differences between the TRE and control groups at baseline (i.e., randomization was not successful for these two variables in this fairly small study, which can happen), and the supposed effect of TRE may be what we commonly refer to as regression to the mean. I therefore considered the results from this trial, but by itself, I would not give it a lot of weight because of this limitation.
For example, the trial shown in the table above by Jamshed and colleagues was a small cross-over design RCT in which people who were insulin-resistant, but glucose-tolerant, were studied. They were given all of their food, twice for four consecutive days, separated by a wash-out period. In one study period, they ate in their normal eating window, and in the other study period, they ate the exact same food, including the same number of calories, in an early time-restricted eating window of 6 hours. At the end of just four days on 6-hour-TRE, their fasting glucose and insulin levels were measurably reduced, and HOMA-IR was reduced, meaning their insulin sensitivity had improved. The participants also wore CGMs, and investigators observed lower average glucose levels and a substantially reduced amplitude of glycemic excursions, which means that whenever glucose levels rose, they rose much less substantially. Given that participants ate standardized meals during these four days, that finding would suggest improved glucose tolerance.
A similar study design was used by Sutton and colleagues. They enrolled participants with prediabetes, meaning they were insulin-resistant and moderately glucose-intolerant. They again conducted a cross-over design trial in which each participant completed each of two 5-week intervention periods. They received all of their meals in both intervention phases. In one phase, they consumed these meals in 6 hours early in the day, meaning they followed 6-hour early TRE. In the other, they consumed the meals spread out over 12-13 hours (see study design below). Because in this trial, the same participants consumed identical meals with identical calories, this is a true investigation of TRE only, because body weights and fat mass didn’t change differentially between these two intervention periods.
At the end of five weeks, these investigators conducted a 3-hour OGTT. They found that glucose tolerance did not differ between the early TRE phase and the control phase. See in the figure below how the blood glucose response was almost exactly the same at the end of the two intervention periods. However, participants required a lot less insulin at the end of the TRE phase to regulate their blood glucose levels, and so their insulin concentrations were a lot lower throughout the test after the TRE phase. This means that they had become more insulin-sensitive. These authors also report that beta-cell function was improved after early TRE. What we can confidently conclude from this very nice trial is that TRE improves both insulin sensitivity and beta-cell function independent of any changes in body weight and fat mass. However, these improvements in insulin sensitivity and insulin secretion did not translate to better glucose tolerance, and that is a bit of an unusual and unexpected finding. Usually, improvements in insulin sensitivity of this magnitude alone would be expected to improve glucose tolerance in people with prediabetes.
Now, to make this even more confusing, we have another study by Martens and colleagues in which six weeks of 8-hour TRE did improve glucose tolerance. They conducted a 2-hour OGTT, and found that the area-under-the-glucose curve was significantly lowered by TRE compared to the normal feeding control (see figure below). Again, this is independent of weight change.
What makes this confusing is that glucose tolerance changed in this study, but not the main determinants of glucose tolerance, insulin sensitivity and insulin secretion. Now, the big difference was that participants in the study by Sutton and colleagues were quite insulin resistant and glucose intolerant, whereas participants in this study here by Martens and colleagues had fairly normal glucose tolerance and insulin sensitivity. Take a look at the differences in the concentrations of glucose and particularly insulin in the figure below.
It would certainly be nice if the results were intrinsically a bit more consistent, but what we can say is that 4-10-hour TRE doesn’t have negative impacts on glucose homeostasis in any of these studies, but instead improved different aspects of blood sugar regulation in a weight change-independent way.
Three other studies have compared TRE vs. a control group that kept body weight stable experimentally, and by and large, they confirm these findings. So we can conclude that independent of changes in body weight and fat mass, TRE improves insulin sensitivity in people who are insulin resistant, and also glycemic control in people with elevated glucose levels. There may also be weight-loss independent improvements in beta-cell function. Surprisingly, glucose tolerance does not seem to be consistently improved. We have limited data on this, with two studies showing improved glucose tolerance while one study shows no difference in glucose tolerance, so there is some uncertainty. Our best estimate at this time is that there MAY BE an improvement in glucose tolerance.
Through which weight loss-independent mechanism may TRE improve insulin sensitivity, insulin secretion, and glycemic control? The truth is we don’t know. I would have suspected that the quality of the diet improves when people follow TRE, because now they can no longer have snacks such as chips, cookies, ice cream, or alcoholic beverages after dinner. Now, diet quality may improve on TRE, and that could very well be a beneficial side effect, but that doesn’t seem to be the main mechanism here, because benefits are also seen in studies in which diet quality was standardized.
Other candidate mechanisms could be that adopting TRE addresses issues of circadian disruption, because with TRE we are aligning our food intake more with the natural light-dark cycle. That may have its benefits in its own right, because circadian disruption is known to reduce insulin sensitivity and glucose tolerance. But it could also contribute to better sleep, and because sleep deprivation can cause insulin resistance, this is certainly a potential weight loss-independent mechanism. There are also some data suggesting that TRE may affect the gut microbiome and reduce measures of low-grade chronic inflammation in the body, and some of that may be independent of changes in body weight and fat mass. Inflammation is one of those very strong and acute triggers of insulin resistance, and a dysfunctional gut microbiota has also been implicated in insulin resistance, so I’d say it’s plausible that these could be weight loss-independent mechanisms through which TRE improves blood sugar regulation.
Other Evidence
Now, there are five more randomized controlled trials that have compared the impact of TRE vs. a control group on blood sugar regulation (and these were not mentioned in the YouTube video, by the way). These were not super consistent, but mostly they confirm what we have been discussing so far.
Mostly, these studies compared TRE to a calorie-restricted control group, where both groups lost similar amounts of weight (see table below). In these studies, we usually see similar improvements in measures such as fasting glucose, HOMA-IR, and HbA1c. I would suggest that we don’t see differences in these measures in these particular studies between the TRE and control groups because (a) participants again were mostly reasonably healthy, with either normal blood sugar regulation or only slightly elevated fasting glucose or insulin resistance, and as we discussed, it’s a lot harder to detect intervention effects in such people; and (b) the fairly significant weight loss in both groups introduces a lot of variability in response, and that variability again makes it more difficult to detect intervention effects. It is also fair to say that most of these trials were not designed to detect effects of TRE on glucose homeostasis, and therefore their ability to detect differential effects may have been limited for the reasons outlined above.
Scientific Studies Not Considered
Many more studies have been published that reported on the effects of TRE on one or several measures of blood sugar regulation, but that were not considered because they did not meet standards for scientific rigor. Of note, our approach here was to review the study design and reporting first, and to not let the scientific findings of the study impact whether the study was considered further.
Several studies were excluded from consideration because they did not include an adequate control group. These included the studies by Gabel et al. (2018), Lowe et al. (2020), Wilkinson et al. (2020), and He et al. (2022). An adequate control group is important because of what is often called hidden time effects that may occur in study participants that have little or nothing to do with the nature of the intervention. Specifically, participants enrolled in nutrition studies often want to lose body weight, and a common observation is that they lose body weight even if they are randomized to a control group. It is, therefore, imperative that changes in an intervention group are compared to an adequate control group. Lowe et al. (2020) did include a control group. However, the study design made it possible for control participants to also follow TRE and still be compliant, and we, therefore, made the decision not to consider this trial. Similarly, He et al. (2022) compared TRE vs. a low-carbohydrate diet vs. TRE plus a low-carbohydrate diet. Unfortunately, without a control group that did not receive any intervention, the results from this trial are difficult to interpret. What we can say is that 6h-TRE, when added to a low-carbohydrate diet, leads to a greater loss of body weight and visceral fat than a low-carbohydrate diet alone, and that greater weight loss was associated with greater reductions in insulin resistance (HOMA-IR).
The study by Schroder et al. (2021) was not considered because participants were not randomized, but volunteered for the TRE intervention.
The study by Phillips et al. (2021) was not considered because it tested a version of TRE that had a very long eating window of 12 hours, which is similar to the control eating window in many studies and similar to the normal eating window of many people.
Lastly, two studies were excluded because of the insufficient or intrinsically inconsistent reporting of study findings or statistical analyses. These included the studies by Feehan et al. (2023) and Meessen et al. (2022).
A review of the principal findings of these excluded studies did not suggest that their inclusion would have meaningfully changed the overall conclusions from this analysis.
What Is Better for Blood Glucose Regulation: Early or Late TRE?
Now, do we see these improvements in insulin sensitivity and glycemic control more with early TRE where our eating window is early in the day, or with late TRE where maybe we are skipping breakfast? Most of the studies that we have discussed so far tested early TRE, and so we have the most evidence that the early form of TRE is effective. We also have five studies that have directly compared early vs. late TRE. Two of these (Zhang et al. 2022 and Xie et al. 2022) found greater improvements with early TRE, whereas the other three (Hutchison et al. 2019; Deng et al.; 2024; Queiroz et al. 2023) found no difference in measures such as fasting glucose, fasting insulin, HOMA-IR, HbA1c, or the area-under-the-curve for glucose or insulin in an OGTT.
I’d say that if you are insulin resistant or have high blood glucose levels and you’d like to try TRE, maybe try early TRE first if you don’t have a preference, given that we have more evidence that early TRE works and some modest evidence that early TRE may be better. At the same time, if early TRE is not feasible for you, then I see good evidence that late TRE could also work for you. The data are not fully conclusive in this regard, but my best estimate is that late TRE may have the same weight loss-dependent benefits for blood sugar regulation as early TRE, but that early TRE may have more pronounced weight loss-independent benefits.
One-Meal-a-Day (OMAD): A Special Case
Finally, let’s talk briefly about an exception to all of these findings, and that is related to people who adopt one-meal-a-day, or OMAD, the most extreme version of TRE. We do have solid evidence that following OMAD for 8 weeks leads to a reduction in insulin secretion, specifically in the first-phase insulin response, and that reduction in insulin secretion reduces glucose tolerance, and quite substantially so. This increase in the blood glucose response shown in the figure below is a whooping effect.
As a reminder, after a meal, insulin is secreted in two phases. In the first phase, pre-formed insulin stored within the pancreatic beta-cells is secreted into the blood. That happens within a few seconds of eating carbs, and this first-phase insulin helps prevent an early blood glucose spike. In the second phase, which starts several minutes after the meal has been initiated, the pancreatic beta-cells make insulin from scratch, from DNA to RNA to protein, so that’s why this takes longer.
In a way, think about insulin production as a factory with an attached warehouse. The first-phase insulin response consists of the insulin that is sitting in the warehouse, ready to be dispatched at any moment. The second-phase insulin response is where the factory waits for the order to come in and then starts the machines to begin production.
Because in the first-phase insulin response, we rely on pre-formed insulin, any factor that reduces the amount of pre-formed insulin that is present in the beta-cells will also lower the first-phase insulin response. That includes any condition where we are not eating carbs for a while, including low-carb meals or longer fasting, as apparently, in OMAD.
It is impossible to know from the available data how long the eating window has to be to not trigger such a reduction in the first-phase insulin response. We see it in trials of OMAD, but not in TRE-studies with eating windows of 4 hours or longer, so my best estimate is that a fasting window of at least 20 hours is required to trigger a meaningful reduction in the first-phase insulin response and glucose tolerance.
The data suggest that if you want to do OMAD or any TRE with an eating window shorter than about 4 hours, it may be a good idea to make the diet lower in carbs while you are doing one of these fairly extreme forms of TRE. That is because a lower first-phase insulin response is much less relevant if your meals have a low glycemic load, that is, fewer carbs and, specifically, fewer highly glycemic carbs. If you don’t want to follow a low-carb diet, then my suggestion would be to do TRE with an eating window more in the 6-10-hour range, as the reduction in insulin secretion and glucose tolerance was NOT seen with these longer eating windows.
Summary & Conclusions
Here is what the cumulative evidence tells us:
First, if you have normal or close to normal blood glucose levels and insulin sensitivity, adopting late TRE will probably not lead to any improvements in your blood glucose regulation unless you lose a lot of weight, whereas early TRE may lead to modest improvements even with minimal weight loss.
Second, adopting an eating window between 4 and 10 hours per day improves insulin sensitivity in people with insulin resistance, and glycemic control in people with elevated glucose levels. To some degree, these benefits are partly due to the loss of body weight and fat mass triggered by TRE, but to some degree, they seem to be independent of changes in body weight and fat mass. TRE also seems to improve insulin secretion and MAY improve glucose tolerance, but we have less certainty supporting an effect of TRE on these endpoints.
Third, the totality of the evidence suggests that early TRE benefits blood sugar regulation more than late TRE, even though late TRE also seems to have some benefits.
Fourth, forms of TRE with a very short eating window, such as one-meal-a-day or OMAD, reduce the first-phase insulin response and glucose tolerance. This would make it advisable to avoid large meals with a very high glycemic load when following TRE with an eating window of less than 4 hours.
I would encourage you to find a version of TRE that works for you and that you can happily live with in the long term. Because those habits that we can stick with have the potential to really exert their positive impact on our health. And in this case here with TRE, I feel that it’s an excellent example of not letting perfect be the enemy of the good.
In other words, early TRE may be better than late TRE, but late TRE is still better than eating all day and late into the night.
Take care.
References
- Moro et al.; Effects of eight weeks of time-restricted feeding (16/8) on basal metabolism, maximal strength, body composition, inflammation, and cardiovascular risk factors in resistance-trained males. Journal of Translational Medicine 2016; 14: 290.
- Brady et al.; Effects of 8 wk of 16:8 time-restricted eating in male middle- and long-distance runners. Official Journal of the American College of Sports Medicine 2021; 53: 633-42.
- Chow et al.; Time-restricted eating effects on body composition and metabolic measures in humans who are overweight: a feasibility study. Obesity 2020; 28: 860-9.
- Bantle et al.; Time-restricted eating did not alter insulin sensitivity or beta-cell function in adults with obesity: a randomized controlled study. Obesity 2023; 31: 108-115.
- Kotarsky et al.; Time-restricted eating and concurrent exercise training reduces fat mass and increases lean mass in overweight and obese adults. Physiology Reports 2021; 9: e14868.
- Zhang et al.; Randomized controlled trial for time-restricted eating in overweight and obese young adults. iScience 2022; 25: 104870.
- Xie et al.; Randomized controlled trial for time-restricted eating in healthy volunteers without obesity. Nature Communications 2022; 13: 1003.
- Moro et al.; Twelve months of time-restricted eating and resistance training improves inflammatory markers and cardiometabolic risk factors. Medicine & Science in Sports and Exercise 2021; 53: 2577-85.
- Che et al.; Time-restricted feeding improves blood glucose and insulin sensitivity in overweight patients with type 2 diabetes: a randomized controlled trial. Nutrition and Metabolism 2021; 18:88.
- Cienfuegos et al.; Effects of 4- and 6-h time-restricted feeding on weight and cardiometabolic health: a randomized controlled trial in adults with obesity. Cell Metabolism 2020; 32: 366-78.
- Cienfuegos et al.; Changes in body weight and metabolic risk during time-restricted feeding in premenopausal versus postmenopausal women. Experimental Gerontology 2021; 154: 111545.
- Kaercher Kramer et al.; The impact of time-restricted eating on beta-cell function in adults with type 2 diabetes: a randomized controlled cross-over trial. Journal of Clinical Endocrinology & Metabolism 2024; doi: 10.1210/clinem/dgae594.
- Jamshed et al.; Early time-restricted feeding improves 24-hour glucose levels and affects markers of the circadian clock, aging, and autophagy in humans. Nutrients 2019; 11: 1234.
- Sutton et al.; Early time-restricted feeding improves insulin sensitivity, blood pressure, and oxidative stress even without weight loss in men with prediabetes. Cell Metabolism 2018; 27: 1212-21.
- Ravussin et al.; Early time-restricted feeding reduces appetite and increases fat oxidation but does not affect energy expenditure in humans. Obesity 2019; 27: 1244-54.
- Martens et al.; Short-term time-restricted feeding is safe and feasible in non-obese healthy midlife and older adults. GeroScience 2020; 42: 667-86.
- Turner et al.; The effects of time-restricted eating versus habitual diet on inflammatory cytokines and adipokines in the general adult population: a systematic review with meta-analysis. American Journal of Clinical Nutrition 2024; 119: 206-20.
- Maruthur et al.; Effect of isocaloric, time-restricted eating on body weight in adults with obesity. Annals of Internal Medicine 2024; 177: 549-559.
- Domaszerwski et al.; Effect of a six-week intermittent fasting intervention program on the composition of the human body in women over 60 years of age. International Journal of Environmental Research and Public Health 2020; 17: 4138.
- Domaszewski et al.; Comparison of the effects of six-week time-restricted eating on weight loss, body composition, and visceral fat in overweight older men and women. Experimental Gerontology 2023; 174: 112116.
- Stote et al.; A controlled trial of reduced meal frequency without caloric restriction in healthy, normal-weight, middle-aged adults. American Journal of Clinical Nutrition 2007; 85: 981-8.
- Carlson et al.; Impact of reduced meal frequency without caloric restriction on glucose regulation in healthy, normal-weight middle-aged men and women. Metabolism Clinical and Experimental 2007; 56: 1729-34.
- Pavlou et al.; Effect of time-restricted eating on weight loss in adults with type 2 diabetes. A randomized clinical trial. JAMA Network Open 2023; 6: e2339337.
- Lin et al.; Time-restricted eating without calorie counting for weight loss in a racially diverse population: a randomized controlled trial. Annals of Internal Medicine 2023; 176: 885-95.
- Queiroz et al.; Cardiometabolic effects of early v. delayed time-restricted eating plus energetic restriction in adults with overweight and obesity: an exploratory randomized clinical trial. British Journal of Nutrition 2023; 129: 637-49.
- Manoogian et al.; Time-restricted eating in adults with metabolic syndrome. A randomized controlled trial. Annals of Internal Medicine 2024; published online ahead of print on October 1, 2024.
- Peeke et al.; Effects of time restricted eating on body weight and fasting glucose in participants with obesity: results of a randomized, controlled, virtual clinical trial. Nutrition & Diabetes 2021; 11: 6.
- Jamshed et al.; Effectiveness of early time-restricted eating for weight loss, fat loss, and cardiometabolic health in adults with obesity. JAMA Internal Medicine 2022; 182: 953-62.
- Steger et al.; Early time-restricted eating affects weight, metabolic health, mood, and sleep in adherent completers: a secondary analysis. Obesity 2023; 31: 96-107.
- Thomas et al.; Early time restricted eating compared to daily caloric restriction: a randomized trial in adults with obesity. Obesity 2022; 30: 1027-38.
- Wei et al.; Effects of time-restricted eating on nonalcoholic fatty liver disease. JAMA Network Open 2023; 6: e233513.
- Liu et al.; Calorie restriction with or without time-restricted eating in weight loss. New England Journal of Medicine 2022; 386: 1495-504.
- Andriessen et al.; Three weeks of time-restricted eating improves glucose homeostasis in adults with type 2 diabetes but does not improve insulin sensitivity: a randomized controlled trial. Diabetologia 2022; 65: 1710-20.
- Jones et al.; Two weeks of early time-restricted feeding (eTRF) improves skeletal muscle insulin and anabolic sensitivity in healthy men. American Journal of Clinical Nutrition 2020; 112: 1015-28.
- Liu et al.; The effect of early time-restricted eating vs later time-restricted eating on weight loss and metabolic health. The Journal of Clinical Endocrinology and Metabolism 2023; 108: 1824-34.
- Deng et al.; Effects of time-restricted eating on intrahepatic fat and metabolic health among patients with nonalcoholic fatty liver disease. Obesity 2024; 32: 494-505.
- Hutchison et al.; Time-restricted feeding improves glucose tolerance in men at risk for type 2 diabetes: a randomized cross-over trial. Obesity 2019; 27: 724-32.
- Gabel et al.; Effects of 8-hour time restricted feeding on body weight and metabolic disease risk factors in obese adults: a pilot study. Nutrition and Healthy Aging 2018; 4: 345-53.
- Lowe et al.; Effects of time-restricted eating on weight loss and other metabolic parameters in women and men with overweight and obesity. The TREAT randomized controlled trial. JAMA Internal Medicine 2020; 180: 1491-9.
- Wilkinson et al.; Ten-hour time-restricted eating reduces weight, blood pressure, and atherogenic lipids in patients with metabolic syndrome. Cell Metabolism 2020; 31: 92-104.
- He et al.; Time-restricted eating with or without low-carbohydrate diets reduces visceral fat and improves metabolic syndrome: a randomized trial. Cell Reports Medicine 2022; 3: 1007777.
- Andriessen et al.; Three weeks of time-restricted eating improves glucose homeostasis in adults with type 2 diabetes but does not improve insulin sensitivity: a randomized controlled trial. Diabetologia 2022; 65: 1710-20.
- Schroder et al.; Effects of time-restricted feeding in weight loss, metabolic syndrome and cardiovascular risk in obese women. Journal of Translational Medicine 2021; 19:3.
- Phillips et al.; The effects of time-restricted eating versus standard dietary advice on weight, metabolic health and the consumption of processed food: a pragmatic randomized controlled trial in community-based adults. Nutrients 2021; 13: 1042.
- Feehan et al.; Time-restricted fasting improves liver steatosis in non-alcoholic fatty liver disease – a single-blinded crossover trial. Nutrients 2023; 15: 4870.
- Meessen et al.; Differential effects of one meal per day in the evening on metabolic health and physical performance in lean individuals. Frontiers in Physiology 2022; 12: 771944.
3 Responses
Wonderfully written and informative. Thank you so much for this.
I went from ~33.5 BMI in February this year, to ~20 BMI by maintaining at least 16:8 early TRE and caloric restriction. It appears that I was right when it comes to early vs late TRE but I’m thankful that I read your section about OMAD. I probably shouldn’t have pushed OMAD with a meal mostly consisting of rolled oats and fruits given my family history of type 2 diabetes. I will use the advice from this post and spread my TRE to at least 4 hours. I knew OMAD would be short term and that I will return to breakfast and early lunch at least. Luckily with BMI in normal range I had a fasting glucose of 4.71 mmol/L, triglycerides 0.81 mmol/L and LDL-C 2.26 mmol/L in July so I probably didn’t mess up my blood sugar control and not having a lot of visceral and ectopic fat should have helped. After reading this post it seems that TRE isn’t going to help me much like with athletes study, but it’s been my habit for the past 8 months so I will continue to use it as a lifelong health pursuit without resorting to OMAD.
Enough about me, just wanted to share my sincerest gratitude for the work you do and for making it available to everyone freely. I’m sure that I speak for many of us out there on our journey to long healthy lives. I am humbled by your non biased approach and scientific rigor.
All the best! Greeting from Belgrade, Serbia 👋
Thanks for the kind feedback, Alex. And congrats on your health journey. What a success!
I do agree with you that continuing TRE may be a great way for you to maintain the weight loss you have achieved. And, by the way, do expect a bit of a rebound. Such a major weight loss in such a short period of time is almost impossible to maintain forever. I am not saying this to discourage you, but to make sure you have the right expectations and don’t beat yourself up if you do regain some of the weight you lost. And truth be told, if you adhere to your new, healthier habits going forwards, your health is going to be a lot better than what it was even if you do regain a few pounds. If I was you, I would now focus heavily on finding a way of eating and not-eating (TRE) as well as other lifestyle habits (exercise, sleep, light exposure) that I can happily live with long-term and that can help me maintain most of the health benefits and weight loss I have achieved.
Warm wishes,
Mario
Thank you so much Mario 🙂
You have raised a great point regarding rebound and I appreciate it a lot. That’s exactly what my priority now is, figuring out how to maintain healthy habits and weight but with realistic expectations and not getting discouraged with some rebound. I have learned recently about Kevin Hall’s research regarding metabolic slowing with extreme weight loss and I think that I have caused something similar to myself. I just have to figure out how to balance it out and I have enough leeway to still stay in healthy BMI range (I know BMI is just one number and health is way more than that).
Nutrition science is fascinating topic and entirely new to me. I’m in completely different field (IT). I’m going to read remainder of your blog posts to learn how to incorporate best of our current scientific understanding in my life.
Best wishes, looking forward to your future work!