REPORTER

Fasting Generally

​

Who would benefit from fasting and why?

Virtually everyone who wants better health. The best research we have suggests that unless you have an extremely rare disorder that makes it dangerous to go several hours without food, a fast will induce metabolic repairs that will make you healthier and might help you live longer.

 

What’s the best way to fast?

There are two main types of fasting: daily fasting and prolonged fasting. Just about everyone would benefit from a healthy fast each day, and most people would probably benefit from a prolonged fast of a week or more every once in a while. For the majority of people, the prolonged fast should be done under medical supervision, but the daily fast can be done on your own. That said, if you’re taking medication, scientists and doctors recommend you first consult with a doctor knowledgeable about fasting to see if your dose needs to be tweaked during your fasting hours. A diabetic, for example, may need a lower dose of insulin while fasting so as not to run the risk of hypoglycemia.

​

Daily Fasting

​

What is daily fasting, and why would I want to do it?

Most people don’t realize it, but they already fast every day. When you stop eating overnight (or during the day if you’re a shift worker), you’re fasting until you take your first calories or caffeine the following morning. The crucial question that researchers only recently thought to ask is whether the length of the fast affects our health. What they’ve discovered is that people who fast longer overnight are healthier, evidently because a longer fast gives the body more time to make repairs.

 

Each night the body does a great deal of mending, reconstructing, and tidying up—work that it mostly can’t do during the day because it’s busy processing the nutrients from our meals, firing all the neurons that make up our thoughts, supplying our muscles with the energy to move about, and doing everything else that adds up to our waking lives. So out of necessity the body puts off many of its deepest repairs until night when we’re no longer eating or stirring around.

 

The overnight repairs are important indeed. To take just one example, every night our gut has to patch vast sections of its whisper-thin lining that got abraded as food slid across it during the day. If the lining isn’t fixed, we can get leaky gut syndrome, in which food, bacteria, and viruses leak out of the gut and into the abdominal cavity, where they cause all sorts of havoc. To take another example, our brain has to clear out the waste products of the day’s countless chemical reactions each night. If it doesn’t, we can develop ghastly neurological disorders like dementia. Our bodies make untold numbers of repairs like these night after night to keep us free of cancer, cardiovascular disease, diabetes, kidney disease, and on and on, and when they have more time for repairs, the healthier we are—which is why a longer overnight fast helps.

 

How long is the ideal daily fast?

Most scientists would say at least twelve hours and maybe up to about eighteen hours. Researchers recently discovered that the body doesn’t get to work on most of its deeper repairs until about six hours after we’ve eaten our last calories. They think the body waits because there’s a metabolic cost to switching from eating and processing nutrients to fasting and making repairs. Apparently the body doesn’t want to make that costly metabolic switch if we’re just going to force it to switch back an hour or two later by putting more food in our mouths. So about six hours after our last calories, the body seems to be confident enough that we’re done eating to begin its repairs in earnest, albeit at a slow rate. With each hour that passes, the rate increases, as if the body is gaining more confidence that we’re really fasting and won’t disrupt its work by taking more food. By twelve hours after our last calories, the rate of repair doubles with each extra hour of fasting.

 

Unfortunately, most people rarely go twelve hours a night without food. Studies have found that people in the developed world often eat across fourteen or fifteen hours a day. A typical person who finishes her bedtime snack around 10 p.m. and breaks her fast the next day with coffee at 7 a.m. is fasting only nine hours a night, which means she’s getting just three hours of modest repairs and never reaching the exponential rate of repair. If, however, she took her last food at 6 p.m. and broke her fast at 9 a.m., her fifteen-hour fast would give her nine hours of repairs, three of them in the exponential phase. Scientists think those extra hours of repair will keep her healthier over the course of her life. We don’t yet have long-term human trials to confirm (or disprove) that assumption, but short-term human trials and long-term studies of lab animals point strongly in that direction.

 

Although researchers don’t know for sure how long the ideal daily fast should be, human studies generally show health benefits after about twelve hours of nightly fasting, and the benefits increase markedly with every hour beyond that. Eating windows of as few as six hours have yielded excellent improvements in biomarkers that predict long-term health. My sense from the researchers I’ve talked to and whose studies I’ve read is that most think the ideal daily fast lies around sixteen or eighteen hours, for an eating window of eight or six hours. A few researchers think the ideal fast might be as long twenty hours, but most think cramming an entire day’s worth of food into your stomach in a four-hour window might exact too high a toll on the body. Most trials of time-restricted feeding (TRF) have an eight- or ten-hour eating window because researchers have found that almost all volunteers can eat comfortably in those times. For the same reason, many researchers themselves eat in an eight- or ten-hour window. I’ve eaten in a six-hour window for a couple of years now and have found it both very easy and, I believe, great for my health, as I discuss in chapter 16 of The Oldest Cure in the World.

 

Can a longer overnight fast keep you slimmer?

Probably. Mostly this seems to be because people who narrow their eating window tend to eat fewer calories—and they aren’t any more hungry than people who graze all day; in fact, studies show they may be less hungry. A longer fast also helps the body burn a little more fat overnight, and the reason is similar to that for overnight repairs. To oversimplify just a little, the body has two modes with regard to fat—fat-making and fat-burning—and it can be in only one of them at a time. During the day, the body is mostly in fat-making mode in order to store fat from the foods we eat. When we stop eating long enough, the body switches to fat-burning mode and burns some of its fat stores. As with the overnight repairs, this burning of fat doesn’t start in earnest until about six hours after our last calories. And just as with repairs, the fat burning starts to increase dramatically after about twelve hours of fasting. The person in my example above who fasts only nine hours a night is not only limiting herself to three hours of repairs and no exponential repairs; she’s also limiting herself to three hours of fat burning and none of the “overdrive” burn. But if she fasted fifteen hours, she’d get nine hours of repair and fat burning, including three hours in overdrive.

 

Does it matter when our eating window is?

Yes, the timing is crucially important, with an earlier window almost certainly healthier than a later one. A great many studies show that our bodies were built to process nutrients most efficiently early in the day and that good health flows from eating in sync with that biology. One reason is that by late afternoon our circadian rhythms force our digestive organs and other nutrient-processing apparatus to slow down. We can of course still eat late in the day or at night, but when we do, the food lingers longer in the stomach and gut, glucose dawdles in the bloodstream, and the nicks and dings that we suffer as a result add up over time. Eating at night appears to be especially damaging, as it disrupts some of the vital overnight repairs I mentioned above. I discuss this at some length in my book.

 

In the best trials we have so far, volunteers have been healthiest when eating more of their calories earlier in the day and taking few or no calories after the late afternoon. The science is still young, and it would be premature to trumpet a specific window with certainty, but the eating windows that have produced the best health results in humans tend to be around 9 a.m. to 3 p.m., or 8 a.m. to 4 p.m. That said, the best window for any particular individual will vary slightly. It depends on factors like when you wake up and, after waking, how long it takes your hormones to get your food-processing machinery to crank into gear. To account for individual variability, scientists usually instruct their volunteers to start eating one to two hours after getting out of bed in the morning.

 

Trials of later eating windows, like noon to 8 p.m., have shown inconsistent benefits, and still later windows, like 4 p.m. to 8 p.m., have shown almost no benefit and possibly some harm.

 

So are you saying it’s healthier to skip dinner rather than breakfast?

Yes. Or rather, move dinner earlier in the day. Most people who practice time-restricted feeding make the mistake of skipping breakfast. I certainly did. But because evolution has primed our bodies to process foods most efficiently in the morning, we skip breakfast at our peril. The best evidence we have says good health flows from putting dinner in the middle of the day, as many people around the world do. I eat my big meal—what most people would call dinner—around noon or even earlier, with a smaller snack around 2 p.m. Whenever you eat your last meal, it’s unhealthy to snack later in the evening. The Salk Institute’s Satchin Panda, one of the world’s leading experts in the field, says eating at night is one of the most harmful metabolic assaults you can make on your body.

 

I’ve been eating dinner for decades. If it’s so bad, how come I’m not sick?

The damage that comes from eating late, like the damage that comes from smoking a single cigarette, is small and progresses to disease only after doing it many times. A diagnosable disorder can take quite a long time to manifest. That said, people who narrow their eating window often notice within a few days that they feel less sluggish and more alert, just as people who quit smoking often notice right away that they can breathe more easily.

 

But here’s another point to consider: Are you sure you aren’t sick? Do you, like half of all American adults, have high blood pressure? How about high cholesterol? Although many doctors will tell you these and similar disorders are a normal result of aging, they aren’t. Hypertension and hypercholesterolemia are diseases, almost always avoidable through good diet and other habits, and they usually signal worse trauma to come, like heart attack, stroke, diabetes, kidney disease, dementia, cancer, and much more. If you have such a disease, you’re sick, and it’s possible that the timing of your eating may be contributing to your sickness. The good news is the power to reverse disease often lies in our own hands. As I discuss in my book, most people can get healthier with a few lifestyle changes, chiefly by eating plants, and fasting can often help.

 

Does it matter what I eat during my window?

Absolutely. A narrow eating window can overcome some of the harm of a bad diet, but we have no evidence that it can completely counterbalance unhealthy eating. No matter when you eat, the best research suggests a healthy diet—one that is mainly or, better still, entirely of minimally processed plants—is needed for optimal health. For more information, see “Sources on Diet” at the end of my book.

 

In The Oldest Cure in the World, you base a lot of your argument for a narrow and early daily eating window on a couple of studies from a single lab. Are there any other studies?

At the time I wrote the book, the studies into early time-restricted feeding (eTRF) by Courtney Peterson, now of the University of Alabama at Birmingham, were just about the only rigorous human studies on daily fasting and certainly the only rigorous human studies on eTRF. Since the book has gone to press, scientists elsewhere have published results from careful-seeming eTRF trials that have strongly corroborated Peterson’s findings. In October of 2021, for example, researchers at Tianjin Medical University in China reported that when they asked type-2 diabetics to eat for twelve weeks either in a ten-hour eating window from 8:00 a.m. to 6:00 p.m. or at whatever hours they liked, those who ate in the ten-hour TRF had better functioning beta cells (the insulin-making cells of the pancreas), less insulin resistance, lower triglycerides, lower total cholesterol, and lower LDL cholesterol (the most harmful kind). As in Peterson’s studies, these were impressive changes from merely changing when, not what, the volunteers ate.

 

In another study published in February of 2022, researchers at the Chinese Academy of Medical Sciences asked healthy volunteers to eat for five weeks in one of two eight-hour windows. Half the volunteers ate in an early window (eTRF) while the other half ate in a midday window (mTRF). The early volunteers were allowed to choose an eight-hour window between 6:00 a.m. and 3:00 p.m. (the slight flexibility in the window was meant to accommodate different wake-up times), while the midday volunteers chose an eight-hour window between 11:00 a.m. and 8:00 p.m. The eTRF group saw improvements in a range of biomarkers—insulin sensitivity, fasting glucose, total body mass, adiposity, inflammation, and diversity of gut microbes—whereas the mTRF group improved very little (chiefly in insulin sensitivity, and even then not as much as the eTRF group).

 

Both of these Chinese studies were randomized controlled trials, so they added considerable weight to Courtney Peterson’s finding that eating earlier, with the last calories in mid-afternoon, significantly improves health.

 

But aren’t there studies that say time-restricted feeding (TRF) doesn’t work?

Yes, researchers behind two prominent studies have claimed, or the media have claimed for them (often vociferously), that TRF doesn’t work. But to my eyes both studies had deep flaws and don’t refute TRF at all. Let’s look at each.

 

The flawed UCSF TRF study of 2020

In September of 2020, a team led by researchers at the University of California at San Francisco published the results of a trial in which they divided overweight or obese volunteers into two groups, one of which ate for twelve weeks in a midday TRF window of noon to 8 p.m., the other eating whenever they wanted. The biomarkers of the mTRF group improved hardly at all. The study so shocked the lead researcher, UCSF’s Ethan Weiss, that he stopped eating in the mTRF window that he had adopted several years earlier. “There is no benefit to eating in a narrow window,” he said flatly to one reporter.

 

But the facts, including the facts of his own study, don’t come close to supporting his statement. Here are five pertinent facts:

 

First, it’s probably true, as the Chinese Academy study from February of 2022 suggests, that eating in a later window (like noon to 8 p.m.) is less healthy than eating in an earlier window (like 8 a.m. to 4 p.m.). But as we’ve seen (above and in chapter 16 of my book), eating in an early window has been proven in multiple randomized controlled trials to have impressive benefits. So even if it’s true that a later eating window doesn’t improve human health, Weiss’s condemnation of all TRF windows is unjustified at best, sloppy at worst.

 

Second, it may not even be true that a midday eating window confers no health benefits. I say this for two reasons. First, most of the mTRF volunteers in the UCSF trial may not have been truly fasting during their fasting hours. Why do I think so? Because during their daily fasting period they were allowed to drink caffeinated beverages, and caffeine probably disrupts a fast. The research into this is young, but the Salk Institute’s Satchin Panda, who, to repeat, is one of the world’s foremost researchers in the field, believes that a single caffeinated drink, even a non-caloric one, likely jerks the body out of some aspects of fasting metabolism. We don’t know all the ways caffeine might alter a fast, but as Panda wrote in his 2018 book The Circadian Code, “Even if you have your morning coffee by itself, without breakfast, it still counts as the moment when you break your overnight fast.”

 

So although Weiss and colleagues claimed (and no doubt believed) that their mTRF volunteers were fasting for sixteen hours a night, if their volunteers took caffeinated coffee or tea, a zero-calorie energy drink, or a calorie-free caffeinated soda in the morning, they were in all probability breaking their fast to some extent after only ten or eleven hours. We don’t know how many of the mTRF volunteers did so, but since 80 to 90 percent of American adults drink caffeinated beverages regularly, we can safely assume quite a few did, which compromises the study’s findings.

 

Sadly, the researchers’ failure to account for a basic fact of fasting metabolism—that caffeine is apt to disrupt a fast—is a historical commonplace among doctors and scientists doing fasting research. Many scientists who have dabbled in fasting research over the years have been only minimally learned in the science of fasting and have often been wholly ignorant of its clinical practice. It’s the rare fasting scientist who has consulted a doctor experienced in supervising fasts—an expert who might have spared the scientists some of their blushes. Had the UCSF team bothered to talk to, say, Alan Goldhamer, a doctor who has fasted tens of thousands of patients at the TrueNorth Health Center just an hour north of San Francisco, he might have warned them of the dangers of a caffeinated “fast.”

 

The other reason I say a midday TRF window may confer health benefits is that the UCSF trial actually found modest health improvements in one of two subgroups who ate in the midday window. The volunteers in the trial came from two pools. One pool lived in or near San Francisco, were supported in person by the research team, and got a little healthier. The other pool lived elsewhere, participated in the trial almost entirely via a mobile app, and didn’t get healthier. We don’t know for sure why the first pool got healthier and the second didn’t, but other researchers, including Courtney Peterson, have speculated that the app pool, lacking in-person support, may not have been able to stick as closely to the TRF window as the better-supported San Francisco pool. If so, the reason the app pool didn’t benefit from mTRF was probably because they simply weren’t doing it.

 

Here’s a third fact about what was wrong with the UCSF study: the UCSF researchers told the control group and the mTRF group to eat in different ways, which further compromised the results. Specifically, the researchers told the mTRF volunteers to eat however they liked but told the control group to eat three structured meals a day and snack lightly if they needed to. The mTRF volunteers, if they were like most Americans, would have grazed throughout their eight-hour eating window. Because research suggests grazing is less healthy than eating in distinct meals, the scientists’ instructions may have set the mTRF group on a course for worse health while setting the control group on a course for better health. (I should note, however, that we don’t know whether the troubles of grazing throughout a sixteen-hour day would also show up when grazing in an eight-hour window.)

 

The contrast between this aspect of the UCSF study and Courtney Peterson’s studies could not have been sharper. Peterson ensured an apples-to-apples comparison by requiring both her eTRF group and her control group to eat identically. Both groups took all their food in three meals, and at each meal they ate the same food as each other. Her team also prepared every morsel that went into their volunteers’ mouths, so there was no doubt they were eating as identically as possible, whereas the UCSF study left volunteers free to eat whatever they made or bought themselves. Consequently, when Peterson’s volunteers got healthier on eTRF, she knew the difference was due to eTRF, not to variations in how or what her volunteers ate.

 

Fourth, Weiss seems to have overstated the case when he said the mTRF volunteers in his trial lost a concerning amount of muscle mass. Set aside for the moment that many of his mTRF volunteers, drinking caffeine and without in-person support, may not have truly done mTRF. The trouble with Weiss’s statement is that while the mTRF group did lose a small but significant amount of lean mass (as determined by a DXA body scan), lean mass isn’t the same as muscle mass. Lean mass consists of many bodily components, like organs and water, not just muscle. I assume this is why the text of the USCF study doesn’t actually say the mTRF volunteers lost muscle, only that they lost lean mass. It also says, “the DXA analysis of lean mass did not factor in muscle hydration, so it is possible that changes in hydration could confound the lean mass calculations.” That is, the lost lean mass could have been mostly water. But even if much of the lost lean mass was muscle, the amount lost was well below the threshold for clinical concern—not exactly the strongest evidence for abandoning a useful therapy.

​

Unfortunately, the kind of analysis I’ve laid out here didn’t find its way into media reports on the UCSF study, many of which implied or said outright that all forms of TRF were of no use to human health.

 

The flawed China Southern TRF study of 2022

In April of 2022, a second study generated another round of bad press for TRF. A headline in the New York Times was typical of the lot: “Scientists Find No Benefit to Time-Restricted Eating.” In fact, the scientists behind the study said no such thing. A truer, if less sexy, headline would have been “Obese People Improve on a Calorically Restricted Diet, But They Don’t Improve Further When Also Eating in a Restricted Window.” In the yearlong study from China’s Southern Medical University, obese people were asked to eat a calorie-restricted diet either at whatever hours they liked or in a daily feeding window of 8 a.m. to 4 p.m. Both groups lost weight and enjoyed improvements in their biomarkers like lower cholesterol and lower blood pressure, but for the most part the TRF group didn’t do any better than the non-TRF group.

 

The findings, while useful, weren’t a large surprise, at least not to people who know something about nutrition research. Caloric restriction, if severe enough, is one of the most potent—maybe the most potent—of interventions for the health of humans and nearly every other creature on the planet. Calorically restricted animals consistently live much longer with less disease, so it’s not a big shock that volunteers who were undertaking serious CR and already getting impressive health benefits weren’t additionally helped by TRF.

 

Then again, we can’t be sure they weren’t being helped by TRF because the study had some hefty flaws. The heftiest was that the non-TRF group may in fact have been doing TRF. If so, the TRF group wasn’t being compared to free-eating controls but rather to a TRF group much like themselves. The reason the non-TRF group may have been doing TRF is that those volunteers apparently weren’t told to spread their eating over the entire day. Instead, they seem to have been told just to eat when they wanted to, and since they reported before the trial that they usually ate all their food in about ten hours, they probably weren’t eating in a longer window during the trial. (The study didn’t say when they ate during the trial.) If so, it’s no surprise that the eight-hour TRF group didn’t enjoy major improvements compared to the ten-hour TRF group, although one might have expected slightly more improvements than the study found.

 

This brings us to the next flaw. The non-TRF group may have been eating in a window even shorter than ten hours. These volunteers were, after all, people who habitually overate but were asked in the study to cut several hundred calories a day from their diet. When calories are cut so steeply, dieters don’t usually spread their eating over the same number of hours as before. They wake up hungrier and often take a greater share of their calories earlier. I can only speculate, but a reasonable hypothesis is that overeaters who normally ate in a ten-hour window might eat in a nine- or even eight-hour window when their calories were cut by a quarter. If so, there may have been almost no difference in the eating windows of the two groups.

 

Finally, the TRF volunteers in the China Southern study, like the mTRF volunteers in the UCSF study, were allowed to drink caffeinated beverages during their fasting hours. Since the majority of Chinese adults drink caffeinated tea with some regularity, a portion of the TRF volunteers may not have been fasting each day for as long as the researchers imagined. If they took their tea a few hours before or after their eating window, they may have been practicing not an eight-hour TRF but a ten- or even twelve-hour TRF.

 

If you can explain what’s wrong with these studies, why can’t other reporters?

A lot of obstacles stand in the way of precise science reporting, and when the topic is complex, it’s the rare reporter who has the institutional support to do it thoroughly enough to get the story just right. One of the biggest structural problems is that even at large papers like the New York Times, reporters have to cover a broad swath of fields and can’t immerse themselves deeply enough in any one field to know its particular pitfalls. They get a press release from a scientific journal with a sensational headline—“Fasting Method Practiced by Millions Is a Flop”—and they run with it. Almost all reporters will, of course, call at least one or two experts to get some perspective on the study, but sometimes they’re talking mainly to the experts the journal points them to, or to the most quotable but not necessarily best informed experts, or to the experts whose opinions match their own biases. In any case, these chats only sometimes give reporters a broad enough understanding about what the study might have gotten wrong.

 

In addition to structural problems like these, there are other, more malign problems, mostly having to do with clickbait and bias. Stories that yield sexy headlines (“Hot Trend Debunked”) entice reporters and editors for the obvious reason: more clicks mean more advertising revenue. Then, too, all reporters have their own biases, never more so than where food is concerned. Like the rest of us, most reporters eat unhealthily and don’t particularly jump for joy over evidence that suggests they would do well to change, and they also know that many of their editors and readers aren’t eager to hear they should consider changing the way they eat. When it comes to early time-restricted feeding, I suspect few reporters want to bear the bad news that eating dinner at the usual hour may be taking years off our lives.

 

Bias in favor of unhealthy habits is rampant in dietary reporting. Just look at the many stories claiming that butter and bacon are healthy (neither one is), that drinking a little alcohol each day is healthier than drinking none at all (it’s not), or that being a little overweight is healthier than being slender (it’s not). It’s the rare reporter or editor who will tell you the “science” behind such claims is deeply flawed. So while it may be frustrating that many of the same phenomena are playing out with fasting, it’s not a big surprise.

 

Is there any reason to be concerned about daily fasting?

Probably not, but some scientists have raised one potential area of concern. A few observational studies from the 1970s and 1980s noted that people who regularly fasted for more than twelve hours a night were at higher risk of developing gallstones than those who fasted for shorter periods. Those findings, however, were only correlations, and no causation has yet been found between fasting and gallstones. Probably the people who developed stones did so from unhealthy habits like eating too much cholesterol. Still, since most people eat too much cholesterol, it’s reasonable for TRF researchers to keep an eye on stone formation.

 

That said, to my knowledge, not one of the dozens of interventional trials of TRF (as opposed to those older observational studies) has documented a single gallstone in humans or other animals. Additionally, over the past century, multiple fasting doctors reported that prolonged fasting (which, admittedly, is different from daily fasting) could dissolve gallstones and kidney stones. For reasons such as these, the great majority of fasting scientists and fasting doctors confidently recommend that virtually every adult eat in a window of 12 or fewer hours, down to about 6 hours. Experts say children who aren’t nursing can safely eat in a window of 12 hours, just as children all over the world did for many thousands of years before electric lighting extended our eating windows. We may yet learn that it’s also safe, and maybe even healthier, for children to eat in fewer than 12 hours, but that hasn’t been tested yet. By contrast, scientists are certain that children who are nursing or of nursing age, as well as other people with a biological necessity to eat frequently, should not have their eating window limited at all. They should eat whenever they want. To my knowledge, the question has not been tested in pregnant or nursing women, but scientists believe that since virtually everyone used to eat in a 12-hour window, pregnant and nursing women should be fine doing so, provided they’re getting all their calories during that window, as they should be able to do.

​

One important caveat is that if you’re on medications, experts recommend you talk to your doctor first to see if the timing of those medication needs to be changed. If you’re on a high dose of insulin, for example, taking it during your fasting period could make you hypoglycemic, so you may need to adjust the timing of your medication. Once the timing is adjusted (if any adjustment is needed), experts say it’s safe to do a daily fast.

 

Is intermittent fasting the same thing as daily fasting?

Intermittent fasting is a vague term that different people use to mean quite different things, from daily fasting to weekly fasting to prolonged fasts of several weeks. The vagueness is one reason I avoid the phrase. Most people use “intermittent fasting” to mean either time-restricted feeding of the kind I’ve been talking about above, like fasting sixteen hours per day and eating across the other eight hours (16:8 fasting) or a pattern of eating normally five days a week and fasting the other two days (5:2 fasting).

 

Another reason “intermittent fasting” is a bit of a pointless term is that all fasting is intermittent. The alternative, continuous fasting, would lead to starvation and death, so “intermittent” is redundant when used with “fasting.”

 

Are 5:2 fasting and alternate-day fasting (ADF) healthy?

There’s a case to be made for certain forms of 5:2 and alternate-day fasting (ADF), and there’s a case to be made against other forms of them. There are couple of ways to do 5:2 and ADF. One way is to eat no food whatsoever on fast days, while another is to eat a reduced number of calories, like 500 to 800 calories, on “fast” days. We don’t have data to say conclusively, but it would surprise me if going completely without food for two or more days a week proved healthy. The reason is that our circadian rhythms are set in part by when we eat, so if you’re constantly shifting your eating pattern back and forth between eating and not eating for a full day, you may be jerking your circadian rhythms around too much. Again, Salk’s Satchin Panda is one of the leading experts on this, and he has concluded that subjecting your body to these sorts of shifts is like putting yourself through jet lag every few days. My own miserable experience with thirty-six-hour fasting every other day for a month (see chapter 7 of my book) made me a believer in Panda’s conclusion.

 

On the other hand, eating lighter for two or three days a week (in contrast to eating nothing) may be quite helpful, especially if you’re not eating healthily enough the rest of the time. Short-term human trials have shown significant metabolic benefits from this pattern of eating. Similar trials have shown benefits in eating lightly for four or five consecutive days once a month. To keep from throwing yourself into the circadian jet lag that Panda warns about, people who eat this way should probably maintain a consistent eating window across both “fast” days and feast days.

 

If you eat in a 5:2 or ADF pattern, can you eat whatever you like?

No, but you might hear otherwise from some people who should know better. One of my biggest concerns with 5:2 and ADF is that some of the proponents of these ways of eating claim if you follow the rules for “fast” days strictly, you can eat whatever you like on feast days. They believe the health benefits that you accrue on your light-eating days will protect you from wanton feasting. Krista Varady, a nutrition researcher at the University of Illinois at Chicago and a leader in the field, is one of the most prominent voices behind this claim. (Check out the sprinkled donuts, pepperoni pizza, and cheeseburger on the cover of her book The Every Other Day Diet.) Unfortunately, I’ve never seen anything like good data to support the argument, which strikes me as terrifically injudicious. The data we do have strongly suggests that those who eat Big Macs on their feast days will be a lot less healthy than those who eat salad.

​

Does drinking black coffee, tea, or other noncaloric caffeinated beverages break a fast?

Caffeinated drinks almost certainly disturb a fast, but scientists don’t yet know precisely in what ways or to what extent. The researcher who perhaps knows the most about this is the Salk Institute’s Satchin Panda. You can see him in this clip discussing the effects of caffeine on a fast. What we know with certainty is that even a modest amount of caffeine, particularly in the morning, resets our circadian clocks. But to what extent morning caffeine also fires up our metabolism and pulls us out of fasting is, so far as I know, still uncertain. Panda’s best guess (and I concur) is that because caffeine is processed through our digestive system, it interrupts fasting metabolism to some extent, but it’s unlikely to be as much of an interruption as, say, eating a meal. Please note that this answer applies only to daily fasts of up to about eighteen hours. During a prolonged fast of days or weeks, it may be dangerous to drink caffeinated drinks, which most fasting doctors say should be entirely avoided.

​

Does taking a pill, vitamin, or supplement break a fast?

I’m not aware of research on the topic, but I’d make an educated guess that a similar principle holds to the one I’ve stated for caffeine, above: On the one hand, if your digestive system is processing the pill or supplement, and the pill or supplement is changing important biomechanical processes (as caffeine certainly does and some pills or supplements do), the pill or supplement is probably interrupting the fast to some extent. On the other hand, if the pill or supplement is noncaloric, it’s likely not a drastic interruption. Please note that this answer, like the one above, applies only to daily fasts of up to about eighteen hours. For reasons that I discuss at length in the book, it can be dangerous to take pills, vitamins, or supplements during a prolonged fast of days or weeks and, if done at all, should be done only after consulting with and being supervised by a medical practitioner experienced in overseeing prolonged fasts.

​

I’m diabetic. Can I do a daily fast?

Scientists and doctors who specialize in fasting say daily fasts are safe for diabetics. But they recommend that all people who take medications first consult with a doctor knowledgeable about fasting to see if their dose needs to be tweaked during their fasting hours. For diabetics, taking too much insulin while fasting could push them into hypoglycemia. Such people might need to take a lower dose of insulin while fasting.

​

Diabetics may also want to know that prolonged fasting can often reverse diabetes, particularly type 2 diabetes. Such fasts should always be done under the supervision of a doctor skilled in fasting. For a list of a few such doctors, see the answer in this FAQ to the question How do I find a clinic that will supervise a fast? (below). Eating a diet of minimally processed plants can also reverse diabetes, although more gradually than prolonged fasting can. For more information of reversing diabetes through diet, see the excellent work of Dr. Neal Barnard. For example, this video: A Nutritional Approach for Reversing Diabetes Naturally. Or his book: Dr. Neal Barnard’s Program for Reversing Diabetes.