Here’s what happened when I stopped eating dinner for 3 weeks, continued to work out in the mornings, and tracked my blood glucose levels.
Fasting was hard to wrap my mind around, even if it means skipping only one meal a day. Fasting calls to mind starvation, eating disorders, deprivation, malnutrition and people who suffer from fanaticism or other mental illness. Fasting is starting to gain acceptance in fringe groups of health and nutrition geeks, but even though I am a nutritional iconoclast myself it was very hard to break my mind free enough to question the “facts” I knew about
- “starvation mode” ruining your muscle mass and metabolism
- that eating several meals a day and keeping your blood glucose up was the key to avoiding binges.
- That you will become malnourished if you don’t eat a wide variety of foods every day.
I acknowledge that it is hard to put aside everything that we have been taught about how skipping meals is “bad,” but I ask you to shelve your skepticism for the next few thousand words.
There is a growing body of evidence that taking a break from food for a few days, or even skipping meals regularly is actually really good for you. Fasting has been linked to slowing or reversing signs of aging, reversing insulin insensitivity and type 2 diabetes, reducing inflammation, increasing growth hormone, decreasing the risk of Alzheimer’s and dementia, increasing brain function and memory, and of course reducing obesity. Probably a bunch of other stuff too.
Once I was able to get over the hump of skepticism, the idea of fasting for health just made a whole lot of sense. Let’s start with the idea that missing meals is bad for you. The time that humans have had access to a reliable food source at all times of the day, 365 days of the year has been only the last like 80 years (even after World War 2 starvation was still a concern among the poor in the United States). So the idea that bad things will happen if you’re not eating every 2 hours is absurd. If we weren’t meant to miss meals, then we wouldn’t have a mechanism for storing energy for when food isn’t available: fat.
Think of it this way: Oxygen is something we need a constant supply of. It only takes a couple of minutes to use up the oxygen that the body can store, and then we die pretty quickly after that if we don’t get more. Other mammals like seals and dolphins have an extra bladder for air storage, so they can spend minutes or hours underwater without a new supply of oxygen, but we don’t. But because we have fat, we don’t need a constant supply of food like we need oxygen. We can store enough food energy to last us weeks or even months. The world’s longest total fast was 382 days, but there are plenty of well known stories about people going weeks without food for political or spiritual reasons, or in survival situations. Think about good ol’ Jesus Christ. Things didn’t turn out so great for him, but it wasn’t the 40 days and nights of fasting that killed him.
And if it is perfectly normal, safe, and even healthy for humans to miss meals sometimes (or even frequently), then where do I get off thinking that I, –special delicate snowflake that I am — should not be subjected to the discomfort of missing a meal here or there. That argument holds no more water than the idea that someone can’t wake up early to exercise before work and then complains about how they’re not ready for bikini season again this year. Yeah, exercising self-discipline is tough and it sucks sometimes, but someone who doesn’t have any self-discipline and always takes the path of least resistance is never going to see much success in any aspect of their life. I’m not saying that everyone should withhold food from themselves, or much less that fasting should be some sort of punishment or penance. I’m just saying that my body was giving me some signals to slow down on the rich and sugary foods, and it wasn’t going to kill me to skip a few meals while I try to figure out what makes me feel best.
Once I convinced myself that missing dinner wasn’t a physiological emergency and that it didn’t have to be a big deal not to eat after work, it was liberating. I had so much more free time in the evenings. I could save hundreds of dollars a month on groceries. I wasn’t obsessing about the balancing act of having a satisfying meal without overeating. But weren’t you hungry?! For awhile I felt hungry as I went about my routine, but that did eventually become less unpleasant as I got used to it. Now, a month into it, I don’t even get hungry in the evenings.
The turning point was when I realized that I often went through my day with slight cues that I was falling behind on some physiological need or another without giving it much thought. For example, I think nothing of getting through a day after a lousy night’s sleep. Sure, I might feel a little foggy or drowsy at a few moments during the day, but I very rarely give being tired much thought or modify my routine because of it. Sometimes you’re just tired. That’s how life is. There will always be another chance to get a better night’s sleep tonight, I don’t need to climb under my desk for a nap the second I start to droop. I often go days with itchy dry skin, but never take the trouble walk to the bathroom (which is closer than the kitchen) to put on some lotion and relieve it. But if I felt the slightest twinge of hunger between meals it would send me running to the kitchen for a snack, and woe to anyone who got between me and food once I’d set my mind to eating. It was insane. Why did I think that I needed food so badly?
The funny thing was that now that I was wearing my CGM, I could see that the times when I felt hungry (usually only about 2 hours after a meal), my blood glucose was usually fine, or even still slightly elevated from my last meal. And yet first thing in the morning when I had gone 16 or more hours since my last meal, I usually felt a-okay. My hunger was no more a “need” than a smoker “needs” to have a cigarette every time they leave a building. Feelings of hunger are really just the product of habit, and habits can be changed. Once I hadn’t eaten after 2pm in a couple of weeks, I stopped feeling hungry after 2pm.
With the continuous glucose monitor I could see that my blood glucose didn’t start dropping a couple of hours after each meal and then continue to drop continuously until I was in the “red zone.” It would bump up after a meal, peaking after 2-4 hours, and then it would slowly work its way back down to my usual 60-80 mg/dl and stay there for hours.
On the other hand, there were other times when my blood sugar dipped low enough that my phone beeped insistently with a very rude alarm. One Saturday morning I had a 7am meeting with my boss. I wasn’t particularly hungry, but since it had been about 17 hours since my last meal and I didn’t know how long the meeting was going to last, I ate a banana. My blood glucose had a brief bump from the straight carbs in the banana, but the insulin reaction that followed pushed my glucose levels lower than they had even been overnight when I hadn’t eaten in hours. About an hour into my meeting, my phone (which was set to silent) started beeping with a very loud and very insistent alarm to tell me that my blood sugar was in the mid-50’s and dropping, and I had better eat something right away. The thing was, I felt fine. A little “light” maybe, but I felt focused, motivated, and not at all irritable despite being called into work early on a weekend.
One thing that did drive me batty was that the finger sticks that I had to do two times per day to calibrate the CGM almost always came in about 15 points higher than the CGM readings. So it’s probable that my levels weren’t as low as my histrionic alarm was telling me. I have had episodes of true hypoglycemia before (have had them ever since I was a kid). I know the weak feeling, shaking hands, dizziness, cold sweat and superlative hangriness of a truly low blood glucose, and this wasn’t even close. What true hypoglycemia is like: one day on a 6 hour drive back home from a bike ride in the Sierras I had skipped breakfast, and then eaten the only thing that I had in the car for lunch: a party sized bag of Starburst. Three hours later, I was standing in a CVS when all of a sudden I got dizzy, my skin got clammy, and I felt like my vision was closing in. I wasn’t sure I could stay standing for long enough to bring whatever snack I could grab to the register. I forget what I bought, but it was gone by the time I had walked 2 doors down to Whole Foods. I then put myself into debt at the hot foods bar with giant, desperate heaping spoonfuls of every rich food I could find.
Given my body’s reaction to sugar described above, it probably helps that I had also removed sweets from my diet before beginning this experiment because I’m not the kind of adult that ever grew out of wanting to have cake and ice cream for breakfast. And the very fact that I had spent so much money and taken the effort to stick this thing in my body using a very scary needle was also enough incentive to keep me making healthy food decisions. I know the recipe for hypoglycemia and thought about purposely throwing myself into an episode just to watch my glucose numbers through the process. But I had had such an easy time of taking myself off of sweets that I thought purposely triggering an uncontrollable binge wasn’t worth the risk of falling off the wagon.
We all understand logically that riding the glucose-insulin rollercoaster can’t be good for our mental health and behavior, but I was shocked at how painless cleaning up my diet had been this time. In this initial phase, I gave myself permission to eat whatever I wanted at breakfast and lunch, just as long as I didn’t eat sweets and I didn’t eat between 2pm and after my morning workout. But since I wasn’t eating the really awesome stuff like ice cream, cake, candy or pastries, I didn’t really see a big reason to go overboard on lentil stew, oatmeal with fruit, rice and beans, or a cheese omelette. Because my grocery trips were boring, I didn’t think of my home meals as an obligation to eat everything that I had in the house. Even restaurant food wasn’t all that interesting if I wasn’t going to have a big blow-out and follow it with a rich dessert. I stopped obsessing about food, which was a huge stabilizing factor on its own. I was able to focus at work, my coworkers didn’t irritate me, and I didn’t throw myself a pity party every time things didn’t go my way. The depression that I had been carrying around with me as if it were a permanent part of my personality started to lift. I was floating around on the pink cloud that comes when you first get free of an addiction.
What did my blood glucose do when I worked out?
When I began this experiment, I was sick as a dog with a massive chest infection so I didn’t do my first workout until almost a week into my science-ish experiment, and it took another week for me to get back up to speed. I was disappointed that I didn’t get much data when I was at my pot-induced, sugar-binging, working-out-2-hours-per-day worst, but what can you do?
Before we dig in to what my blood glucose did during exercise, I think it’s worth it to do a shallow dive into how your body uses fuel during exercise. The majority of energy that muscles use to fuel exercise comes from carbohydrate (as glucose) and fat. Your body has ways of storing both, but the biggest difference is in amount vs. availability. You can only store about 90-120 minutes of activity worth of carbohydrate in your muscles and liver (when carbohydrate is in its storage form, it’s called glycogen). You can store unlimited amounts of fat, but the problem is that fat takes longer to break down and use. You can think of it like your carbohydrate being your checking account, with enough money for a short period of time available immediately and refilled often. Your fat is more like your life savings tied up in an investment account somewhere; you can get at it, but the transfer takes a few days, so you’re only going to tap that account for periods of big spending.
When I hear people talk about burning fat or carbohydrate, they think of it like a toggle switch: you are either burning one or the other. In fact it’s more like a dimmer: when your energy demands are low, your body meets most of your needs with fat. But when energy demands are high, such as when you’re exercising, more and more carbohydrate comes in to cover that need. Let’s illustrate with a very simplistic example with made-up numbers. Let’s say that you can generate about 4 calories per minute from fat but you burn 5 calories per minute walking and 10 calories per minute running. If you’re running you’ll need 6 extra calories from carbohydrate to meet your 10 calorie/minute energy needs and only 40% of your energy will come from fat. However if you’re walking, you’ll only need to burn 1 extra calorie from glucose to meet your energy needs so 80% of your energy is coming from fat (4 out of 5 calories). This is where the “fat burning zone myth” comes from. It’s called the “fat burning zone” because a larger percentage of your calories are coming from fat, and it’s called a “myth” because it’s the same number of absolute calories coming from fat in both activities, but you burn more total calories at a higher intensity. Note that this is a simplified version of what’s going on, but you get the idea.
You can train your body to be more efficient and burn more fat for fuel by doing a lot of low intensity training and very, very slowly increasing the intensity of your workouts so that your body gradually is able to use more and more with fat before calling in carbohydrate reinforcements. But the dimmer can also go the other way, and you can turn down your fat burning by having too much glucose floating around in your bloodstream that your body wants to burn up before it turns to the fat, or by having too much circulating insulin that inhibits fat burning and tells your fatty acids to stay where they are.
I was surprised to see that in my first light workouts (short runs and brief strength sessions), my blood sugar reacted very little. Later, as I got back to my normal routine of hour-long morning runs and riding my bike to work when weather permitted, I found that I could pretty much count on the rude low glucose alarm to interrupt me after 15 minutes and to keep interrupting me for the remainder of the workout. I wish I had more data to learn exactly what this means. It could mean that I had so much insulin floating around shoving glucose into my cells that when I exercised and my cells started sucking even more glucose from the blood stream, there was hardly any left to circulate. I imagine this like a scenario where my credit card bills are so high every month, sucking my account dry, that I don’t have enough income to cover groceries. Let’s call this the “too many automatic withdrawals, not enough income” scenario.
An alternative explanation: maybe my stored glucose was dumping into my bloodstream in adequate amounts, but my cells were sucking it up and using it so fast that usage was very close to production without insulin messing with the balance at all. Like a checkbook that balances to zero every month not because you’re in debt, but because spending matches income.
Or a third explanation, maybe this was the explanation for why my performance had gone through such a precipitous decline a couple of years ago. Maybe I was so metabolically fucked from always having my fat metabolism downregulated from too much insulin floating around all the time from my constant meals, that I just couldn’t make fat available as a fuel source and my body had essentially forgotten how to metabolize it/gotten really inefficient at it, leaving glucose as my only substrate, which just couldn’t cover the energy demands of higher performance. Let’s call this the “Scrooge” scenario, where my body gets used to cramming money in savings, but refuses to spend that and instead lives like a pauper, limiting my overall performance.
What surprised me was that I actually felt totally fine as my urgent low glucose alarms beeped away. One morning I got my first low glucose alert 2 miles into an 8-mile run. I finished the run, feeling great the whole time. I drove home and showered, all with glucose levels in the mid-40’s without incident before I ate some breakfast.
It is likely that my fancy CGM wasn’t actually very accurate. In my twice-daily finger stick calibrations, my manual measurements frequently came in 15 to 20 points higher than the CGM values. While I wanted absolute precision, I knew that this was the best I was going to get, so I would have to take the absolute values with a grain of salt and just watch the shape of the daily graph for trends.
What happened to blood glucose on a long run?
Once I noticed that my glucose seemed to reliably drop into the 50’s and even 40’s for the duration of a run without me feeling like I was wilting, I was eager to find out what it would do on a long run. What did my blood glucose look like when I bonked? For various reasons that have nothing to do with this experiment I decided to structure my first long run in two parts: I would begin with a 10-mile hilly trail run with the dogs, and then return home and run on the treadmill until I started to hit the wall.
I had been having connectivity issues with my sensor earlier in the week, but I re-paired it with my phone the night before and all seemed to be going well. When I did my regular finger stick calibrations, I was pleased to see that both the manual measurement and the CGM seemed to finally agree to within a couple of points. I started running with values in the mid-60’s (low based on published norms, but normal for me after a cup of tea). So far so good.
Fifteen minutes into the run, I started getting the familiar alarm. I was running on singletrack in the dark with the two dogs, so it took me awhile to reach a spot where I felt safe to tear my eyes away from the beam of my headlamp to look at my glucose levels. Motherfucker! The damned thing had lost the sensor again. Once the app decides that the sensor is gone, it takes 2 hours to re-pair and still won’t start tracking until you calibrate it with a finger stick. So this entire run would go unmeasured. Dammit!
Including potty breaks, photo stops, and leash-untangling breaks, it took just under 2 hours for us to run 6 miles to the top of the mountain and 4 miles back down. I got to the car feeling slightly grouchy and rumbly in my tumbly, but the grouchiness was just as likely due to the frustration of the glucose monitor not working and the patience of spending 2 hours trying not to trip over dogs.
As soon as I got home from the 20 minute drive I manually tested myself and got a reading of 62. Then I started running on the treadmill. Twenty minutes later I hopped off and took another reading. Usually I need to really squeeze to get enough of a blood bubble to take a reading, but with my heart rate up the blood was covering my finger tip in no time. Apparently it’s possible to overwhelm a test strip with too much blood. After a couple of unsuccessful tries, the bleeding finally slowed down enough for me to get a good reading: 96 mg/dl. My blood glucose had gone up, even in my third hour of fasted exercise.
After another 20 minutes (now with over 14 fasted miles behind me), my glucose level was still 76 mg/dl. I was starting to get a little tired of running, but my legs still felt okay.
In the next 20 minute block I started to get that out of breath feeling at moderate speeds that I recognized as the first stages of a bonk. This time my glucose tested at 99 mg/dl. Oh for heaven’s sake! How long was I going to let this go on? I had already run over 16 miles, hadn’t eaten in 20 hours, half the day was gone, and I wanted to be done and eat something already. I cooled down and called it a day.
Thirty minutes after my workout, my glucose levels were still at 88 mg/dl, and another 30 minutes after that (moments prior to eating), they had settled in to 73 mg/dl (roughly the same level as when I had woken up that morning). There were no signs of imminent death. Despite feeling a little harried, my body felt fantastic and I probably could have run several more miles without incident or food if I had turned down the intensity slightly.
By chance a few months ago I had found that I felt better on my long runs when I did them without nutrition or pre-run breakfast. However, I attributed my good performance to a pot-fueled munch-o-rama the night before and so made sure to eat like a garbage truck before each long run leading up to my last marathon. I bonked on none of these fasted long runs, even though short months before I had had trouble running longer than 8-10 miles without falling to bits. But on this run I had skipped the pre-run binge and was still feeling as good as I used to before a mystery ailment took my speed away. I will talk more about this in my next section: conclusions.