In this podcast with Dr. Heidi Pak (@pak_heidi) we primarily talked about the difference between fasting and time-restricted feeding, how and why it works and also the efficacy of caloric restriction (CR) in different mouse strains and species. It was a great podcast, very fun as well.
Other topics that we touched on included sleep, the postdoc shortage, hardships of being a grad student, the healthspan vs lifespan debate and some more. We also came up with some less-than-serious nutritional recommendations:
"I guess the diluted [fasting] diet [I studied] is the equivalent of eating celery" (Heidi Pak)
If you want to know why we were talking about celery and why Heidi wants to put pigs on a weight loss and fasting diet, then this podcast is for you.
Heidi Pak – short bio
Heidi did her PhD with Dudley Lamming at the University of Wisconsin–Madison where she studied fasting, time-restricted feeding and caloric restriction in relation to aging. She now works as a postdoc for Joseph Takahashi, continuing her work on fasting and moving into the relatively new field of circadian rhythms.
What is caloric restriction?
Also known as dietary restriction is the practice of restricting the caloric intake of mice, usually by 30%. This extends the lifespan of many, although not all, mouse strains considerably. CR is hailed as one of the biggest breakthroughs in biogerontology and for good reason. If it worked in humans; or if we could find a drug that mimics CR in humans it would be the biggest biomedical revolution since the invention of antibiotics!
The discovery of CR is often attributed to seminal work by Clive Mccay in the 1930s. However, what made CR big was not the work by McCay showing that food restriction from birth stunts growth and extends lifespan, rather it was protocol improvements. For example, Walford and Weindruch found in the 1980s that adult-onset CR can work reliably in many strains if it is introduced gradually.
This podcast episode is about another such protocol improvement. The disambiguation of fasting and caloric intake through modern feeding protocols and enabled by 24/7 monitoring of food intake, locomotion etc.
Using those new feeding protocols Heidi and others found that fasting contributes to the lifespan and health benefits of CR.
What is fasting and time restricted feeding?
The idea behind fasting to slow aging was that after the body has digested most of the food, when hunger starts setting in, an organism might be in something akin to a CR state. Therefore, imposing periods of fasting could reap some of the health benefits of CR without the drastic weight loss, constant hunger and other side-effects normally associated with CR. While plausible this theory had many problems, not least the fact that it seemed to contradict the work of Walford and Weindruch who suggested that gradual onset CR was more beneficial in adult mice than starting the protocol abruptly. It stands to reason that alternating periods of fasting and feasting could be stressful to the body, just like the abrupt CR protocols used in the past, and thereby shorten your life (or that of a mouse).
We needed evidence to settle this debate, or at least start addressing some of the major points. Before we can understand the evidence, let us take another look at what fasting exactly is.
Defining fasting is not easy since the cutoff when “fasting” starts is somewhat arbitrary and that can be a problem when reading papers and articles on this topic. When reading about fasting research it is important to keep in mind that a 16hr fast and a one-week water fast could be reported as the same thing by the media, even though the physiologic effects will be very different. Importantly, it could well turn out that certain eating windows are much more beneficial than others. Maybe the water fast is healthier, or maybe it is not. We do not know what will work in humans yet.
Intermittent fasting and time-restricted feeding mean similar things in mice. Although generally fasting refers to on-off feeding schedules and restricted feeding refers to shorter and variable eating windows within a day.
Little progress was made on this matter and some people, like me, remained convinced that fasting has no benefits or that we “simply do not know”. Early studies using intermittent fasting protocols were always biased by weight loss. Since the fasted mice lost weight, even if it was less than during CR proper, it was hard to attribute the health benefits to fasting rather than weight loss. There were also studies suggesting that fasting is not an important component of CR (Nelson and Halberg 1986).
However, one observation pointed out in the work by Heidi Pak always made a strong case for the importance of fasting. While, yes, intermittent fasting induces CR; CR actually induces a form of intermittent fasting! Apparently, the CR data itself was always potentially biased by the effects of fasting.
From Mitchell 2019, eating windows indicated for caloric restriction (CR), control (AL) and meal fed mice (MF).
As you can see from this image taken from a typical fasting study, mice tend to self-impose a fasting window when they are under CR. They are so hungry that they gorge on their food early during their waking hours and then fast for many hours afterwards.
Building on this finding Heidi Pak and many others designed protocols to test this further.
How do we test the effects of fasting?
Several protocols exist to induce fasting without weight loss. One is caloric dilution, the other is time-restricted dispensation of food using a machine, or training mice to eat during a specific time window. All of them have advantages and disadvantages.
Caloric dilution involves diluting the diet with cellulose that cannot be digested and is one of the approaches that Heidi used in her research. While a very smart idea, I am not the biggest fast of the caloric dilution approach as I have written on my blog when I was first reviewing Heidi’s paper.
See here for more information why I think this way: http://biogerontolgy.blogspot.com/2021/12/caloric-restriction-and-fasting-does-it.html
Be that as it may, I do not want to be overly critical of the research here. As scientists we tend to be overly skeptical of new research. Fact is, modern fasting research was a big breakthrough in the field of CR because it provided conclusive evidence that fasting contributes at least to some of the health and longevity benefits of CR as we discuss in the podcast. This is very exciting because fasting is much more tolerable than CR in humans, thus could see actual application in the real world.
Another strategy to study fasting is removing the fasting window that CR mice self-impose to isolate the effect of calories. This means forcing CR mice to eat multiple times throughout the day. Such studies generally find that fasting is one but not the sole contributor to the health benefits of CR.
This is where it gets complicated – circadian timing
Recent work has shown that not only is fasting one component of CR that contributes to the lifespan extension, there is also a component of circadian alignment. The recently published work by Acosta-Rodríguez et al. 2022 from the Joseph Takahashi lab demonstrated this nicely. This paper has some of the most beautiful figures I have ever seen so highly recommended to check it out.
Here is a very good video explanation of the paper: https://www.youtube.com/watch?v=q5ZN_mT-xbw&t=12s
During the podcast we briefly talked about this work and I mistakenly mentioned that mice in that paper were fasted during the night. In fact it is quite a bit more complicated than that. Night is the active period for mice and it is generally beneficial for them to eat during this period. The above paper actually tested two groups fed at night. One with a 12 hr feeding window and one with a 2 hr feeding window during the day or during the night period. The latter was the most beneficial for lifespan, meaning the mice will be fed at the beginning of the dark period, will quickly finish their food and then will fast for most of the night and the day.
Why humans are not two-legged mice
Even though mice are my favorite animal model and I consider myself a mouse researcher, I have to admit that mice have their limitations as a model of aging. Let us first talk about the benefits of mice, though. Running mouse studies is orders of magnitude cheaper than running human studies, it is more ethical, the mouse is well-understood as to its genetics and longevity, it is amenable to transgenic manipulation and so on. In contrast to nematode worms and fruit flies, the other two popular aging models, mice are closer relatives of humans; they are real, breathing mammals with many of the same traits as humans. They get cancer, they like to cuddle with other mice and sometimes fight with them as well.
The fact that mice are so similar to humans, by the way, is also a reason why we need to treat them well and keep them healthy. Only data from healthy mice will translate to healthy humans. In the podcast we briefly discussed how important it is to keep mice healthy, which is reflected by long lifespans of the control group.
What is the problem with mice then? Well, they are unusually “fragile”, as we have alluded to in our podcast with Peter Fedichev. Meaning, they are unusually short-lived for their size, and unusually prone to cancer, whereas humans are exceptionally long-lived given their body mass and size.
Another limitation is the small size and high metabolic rate of mice. This precludes many surgical interventions and has led to a lot of spurious findings. In this context I like to tell the story of hydrogen sulfide induced suspended animation, which would be also very interesting to aging researchers, if it worked. However, as it turned out this was an intervention that only worked due the physics of small body sizes (Asfar et al. 2014). This is where other animal models come in handy to validate the mouse data. We were talking about pigs as one option during the podcast.
It is difficult to overstate the differences in metabolic rate between mice and humans. A mouse heart contracts at an unbelievable rate of 500 to 700 beats per minute and their energy needs are commensurate with their fast metabolism. Although a mouse may not eat much in absolute terms, per gram of body weight it eats much more than a human, consuming a staggering 10% of bodyweight every day. Imagine eating 8kg of food every day! If you starve a mouse for a couple of days, it will die. Humans in contrast can go on for months without food.
Nevertheless, we can learn a lot from mice. We learned that fasting can work to extend health and lifespan, in principle. Nevertheless mice cannot tell us whether fasting will work in humans.
Perhaps there is an eating window in humans that can mimic mouse time restricted feeding? No one knows whether this will have to involve a week of severe fasting, maybe 2 weeks of moderate fasting or whether the currently used protocols in clinical practice employing moderate fasts are enough.
Hopefully future research by Heidi and others will help us to understand this issue.
References and notes
Acosta-Rodríguez, Victoria, et al. "Circadian alignment of early onset caloric restriction promotes longevity in male C57BL/6J mice." Science 376.6598 (2022): 1192-1202.
Nelson, W. & Halberg, F. Meal-timing, circadian rhythms and life span of mice. J. Nutr. 116, 2244–2253 (1986).
Asfar, Pierre, Enrico Calzia, and Peter Radermacher. "Is pharmacological, H2S-induced'suspended animation'feasible in the ICU?." Critical Care 18.2 (2014): 1-8.
Mitchell, Sarah J., et al. "Daily fasting improves health and survival in male mice independent of diet composition and calories." Cell Metabolism 29.1 (2019): 221-228.