It is sometimes useful to consider things from an evolutionary standpoint. We can look back at early humans and make some general recommendations. Granted, there is little or no proof, but the exercise is still useful and interesting.
There is often great debate about whether we should be eating constantly, or occasionally. For example, Dr. Yoni Freedhoff recommends, in his book, The Diet Fix to eat as soon as you get up and then every 2.5 hours or so during the day. On the other hand, Intermittent Fasting proponents would say that it is quite sufficient to eat once a day or even once every other day. So what’s the truth? First, let me say that there are people who use both systems and do well. But which system makes more sense?
Let’s consider our cousins – the omnivorous wild mammals. It is virtually unheard of, in the natural world, to require feeding 3 times per day, every day in order to stay healthy. Most omnivorous large mammals eat considerably less frequently than that. Obligate herbivores, because of the low caloric density of their food, often require constant grazing – think cows and sheep. Grass, for example has very low caloric density. Much of the grass is indigestible and passes through the cow to exit as manure.
Most carnivores, such as lions and wolves will eat only several times per week or even several times per month. Sometimes this is because food is scarce, but even in times of plenty, it’s probably because food is not so easily available. Catching a zebra is much harder than catching a bag of cheetos. This also likely has something to do with caloric density, since most of the animal foods are absorbed by our bodies.
We’ve all seen those TV shows with lions and tigers all around a herd of zebra sleeping away in the hot African sun. Well, those lions were not hungry and therefore did not eat. One meal per week seems to do just fine for them. If a hippo carcass happens to wash up on shore, sure,they’ll eat. So, understand that eating several times per day is not a necessity for omnivores and carnivores. We are not solely driven to eat by nutrient deficiency.
Physical and mental capacity is not impaired by a lion’s week long ‘fast’. If fasting made them sluggish and stupid, well the lion species would not have survived very long. No, the long interval between meals does not impair them in any significant manner. They ate a large meal – storing much of the calories in their bodies and then are using these stored calories to survive. It’s normal.
Mammals have adaptations that allow them to survive with an intermittent food supply. That is, the body has a way of storing food energy, so that lion can eat once a week. This goes for humans as well. The main way to do this is to store glycogen in the liver (stored sugar) and then to store triglycerides in fat tissue. When you eat, you are putting food energy into your stores. When you fast, you are pulling food energy out. It’s inconceivable that mammals are designed with this amazing system for storing food energy and yet it is still necessary to eat every couple of hours to stay healthy. That’s like building an amazing pool and spa, and then arbitrarily deciding that you can’t get wet after all.
Hunter-gatherer societies, as well as wild animals virtually never got the problems of obesity, diabetes of cardiovascular disease, even during times of plenty. It is estimated that animal foods provided about 2/3 of their calories. So, for all the modern teeth gnashing about meat and saturated fats, it seems that our ancestors had little problems eating them. It should also be noted that many societies ate carbohydrate based diets (eg. Kitavans and Okinawans) and also had no problems with obesity. It seems to be a modern problem, and I suspect that refined grains and sugar plays an overwhelming role here.
Things started to change about 10,000 years ago with the agricultural revolution. Early man started to farm instead of hunt, which led to a greater reliability of food allowing a typical pattern of eating 2-3 times per day. Even with that, there was little obesity until relatively recently (1970’s USA).
So, it is certainly possible to eat meat and have little diabesity. It is also possible to eat carbohydrates and have little diabesity. The problem, (Nutritionism’s Greatest Blunder) is focusing obsessively on macronutrient content (how much fat, how much carbs). It’s the insulin response that matters, not the macronutrient breakdown. The toxicity lies in the processing, not the food. So highly refined and processed grains and sugars, as well as vegetable oils are the problem, not carbs and fats.
Circadian rhythms are predictable, 24 hour self-sustained changes in behaviours, hormones, glandular activity etc. Most hormones of the body, including growth hormone, cortisol and parathyroid hormone are secreted in a circadian rhythm. These rhythms have evolved to respond to differences predominantly in ambient light determined by the season and time of day (which governs food availability). These patterns are seen in virtually all animals from flies to humans, and it is estimated that 10% of a given organism’s genes show circadian changes.
The master circadian clock is the suprachiasmatic nucleus (SCN). It is believed that food was relatively scarce in Paleolithic times are predominantly available during daylight hours. This is mostly because humans hunt and eat by day and once the sun went down, well, you just couldn’t see the food in front of your face. Other animals are nocturnal and may very well have circadian rhythms more suited to eating at night, but not humans.
So, is there a difference between eating during the day and eating at night? Well, the studies are few, but perhaps suggestive. One very interesting study compared the effect of eating a large breakfast versus a large dinner. While there are many association studies, this is one of the few intervention studies done in humans as opposed to mice. Most have favoured eating breakfast, or eating earlier in the day, although most studies have too many confounders to be truly useful.
So what this study did was to randomly assign two groups of overweight women to eating a large breakfast (BF group) or a large dinner (D group). Both ate 1400 calories/day, and the macronutrient composition of each diet was matched – only the timing of the largest meal was changed. While both groups lost weight, the BF group was clearly superior for both weight loss and waist size (important measure of visceral fat) by almost 2.5 times (-8.7 kg vs -3.6 kg).
So why there such a huge difference in weight gain? Well, this further graph may explain things a bit. The graph shows the insulin response to meals. The BF group had more insulin in the morning while the D group had more at night, as expected. However, by totalling the Area Under the Curve (AUC – graph to the far right) you can see that overall, the dinner group had a much larger rise in insulin. This is fascinating. The same total calories led to more insulin secretion simply based on meal timing.
An earlier, smaller 1992 study had shown much the same thing. In response to the same meal given either early or late in the day, the insulin response was 25-50% greater in the evening.
Weight gain, of course is driven by insulin. So, while the carbohydrates and calories were identical in both groups, the insulin response was not, translating into more weight for the D group. This illustrates the very important point that obesity is a hormonal, not a caloric imbalance. This study has profound implications over meal timing. There is certainly the well known association of night shift work and obesity. However, this may also have to do with the increased cortisol response due to disturbed sleep.
Now, this does not necessarily mean that you must eat a large meal as soon as you wake up. But it means that perhaps eating a large meal in the evening (after the sun goes down) may cause a much larger rise in insulin than eating that same meal during daylight hours. The problem with breakfast is generally that we are in a hurry in the morning and tend to eat very highly refined carbohydrates (toast, cereal, bagels etc) which tend to also stimulate insulin severely. But waiting until noon to have a large lunch as your main meal seems to be a good solution. This also avoids the ‘rushing out the door’ or ‘grabbing a muffin’ sort of response to the exhortation to ‘eat breakfast – it’s the most important meal of the day’.
Folk wisdom, of course, also advises to avoid eating large meals in the evening. The reason offered usually is something along the lines of “If you eat just before bed, you don’t get a chance to burn it off and it will all turn to fat”. Maybe not technically true, but perhaps there is something here. Eating late at night seems to be especially obesogenic.
There is also a natural circadian rhythm to hunger. After all, if it was simply due to food intake, we would consistently be hungry in the morning after the long overnight fast. But personal experience and studies confirm that paradoxically, hunger is lowest in the morning. This is ‘paradoxical’ because the morning time meal follows the longest period of the day without food. Breakfast is typically the smallest, not the largest meal of the day. This indicates that there is a circadian rhythm that is independent of the eat/fast cycle.
Ghrelin, the hunger hormone, shows a marked circadian rhythm with a low at 0800. Interestingly, with fasting, ghrelin peaks at day 1-2 and then steadily falls. This aligns perfectly with what is seen clinically, where hunger is the worst problem on fasting day 1-2. Many people on longer fasts report that hunger typically disappears after day 2.
Hunger typically falls to its lowest level at 7:50 am and peaks at 7:50 pm. This applied to almost all foods. Interestingly, vegetables show no circadian rhythm in desire to eat. I don’t really know what this means but I don’t think it’s because vegetables are not delicious.
Understand once again, that these are natural rhythms that are inherent in our genetic makeup. If you take away all external stimuli, these rhythms still persist. What does it mean that hunger is lowest in the morning? One implication is that hunger is not so simple as ‘the longer you don’t eat, the more hungry you’ll be’. No, there are many more subtle inputs and hormonal regulation of hunger plays a key role.
However, the studies are conflicting. NHANES data on evening eating failed to show any association between late eating and weight gain, as might have been predicted. Nevertheless, the possibility that eating during daylight hours results in less insulin secretion must be considered.
So, what’s the practical implication? At 0800 in the morning, our hunger is suppressed actively by our circadian hormonal rhythm. It seems counter-productive to force oneself to eat. What’s the point? Eating does not produce weight loss. Forcing ourselves to eat at a time when we are not hungry is not likely to be a successful strategy.
However, eating late at night also seems to be a poor strategy. Hunger is increased maximally at approximately 7:50 pm at the same time that insulin will be maximally stimulated by foods. This means higher insulin levels for the same amount of food intake. This higher insulin level will naturally drive weight gain. This is the typical pattern of eating in North America, where dinner is the main meal. This is mostly driven, not by health concerns, but by the hours of the working and school day. This also leaves shift workers at a particular disadvantage. They tend to eat larger meals later in the evening, leading to higher insulin.
So the optimal strategy seems to be eating a large meal in the mid-day – sometime between 12:00 and 3:00pm and only a small amount in the evening hours. Interestingly, this is the typical traditional Mediterranean eating pattern. They have traditionally eaten a large lunch, followed by a siesta and then a small, almost snack sized ‘dinner’. While we often think of the Mediterranean diet as healthy due to the foods, the timing of the meals may also play a role.
One final word of advice – We should DEFINITELY all take siestas. Even better, take a siesta by the lake in a hammock.