How can bitter foods be good for us when they taste so bad? – Resolving the paradox

Laying out the problem

Our recent post on bitters left me with a lot of questions.

If bitter tastes indicate the presence of toxins and thereby help us avoid poisonous foods, why do they stimulate such positive physiological responses? Why would some of those responses protect us from metabolic diseases like diabetes and cancer? If bitter taste is merely a warning to avoid a particular food, then why do many traditions revere bitter foods? How do we explain why adults develop a taste for bitter foods that as children they found repulsive? Why does folk law say “Good medicine always tastes bitter”?

After a lot of pondering I think I’ve got an answer but to make sense of it I need to lay out what I see as the relevant parts of the puzzle first.
Read time: 16 minutes (3100 words)

Why do bitters protect against the diseases of civilisation?

In the Western dietary pattern avoidance of bitter foods in favour of highly palatable sweet, salty and savoury flavours is associated with hyperphagia (overeating) and the consequential diseases of civilisation, including diabetes, heart disease, cancer and neurodegenerative conditions.

Significantly, bitter compounds in foods have been shown to directly tackle each of these metabolic diseases at a fundamental level by switching on cellular processes that lead to appetite control, glucose regulation, cancer inhibition, and reduced inflammation. It is almost as if bitter foods were designed as a cure for the Western diet.

This indicates a very important physiological role for bitter foods. But this is strikingly at odds with the fact that bitter tastes are universally perceived as repellant. Why has the body evolved the need to ingest foods that are repugnant to it? and why do our cells only switch on these health-promoting responses when potential toxins are detected in the diet?

Bitter compounds are not nutrients

Another fact that I think is relevant to the puzzle, is that these bitter compounds are not really nutrients; they provide neither calories nor basic building materials (macronutrients), nor co-factors for metabolism, vitamins or minerals (micronutrients).

Their only physiological role appears to be signalling. Our body uses them to determine the direction of cellular activity. So when bitter substances are presented to the cilia in the lungs waft more, and airways relax and open, cancer stem cells divide less and die off more; appetite signalling says “I’ve had enough” sooner.

Why would this be? Why not activate these processes without having to ingest bitter substances first? Why can’t we eat a nutrient-replete diet and skip the bitters? What is it to our evolved metabolism that it has to have a bunch of otherwise useless bitter compounds around before it looks after itself?

Hormesis and bitters

Some of the healthiest diets studied to date, including the Mediterranean, Hunter-Gatherer and Okinawan diets, may owe most of their success to the presence of bitter principles and other plant-based ‘toxins’ that exert their effect through hormesis.

Indeed, a paper published last month in Nutrition Review suggests just this:

 the Mediterranean diet can be conceptualized as a form of chronic hormetic stress [Martucci M et al, 2017, Full Text]
 

There has been much speculation about which elements make these diets healthful. If the hormetic explanation proves to be correct, then most of their benefit may derive not so much from what they leave out but what they include: a wide range of herbs, spices, berries, seeds and nuts, fruit and vegetables, all of which contribute to bitter receptor activation.

Hormesis provides a powerful explanation about how bitter compounds affect us, but it does not explain why we have this response. Remember: it’s not the bitter compounds doing this to us, it’s our body that has evolved to respond this way. Why this should be so is part of the puzzle.

An overarching theory of bitters should explain why hormesis evolved.

Summing up the challenge

There are several observations that need squaring, and they all need to make sense in evolutionary terms as well:

1. Bitter flavours are associated with compounds in plants that are toxic at high doses. Bitter tastes invoke aversion.
2. Adults, but not children, develop a taste for bitter foods. Many traditional diets revere bitter flavours.
3. Eating bitter foods activates healthful biological responses that go to the root of the major diseases of our time.
4. Bitter foods contain no essential nutrients, so appear to have no dietary value.

I’ve been looking at these for several days and scratching my head a lot. Whilst it’s quite easy to generate hypotheses that explain two or three of these, finding a satisfactory overarching theory that explains all of them is not so easy. But after a lot of thinking, I may have come up with an answer, and that answer comes from proper paleo thinking.

What are bitter compounds signalling?

As I read around this topic of bitters a question kept popping into my mind: If bitter compounds are cellular signals, what are they signalling? The first answer is “the presence of dietary toxins” and indeed many of the identified cellular responses involve up-regulation of detoxification pathways. This makes sense of several bitter responses:

• Appetite suppression = “don’t eat more than you need to – you don’t want to go poisoning us”
• Increased cilia wafting in the airways = “get those toxins out of my lungs”
• Gut cells pumping toxins into the intestines = “throw out those toxins I just absorbed”

Each of these makes sense as a response to the potentially toxic nature of bitter compounds. But the trouble with this is that it doesn’t explain some of the other responses, for example inhibiting cancer, reducing inflammation, enhancing sperm production (yes bitters play a role there too!). If removing toxins was the whole story here then why would it be unhealthy to eliminate bitters from the diet altogether?

Bitters, I suggest, are signalling something more significant. Something left over from our evolution. I think they are signalling the success of the nutritional acquisition strategy along a binary axis: hunter ⇔ gatherer. And I think the following argument demonstrates the power of the evolutionary discordance principle, which posits that diseases of civilisation are caused by a mismatch between the environment we find ourselves in today and the one in which we evolved.

I’m going to sum up my ideas with a graphical representation then go on to explain how it all fits together.

Hunter Gatherer Food Choices

The axis I have laid out above arranges the major hunted and gathered foods by energy density. Honey is the most energy dense food, followed by meats, then starchy tubers, then berries. Lowest are leaves in which I include stems, shoots and other aerial plant parts.

(This axis is not meant to be definitive, and there are arguments for placing certain foods higher or lower on this scale, but the broad thrust of it is sufficient for the argument I am laying out.)
This pattern divides nicely into hunting on the left and gathering on the right. Although honey acquisition might be considered a form of gathering it is usually collected by hunters (men) and has a very high status similar to or even greater than that of hunted game.

Surveys of hunter-gatherer food preferences follow a very similar pattern, with only the position of tubers and berries being contentious [see our post on Hadza food preferences]. (There is also an argument for exchanging their position on this axis in terms of energy density as in some cases tubers may actually have a lower energy density than berries, however, such tubers may be available for more of the year and in larger quantities than berries.)

Another gradient that is found from left to right across this axis is the presence of toxic compounds. These are very low on the honey and meat end but increase the further one moves to the right-hand side.

This pattern of toxins is due to the defence strategies associated with each food item. In the case of honey, the defence is external in the form of the bee stings. Prey animals rely primarily on mobility to avoid being consumed, so have very few defence toxins in their flesh. (Some frogs and fish are poisonous, and some birds accumulate poisons from their diet, but practically all other vertebrates are edible). Plants, on the other hand, being unable to move put considerable energy into the production of toxins – insecticides, fungicides etc.

Honey and meats are strong stimulators of sweet and umami taste sensors. Moving left to right, sweet and umami compounds decrease and bitter compounds increase.
So how do all of these factors tie together?

Paleo principles

Homo Sapiens evolved, like most animals, in an environment where food quality and availability varied considerably (see box 1). There would have been times of plenty and times of scarcity. Changing patterns of climate, weather, competition, seasonality and serendipity would ensure that finding enough high-quality food was never easy for long. Hunter-Gatherer tribes have fall-back foods – such as the Hadza’s low-calorie tubers – which are least preferred and only relied on when better food is unavailable.

When a large herbivore was captured, or honey was found these were important foods – high in calories and macronutrients. Because of their nutritional value, these foods had high status and desirability, and the taste sensations, understandably, evolved to be highly rewarding.

But in lean times, wild seeds, berries, tubers and leaves would make up a higher percentage of the diet. This would require eating more bitter, less palatable foods, which would be higher in secondary plant metabolites (toxins) than in times of plenty. Bitter taste receptors evolved to register displeasure, partly because they indicate the presence of toxins, helping us to choose the least toxic foods from this category, but also as a reminder to go find some macronutrient rich foods.

So far, that explains the palatability side of the taste receptors, but it does not explain why the lean-times bitter-diet should signal so many beneficial processes around the body. To understand that we need to look at what cells can do.

For the sake of simplicity, I’m going to posit that humans can only be in one of two metabolic states, although in reality there is almost certainly a spectrum between the two.

State 1: Growth (Make hay while the sun shines)
When macronutrients are abundant (feast) the body needs to grow. This means building muscle, cell division, laying down fat, increased fertility for reproduction. This drives the species forwards in terms of physical strength and increased births and buffers the body against future lean times. During this state, cells replicate but are more vulnerable to copy-errors and unchecked mutations. Detoxification pathways are down-regulated and there is a heightened risk of oxidative stress.

State 2: Repair (fasting state)
When there are few macronutrients available cellular activity switches to a repair and conserve mode. Food might be scarce for a long time. The body puts its efforts into maintaining the individual ready for the next time of plenty by down-regulating growth, increasing cell repair and defences, up-regulating anti-oxidant systems, reducing fertility, and being more efficient with limited resources.

There are multiple lines of evidence for this bi-phasic response. Calorie restriction and fasting increase the life span of test animals. However, if caloric intake falls too low, such as in anorexia, then fertility takes a dive: menstruation will cease. Malnutrition in childhood causes stunted growth and delays puberty, whereas in countries with excessive food availability puberty can take place unusually early.

The biphasic response to feast and famine is also evident in some key cellular pathways which act as nutrient sensing switches:

• Nrf2 switch, responds to a huge range of phytonutrients, up-regulating the body’s natural anti-oxidant defences, detoxification pathways and improving mitochondrial function [See Pall & Levine, 2012 for a good overview]. Over 100 genes are controlled by Nrf2.
• AMPK / mTor switch, responds to calorie, carbohydrate and proteins, pushing cells towards growth when macronutrients are abundant, and towards repair and defence when they are limited. [For a brief overview see ScienceDaily, 2017].

What we are seeing with bitter receptors seems to be another switch type mechanism.
So from an evolutionary perspective, when the body detects bitter foods it interprets that as indicating lean times (because, hey, who would eat bitter herbs when there is bison on the menu?) so it responds by putting more resources into repair and maintenance. Hence research showing bitter receptors inhibit cancer, appetite, and fat storage.

In the last 150 years, Western diets have shifted towards increased consumption of macronutrient-dense, sweet, salty and savoury foods with few bitter flavours. The result has been to push cellular activity almost exclusively toward the growth phase with little opportunity for repair and restore. Diseases associated with this dietary imbalance are all essentially metabolic: obesity, diabetes, auto-immune and cancer.

Putting it all together

• Over evolutionary time, the human body has had to respond to extremes: feasts and famines.
• Thinking of bitter receptors as sensors of the nutritional environment explains many of their apparently contradictory properties.
• A high intake of bitters provides a biological signal that food is limited and so it is time to play it safe: switch on the protective mechanisms.
• A low intake of bitters signals that high-quality food is in abundance: time to grow, reproduce, store fat for a rainy day.
• Children’s aversion to bitter foods is appropriate because they are still growing. Once they reach adulthood a taste for bitter flavours develops which helps reduce growth signalling and increases repair and maintenance.
• Western dietary patterns do not include sufficient signalling from bitter compounds to flip the switches into repair and restore mode.
• Diseases of civilisation appear to be the result of a mismatch between our evolved feast/famine response and the modern highly palatable foods.
• Fasting, calorie restriction and short-term vegan diets mimic times of food scarcity, producing a range of well-documented health benefits in accordance with these ideas.
• Increasing the use of bitters through the use of herbs, spices, fruit and vegetables, extra-virgin olive oil, red wine, cocoa, tea and coffee may provide a sustainable healthful eating pattern such as seen in the Mediterranean diet, offering the best of both worlds.

An evolutionary feast-famine model of nutrient sensing and disease

In the above diagram each curve represents the health benefits that can be obtained from (a) macronutrients and (b) phytonutrients (including bitter components). Their peaks represent a presumed optimum benefit, whilst the low points represent negative health effects due to excess (extreme left and right above) and deficiency (inside edges)

There are three zones representing healthy and unhealthy eating patterns. As we have seen, for most of our evolution, hunter-gatherer food availability has moved back and forth between feast and famine repeatedly passing thought the healthy zone.

When feeding becomes stuck in one mode chronic stimulation of nutrient receptors promote disease.

Chronic Feasting Zone

This represents a dietary pattern dominated by macronutrients and refined foods that are inherently low in phytonutrients. It is characterised metabolically by chronic stimulation of mTor and sweet/umami signalling, at the expense of bitter and Nrf2 signalling. This is typified by the Western diet, and associated with the ‘diseases of civilisation’ (hyperphagia, obesity, metabolic syndrome, heart disease, diabetes and cancer).

Chronic Fasting Zone

This region includes issues arising from a chronic lack of macronutrients (starvation/malnutrition) and/or where individual phytonutrients exceed their hormetic benefit and start to exert detrimental (toxic) effects.

Starvation is rare in Western countries but occurs in anorexia nervosa. Malnutrition is more common, with vitamin and mineral deficiencies being widespread (e.g. iodine, vitamin C, D, K2, B12). However, these can arise for many reasons which have nothing to do with the feast/famine signalling model being proposed here. B12 deficiency in vegetarians might be considered an example.

Phytonutrient excess can lead to toxicity. Examples include kidney stone formation from excess oxalates, carotenaemia from excess carotenes, inhibition of mineral absorption from phytates (particularly in marginal diets and cereal-based weaning), iodine deficiency (hypothyroidism, goitre, cretinism) from excess cruciferous family. Chronic cyanide poisoning from an excess of some fruit seeds – such as apple, apricot and cherry pits (also tropical ataxic neuropathy from cassava). Photo-dermatitis, where the skin produces lesions when exposed to UV sunlight, from an excess of furocoumarins found in Rutaceae (e.g., citrus fruits) and Umbelliferae (e.g., parsnip, parsley, celery, carrots) families. [For a comprehensive review see Naturally Occurring Food Toxins, Laurie C. Dolan, 2010.]

Healthy Zone

In the healthy zone macronutrients (feast) are balanced by the presence of phytonutrients (famine) signalling. Sweet/umami receptors and bitter receptors are stimulated. Nrf2 and bitter signalling ensure protective mechanisms are operating within the region of beneficial hormesis, preventing the development of metabolic dysfunction.

Traditional healthful diets such as the Mediterranean and Okinawan diets are examples of successfully combining both forms of signalling. However, the composition of such diets is presented as constant over time. The temporal variability seen in hunter-gatherer feast/famine cycles may have additional benefits.

For example, constantly high levels of phytonutrients in the diet can lead to problems, as expressed in the following excerpt, and the author recommends varying intake throughout the day.

 …very high chronic, long-term Nrf2 elevation can produce pathophysiological effects like almost any regulatory effect taken to extreme. Therefore, one needs to take care not to raise Nrf2 levels too high for too long. One way of minimizing any pathophysiological effects is to vary the levels of Nrf2-raising agents in the body at different times of the day.

  Pall & Levine, 2012 

The pendulum image in the healthful zone above, indicates variations in the balance (feast/famine) over short, medium and/or longer periods, mimicking hunter-gatherer adaptations to variations in food availability.

There is little research yet to indicate what might be an optimal strategy in this regards, but what is known is that fasting, intermittent fasting, calorie restriction and short-term vegan dieting have many documented benefits in the context of the Western dietary pattern.

Historically, our own culture included temporal variations in diet, partly as a result of seasonal food availability, but also through cultural observances such as feast days, Lenten abstinence – when rich foods were given up, and a prohibition on eating meat on Fridays. Furthermore, traditional medicine has emphasised the use of bitter herbs and fasting in certain conditions. One wonders to what degree the loss of these traditions contributes to modern chronic disease burden.

End Note

Understanding the role of dietary bitters as being a signal for times of famine, in which the body engages in cellular repair, protection and detoxification provide a rationale for maximising their benefit in the dietary pattern.

Whilst our aversion to bitter flavours makes sense as an evolved mechanism to avoid toxins, the more interesting aspect of bitter foods is their health-promoting properties at sub-toxic doses, especially the idea that they can prevent or reduce the metabolic imbalance underlying the most burdensome diseases of our times.

The Western dietary pattern, it seems, is in desperate need of some bitter medicine.