What follows is a summary of a presentation I gave to a symposium on the future of food. In attendance were farm producers, food processors, food retailers, and research scientists.
I believe that the development of nutrigenomics will prove to be the most important development in food production and preparation since the introduction of fire allowed our ancestors to eat cooked food instead of eating it raw. That’s a sweeping statement, so let me see if I can substantiate it.
Today, the idea of food as medicine, functional foods, and nutriceuticals is being accepted and assimilated into popular culture. People buy blueberries and drink tea because they’re supposed to contain anti-oxidants, and even when they don’t know what anti-oxidants are, they know that they’re supposed to be good for you. Some men eat pink grapefruit because it’s supposed to help avoid prostate problems. And people drink wine or beer because it decreases the likelihood of heart attacks and heart disease (although there might possibly be other reasons why people drink as well).
Sometimes this interest in food as more than fuel for the body goes off into blind alleys, such as the “zero carb” fad diets of several years ago, but the trend towards functional food is now firmly established. There is, however, a problem in that people don’t understand why they keep getting mixed signals about whether a particular food is good for them or not. Part of the reason for this is our culture encourages short-attention spans, and sound-bite explanations for everything. Therefore, something that leads to greater complexity goes against this trend, which makes it harder for it to be appreciated. This is a battle that is not going to go away, but will continue to be fought, probably forever, but certainly for the rest of your careers. And you will spend your working lives in stretching to find analogies to explain why one person can eat cheese without having it affect their cholesterol, whereas their spouse or neighbor can’t, as well as explanations for other genetically-related effects of food. The reason why you’ll continually stretch for analogies is that there really aren’t many good ways of explaining these things in sound-bite formats, and most analogies are too simple for something as complex as genetics. The result is that public understanding of food’s effects will continue to fall short of the reality.
Another major factor that affects where we are today in dealing with food is that this field is starting to move at computer speeds. Ever since Craig Ventnor’s work on the Human Genome Project showed us that computers could do data processing even faster and cheaper than grad students, the entire bioscience field has started shifting into overdrive, at IT speeds instead of in vitro speeds. This is creating its own problems, because it means we are building mountains of data for which we do not yet have good research tools. Traditional statistical analysis tools are inadequate for this kind of massive, multi-dimensional, multivariate research.
The Four Dimensions of the Future of Food
So, with that as background, let me introduce what I see as the keys to the future of food: a four dimensional matrix of what will motivate individuals to select and buy foods for functional, as well as aesthetic and nutritional purposes.
The first dimension is demographics, starting with who’s paying attention to the importance of food. We know that the body’s natural resilience declines as we age, and I suspect that we will find that this applies to our ability to tolerate foods that aren’t quite right for us. Accordingly, as we age, we should hew more and more closely to our bodies’ actual needs in order to attain optimal health. This means that as we age, we can tolerate fewer Big Macs, and should eat more fiber, for instance.
If you look at health care costs, you will see that health care costs per person, per year, tend to bounce around a little, yet are reasonably stable from around the age of 2 until about age 55. After that, they start to increase almost exponentially. Today, the boomers are roughly between the ages of 42 at the low end, and 62 at the high end. This means we have the largest generation in history about to move into the high-rent district of health care.
This also means that we have a rapidly aging population that is starting to pay more attention to what they eat, and how their food directly affects their health and sense of well-being. As a result, we are moving into an era where, first, we have the attention of an important and growing group of people (aging boomers), and second, the costs of ignoring custom-tailored nutrition will explode, both for the individual, and for society and governments.
The end result of this is that the marketplace should become more receptive to the health and dollar value of appropriate nutrition. But there’s another area of research related to age that is going to impinge on research in nutrigenomoics as well: extended life expectancy and expanded life span. There are researchers who are talking about significant increases in life expectancy over the next 20 years, potentially on the order of an increase of 50%, which would take life expectancy to somewhere in the vicinity of 120 years. Then there are the extremists, such as Ray Kurzweil and Terry Grossman in their book, Fantastic Voyage: Live Long Enough to Live Forever. They literally believe that someone who can manage to live for another 20 years may be able to live forever – or at least until their money holds out, which isn’t quite the same thing.
Regardless of who’s right, we have no precedents for life expectancies like this, and hence we have no idea if physically hale and active 100-year olds have the same nutritional needs as hale and active 50-year olds. All we have are suspicions based on anecdotal evidence, without statistical rigor. We don’t really know what the genetic differences between two such people are – or will be. This does not directly affect nutrigenomic research right now, but will eventually do so as longevity research progresses and we have more and more people living to triple-digit ages.
The 2nd Dimension: Genetic Differences
The 2nd dimension that defines the future of food is genetic differences or markers. I’m fascinated that we are discovering that an individual’s biomarkers can change with environmental stimuli or triggers. For example, I’m a celiac, but had no symptoms of celiac disease until I was in my late 30s. It may be that I contracted this genetically-linked autoimmune disease because I had a genetic predisposition, and was exposed to some kind of environmental trigger. If that’s possible (and we don’t know at this time), then our nutritional needs may change not just with age, but also with exposure to specific environmental factors that alter our gene expressions.
One research group that is working in oncology believes that the genome isn’t a blueprint, as is popularly thought, but rather a finite state or Turing machine (i.e, a computer). That means if you provide different inputs (environmental stimuli, potentially including nutrition), you can get different results.
We already know that it’s true that “different strokes for different folks”, that subtle genetic differences from one person to another can have dramatic consequences, but that’s only vague hand-waving compared to where we are headed. The race is on for the $1,000 DNA sequence for an individual, pushed in part by government funding, and in part by an X-prize of $10 million for the first group to do it repeatedly. There is also talk of pushing towards $100 as the price tag for DNA sequencing for an individual.
Once we have a comprehensive DNA sequencing mechanism at a price health care payors or individuals are willing to pay, it becomes a whole new ballgame in nutrigenomics, one with an embarrassment of statistics on the players. We will have problems deciding what to do, and worse, how to do it. Data reduction is already a problem – now multiply that by the thousands or millions of people who will have detailed genetic information available, and the problems of performing satisfying research become legion. It’s going to take an awful lot of work to extract it, plus techniques and knowledge we don’t yet have. But then think of the gold in them thar hills – starting with all of the currently unknown effects and consequences of the interplay of genetics and environment, of different forms of food, and nutrition from different sources. This isn’t just a mountain of data we’re talking about – it’s a mountain of gold.
What will happen, of course, is that we will start with coarse-grained approaches that apply to large groups, such as people of a common racial or geographic background, and gradually become more and more fine-grained, focusing on smaller and smaller groups. Eventually we will get down to the level of being able to tailor specific nutritional profiles for specific individuals that will optimize their health. We’ve already seen this in the manufacturing trend towards mass customization vs. mass production. We know that mass customization can produce superior results at lower prices than mass production, as with Dell customized computers (before they lost their way) compared to Compaq mass produced computers. Translating mass customization to the food production, processing, distribution, and retailing industries will require a massive re-engineering of processes – but that, too, is probably where we are going.
There are probably limits to this. I doubt if we’re going to hire specific farmers to grow specifically modified foods for particular individuals. But short of this, I can’t clearly see where customization will stop – and it’s going to matter a lot to the food industry. Likewise, we can’t have JIT farming, but can have JIT, customized processing of (relatively) common food components. As we develop data on populations, we will be able to modify, grow, or produce specific foods for smaller and smaller groups of people. As one contemporary example, consider the growing demand for certified organic foods, which started as a tiny niche market, and is now largely mainstream. This move towards developing foods for smaller groups is all a gradual process: as we develop more awareness of how specific nutritional components affect specific individuals, we will gradually move towards more finely tailored foods. Moreover, along the way, we are defining specific and valuable niche markets that can be served at premium margins, and with higher costs of entry for competitors. If that doesn’t sound like an enticing commercial proposition, I’m not sure what does.
And, by the way, as John Kelly said to me, this industry seems to be at the stage that the IT industry was around 1985. If that analogy is correct – and I believe it is – than the changes will happen slowly at first, and then faster and faster. In fact, not only will the rate of change accelerate, but the rate of acceleration will increase, just as it has in IT. What this means is that if you’re in the food industry, you’d better not wait too long before deciding whether you want to undertake this kind of massive change, or you will be left way behind on an exponential curve.
The 3rd Dimension: Price
The 3rd dimension of the future of food is reasonably simple: price. People tend to choose cheaper foods, and specialty foods tend to be more expensive. Using me as example again, when I buy a loaf of gluten-free bread, it costs me $6 a loaf. If you’re battling it out for market share selling bread at a buck-fifty, this sounds mighty attractive, and more and more producers will look for ways of entering such a premium market with reasonably costs.
This natural attraction for high-margin goods will produce pressure on food retailers and producers: As consumers become used to the idea of eating specialty foods, and actively managing their health and wellness using food as an critical ingredient, demand will swell, and consumers will look for better sources for what they want. Again, using my own experience, in the late 1980s, when I was first diagnosed, gluten-free products were scarce and of questionable quality. I would occasionally go to the specialty store at Sick Kids hospital to buy pasta, and just didn’t eat bread, pizza, cookies, or anything like that at all because I couldn’t get them regularly. Then Kingsmill rice bread started to become available in my local supermarket, but it was only white bread, the quality wasn’t very good. It gradually got better, but I also found new suppliers who produced products I liked better, with more fiber and better texture, and I now have a much broader range of products available, including muffins and baguettes. Indeed, Pizza Pizza, a local pizza chain, now produces a gluten-free pizza that I can get delivered to my door. Indeed, Pizza Pizza behaves as if they’ve hit the jackpot with gluten-free crusts, and it’s now available at most of their locations. However, gluten-free foods are still very expensive: I still pay $6 for a loaf of bread, and almost $20 for a medium-sized, gluten-free pizza delivered to my home.
Not everyone will be willing to pay above the commodity price for specialty foods, and their willingness, or lack of willingness, to pay will be one of the determining factors of who will form the market for genetically-appropriate food. Indeed, I would suggest that the future of genetically-appropriate foods will follow a path very similar to that of certified organic food today. Not everyone will buy it, preferring cheaper alternatives, but some will, and that percentage will probably grow until genetically-appropriate foods become mainstream.
The 4th Dimension: Public Understanding
The 4th, and most unpredictable dimension, is public awareness and understanding. This is related to price, but it’s more than that. The first part of this is awareness: do people know about the benefits of genetically-appropriate foods? Part of the problem here is that the people in this room, and your colleagues elsewhere, have done a poor job in explaining why it’s important. That’s not actually your fault: it’s very difficult to explain the importance of genetic variations to people who think of DNA as a kind of small string with some flags on it. Our culture encourages the 10-second sound bite (never mind 30-seconds; that’s passé). This makes it a very difficult problem, and one that you will be fighting for the balance of your careers. Indeed, you see this problem all the time it in the popular press: there are regular reports that seem to say that scientists can’t make up their minds about whether a given food is good for us, or not. First butter is bad for you, then it’s good for you. Then cholesterol is bad for you; and then it’s good for you – if it’s the right kind. Or it’s bad for you if you’re the wrong kind of person. It gets complicated, and people tune out, and blame the scientists. Most people don’t pick up on the fact that something that’s good for one person may not be good for another. Worse, they have no idea of how the scientific method works, what statistical significance is, or any of the important qualifiers in any valid scientific study. This is why scientists regularly get beaten up by ax-grinding politicians in debates over scientific issues. Climate change is a classic example of this, where George W. Bush (and others) exploited the reluctance of climatologists to make categorical statements, and instead promoted their biased view that climate change isn’t happening, and therefore there’s no need to do anything about it. That’s an unfortunate part of telling the truth: it puts you at a short-term disadvantage.
However, if it were my job to raise public awareness of what genetically-appropriate food was and meant, I would publicize the hell out of the old saying “One man’s food is another man’s poison.” This is what is known as employing innovative familiarity, much like having digital cameras that go “click”, even though they don’t have a shutter.
We Need New Analytical Tools
Next, I want to mention, more or less in passing, one of the critical issues in research: information overload. Traditional data reduction tools, such as statistical analysis, can’t do the job. They’re too slow, and they don’t manage ambiguous and multivariate analysis very well. One of the reasons why I raise this issue is that there is a range of tools arising that may produce superior results. I want to mention one in particular, although there are others. I know about Genetic Programming (or “GP”) because my brother developed and patented a system for GP, and is using it to perform research into diagnostics and treatments in oncology. And, by the way, he has no interest in working in nutrigenomics, so there’s no point in asking him.
Let me start by defining it: genetic programming is a machine-learning technique where the solution to a proposed problem evolves by reinforcing success. In other words, GP uses the idea of natural selection to discover solutions. Those solutions that work best are combined to discover even better hybrids, much as cross-breeding horses, for example, can create offspring that are faster and more robust than their parents. GP is not an artificial intelligence system. There is no attempt to mimic human reasoning. GP’s advantages are that it:
• solves the problem of too much data;
• integrates large & diverse data sets;
• facilitates the unbiased discovery of key factors, especially where the key factors are unknown;
• ignores factors that turn out to be irrelevant or unimportant; and
• creates human-readable models which can serve as indicators for further research.
Moreover, GP is a non-linear method, which means it can be dramatically faster than conventional analysis techniques, especially for multivariate analysis involving large amounts of data. Indeed, large amounts of data tend to produce more robust results.
The Commercial Prospects Ahead
Finally, let me finish up by talking about the commercial prospects for the future of food. There’s a global gold rush on right now to find ways of making money out of genetic information generally, and nutrigenomics specifically. There are going to be lots of market niches opening up at all stages of food production, processing, and retailing, but as they become proven and mainstream, the global economy will attract more and more competition to any given niche. Of course, higher competition will also drive down prices, which will increase consumer acceptance in the time-honored manner. Hence, it may well be worth sticking to a niche as it goes mainstream – at least until it reaches the level of commodity pricing and competition.
Indeed, if there is a common theme for all companies that are exposed to global competition (and that’s everyone involved in producing a product, and many involved in producing services as well), it is that you will never again live in a status quo world. You will always have to be prepared to abandon one market niche in favor of a newer one, or one that requires more knowledge or that adds more value. Therefore the approach most likely to work is to continue to pioneer – and charge premium prices – for specific sub-populations, harvest the benefits until competition makes the niche less attractive, then move on to other, more tightly defined populations. You will never be able to stand still again.
But let me end on a positive by repeating something I said at the outset: The development of nutrigenomics is, in my view, going to be the most important development in food production and preparation since we started eating cooked. Or, to put it more directly: Ladies and gentlemen, it’s raining soup. Grab a bucket and get out there.
© Copyright, IF Research, March 2009.