The low-carb, high fat diet debate and deviant thinking

The low-carb high fat diet debate: Three videos, and thoughts on polarized views and 'deviant' thinking

One of the hot topics in exercise science and diet is the low-carb, high-fat diet concept, now backed vocally by Prof Tim Noakes in South Africa.  In December, he and Prof Jacques Roussouw debated the dietary guidelines with respect to cholesterol and its impact on heart disease, and a video of that debate can be found below.  The low-carb diet echoes other topical issues in exercise science, perhaps most notably barefoot running, in that it polarizes opinions between two opposing camps.  It then strays into 'rules' and over-simplifications, which are arguably incorrect.  Here's how to 'pole-spot' and embrace complexity, along with the low-carb debate

- Ross Tucker

In the course of a debate on doping and cycling a few years ago, a certain well-known exercise physiologist who had tested and defended Lance Armstrong publicly dismissed Jonathan and I as "newly-minted scientists".  True, of course, since we had both obtained our qualifications within four years of him writing those words.

He intended it as disrespectful at the time, suggesting ours was an opinion not worth listening to because we did not have 300 years of experience (and about as many conflicts of interests, I'd add) behind us.  I always viewed "newness" as a distinct advantage, because it brings with it some aspect of novelty, a new way of approaching an old problem.  That's often lacking in science and in many areas of life (coaches, managers, I'm looking at you!), and as I've evolved from newly minted to (recently?) minted, I've come to recognize that progress usually comes from forcing a novel view.  

Deviant thinking and innovation

I recently spoke to a group of financial consultants about the lessons I have learned about high performance teams from my involvement with sports teams and athletes, and one thing that I tried to re-inforce, in business and in sport, is that progress is the result of so-called deviant thinking.  By "deviant", I mean that person who pushes back against convention, who asks the apparently ridiculous questions and forces others to rethink their positions of comfort.  Deviants make us anxious, but they also drive innovation.

If we are allowed to drift along with the current, we never challenge paradigms.  Jonathan and I were both fortunate that our post-graduate training was overseen by Prof Tim Noakes, who is not newly minted but has retained the capacity to challenge current beliefs.  He is a scientific "deviant", in the most complimentary sense of the word.  In so doing, he has driven a change in perceptions around fluid intake and dehydration during exercise, and also has contributed to our understanding of fatigue and the role the brain plays in performance regulation.  These topics were, respectively, the subject of Jonathan and my PhDs, and so we have inherited this desire to push back against convention, hence the existence of, and many of the approaches and articles on, this website.

The low-carb high-fat diet debate

The latest area of Noakes' interest is diet.  Specifically, he is a vocal proponent against carbohydrates and processed food, arguing for a high fat, low carb diet.  In South Africa, it is impossible to give a presentation on exercise and health without some member of the public asking about Noakes' dietary views and their implications for exercise, weight loss and health.

However, it is not a topic whose specific content and details I am comfortable dealing with.  I am not an endocrinologist, nor a cardiologist, nor a dietician.  I understand the basics, but in the same way that my driver's license does not entitle me to tell Michael Schumacher, Sebastian Vettel or Jimmie Johnson how to drive, I would not presume to educate or correct the experts on diet and cardiology - I might ask them a few pointed questions, of course, and challenge their thinking, but there's a line that I wouldn't cross in terms of dictating to them.  I have not dealt with people struggling to lose weight, and have not encountered the very real, practical challenges they face.  I do not have a lifetime of expertise evaluating research studies on heart disease, though I can appreciate how many 'holes' exist in current thinking.  Nor have I devoted any length of time to evaluating the respective sides of this particular debate.

And so I won't delve into specifics, at least not now.  However, in order to make the debate as widely accessible as possible, which is important, I want to share with you three videos.  They are taken from the University of Cape Town's Centenary Debate, held last year in December, where Prof Tim Noakes and Prof Jacques Roussouw debated various aspects of the high fat diet.  The focus is very much on cholesterol and its links with heart disease.  The videos are long, but worth watching when you have the time.  I'd love your feedback, your thoughts on who "won" the debate and what it means for our understanding.  Those videos are at the bottom of this post.  

Thoughts on scientific concepts and complexity

But first, my view on this whole debate, without delving into the specifics.  My biggest "objection" as it were, is not to the content of the debate, but rather the manner and justification for each side's respective positions.  Below is part of a presentation I gave to the public last year, and in it, I mention two examples of how scientific progress and application to the public can be undermined by the natural, human desire to simplify the message and adopt a polarized view of what are actually very complex concepts.  (Click here if you are reading this in an email)

The first is the 10,000 hour concept for expert performance - a great theory, wonderful to motivate parents and young athletes about the value of training, but a pretty useless theory in practice - in sport, it hardly ever applies.  The second is barefoot running, which has been taken and transformed into a cure for everything without any evidence.  

Polarized science, rules and a wildly swinging pendulum

The result of these kinds of debates is a polarized science, one where the pendulum swings wildly from one extreme to the other.  We go from "Practice is the only thing" to "Genes are the only thing" and back.  Or from "Barefoot running will prevent all injuries" to "Barefoot running is a fast-track plan for physical therapists".  Neither is true as a "rule", though within any population, there will be those who succeed at the extremes, and those who fail.  That of course introduces a huge confirmation bias, because every success story is held up as "proof".  It also leads to cherry-picking, because anything not supporting the pole has to be ignored.  Those who advocate for those polarized positions must recognize that they are pulling everyone to the sides, where they may not belong. 

The same is true for diet and metabolism.  The reality is that we are dealing with complexity in physiology that can't be explained by one theory, and an obesity problem that does not have one solution.  Biological complexity dictates that what works for one will not work for another, and that's what coaches figure out very early with athletes, and dieticians learn empirically with clients.  The idea that shoes are bad is just as wrong as the idea that shoes are essential, because in any population, either could be true for some people.  These kinds of over-simplifications are damaging because they polarize understanding in a way that benefits few, introducing dogma that is then disseminated to the detriment of many.  And that is the point I make in the presentation above.

So, how is this relevant to diet?  Well, the same things I see from the barefoot debate appear to be happening in the dietary debate.  Conventional wisdom is challenged, and rapidly leads to the formation of two opposing camps, whose idealism is so at odds that the poor people in the middle, who are ultimately the "end users" of the information, are caught in a figurative stretching rack, being pulled in opposite directions by 'extremists'.  If it is difficult for experts to agree, then imagine how complex it becomes for those in the middle.

The problems at the poles

Science is never black and white.  It's one of the first lessons I learned, and have relearned weekly since.  Does dehydration impair performance and health? Is fatigue the result of chemicals in the muscle?  Is barefoot running safer?  Does cholesterol lead to heart disease?  There is no such thing as a straight-forward answer to any of these questions, and so a single extreme view is invariably wrong.  

It then becomes more about "how" the message is communicated, and not "what" is being said.  That is, the content of the deviant view almost always has value - the barefoot running concept, for example, may be incredibly helpful to many runners and I would strongly support that everyone take something from it.  On the other hand, there may be people who simply cannot succeed barefoot.   

What then tends to happen is that the polarized camps become almost obnoxious about their view, blaming everything but their view for the obvious failure to succeed 100% of the time.  If you are injured running barefoot, it's your fault, for instance, and its advocates seem to show no awareness that they are making exactly the same mistake as they accuse shoe companies of making before them.  It is that aspect of the debate that is most off-putting, and I find the same true in the carbohydrate debate.  The justification for a given position becomes more and more 'radical', and eventually, it is based on anecdotes, resembling a series of TV infomercials promising "more".  

Polarization also introduces a risk of weak scientific interpretation, and I've seen examples recently where an association study is dismissed as weak and unreliable when it suggests that carbs are important, only for the same type of association studies to be used as "proof" when they support the desired viewpoint a day later.

Ultimately, there is without doubt truth in any deviant view, but there is also a problem with the idea that the scientific pendulum should swing all the way from its current position to an entirely new one.  With respect to the carbohydrate debate, there is no question that Noakes has, like those advocating for it before him, contributed to many success stories and positive changes as a result of diet.  And by opening up the kind of debate you see below, he has potentially created enough "scientific anxiety" that it will stimulate a whole new area of research that will ultimately help advance our understanding of how INDIVIDUALS respond to different macro-nutrients, and hopefully reduce the obesity epidemic we face.

But in all this, and in debates about shoes vs barefoot running, and talent vs training, and methods of training and so on, don't feel compelled to pull the pendulum to the other extreme - remember, that's what you're suggesting others have done wrongly before!  There's no such thing as "we were 100% wrong before".  We just weren't 100% right, and it's the contribution of deviants who help us see that.  But stay away from the poles.

UCT Centenary Debate: the cholesterol debate

Right, so below is the debate.  It's one long video, divided into three chapters:

1. Prof Tim Noakes introduces his theory (35:43)
2. Prof Jacques Roussouw responds (43:21)
3. Questions and discussion with the audience (51:05, but probably the most interesting aspect)

To compliment the video, you may also want to view the presentations (it's not filmed very well, have to say):

View Prof Noakes' presentation
View Prof Roussouw's presentation

Feel free to comment and share your views.  Again, I'm not going into specifics, it just isn't my place and feels disrespectful to people who arguably know more than I do about this, but gladly debate the manner of the debate and the way ideas are communicated!

Ross

Long-term athlete development: Foundations & challenges

Long-term athlete development: Foundations and challenges for coaches, scientists & policy-makers

The Long-Term Athlete Development (LTAD) model is a physiological framework proposed to manage the focus, volume and type of training applied to athletes as they develop through adolescence into adulthood.  There remain a number of question marks against the foundations of LTAD, though it provides a sound framework for sporting development.  It does however introduce a number of practical challenges, and its success requires that coaches recognize the potential barriers and conflicts.  These are described below in a presentation and summary of LTAD.

- Ross Tucker

I'm currently in Dublin, at the invitation of the International Rugby Board, to present at their biennial Coach Education Workshop.  Topics include rugby safety/risk (as my tweet last night illustrated), the professionalization of coaching, Sevens rugby, and the topic for which I was invited, Long-Term Athlete Development (LTAD).

LTAD has become something of a staple for sports federations, coaches and administrators - in South Africa, there is a veritable alphabet of LT_Ds, including Athletes (LTAD), Participants (LTPD), Coaches (LTCD), with touted "Fs" (facilities), "Ms" (mentors)" and "Rs" (for resources) in the pipeline.

In any event, LTAD is an interesting starting point for discussion, and this is not exclusively for rugby coaches.  Of course, my talk was targeted at rugby, but the same conceptual framework has been applied to every sport, and it's worth debate between scientists, coaches and managers.  I take a rather strategic view of this - the specifics have been debated elsewhere, and I think the big picture matters more than the minute details.  It's a question of management strategy as much as it is science.

Below is my presentation.  Obviously, it lacks the narrative of me talking through it, but I hope that it makes at least some sense without it.  Once you've gone through it, I have included some thoughts below, in bullet point form, to sum up the key points regarding LTAD and the challenges facing its implementation.

Here is that presentation (If you are viewing this as an email, click here for the presentation):



Key points:  

Introduction

1. The purpose of any framework, be it LTAD or any other, is to drive the allocation of resources in an effective and efficient manner.  These resources, human, financial or structural, are finite, and the decision must be made about where to invest.  Talent ID for sport is a relatively simple question - it asks "Where is the athlete today, who will represent our country in 15 years' time?"  The simple question however has a complex answer, because we need to find him and then develop that potential talent.  That requires some important decisions about who does what within a sporting system?  How those resources are allocated is the crux of LTAD.
   
   
2. Talent Identification cannot be formulaic.  The biggest oversight or error is to view frameworks such as this as "formulas" for success.  Sporting success is multi-factorial, and too complex to obey a single formula.  As a result, we look in hindsight at what worked and create models to apply with foresight, but the mistake would be to become too prescriptive or literal.  For every rule, there are exceptions, which probably means there is no rule.
   
   The model of Vaeyens et al, presented above, explains the factors associated with sporting excellence.  "Giftedness" or innate abilities, along with chance, are recognized as a significant elements, and catalysts including environment and intra-personal characteristics are crucial.  Talent Identification and Talent Development contribute to optimizing these elements or catalysts.
   
   
3. Within a given sport, there exists a pipeline that takes young athletes to higher performance levels with age.  A number of questions need to be answered - how young and how old? How are resources applied? What role to coaches and competition play?  The volume of the "base" drives the required efficiency in order to achieve the same pinnacle.  The answer to these questions is proposed by the LTAD model of Balyi.  In the presentation, I depict a summary of the excellent Irish Rugby Football Union model, "Six to Six Nations".  A google search will reveal many other similar applications of LTAD for various sports
   
   LTAD: Foundations and concerns
4. LTAD divides the path from a child to a professional into a number of stages.  For each stage, the focus, the role of the coach and the responsibilities of the player vary.  In the Irish Model, and generally true of LTAD, serious competition is delayed until after adolescence.  This has repercussions for where the sport fits in within society, and the management of various stakeholders.  In fact, it is this balance between competition (which society demands) and the LTAD proposed delay in competition that is likely to be the source of most "tension", and thus failure, within the model.  This is explained later
   
   
5. While conceptually sound, it must be borne in mind that LTAD has not been conclusively proven.  Two of the foundations I discuss briefly in the presentation are the "Windows of opportunity" concept, and the 10,000 hour concept.
   
   Re the windows of opportunity, the issue is not so much that they do not exist (though there is some academic debate on this point), but rather the literal or wrongful interpretation of them to lead to neglecting other attributes.  In the presentation, I mention two examples of this - physical literacy and aerobic development.  For a more comprehensive review, see Ford et al.  This is a typical example of applying the concept as though it is a formula - the value is not in being specific, but in understanding principles.
   
   In this regard, there are three core concerns, which I explain in the presentation.  The first is that it can become too prescriptive.  Always remember that science loves averages and "typical" patterns, but not many individuals are average or typical.  As a result, if a coach tries to apply LTAD principles based on the average, there is a danger of "writing off" any young athlete who doesn't adapt, or obey the 'science'.  The second is that it's too literal, as explained.  And the third is that LTAD can become a real burden because of its extended period of responsibility for the coach or sport.
   
   Early exposure, relative age and 10,000 hours
6. It seems quite clear that early exposure is important, but if we select talent too early, we run the risk of making mistakes and voluntarily cutting our player pool down by an enormous amount.  The best illustration of this is the relative age effect, where coaches of young children confuse maturity with ability, and so when picking players for teams, make the error of picking relatively older players.  A large number of potentially great athletes are thus neglected and never receive opportunities reserved for those "lucky" enough to be born at the right time of year!
   
   This is a significant challenge for sports, and may require a rethink about 
1. The age at which we begin to select teams, and;
2. The level of coaching we provide to our best young players compared to those who don't quite make the cut - I'd argue that the best coaches should perhaps be allocated to the second best players at this age
   
   Having said this, there is some evidence that this relative age effect disappears in adults, which is really interesting, and may suggest that once you have early exposure, and once all the physical differences between early and late developers are ironed out, something else predicts long term success in sport.
   
   
7. The 10,000 hour concept for success, popularized by Gladwell in "Outliers" and Syed in "Bounce" has very little merit if applied literally to sporting success.  What it does do is provide a compelling argument that practice helps performance, but this is so obvious it doesn't really need to be said to a group of coaches!
   
   The 10,000 hour concept owes its existence to a study on violinists, by Ericsson, in which he found that the best players have accumulated 10,000 hours.  What he failed to do was any statistical analysis at all, and the result is that he didn't show that some people become best experts with less, and others fail despite doing more than 10,000 hours.  It took a study on chess players to reveal this - the average time taken to become a master is 11,000 hours, but some did it on 3,000 hours of practice, some haven't succeeded despite 25,000.  Those people effectively disprove the theory, and leave us realizing that a lot of practice is required to get good at things, but to commit to a specific number is a myth.  In fact, I'd go so far as to say that for a coach, one of the best methods of talent ID is to look for responsiveness to training.  If an athlete cannot acquire a new skill or adaptation rapidly, they're not going to become elite.
   
   Delayed high volumes of training predict success
8. Studies have shown that success in the CGS sports (sports where performance is measured in centimeters, grams or seconds) is related to a) delayed specialization, and b) delayed high volumes of training.  In other words, athletes who perform higher volumes of training when younger are less likely to become elite.  Those athletes who delay this increase succeed.  There are a number of possible explanations for this - one is burnout, in athletes who do more when younger.  The other is physiological, and this is the one I've explained in the presentation above.
   
   This is again an argument for delaying the identification and training of young athletes until after the physiological changes associated with adolescence are completed.  It is at 16, not 13, that talent ID and development become more effective.
   
   Physiological determinism and fate: the role of physiology/genes
9. In rugby, as in many sports, physiology plays a crucial role.  Even among a very good group of rugby players, who are the best in the country, there is a small but significant difference in stature and mass in those players who go on to become the very best (Springbok players).  In rowing, one of the more amazing findings I have seen in recent years, shows that elite female rowers have testosterone levels 112% higher than sub-elite female rowers.  One interpretation of this is that if you do NOT have testosterone levels in that range, then no amount of training, no LTAD and no development is going to make you elite - physiology determines your fate.  That's not to say that having high testosterone levels ensures success, but it is a crucial requirement.
   
   
10. Applying this to rugby, I look at the data of how many rugby players have played at the highest level in South Africa at the age of 13, the age of 16 and the age of 18.  Turns out that the conversion of good 13 year olds to good 16 year olds is relatively poor - only 31.5%.  From 16 to 18, it's much better - 76% of young players who play at U/16 level also play at U/18 level.
    
    Collectively, what this means is that if you are good enough to play at U/13 level, the chance that you'll make to U/18 level is basically 1 in 4.  Not too good.  If you make it to U/16 level, there's a 3 in 4 chance that you'll get to U/18 level.  Much better.
    
    This has the same significance as the relative age effect and the finding that delayed high volume predicts success - it says that the more you can delay the selection of talent, the more efficient your system.  What it does not do, more profoundly, is tell you the fate of all the players who were NOT selected at those younger age groups.
    
    The competition conflict - LTAD vs society's accepted norm
11. Attitudes to competition provide the greatest barrier to successfully implementing LTAD.  If the competition structure places any priority on winning at the junior level, then it directly conflicts with the fundamental of LTAD, which is to delay the importance of winning until after adolescence   In South Africa, we have a competition-driven system - compete at 13, compete at 16, compete at 18, and the best come through.  It has certainly produced excellent seniors, but may lack efficiency, and possibly, may be detrimental because it 'writes off' a good deal of talent at a young age.
    
    In this kind of competitive model, early maturation is encouraged, and bigger, stronger, faster players are rewarded at a young age.  They may simply be the early developers.  The result is that once differences are ironed out, they no longer possess an advantage, and the system 'pays' for neglecting those players who would've gone on to become equally large, strong and fast, but potentially with other attributes.
    
    Five key challenges
12. Given this reality, there are five key challenges facing LTAD:
1. How do you identify talent without either destroying it or neglecting it?  Talent is destroyed when it is chosen for the wrong reasons.  If you pick players at 13 based on size, speed and strength, you pick a temporary advantage.  But because it is rewarded by the competitive system, it never needs to develop other attributes.  Talent is neglected because late developers often do not receive a look in, and are lost to the sport early because of the way the system has been created.
   
   
2. How do you maintain healthy competition without providing a conflicting message to coaches?  You cannot create and implement LTAD which says "delay competition", and then have annual competitions for 10 or 13 year olds, the results of which are crucial to future success as a player.  That is a mixed message, and the coach will always go with performance.
   
   
3. How does a sport embracing LTAD affect that sport's standing in society?  The reality is that sport is a big deal, even from young ages.  Here in South Africa, high schools look for young children with athletic potential and offer scholarships and potential career paths.  At a young age, good athletes are virtually professional and society has come to accept this as "normal".  Implementing LTAD challenges that, and if the entire environment does not also do the same, then it creates a conflict between one sport and another, and even within a sport.
   
   For example, I work with SA Sevens, and we are looking at driving the specialization of players to become Sevens players from a younger age.  We are not going down to the 10-year olds, but it illustrates that because players themselves are finite, they are the subject of competition.  Imagine rugby implements LTAD and football does not - a good number of young players, perhaps forced by parents, will move towards football.  There is a degree of "security" in early specialization, however wrong that perception may be.
   
   
4. Who are the other stake-holders in LTAD?  It's simply not reasonable to suggest that one sport have an LTAD programme from 5 up to adulthood.  As mentioned, it's unnecessary because you don't need 10,000 hours to begin with, and it's also costly and potentially crippling to place the entire burden on each sport.  Therefore, you recognize that other stakeholders, such as parents and government, also play a crucial role, particularly early on when you actually don't want players to specialize, but rather engage in a number of different sports, learning a range of skills and abilities.  This is perhaps the key concept for LTAD.
   
   
5. How do we change mindsets?  In all of this, it's important to recognize that sporting systems, countries, federations, have a certain inertia.  They are giant, sometimes slow-moving bodies and if you stand in the way, you get flattened.  Therefore, to successfully implement LTAD, you must address the mindsets and begin to 'nudge' them in a different direction.  Failing this, LTAD, or any other similar plan, is nothing more than a fantasy of "best-case", and won't work in the real world.  It will take brave leadership to change the competition structure, for example, and to adopt a no compromise attitude towards youth talent ID and selection, based on current principles.  I doubt many will have the stomach for the fight, but that may be what it takes.
   
   Alternatively, we can accept a 25% success rate as good enough, which is fair.  But that cannot co-exist with excessive competition, as many of the rugby nations here in Dublin have discovered.

Conclusion

A sound concept, LTAD introduces a number of challenges at a system or management strategy level.  It also has some debatable physiological concepts, but the debate on those is perhaps too academic and thus not relevant for coaches.  It should not be taken too literally or prescriptively, but rather regarded as a framework to guide decisions.

The big picture is where it is far more complex.  Certainly, in South Africa, we will have to grapple with whether competition at young ages is the best way to achieve senior success, or whether it is worth the aggravation to change this.  There is no evidence, because there are no long-term prospective studies, that help us 'guess' how elite performance would change as a result of policy changes at the junior level.  However, it seems reasonable to hypothesize that if youth selection is delayed, and if the pool of available talent is kept large for long enough, we will see more viable prospective talents and thus better performance.

Coaching is often referred to a mix of art and science, and LTAD is similar.  There is no single path, and this is a debate likely to extend well into the future.  Feel free to weigh in below.

Ross