All doctors (and probably most non-doctors) have heard of Burkitt’s lymphoma, a type of cancer found primarily in children living in malaria-endemic areas in Africa. Denis Burkitt was the first person to describe the disease, and also the first person to propose that there was an environmental cause (now known to be simultaneous infection by both malaria and Epstein-Barr virus).
Most doctors probably don’t know that Denis Burkitt is also almost single-handedly responsible for the now widespread belief that dietary fibre is an important part of a healthy diet. Interestingly, Burkitt developed his ideas about dietary fibre after corresponding with a less well known doctor called Thomas Cleave (who hasn’t been allowed to give his name to any diseases).
Cleave was interested in the connection between diet and disease, and had noticed how the transition from a traditional diet to a modern diet, rich in refined carbohydrates, was associated with a massive increase in a large number of diseases, including cancer, diabetes, heart disease, and obesity. He even wrote a book on the subject. Burkitt was introduced to Cleave in the late 1960’s by epidemiologist Richard Doll (himself famous for discovering that smoking causes lung cancer).
Burkitt was deeply affected by Cleave’s ideas, and in particular his conception that all the “diseases of civilization” had a single underlying cause, but he took the data and went off in a different direction with it. While Cleave believed that it was the refined carbohydrates in the modern diet that were causing harm, Burkitt came to believe instead that it was the absence of dietary fibre that was responsible. Since refined carbohydrates are by definition low in dietary fibre, the two things track together perfectly, and it becomes almost impossible to say from observational data which is the causative factor and which is the confounder.
There was a big fly in the ointment of Burkitt’s hypothesis from the start, however, and that was the fact that the Maasai tribespeople in Kenya and Tanzania, who lived on a diet consisting almost entirely of meat, milk, and blood, showed none of the diseases of modern civilization, even though they had virtually no fibre in their diet. Burkitt, who spent much of his career in Africa, was well aware of this fact, but seemingly chose to ignore it because it didn’t fit his hypothesis.
The world of nutrition was at this time (the late 1970’s) focusing increasingly on dietary fat as the cause of modern diseases (based on atrociously low quality evidence and forceful lobbying by diet-heart hypothesis originator Ancel Keys), and Cleave’s hypothesis was inconvenient, because telling people to cut down on both fats and carbohydrates wouldn’t work – people had to eat something. Burkitt’s fibre hypothesis could however be made to fit together with the diet-heart hypothesis without too much trouble. The two were thus wedded and came to dominate dietary advice for the next couple of decades.
That is how breakfast cereals came to be considered a health food, and why we’ve all been told to increase our intake of dietary fibre. Anyway, it’s now a couple of decades later. One would think that by now there would be plenty of data from actual randomized trials to tell us whether we should be eating more dietary fibre or not. Unfortunately we’re still to a large extent stuck with crappy and confounder-riddled observational studies that cannot separate the presence of refined carbohydrates from the absence of dietary fibre, and that therefore cannot actually say anything about what causes what.
The Cochrane Collaboration tried to do a systematic review and meta-analysis in 2016 to look at the state of the evidence when it comes to the ability of dietary fibre to prevent cardiovascular disease. They found 23 randomized controlled trials with a total of only 1,513 participants. Most of the studies ran for only 12 weeks (the minimum length of time the reviewers had set for inclusion in the review, since short term effects are meaningless from a public health standpoint), and the longest ran for only six months. That’s why I say they tried – there simply isn’t enough data to draw any firm conclusions about what effect dietary fibre has on cardiovascular disease risk. 1,513 people followed for a few months provides far too little data to be able to say anything certain.
With that being the case, the reviewers decided to look at surrogate markers for cardiovascular risk instead of looking at hard outcomes like heart attacks and deaths. In other words, they looked at blood lipids and blood pressure. They included trials of both dietary interventions (i.e. that had people eat more food rich in dietary fibre) and trials of fibre containing supplements. I prefer the fibre supplement trials, since the dietary intervention trials have the same confounding issue that the observational studies have – i.e. that an increase in intake of foods rich in dietary fibre virtually always also means a decrease in intake of foods rich in refined carbohydrates. The supplement trials are also at lower risk of bias, since they can be placebo-controlled and thereby blinded. But I’ll report what the diet modification studies showed too, for the sake of completeness. In total, there were fifteen fibre supplement trials and eight diet modification trials.
Let’s get to the results. The fibre supplements were associated with a marginal 0.04 mmol/L reduction in LDL (“bad cholesterol” – yes I know that description is dumb and technically incorrect) that was just about statistically significant. The effect was similar (0.03 mmol/L) in the diet modification trials. For an average person with an LDL of 4 mmol/L this would represent a reduction of just 1%, in other words nowhere near enough to be expected to have any noticeable impact on cardiovascular disease risk.
And LDL is anyway a poor predictor of risk of cardiovascular disease. A much better predictor is triglycerides. The fibre supplements were associated with a 0,03 mmol/L reduction in triglycerides (not statistically significant), while diet modification was associated with a 0,02 mmol/L increase in triglycerides (also not statistically significant). Considering that the average person has a triglyceride level of around 2 mmol/L, this amounts to no noticeable effect on triglycerides whatsoever (and when the supplement and diet modification trials were meta-analyzed together, the difference between high-fibre and low-fibre was exactly zero mmol/L).
When analyzing blood lipids, the final piece of the puzzle is HDL (“good cholesterol”). While you want LDL and triglycerides to be low, you want HDL to be high. Unfortunately, fibre was associated with a reduction in HDL – 0,04 mmol/L in the supplement trials and 0,03 mmol/L in the diet modification trials. Fortunately, that difference was again so small as to be utterly inconsequential.
Ok, so the overall picture is that fibre doesn’t meaningfully impact blood lipids one way or the other. What about blood pressure?
Both the supplement studies and the diet modification studies reported a 2 mmHg reduction in systolic blod pressure and a 2 mmHg reduction in diastolic blood pressure. Considering that the average person has a systolic blood pressure of around 130 and a diastolic blood pressure of 80, this again amounts to such a marginal difference that it’s not going to have any noticeable impact whatsoever on an individual’s risk of cardiovascular disease (in other words, the story when it comes to fibre is the same as the story when it comes to salt – the impact of diet change is far too small to have any noticeable impact on an individual’s heart disease risk).
So the evidence to support the notion that fibre is “heart healthy” is weak. Kellogg’s should definitely stop marketing bran flakes as a “heart healthy” food.
Fifty years ago, when Denis Burkitt started researching fibre and it’s possible health benefits, the first thing he focused on was it’s potential to prevent colon cancer. The notion that fibre might have a role in preventing colon cancer makes a lot more intuitive sense than the notion that it might prevent heart disease, for the simple reason that fibre doesn’t move from the intestine in to the body proper (technically the contents of the intestine are considered to be outside the body), but fibre does have various effects on the intestine, not least of which is the fact that it interacts with the bacteria that reside in the colon.
Colon cancer is thus a good test case for the many claims made about dietary fibre’s health benefits. The Cochrane Collaboration carried out a systematic review in 2017 that looked at the ability of dietary fibre to prevent colon cancer. The review included randomized trials of people who had had polyps removed and that then followed them over time to see if they developed new polyps and/or colon cancer.
Colon cancer usually progresses in an orderly fashion, beginning as a polyp that over time (if you’re unlucky) progresses to full blown cancer, so studies that want to determine risk of colon cancer progression can usually get away with looking at whether new polyps develop rather than having to wait and see whether the participants develop cancer (which saves time and allows for shorter, smaller studies).
Five trials were identified, with a total of 4,798 participants. The average age of the participants was around 60 years at the start of the studies, and they were followed for two to four years. As with the heart disease studies, there was quite a bit of variation in terms of the intervention used, with four trials providing dietary supplements while one attempted diet modification. The trials were for the most part able to at least double people’s fibre intake.
Ok, let’s take a look at the results.
Over the course of follow-up, participants in the high fibre group were 4% more likely to develop at least one new polyp in their colon than participants in the control group, although the difference wasn’t statistically significant. Hmm. Odd. We’d have expected at least some signal of benefit. The trend definitely shouldn’t be towards harm.
But polyps are really just a surrogate marker, like blood pressure is when it comes to heart disease. What we really want to know is whether the high fibre diet protects against colon cancer. Two of the studies were big enough to provide data on this more meaningful outcome.
In these two studies, participants in the high fibre group were 170% more likely to develop colon cancer than participants in the control group. Yes, more. Not less. That difference was statistically significant. So… that’s strange. Admittedly, this result is based on a few thousand participants followed for a few years. It could be wrong. But what it means is that the highest quality evidence currently available suggests that a high fibre diet might actually increase your risk of colon cancer, not decrease it.
How do we square this finding with the observational data that shows a decrease in colon cancer risk with a high fibre diet? As mentioned, the observational data is heavily affected by confounding variables, not least of which is the fact that a high fibre diet usually means a diet low in refined carbohydrates. These results support the notion that Burkitt was wrong and Cleave was right – that the harms associated with a diet rich in refined carbohydrates are due to the presence of refined carbohydrates, not due to the absence of fibre.
So, what can we conclude from all this? Does fibre prevent the so-called “diseases of civilization”?
Well, maybe. That is certainly the impression you would get if you look at the observational data, which find a correlation between a low fibre diet and pretty much any chronic disease you care to look at. The randomized trials that have been done have however for the most part failed to show evidence of a benefit of increasing intake of dietary fibre.