Tuesday, May 4, 2010
The Fiber Hypothesis
Today's question is: What if you know that the dietary-saturated-fat-and-cholesterol hypothesis doesn't work very well to explain heart disease, but at the same time you don't want to admit that eating too many refined carbohydrates might be the cause?
Answer: You put the blame on fiber. Or rather on not eating enough fiber.
In 1972, Peter Cleave tried to explain to a U.S. Senate Select Committee that when primitive cultures adopted Western eating patterns, they also began to experience the diseases of Western civilization, including diabetes, heart disease and hypertension. Cleave pointed out that the fat and cholesterol hypothesis of heart disease did not explain this transition, but that the adoption of a diet rich in refined carbohydrates did account for it rather elegantly. The Senators reached the only logical conclusion. They refused to believe Dr. Cleave. Dr. Ancel Keys had so completely won the argument that dietary fat was the cause of heart disease, that any alternative hypothesis had to be rejected out of hand.
The Senators were left with the problem of how to explain the increased incidence of heart disease in transitioning cultures. Enter a famous medical missionary, Denis Burkitt. While working in Uganda, Dr. Burkitt had noticed that Africans produced several times more feces than people in Western countries. He hypothesized that the presence of dietary fiber produced the absence of the diseases of Western civilization. Burkitt collected over 800 anectodal reports showing that primitive peoples ate high fiber foods, while Westernized cultures tended to eat foods that were nutritionally dense and low in bulk, not providing enough volume to allow the intestines to remain healthy. This idea made sense to the granola-eating counterculturalists of the time. More importantly, it did not contradict Ancel Keys' diet-heart hypothesis.
Forty years later, the need for high fiber in the diet has become received wisdom. Some studies show that eating more dietary fiber is associated with lower all-cause mortality, for example Dietary fiber intake in relation to coronary heart disease and all-cause mortality over 40 y: the Zutphen Study. Other studies show no relationship between fiber intake and all-cause mortality, including this one, The long-term effect of dietary advice in men with coronary disease: follow-up of the Diet and Reinfarction trial (DART).
For the purposes of low-carbers, several of the studies on the relationship of glycemic load with the risk of type 2 diabetes may be instructive. The glycemic load is the glycemic index of each food eaten, multiplied by the number of carbohydrate grams of that food eaten, summed for all items consumed during a day. In two studies (here in women and here in men), Salmerón et al. showed that the combination of a high glycemic load and a low cereal fiber intake increased the risk of type 2 diabetes when compared with a low glycemic load and high cereal fiber intake. The figure below is taken from the women's study.Looking at the X axis, at all levels of intake of cereal fiber, the relative risk of diabetes decreases as the glycemic load goes from high to medium to low. On the Z axis, at all levels of glycemic load the relative risk of diabetes decreases as the cereal fiber intake goes from low to medium to high.
Let's say that two low-carbers eat an identical number of carbs. One eats high-glycemic foods and has a high glycemic load. The other eats low-glycemic foods and has a low glycemic load. Both of them will need to release insulin to dispose of the carbs, but the first low-carber will have to release insulin in spikes to counteract the rapid rise of his blood glucose, while the second low-carber will be able to get by with a more gradual release of insulin. As Sullivan et al. have shown here, there is reason to believe that insulin spikes contribute to the development of insulin resistance.
How does fiber fit into the equation? The traditional explanation is that fiber fills us up. However, experiments done with caloric dilution show that when low-calorie foods are subsituted for higher-calorie ones, humans are well able to adjust their consumption of food to maintain their customary caloric intake. Another function of fiber is that it slows the absorption of nutrients from food. In other words, the addition of fiber can be expected to lower the effective glycemic index of high-, medium- and low-glycemic index carbohydrates. The relationship of the total amount of fiber to the total number of carbs to the glycemic index is probably quite complex, which may explain why many of the fiber-health studies do not show clearcut relationships between fiber intake and outcomes such as heart disease, type 2 diabetes, obesity and cancer.
In other words, if low-carb is good and low-glycemic index carb is good, the addition of fiber to all of that might be better. Low-glycemic-index foods like broccoli and nuts do tend to contain more fiber, so perhaps the point is moot for low-carbers who are careful about the type of carbs they consume. In any case, there is good evidence that lowering carbohydrate intake and lowering the glycemic index of those carbs is protective against the diseases of Western civilization. The data on the benefits of fiber intake is not overwhelming, so use your own judgment to decide what level of fiber intake might be right for you.