The recent surge of interest in the Paleo diet, based on the speculation that evolution has equipped humankind to thrive on a diet relatively rich in fat and low in carbohydrate, has added new spice to the long-standing debate over the optimum proportion of fat and carbohydrate in our diets. This debate is of substantial importance for anyone seeking to live a long and healthy life, and is of particular importance for endurance athletes who subject their bodies to the rigours of extensive training and require those long-suffering bodies to function with peak efficiency on race day. There are five related mechanisms by which diet is likely to affect health, longevity, response to training and race performance. These five mechanisms are: the capacity to utilise fat in preference to carbohydrate; minimization of sustained elevation of cortisol; avoidance of chronic inflammation; prevention of insulin resistance; and control of body weight. In the post, I plan to examine the evidence regarding the influence of the proportions of fat and carbohydrate in the diet on these five related mechanisms. In my final post in this series, I will examine the evidence for effects on the ultimate outcomes: race performance, health and longevity.
The metabolic challenges facing the endurance athlete
I will start with a brief review of an issue covered in my most recent post: the metabolic challenges that the runner faces if glycogen becomes depleted in the final stages of a marathon. The body’s paramount goal in these circumstances is ensuring adequate glucose to fuel the activity of the brain. Secretion of the stress hormone, cortisol, increases dramatically, with three immediate consequences: cortisol promotes gluconeogenesis in the liver thereby replenishing glucose; it inhibits the function of the glut4 transporters that transport glucose across the cell membrane into muscle cells and other peripheral tissues; and it promotes beta-oxidation of fatty acids which become the main energy source for muscle. This response averts disaster for the brain, but it is not ideal.
Not only does excessive elevation of cortisol have potential adverse long term effects, but there are immediate undesirable consequences. Reliance of fat as the main fuel for muscle have a dampening effect on power output because fat metabolism requires oxygen making it difficult to exceed the limit achievable aerobically, but in addition, as we saw from an examination of the central role that the Krebs cycle plays in both catabolic and anabolic processes, there are multiple other metabolic consequences. Reduced production of pyruvate from glucose in muscle makes it necessary to utilize glutamine to keep the level of the intermediate metabolites of the Krebs cycle topped up. Muscle is the main source of glutamine for other organs and body systems. Depletion of glutamine in muscle leads to a fall in blood levels of glutamine which has adverse effects in the gut, liver, kidney and immune system. In the gut, glutamine serves a diverse range of essential metabolic functions. In the liver, glutamine is a major source of the oxaloacetate required for gluconeogenesis, when low levels of glycogen limit the generation of oxaloacetate via pyruvate derived from glucose. In the kidney, glutamine is the source of ammonia needed for the excretion of acids. Glutamine also plays crucial catabolic and anabolic roles in the cells of the immune system, and a fall in glutamine will exacerbate the direct adverse effect of cortisol on the immune function. Although the body appears prepared to tolerate some loss of immune function during vigorous exercise, overall, it is undesirable to allow glutamine level to fall too far. Therefore, not only is it crucial to start an endurance event with well-stocked glycogen stores, but one of the key goals of endurance training is developing the capacity to utilise fats in preference to glucose at aerobic paces thereby avoiding a state of serious glycogen depletion in long races.
Effect of nutrition on capacity to metabolise fats during exercise.
Many studies, reviewed by Burke and Hawley, demonstrate that a high fat diet promotes the utilization of fats during exercise. However, on race day, the endurance athlete requires not only a well developed ability to utilise fats, thereby minimising the depletion of glycogen stores, but also needs to start the race with glycogen stores fully topped up. Fortunately, the evidence reviewed by Burke an Hawley indicates that switching from a high fat diet to a high carbohydrate diet in the period immediately before the race does not undermine the ability to utilise fats. As an illustration, Staudacher and colleagues demonstrated that a short term high fat diet (6 days; 69% fat) followed by a high carbohydrate diet in the day preceding exercise produced a 34% enhancement of ability to utilise fats during submaximal cycling in a group of highly trained endurance athletes, whereas 6 days of high carbohydrate intake (70% carbohydrate) followed by a further high carbohydrate day, resulted in a 30% reduction in fat utilization. This has encouraged the hope that a “fat adaptation” strategy in which a high-fat, low-carbohydrate diet is consumed for up to 2 weeks during normal training, followed by high-carbohydrate diet during a brief taper in the few days before a key race, might improve performance.
However despite the consistent evidence that such nutritional periodization can achieve the desired enhancement of fat utilization, there is less clear evidence of enhancement of race performance. This might simply be that may other factors influence performance, so large, well-controlled studies are required to allow any benefit for this nutritional strategy to emerge clearly from the inconsistencies due to other sources of variance. Alternatively, it might be that this nutritional strategy has hidden adverse effects. For example, it is plausible that the nutritional strategy might upset hormonal balance is an adverse manner. I will return to the issue of effects on performance in my next post, but first we need to consider the possible effects of nutrition on hormones such as cortisol and insulin.
Effect of nutrition on sustained cortisol levels
The balance of evidence indicates that a very low intake of carbohydrate and high fat consumption, for either a few days or for longer periods, leads to sustained elevation of cortisol. For example, Langfort and colleagues compared the effects of 3 days of a high fat and protein diet (50% fat, 45% protein and 5% carbohydrates) with three days of a mixed diet. They observed no difference in maximal aerobic capacity, but did observe a significant increase in both adrenaline and cortisol before and after exercise. Furthermore, several studies have shown sustained elevation of cortisol after longer periods of high fat diet. For example, in a comparison of three different diets, each administered for 4 weeks to overweight young adults, Ebbeling and colleagues found that twenty-four hour urinary cortisol excretion was highest with the low-carbohydrate, high fat diet (10% from carbohydrate, 60% from fat, and 30% from protein.) Similar effects are seen with more moderate amounts of fat. For example, in a study of runners, Venkatraman and colleagues observed greater pre-test cortisol after 4 weeks at 40% fat compared with 15% fat, but in other respects, the outcome tended to be more favourable, including greater time to exhaustion in the 40% fat group.
The type of fat might matter: in a study of Spanish women, García-Prieto and colleagues found that high saturated fat intake was associated with an unfavourable loss of the normal daily variability in cortisol levels while women who dietary pattern was closer to the Mediterranean diet, with high consumption of monounsaturated fatty acids, showed healthy regulation of cortisol levels. However, it perhaps important to emphasize at this point that when it comes to other long term health outcomes (which we will examine in the next post) there is relatively little evidence that saturated fats are especially harmful. There is little basis for the long-standing demonization of saturated fats in comparison with unsaturated fats. The recent pressure by the UK government on the food industry to reduce saturated fat content of foods is scarcely justified.
Nonetheless, the association between a high proportion of dietary fat and sustained elevation of cortisol is a potential concern. Epidemiological studies demonstrate that sustained high cortisol levels may promote adiposity, insulin resistance, and cardiovascular disease. For example, in a 6-year prospective, population-based study of older adults, individuals in the highest third of 24-hour cortisol excretion had a 5-fold increased risk of cardiovascular mortality, compared with the lowest third.
One of the important mechanisms in adverse long term cardiovascular outcome is insulin resistance, the cardinal feature of type 2 diabetes. The claims regarding the relative harmfulness of fats and carbohydrates in regard to insulin resistance remain a source of controversy. Consistent with the evidence that high blood and tissue levels of fatty acids are associated with insulin resistance a substantial body of historical evidence indicates that high fat diet impairs glucose tolerance. On the other hand, ingestion of carbohydrate leads to increased levels of blood glucose which triggers insulin release, which in turn can result in insulin resistance. It is likely that the answer is not to be found simply in the proportion of energy derived from carbohydrate or fat, but rather in the type of carbohydrate or fat. In the case of carbohydrates, it is likely that high glycaemic index (GI) foods promote insulin resistance. Brand-Miller and colleagues demonstrated that in lean young adults, a meal with a high glycemic load (the mathematical product of the amount of carbohydrate by the glycemic index of the carbohydrate-containing foods) result in higher insulin concentration, than a meal with similar total calories but low glycemic load. At least in individuals at genetic risk, high insulin secretion promotes insulin resistance. Consistent with this evidence suggesting that a diet based on low glycaemic load might reduce insulin resistance in those at risk, Barnard and colleagues demonstrated that a low fat vegan diet with a high proportion of low GI carbohydrates improved the control of blood glucose in individuals with type 2 diabetes more effectively than a low carbohydrate diet.
A large body of evidence, reviewed by Grimble and colleagues, reveals an association between insulin resistance and chronic inflammation. Grimble concludes that the evidence regarding the direction of the causal relationship favours chronic inflammation as a trigger for chronic insulin insensitivity.
Endurance athletes are at particular risk of chronic inflammation, in part on account of the repeated trauma to muscle and other connective tissues associated with training, making the effect of nutrition on inflammation a crucial issue. The issue is complex. On the one hand, as discussed in my recent post on inflammation, a high carbohydrate load can promote inflammation due to the release of the pro-inflammatory fatty acid arachidonic acid in association with insulin secretion from the pancreas. Furthermore, some carbohydrates, especially cereals containing gluten, can impair the lining of the gut, leading to chronic inflammation. However, a high fat diet also carries risk. Omega-6 fatty acids are pro-inflammatory, making it important to have a good balance of omega-3 and omega-6 fats in the diet, yet the typical Western diet is much richer in omega-6 fats.
Thus, both a high carbohydrate diet and a high fat carry risk of inflammation, with evidence suggesting that identifiable components of these diets account for much of the risk. High GI carbohydrates and a -6 to omega-3 fats appear to generate the greatest risk. The traditional Mediterranean diet, containing a moderately high level of fat with near equal proportions of omega-3 ad omega-6 fats; and vegetables with a relatively low glycaemic index, appears to offer a near optimum combination. In a comprehensive review association of dietary patterns with inflammation and the metabolic syndrome (whose key feature is insulin resistance), Ahluwalia and colleagues concluded that healthy diets such as the Mediterranean diet can reduce inflammation and the metabolic syndrome.
While control of body weight is one of the major preoccupations of dieting non-athletes, it is not usually the main preoccupation among endurance athletes simply because endurance training itself promotes weight loss. Nonetheless, even a modest excess of weight has serious implications for endurance race performance, because the energy required to accelerate the body to compensate for the inevitable braking during every stride, and to elevate the centre of mass in order to become airborne, is proportional to body mass. Therefore, the weight of any body tissue that is not performing a useful purpose is a handicap. However, the issue of what tissues perform a useful purpose for the endurance athlete is not entirely straightforward. Muscle that does not contribute to propulsion might be a handicap, while at least some fat is required to sustain balanced hormonal function. The ideal weight for endurance athletes is likely to vary between individuals, but observation of elite athletes suggests it is likely to correspond to a body mass index range between 20 and 23. Alternatively, since excess fat is likely to be a greater handicap than excess muscle, a body fat percentage in the range 5-11 percent for males and a somewhat higher proportion for females, might be a more relevant guide.
While an excess of the ratio of total calories consumed to total calories expended is an important factor in determining the likelihood of weight gain, there has been much debate about the relative merits of low fat or low carbohydrate diets. A recent meta-analysis of 23 trials including almost three thousand participants concluded that both types of diet improved weight and metabolic risk factors, with no significant differences between the two in the reductions in body weight or waist circumference. Nonetheless there were slight but significant differences in some of the metabolic risk factors, with low carbohydrate diets producing a potently healthier increase in high density lipoprotein cholesterol and reduction in triglycerides, but a lesser reduction in potentially harmful low density lipoprotein cholesterol. The authors concluded that low-carbohydrate diets are at least as effective as low-fat diets at reducing weight and improving metabolic risk factors.
Nutrition does indeed have an appreciable impact on the five metabolic mechanisms that are likely to influence both endurance performance and long term health. There is unequivocal evidence that a high fat diet produces an increase in the utilization of fats in preference to carbohydrate, which is potentially beneficial for endurance performance. However, there is also evidence that consumption of a high fat diet over a period ranging from a few days to 4 weeks, results in a sustained increase on cortisol, which is potentially harmful in the medium and long term.
Nutrition also plays an important role in insulin resistance and inflammation. For these two issues, the type of fat or carbohydrate appears to be especially important. High GI carbohydrates and high total glycaemic load promote inflammation and insulin resistance. However high levels of fat in blood and in body tissues are associated with insulin resistance, while omega-6 fats are pro-inflammatory. With regard to weight control, either low fat or low carbohydrate diets can be effective.
Overall, these observations do not provide any simple answer to the question of the optimum proportion of fat to carbohydrate, but do suggest that both fats and carbohydrates can carry risks. It is noteworthy that much of the evidence demonstrating adverse effects is based on studies in which there was an abrupt change to a high proportion of either fat or carbohydrate. In the next post, we will examine the evidence regarding the influence of proportion of fats and carbohydrates on endurance performance and on long term health, before finally drawing practical conclusions based on a synthesis of the evidence.