Monday, October 6, 2008

Practical Low-Carb Diets

There’s a lot of hype on all sides (and there are many more than two) of the low-carb diet wars. Unfortunately, many of the arguments and recommendations are based on bad science or even no science at all. Too many dietary recommendations have been based on “expert” pronouncements rather than sound scientific data and rigorously tested hypotheses.

We’ve been reading about nutrition and making our own dietary decisions based on that reading and our personal experience for about nine months now—hardly long enough to become real experts, but long enough to have developed our own best guesses as to what seems to be right, at least for us.

So here is “David and Cynthia’s Guide to Low-Carb Eating.” Basic principle: cut back drastically on the simple starches and sugars that have become ubiquitous in first world diets (of all ethnicities and nationalities). Period. End of story. Beyond that, the details are secondary and serve mostly to provide guidance as to how to achieve that goal. Some details also have secondary health effects that are important if still secondary.

First, of course, you need to recognize what the major dietary sources of simple starches and sugars are. Sugar would seem to be obvious, but there is a large misinformation campaign out there on the part of advocates of various sugar substitutes that can trip up the unwary. A simple rule of thumb is that if it tastes sweet and isn’t a non-nutritive sweetener (sucralose, saccharin, aspartame, etc.), it’s sugar and needs to be minimized in your diet. Calling it “evaporated cane juice” is just a silly marketing ploy. Honey, rice syrup, maple syrup, etc. may have additional valuable trace nutrients, but the macronutrient is still sugar. Even maltodextrin, which is a long-chain sweet-tasting molecule, that gets classified as a starch instead of a sugar, is still readily broken down by the body to simple sugars (that’s why it is added to sports drinks and “gels”). Don’t be fooled by “100% juice” products and similar products with added fruit juice that use apple, pear, or white grape juices to provide sweetness. These are just alternative sources of sugar; apple and pear are particular bad in that they contain a high percentage of fructose. If you must sweeten your food, use whichever of the artificial sweeteners that you tolerate the best. We don’t really like any of those that we have ready access to and generally opt to simply make do with a lot less sweet taste in most of what we eat, but we do use Splenda (maltodextrin and sucralose) when we feel we need a sweetener. (There are at least a couple of potentially better non-nutritive sweeteners with less objectionable taste and aftertaste: Stevia-based products like Truvia and Acesulfame-Potassium-based products like Sweet-One. These sound promising in written descriptions, but we haven’t yet tried them and don’t presently know where to buy them affordably in bulk. Ordinary table sugar and high-fructose corn syrup remain the sweeteners that are commonly sold at much the lowest net price [price per unit of sweetness].) You also need to limit your intake of fruit, especially juices and dried fruits which tend to encourage large servings. All fruits contain significant amounts of sugar in addition to all the nominally healthy micronutrients (vitamins and minerals) that they are often advertised to contain.

Simple starches include starches from grains and tubers. Most ubiquitous in the American diet is wheat starch (flour, bread, etc., etc.). Rice, oats, barley, quinoa, rye, and other grains aren’t much better (despite the advertised heart-healthy characteristics of oats, for example). Potatoes are the most common tuber with a lot of simple starch. These starches are quickly broken down by the body into simple sugars.

Other foods also contain carbohydrates. Many vegetables contain significant amounts of carbohydrates as either starches or sugars or both. Root vegetables such as carrots and legumes such as peas and beans tend to contain more carbohydrates than leafy green vegetables (spinach, kale, lettuce, etc.) and cruciferous vegetables (broccoli, cauliflower). However, for the most part, you are much less likely to overeat starches from these vegetables, and you get plenty of other good nutrients. On balance, we don’t think you need to particularly avoid the higher-carb vegetables, although some recommendations such as those for the “induction phase” of the Atkins diet, have you minimize their consumption as well.

Dairy products contain carbohydrates too. Lactose is a disaccharide of glucose and galactose, and a typical 8 oz serving of milk or yogurt contains 9 to 17 g sugars (depending on how much dry milk was added or how much fermentation occurred). Hard cheeses generally contain much less carbohydrates. Casein is reputed to induce insulin release, which of course defeats the purpose of the low carb diet: to minimize blood insulin as well as glucose, so that the fat cells are releasing fat for use as fuel, rather than sequestering it safely away for long term storage (making us fatter). But that doesn’t stop us from using dairy products which also contain a lot of protein and fat; we aren’t interested achieving zero carb intake or in seeing how fast we can lose weight, but rather in eating a diet that we can sustain indefinitely to reach our target weights and maintain it, without giving up in screaming frustration and deprivation. We like our dairy products! Curiously, the “paleo diet,” which is a fairly low carbohydrate diet, proscribes dairy products but encourages consumption of a lot of fresh fruit. We remain skeptical of such a recommendation unless you happen to be lactose intolerant.

You probably shouldn’t try to cut out all carbohydrates. How low you need to go to be practicing a “low-carb” diet is something that is not well understood. Our own feeling (read “guess”) is that there is no magic threshold, and the answer may vary widely from individual to individual. Certain individuals seem to be able to thrive on a high-carbohydrate diet for a lifetime, while others clearly do much better on much lower levels of carbohydrate consumption. It probably depends on both your genetics and on how active a lifestyle you lead. For us, cutting carbohydrate consumption to under about 20% of calories seems to work well. Some studies have shown significant benefits of lesser reductions. Others advocate more extreme reductions in carbohydrates. If your goal is to get under 20%, then you don’t need to worry too much about starches in vegetables, and it’s even OK to “cheat” a bit. Go ahead and eat some sweet dessert occasionally or even eat that piece of pizza at the party where no low-carb alternatives are offered. Just try to keep the averages down.

The Atkins organization recommends starting out a low-carb diet with an ultra-low-carb induction phase (less than 20 g per day) for a couple of weeks followed by a low-carb phase which is then further relaxed after target weight levels are reached. We don’t see any clear evidence that such a phased approach makes any particular sense or has any particular benefit other than shock value and initial rapid weight loss to keep motivation high. It may be true that the abrupt change induces fat adaptations more quickly, but we know from studies that it takes at least 2 weeks to biochemically adapt (see Phinney (2004)), and a more gradual adaptation may be less stressful for some people. It seems simpler, and probably just as effective, to simply change to a new long-term diet and stick with it. Too many people treat the Atkins diet as a temporary weight loss diet that can be abandoned after the weight loss is achieved. These people regain much or all of the weight lost. It’s not clear that repeated Atkins inductions are as effective either.

A couple of further remarks on types of carbohydrates are worth making. First, we reiterate the important difference between fructose metabolism and glucose metabolism (see our earlier blog entry on fruit). While fructose produces a lower glycemic response, it’s probably a lot worse than glucose from a weight management point of view. Most sweet foods, including most fruits, actually have about equal amounts of fructose and glucose, so there’s not a lot you can do to minimize fructose in favor of glucose, but if you do have a choice, don’t go for high-fructose foods.

Fructose is metabolized by the liver into triglycerides which are then easily stored as fat. The path from glucose to fat is at least a little more convoluted, but it’s clear that triglycerides track carb intake, a fact that is shocking to most people since we’ve been taught that fat intake is what drives triglyceride levels high. However, this is not so: ingested fat is transported in chylomicrons and is preferentially absorbed by the body tissues, and so disappears from the blood quickly (within an hour typically). Triglycerides, in contrast, are produced by the liver in response to fructose and carb ingestion and released into the circulation for hours after each carbohydrate containing meal. Hence high carb meals result in all-day (and all-night) elevated triglycerides. That’s why people on low carb diets have very low triglyceride levels compared with people eating “normal” high carb diets, despite proportionately higher fat intakes (see, for example, Table 3 of Gardner et al., 2007 and Parks and Hellerstein 2000).

Second, certain starches have come to be known as “resistant” starches that are metabolized more slowly and generate less of a glycemic response (blood glucose spike). For the most part, these are the vegetable starches, and perhaps most importantly, the legumes, in particular. This provides a justification for the standard practice in low-carb cooking of replacing grain flours with legume flours such as soy flour or chickpea flour. These bean flours have less carbohydrate than wheat flour, but still have significant amounts. The fact that the starches in beans are digested more slowly seems to make them less prone to producing a blood glucose spike and less prone to being converted to stored fat. Speaking from personal experience, this substitution works quite well, at least from a nutritional point of view. You can do a one-for-one substitution of soy flour for wheat flour in many recipes. Those foods (like bread) which depend on wheat gluten to provide mechanical strength (elasticity) to a dough don’t work too well with a 100% substitution. In these cases, we use part soy flour and part gluten flour (wheat gluten separated from the wheat flour—use about ¼ to ⅓ wheat gluten by volume). Wheat gluten is mostly protein and is safe to use in low-carb cooking (unless, of course, you happen to be sensitive or allergic to it). You can also effectively substitute fiber such as wheat or oat bran and almond meal or other grated nuts for part of the flour.

What about “whole” grains and fiber? Don’t we need a lot of fiber in a healthy diet? Aren’t whole grains good for you? Not as far as we can tell! And we have no personal bias against whole grains. We like to eat them. Our current best hypotheses regarding whole grains and fibers:
1. To the extent that fiber is good for you it is in the context of high-carbohydrate diets only. If nothing else, adding fiber tends to fill you up with more non-nutritive filler, and you eat less simple starch.
2. Fiber probably also promotes intestinal health in the presence of a large carbohydrate load in the gut. Without the large carbohydrate load, it is less clear that you actually need a lot of fiber. Populations that eat no carbohydrates manage just fine with little or no fiber.
3. The alleged binding of starch to fiber to slow starch metabolism seems to be a myth. The measured glycemic response to whole grains is identical to that of equivalent amounts of refined grains. Perhaps if you swallowed genuinely “whole” grains that are still fully encapsulated in fiber (don’t chew!), you might slow the starch metabolism, but otherwise the starch you do consume is still processed quickly.
4. You may get more trace nutrients from whole grain than from refined grain, but there are better sources for those trace nutrients.
5. The same goes for fruit! Fruit contains a lot of fiber and significant trace nutrients, but it comes with too large a dose of sugar.
6. Fiber is not completely non-nutritive. It is at least partially digested by bacteria in the intestines, and seems to lead to production of short chain fatty acids, which are absorbed to varying extent. What that means nutritionally, I’m not sure. There’s probably not enough net usable calories in fiber to be of much significance. On the other hand, the short chain fatty acids are elongated in the liver to increase the net triglycerides.
7. If you want to eat fiber, go ahead! Focus on vegetable sources which come with a lot higher doses of trace nutrients anyway. It’s probably OK to add wheat or oat bran to your baked goods, but do it because you like it, not because it makes you feel virtuous. But be aware that sources of phytic acid such as grains are known to reduce absorption of minerals and may contribute to osteoporesis and other deficiencies.
8. There aren’t any significant nutritional differences between soluble and insoluble fiber that we’ve been able to identify.
9. When you read nutritional labels, “total carbohydrates” generally include any fiber in the food. It’s probably OK to subtract out the fiber as being largely non-nutritive. But beware of claims of “zero net carbs” or similar for foods that obviously contain simple starch. You can only subtract the fiber content from the total carbohydrate content to the extent that the total includes the fiber in the first place. You can’t add fiber to offset the presence of simple starches.

What about fuel for exercise? Don’t athletes need carbs to provide short-term fuel for peak performance? Probably not! This one is fairly complicated to sort out in detail, though. If you are accustomed to a high-carbohydrate diet and switch, you will likely feel that you can’t get peak performance or even satisfactory performance from your muscles for some period of time while the body’s endocrine system and metabolism readjusts to use the fuels you do provide. For instance, there is evidence that a high fat diet increases mitochondrial biogenesis (through activation of PPARdelta) and induces muscle type switching from fast twitch to slow twitch (from glycolysis to fat burning), and importantly, may induce resistance to obesity. This adaptation period of time may last for 2–6 weeks depending on your particular circumstances, but you will get through it and eventually find that you can switch over to fat burning. The evidence supports low carb diets for endurance sports at least, as fat burning enzymes and mitochondria are increased resulting in carb sparing for when you really need it. Is “carb loading” before competition useful? Probably not, again! Making sure your muscles are fully loaded with glycogen may be valuable for maximal anaerobic efforts and sprints, and possibly for longer events, but do it by tapering your training, not by carb loading. What about right before or during the event (especially for events of more than 2–3 hours duration)? This one is less clear. It’s possible that you can achieve a personal best performance with the aid of some extra quick-release fuel from carbohydrates, but even that could be a myth. While we find that, now that we are adapted to low-carbohydrate-fueled exercise, we can go longer without any refueling, but we still eventually need to refuel. We still tend to use carbohydrates, because that’s what’s usually available. But we do try to avoid or at least minimize the simple sugars and make sure we get some protein and fat, too. We haven’t yet tried to maintain a high level of exercise over periods of more than four hours using mostly fat or fat and protein as fuel. Right now, we’re inclined to think it will work just fine, although it may be necessary to start refueling a little sooner to compensate for slower digestion. Post-event nutrition is important, too, but the key need then appears to be added protein to speed muscle repair and recovery, not carbs.

So then, if you’re not eating carbs, what do you eat? The short answer is fat! While you may not have been eating enough protein on your high-carb diet, it’s pretty hard to replace carbs with only lean protein, and can make you sick (the liver can’t dispose of that much ammonia). As a practical matter, a low-carb diet is almost always a high-fat diet. Anyone who tries to have it both ways by continuing to recommend restricting fat consumption while supporting reduced carbs is deluding themselves. It’s essentially a prescription for extreme total calorie restriction, and no one should maintain severe calorie restriction for more than very short periods (and probably not even that).

What kind of fat should you eat? That’s a much harder question, and we think it will be a few years before we have a good well-researched answer. From a simple macronutritional point of view, it appears that it probably doesn’t matter too much. It’s much more important to your health that you replace calories from simple sugars and starches with calories from fat than that you carefully choose which fats to eat. That’s the curious thing about how the popular press and the attention of the public at large focus on nutritional issues. Millions of people will start taking the latest supplement based on flimsy evidence that it slightly improved some condition or other in a handful of test subjects while completely ignoring the basic macronutrient imbalance in their diets. We will take the micronutrient research a lot more seriously when we finally get the macronutrient recommendations properly understood and straightened out.

That said, it is worth commenting on what some of the major issues are with regard to fat choices. Most obvious, of course, is the saturated fat controversy. As nearly as we can tell, there is no good scientific evidence that saturated fat is bad for you despite decades of pronouncements that it is. (This subject is too big and too controversial for one or two references; perhaps we'll write more about it in the future.) Higher consumption of saturated fat may be correlated with elevated blood lipids (particularly HDL, but that's "good" right?), but those elevated blood lipids are not well-correlated with health problems such as coronary artery disease. The assumption that saturated fat is bad is so ingrained in the culture and training of our “experts,” that one must read all related studies very critically, and understand that in the presence of excess carbohydrate, triglycerides will be elevated in most people, and will be formed primarily of saturated fat. You often find that authors are struggling to reconcile apparently contradictory results with the assumed “truth” that saturated fat is bad. Looking critically at the actual data often shows that, if there is any clear difference at all, the test subject consuming more saturated fat were better off on whatever measure was being monitored. So go ahead and enjoy your butter and meat fats, but limit the carbs when you do. Use lard if you like.

For a while we were taught that the more unsaturated a fat was, the better. Then we learned that omega-6 polyunsaturated fats are bad and so maybe monounsaturated fats were as good or better. Hydrogenated vegetable fats were initially considered a “healthy” substitute for the evil lard and butter, but were later found to be worse. “Trans-fats” have recently become taboo. More recently, there is concern that polyunsaturated fats are easily oxidized and implicated in inflammatory atherosclerosis processes ("oxidized LDL"). We are now focusing on the relative amounts of omega-3 and omega-6 fatty acids in our favorite fats and our overall diets. So what’s a high-fat eater to do? In the absence of good science, we still have to eat! Here’s our current working recommendations and hypotheses (aka best guesses):

1. Go ahead and eat as much saturated fat as you like. It won't become oxidized and will increase HDL levels.
2. Avoid trans-fats and hydrogenated vegetable fats.
3. We tend to favor olive oil, then canola oil, then nut and seed oils in our cooking right now, but we’re open to new data that might shift the emphasis around. We've been tempted to try the rendered goose fat available in our local market, but haven't yet. We’re likely to choose an oil based on its performance or taste characteristics as much as anything and to favor cheaper oils when performance or taste don’t dictate a clear choice.
4. Omega-3-rich fats and oils are probably desirable when and where you can get them. Wild oily fish such as salmon and mackerel are probably the best sources. (Farmed fish may not be, depending on the feed that is used.) If you want to go with supplements, use fish oil or cod liver oil by preference over flax seed. Flax seed and flax seed meal or oil are probably the best readily available vegetable sources of you prefer them or want to use them in addition, but it’s not clear how much of the omega-3 fats in flax are even converted in our bodies to the desirable forms.
5. “Free-range” and “grass-fed” meat probably has “better quality” fat than the factory or feed-lot (corn-fed) equivalent. You may also find that it tastes better. It’s also typically a lot more expensive if you’re not raising your own. If your budget allows, and your local stores offer it, by all means go for it, but don’t worry too much if you find yourself eating mostly the cheap stuff.

Getting the macronutrients generally right is still the most important thing to do. Since fats become a dominant source of nutrition in a low-carb diet, it is apparent that the choice of which fats to eat must become a close second. Absent clear scientific evidence to support a strong emphasis on one or two sources of fat, the best we can conclude for the moment is that you should diversify and consume a least some fat from a lot of different sources.

So what about protein? It’s pretty clear that you do need some protein, and in particular, you need protein that contains certain “essential” amino acids that the body cannot manufacture for itself, including vitamins such as B12 that are difficult to find anyplace else. Animal protein (especially organ meats) is a very rich source of most vitamins (except folic acid and vitamin K). Chronic protein and vitamin deficiency can cause severe malnutrition, poor muscle development, and premature aging. Vegetarians need to be particularly careful since good vegetable protein sources tend to be limited and tend to have much lower protein and vitamin content than meat and fish. From a purely nutritional point of view, humans aren’t really designed to be vegetarians, though if you choose to be so on moral or religious grounds, it is still possible to avoid malnutrition as long as you make sure you get enough of the right proteins and supplement to avoid vitamin deficiencies.

You will also find warnings against eating too much protein, particularly if you have impaired kidney function. For most people, this is not a serious concern in that it’s fairly difficult to actually eat too much protein. (See, for example, Martin et al. (2005).) Perhaps, if you tried to cut both carbs and fat out of your diet, then you could do it, but most reasonable ad libitum diets emphasizing fat and protein are not likely to give you too much protein.

Obviously, there’s no good reason to follow the low-fat crowd to the leanest cuts of meat. You don’t have to remove the skin from chicken and turkey of you don’t want to, and go ahead and leave your beef, pork, and lamb untrimmed if you like the fat. You can also eat cheese and other dairy products for protein as well as eggs, nuts, seeds, and tofu. These tend to come with a generous serving of fat, but that’s OK.

If you’re an athlete, then you may need to consciously increase your protein intake to help build and repair muscle. If you have trouble getting enough protein from your standard dietary sources, you can supplement by using either whey protein or soy protein powders. Whey protein is probably the better choice from a protein quality point of view, and there is some concern that soy protein, if consumed in large quantities, may provide excessive hormones. We use both with an emphasis on the whey protein. Whey protein tends to be sticky and gooey (at least when concentrated), and soy protein isolate behaves more like a flour; both are largely tasteless by themselves. They can be added to anything from eggs to baked goods and drinks. They are often packaged for retail sale in various adulterated (and overpriced) forms with added flavorings and sweeteners. Buy them unadulterated if you can (one on-line source: Honeyville Grain).

One last useful ingredient that we recently discovered is polydextrose. This is a long-chain sugar polymer that is largely indigestible and functions nutritionally like fiber. It is also tasteless (i.e., not sweet by itself). However, it can provide a lot of the mouth feel and body that artificial sweeteners generally fail to do. So far, we’ve tried it in homemade artificially sweetened ice cream and carrot cake with very satisfactory results. (You may need to disperse it in warm liquid first; it tends to clump.) We never met a low-fat ice cream that we thought was worth eating, but as long as you use flavors that adequately mask the objectionable bitterness and aftertaste of your artificial sweetener, you can make a very respectable low-carb ice cream! You can buy polydextrose as a commercial food ingredient, for example, from Honeyville Grain.

We find we don’t really miss most carbs. We don’t crave things like cookies or noodles at all, but occasionally find it necessary to eat a sweet potato or make polenta to quiet the cravings for a few days. It also depends on how much we are exercising. Of late, with our 50+ per week running mileage, it seems less important to cut carbs, but we generally keep them low anyway, compared to the usual endurance athlete's diet). On the other hand, the more we exercise, the less the weight comes off! We're not sure if it's due to eating more carbs or increased cortisol levels from overdoing it of late (three 50K trail races since Aug 23 and a 50 miler planned!).