Eat Your Starches: Why Safe Starches Are Healthy

Last updated:
Join us and over 84,000 people
Our Paleo Guide: Your Guide to Paleo
Safe starches

A Paleo diet emphasizing high-quality meat and fresh vegetables will naturally be lower-carb than a standard American diet based on cheap carbohydrates like wheat and other cereal grains. By excluding foods that contain toxins (like gluten and lectins), a Paleo diet replaces many starches with fat and protein, effectively treating the problems that spring from a carb-heavy modern diet, especially metabolic disorders like diabetes. But not all carbohydrate-rich foods contain toxins or antinutrients, and some are even high in nutrients: sweet potatoes, for example, are a rich source of Vitamin A. This raises the question: do these non-toxic starches have a place in the Paleo diet? If wheat is unhealthy because of the gluten, not the carbohydrates, starches like potatoes and yams ought to be acceptable, and even encouraged. This is precisely the concept that Drs. Paul and Shou-Ching Jaminet set forth in the Perfect Health Diet. Defining a “safe starch” as “starchy food which, after normal cooking, lacks toxins, chiefly protein toxins,” the Jaminets encourage healthy non-diabetics to consume approximately 400 calories per day (or 20% of daily caloric intake) of safe starches, prioritizing glucose over fructose. While acknowledging that a high-carbohydrate diet leads to all kinds of metabolic disorders, the Jaminets also argue that a diet too low in glucose can lead to problems like nutrient deficiencies, kidney stones, and lower mucus production, which impairs the immune system. The concept of “safe starches” has sparked criticism from advocates of very low carbohydrate (VLC) diets like Dr. Ron Rosedale. The differences between the two groups should not be overstated. Both the advocates and the opponents of the “safe starch” principle agree that the body needs glucose – and both also agree that too much glucose is toxic, leading to chronically elevated insulin levels, insulin and leptin resistance, and eventually metabolic disorders like diabetes. The question of “safe starches,” however, reveals important differences, even between diets that are ultimately much more similar to each other than to the standard American approach to food.

Glucose: The Basics

Carbohydrates come in three major forms: glucose, fructose, and fiber. Glucose and fructose are the only two forms that provide energy to your body, since fiber is not digestible (by humans – the bacteria in your gut love it!). Of the two forms of carbohydrate that you can digest, glucose appears to be preferable. Humans need glucose: it supplies energy to the brain, makes up molecules called glycoproteins (including mucins, the building blocks of mucus), and supports immune function. Fructose, on the other hand, is redirected to the liver as fast as possible – your body essentially treats it like a toxin and tries to get rid of it. While glucose is a necessary fuel, fructose is a tasty reward that fruits evolved to deliver so that animals would eat them and spread their seeds around through defecation. Moreover, too much fructose can lead to bacterial overgrowth in the gut, causing malabsorption and other digestive problems. Between the two sources of carbohydrate energy, then, glucose is clearly preferable – and, in fact, necessary. Unless your body is highly adapted to a very low carbohydrate diet, it uses at least 600 calories of glucose every day, more if you’re athletic or fighting off any kind of infection. Your brain alone needs 480 glucose calories daily to function. But your body’s need for glucose doesn’t necessarily mean that you have to supply that glucose through diet. Through a process called gluconeogenesis , your body can create glucose from fat and lactic acid, ensuring that your brain and immune system get the fuel they need even in the absence of dietary glucose. The question therefore arises, which is better: obtaining glucose from the diet, or from gluconeogenesis? Advocates of the “safe starch” theory claim that optimal health requires some dietary intake of glucose, while the critics of “safe starches” argue that dietary glucose is toxic at any level, and that the body is completely capable of synthesizing all the glucose it needs from fat and protein.

The Safe Starch Debate

Sweet PotatoIn late 2011, Paul Jaminet and Dr. Rosedale debated the concept of “safe starches” in a series of admirably civil blog posts, with Dr. Jaminet advocating the daily consumption of starches, as per the PHD recommendations, and Dr. Rosedale arguing that dietary glucose is toxic at any level, and that the idea of a “safe” starch is an oxymoron. The debate revolved around several basic questions, revealing fundamental differences between the PHD and the Rosedale diet. One major point of divergence is that the two diets are designed for different groups of people. Dr. Rosedale, who specializes in treating people with diabetes and other metabolic disorders, considers everyone to be “metabolically damaged to some extent.” His diet is essentially medicinal – because everyone is sick (insulin and leptin resistant) to some extent. Jaminet, on the other hand, does not classify every potential adopter of the PHD as metabolically damaged. He explicitly states that he is not a diabetes specialist: although the PHD is not bad for diabetics, it is designed for people without metabolic disorders. Diabetics may need to adjust their carbohydrate levels to find the optimal amount – and the adjustment depends greatly on the individual. The two diets are, in other words, based on dramatically different premises: Rosedale starts from the assumption that everyone is sick and in need of treatment; Jaminet starts from the assumption that most people are healthy. On a deeper level, Jaminet and Rosedale disagree on the goal of the optimal diet. Rosedale focuses on extending the human lifespan by outmaneuvering nature – explicitly claims that “I have no qualms about saying that what I’m doing and teaching to be healthy is quite unnatural, as it is the quest to be able to live a long, happy, and youthful life after making babies.” Jaminet, on the other hand, focuses on optimizing health by eating as evolutionarily “natural” a diet as possible. In this sense, the PHD is closer to the Paleo diet in principle, although the Rosedale diet may resemble it more closely in practice. On two diets optimized for different groups of people working towards different goals, it is hardly surprising that the specific recommendations differ.

Dietary glucose: toxin or hormetic stress?

The most obvious point of contention in the debate is the level of dietary glucose most beneficial for human health. Rosedale and Jaminet disagree not only about the specific amount of glucose in an optimal diet, but even more basically, about how to best measure the effects of glucose in the first place. Their disagreement reflects the difference in their attitudes: Rosedale, addressing people he considers to be universally metabolically damaged, and focusing on the effect of glucose on insulin, considers glucose intake a dangerous stressor; Jaminet, writing for an audience of people he considers metabolically healthy and focusing on the effects of glucose on blood sugar, views these it as a tolerable, and even beneficial stress. Rosedale claims that any amount of dietary glucose is toxic. His ideal glucose level is as low as possible – while he acknowledges that any diet will include some glucose, he asserts that the less, the better. No starch is “safe:” any amount of glucose “will, at least to some extent, by spiking blood glucose, insulin, and leptin, mimic the stress response.” Synthesizing glucose via gluconeogenesis, by contrast does not create insulin resistance, since the liver knows exactly how much glucose it needs to make, and does not create excess glucose for the body to process. While he acknowledges its existence and utility, Jaminet claims obtaining glucose solely from gluconeogenesis is less than ideal. In his view, “the dose makes the poison” – too little dietary glucose can cause health problems, and too much is certainly harmful, but small doses of glucose do not cause measurable harm and might even contribute to health through a process called hormesis. Hormesis is the same kind of “good stress” that makes intermittent fasting so beneficial: when your body copes with stress, it becomes resistant to that stress, and you “toughen up” in general. Rather than a straight line, Jaminet sees glucose toxicity as a U-shaped curve , with the greatest risk at very low or very high blood glucose levels. He recommends a fasting blood glucose level of around 100mg/dl for optimal health, and proposes that most people can achieve this by eating approximately 400 calories of starch a day (on a 2,000 calorie diet, this amounts to 20% of calorie intake from starch). Rosedale addresses this question when he rephrases the crucial question of the debate as: “Is there a diet (Rosedale’s or Jaminet’s) or glucose (starch) intake that can better maximize the repair/damage ratio that life, health, and youthful longevity depends on, admitting the inevitability of damage from glucose at any level?” While Jaminet based his recommendations for “safe” levels of glucose consumption on levels of fasting blood glucose, Rosedale challenges his premise by arguing that fasting blood glucose is an inadequate measurement of the damage that glucose does to the body, which mainly occurs when high levels of postprandial glucose spike leptin and insulin levels. Eating a 20% carbohydrate diet might lower blood sugar levels, but only at the expense of raising insulin levels, making it worse than useless. In response, Jaminet simply states that he finds Rosedale’s arguments unconvincing, but prefers to let the matter drop. Several other researchers, however, have noted that insulin resistance does not correlate directly with glucose consumption. Most followers of the Paleo diet understand that chronically high blood glucose levels cause frequent insulin spikes, leading to eventual insulin resistance. However, following a VLC diet can also create a kind of insulin resistance – as a response to a scarcity of glucose, not an excess of it. When your body is forced to create glucose through gluconeogenesis, rather than receiving it through diet, every molecule of glucose is precious. Your muscles become slightly insulin resistant as a protective measure, to spare scarce glucose resources for your brain, which needs them most. This is not a sign of diabetes, and is not necessarily reason for concern: as Guynet has it, “a healthy body increases insulin sensitivity in response to increased demand for glucose disposal.” But the fact remains that insulin resistance occurs at both extremes of the glucose consumption scale, with the diet least conducive to insulin resistance probably falling somewhere close to the PHD recommendations for starches. This goes back to Jaminet’s idea of hormesis: consuming glucose raises blood sugar (a stressor), causing your body to adapt and become stronger in response (insulin sensitivity). In sum, these two different recommendations for dietary glucose reflect the two doctors’ different ideas about their readers: hormesis is only beneficial to someone in generally good health and strong enough to withstand the stress; since Dr. Rosedale sees everyone as metabolically deranged, he would logically shy away from recommending an additional stress to an already damaged system. In their treatment of metabolically damaged individuals, the Jaminet and Rosedale approaches are remarkably similar; the main difference is in whether they consider any significant number of people to be “metabolically healthy.”

Are there any risks associated with very low intake of dietary glucose?

Rosedale and Jaminet disagree not only on the optimal levels of glucose in the diet, but the consequences of straying from those levels. In Rosedale’s view, a “deficiency” of dietary glucose is just as illogical an idea as a “deficiency” of dietary cyanide. Jaminet, on the other hand, argues that a scarcity of dietary glucose does indeed have measurable negative effects, especially for the immune system and the thyroid. While he supports ketogenic diets as helpful for people with certain neurological disorders and infections, he maintains that most metabolically healthy people will be healthier with some dietary glucose intake, describing four major symptoms of “glucose deficiency:” scurvy, mucus deficiency, kidney stones, and thyroid problems. Jaminet’s challenge to a VLC diet comes partly from personal experience: while on a VLC diet himself, he developed scurvy, a nutrient deficiency caused by inadequate levels of Vitamin C. Upon reintroducing the carbohydrates that he describes as “safe starches,” his health improved. After more research, Jaminet found that insulin is actually vital to maintaining healthy levels of Vitamin C – when Vitamin C is damaged by oxidation, insulin is the hormone that prompts the glucose transporters to restore it, allowing your body to recover from injuries and infections. Selenium, another mineral commonly depleted by a ketogenic diet, is also essential to maintaining healthy levels of Vitamin C. Thus, Jaminet argues that a VLC diet can create a Vitamin C deficiency, or even scurvy, even in a person whose dietary intake of Vitamin C is perfectly adequate. From his experience developing the PHD and his dietary research, Jaminet also concluded that a VLC diet reduces mucus production, since mucus is primarily made of water and sugars. Mucus is made from glycoproteins; when these proteins are broken down in gluconeogenesis, they are unavailable to be used as mucus. This dries out the eyes, mouth, and intestines, depriving the body of a crucial barrier against irritants and pathogens and increasing the risk of infection and gastrointestinal cancers. During 2 years on a zero-carb diet, Dr. Jaminet personally noticed uncomfortable dryness in his eyes and mouth, and self-experimentation revealed glucose deficiency as the major culprit. He cites other low-carb dieters who have had similar experiences, as well as a study showing that dry membranes commonly occur during starvation, when the body breaks down its stores of protein to maintain glucose levels. Critics of the “safe starch” approach argue in response that the mucus is a response to the irritation that carbohydrates produce in the intestines – Rosedale also claims Jaminet fundamentally misunderstands mucus deficiency as a problem of glucose scarcity: “eating extra glucose will not make mucus any more than taking calcium will make bone. There must be instructions to do so.” Furthermore, Jaminet argues that a zero-carbohydrate diet can dramatically increase the risk of kidney stones – in children put on a ketogenic diet to treat epilepsy, 1 in 20 developed kidney stones, compared to 1 in several thousand among those not in ketosis. Jaminet sees four factors as primarily responsible for this dramatic increase in kidney stones. First, a person on a ketogenic diet must metabolise both glucose and ketones from protein, producing much more uric acid than someone not in ketosis. Since DHAA is not recycled into Vitamin C on a ketogenic diet, it is degraded into oxalate, which must be excreted by the kidneys in a process that consumes electrolytes and water – both of which the kidneys need to excrete uric acid. Thus, the uric acid and oxalate precipitate as deposits, forming kidney stones. The dehydration common on ketogenic diets merely compounds this set of problems. As a fourth measure of glucose scarcity, Jaminet describes a condition called “euthyroid sick syndrome.” In a healthy thyroid, the hormone T3 stimulates glucose transport and utilization. In a person with low levels of dietary glucose, the thyroid stops producing T3 and produces an inactive hormone called rT3 instead, reducing glucose transport and utilization in the body. In euthyroid sick syndrome, these hormones are out of balance, without any malfunction in the thyroid: essentially, low glucose levels mimic a thyroid problem. Jaminet also connects this to high LDL levels. In response, Rosedale argues that the lower body temperature and metabolic rate consequent to reduced T3 are actually beneficial. With caloric restriction, glucose is reduced, therefore leptin is reduced, therefore T3 is reduced. This, he claims, is healthy because it is “part and parcel of genetic expression of increased maintenance and repair.” Responding to Rosedale’s response, Jaminet contends that low body temperatures suppress the immune system, among other signs of suboptimal health. Rosedale responds by criticizing Jaminet’s understanding of why the thyroid is low: “I am not talking about hypothyroidism. I am talking about a thyroid that is purposefully being lowered to enhance the wellness and survivability of that life.” The low levels of thyroid hormone, he argues, are a sign that a VLC diet mimics all the advantages of calorie restriction, without the restriction. Rosedale argues that a purposefully low thyroid does not suppress immune function, since it does not prevent fever temperatures when necessary. While Rosedale does not specifically address the first three of Jaminet’s arguments, except for a brief assertion that mucus deficiency is caused by problems in the body’s enzymatic reactions, not low levels of glucose, he contends in general that “glucose deficiency” is not caused by insufficient glucose. It’s a problem of “improper instructions about what to do with glucose…and improper insulin and leptin signaling.”

Safe starches: Conclusion

Chestnuts, a great source of starch Both Rosedale and Jaminet agree that a low carbohydrate diet is optimal for diabetics or anyone suffering from metabolic syndrome. But while Jaminet sees diabetes as a disease affecting only some people, Rosedale argues that everyone should eat as though they were diabetic, because to a greater or lesser extent, we all are. However, this is simply not the case – a normal person with a healthy metabolism does not need to reduce the stress of glucose metabolism by eating a VLC diet any more than an uninjured person needs to reduce the stress on his knees by using a wheelchair, no matter how helpful that wheelchair might be to someone with a broken leg. Many metabolically healthy individuals feel more energetic and vibrant when they increase their carbohydrate intake to levels roughly in line with PHD recommendations – and far from being automatically obesogenic, some even find that moderate carbohydrate intake can help break through a weight loss plateau. As well as being beneficial in moderate doses, dietary consumption of glucose is in line with the evolutionary basis of the Paleo diet: glucose is an ancient food that our bodies have evolved a specific enzyme (amylase) to digest. Different indigenous cultures have thrived on a broad range of macronutrient ratios – without the slightest idea of what a macronutrient is, or how much of which ones a particular food contains. Even more radically, Kurt Harris suggests that the process of gluconeogenesis is not a sign that dietary glucose is unnecessary, but rather “evidence that glucose is so metabolically important that we have evolved a way to make sure we always have it.” Harris agrees that most people should have no problems eating 15-20% of calories as starch, and that this approach is actually healthier than eating a VLC diet. Furthermore, he claims that “the whole concept of a macronutrient, like that of a calorie, is determining our language game in such a way that the conversation is not making sense.” Pointing out that the body digests different forms of carbohydrates in different ways, he suggests that we focus instead on the nutritional value of foods, without categorizing them by macronutrient ratios that don’t give meaningful information about the actual value of those foods to the body. Several other Paleo practitioners agree: eating nutrient-dense foods that do not contain harmful toxins is healthful and natural; micromanaging the intake of any particular nutrient is not – and therefore, probably not necessary. “Safe starches” such as potatoes (especially sweet potatoes) and other starchy tubers are therefore a perfectly acceptable element of a healthy, evolutionarily-based diet for someone with no (or few) metabolic problems. For athletes or anyone else who performs relatively large amounts of intense exercises, dietary carbohydrates are essential to performance and general well-being. If you feel better eating a moderate amount of carbohydrates, enjoy them without fear.

Safe starches

Safe starches in your diet

If you’re stumped for a way to add safe starches to your diet, take a look at this table from Balanced Bites, which shows the best Paleo sources of carbohydrates, along with their nutritional values. The only major one missing from the table is chestnuts, which are a great safe starch option. The most commonly available sources are potatoes, sweet potatoes, and squashes; to find the slightly more exotic starches like cassava or lotus root, you might need to venture out to an ethnic supermarket or a specialty grocery store. The easiest way to cook any kind of potato is to bake it – poke a few holes in the skin and microwave it on high until it’s soft to the touch (around 4 minutes for a fist-sized potato). If you’re looking for something fancier, try making sweet potato soup with lime or leeks. Sweet potato fries are a delicious classic; you can also mash your potatoes or include them in a frittata, a salad, or a casserole. And even if you don’t have access to the more exotic types, yams and sweet potatoes have an amazing variety: look for them at ethnic markets or try some as a vacation treat. Spicy Sweet Potato Wedges Squashes taste delicious in soup, or simply roasted with spices. Spaghetti squash can replace traditional pasta in any kind of dish – or try it baked with some sweet potatoes and spices for an easy, portable addition to any meal. Pumpkin pie is easy to make Paleo: just leave off the crust, or make your own from Paleo ingredients. Plantains make excellent chips (delicious with salsa or guacamole), or try frying them with cinnamon. They also make a unique addition to an omelet. Serve them up as fritters with some bacon, or cook them with pork for an easy crock pot dinner. Experiment with varieties of these recipes, or try your own – the basic concepts (baking, mashing, stewing with meat) are infinitely adaptable to many different combinations of spices and additions.

P.S. Take a look at our Paleo Recipe Book. It's a Cookbook we've created to help you prepare the best Paleo food. It contains 370+ recipes and covers everything you'll need.

+ Your Guide to Paleo, our visual Paleo Guide, was just released. It'll help you avoid the most common pitfalls and reach your health and weight loss goals faster.

Get your free recipes