The metabolic effects of low-carbohydrate diets and incorporation into a biochemistry course



One of the challenges in teaching biochemistry is facilitating students' interest in and mastery of metabolism. The many pathways and modes of regulation can be overwhelming for students to learn and difficult for professors to teach in an engaging manner. We have found it useful to take advantage of prevailing interest in popular yet controversial weight-loss methods, particularly low-carbohydrate diets. The metabolic rationale behind these eating plans can be linked to glycolysis, the citric acid cycle, lipolysis, gluconeogenesis, ketosis, glycogen metabolism, fatty acid oxidation, and hormonal regulation. When this approach was used in undergraduate biochemistry classes at the State University of New York at Geneseo, students were highly motivated to learn the biochemical principles behind these diets. The following provides information about low-carbohydrate diet plans that will enable professors to speak authoritatively on the subject. History and studies regarding efficacy as well as biochemical metabolic effects are included.

Learning the metabolic pathways is a challenge for students taking a biochemistry course. The amount of information is vast; details of the pathways are often difficult to remember, and the material is often presented in a “one reaction after another” manner that does little to engage students. To link concepts together, facilitate learning, and help students appreciate the relevance of the material, the biochemistry of popular low-carbohydrate diet plans was integrated into metabolism units in an undergraduate class at the State University of New York at Geneseo. All of the students in the class were familiar with these eating plans, in part due to media sensationalism. Some students or their family members and acquaintances had tried the diets. Therefore, students were highly motivated to learn about these regimens on a biochemistry level.

The following provides information on this topic that can be incorporated into lectures or problem sets and can enable professors to speak knowledgeably. Content from the primary literature is cited. General information is derived from a textbook such as Bhagavan's Medical Biochemistry [1].


Obesity is a growing problem throughout much of the world. The 12th European Congress on Obesity reported that 37% of American children, 20% of European children, and 10% of Chinese children are overweight [2]. Obesity is even increasing in parts of the developing world. As a result, weight loss methods are widespread and are part of popular culture. A recent study found that 44% of adult American women and 29% of American men are dieting at a given time [3]. In recent years, one of the most dominant and yet controversial dieting trends has involved plans that restrict carbohydrate intake. Books describing these low-carbohydrate methods have perpetuated best-seller lists for the past three decades, earning millions of dollars for the authors. The most well-known include the Zone Diet [4], the South Beach Diet [5] and the most famous and most vilified, the Atkins Diet [6, 7]. Dr. Atkins's books have sold over 45 million copies and, in fact, the 1992 edition is one of the top 50 best-selling books of all time. The trend is not limited to North America. Frenchman Michel Montignac, for example, has achieved fame with his own best-selling books advocating a low-carbohydrate plan (Eat Yourself Slim [8]).

Authors of these books, many of whom have medical credentials, try to provide a biochemical basis for the plans that they endorse. As a result, laymen have become familiar with some of the rudiments of metabolism that are traditionally taught in biochemistry classes. Ketosis and insulin regulation are now household/around-the-water-cooler topics of conversation. Biochemists frequently find themselves quizzed by friends and colleagues for more information on these subjects. While some adults may be tired of the topic, in part due to a barrage of media coverage, students in general are eager to learn about the subject. More importantly, doing so can help motivate students to learn and can provide additional familiarity with biochemical concepts to help students retain information about the metabolic pathways and their regulation.


The diet plans discussed here have one thing in common; they limit carbohydrate intake partly to control the release of insulin by the pancreas. 11 Insulin is released in proportion to the rate at which glucose enters the bloodstream. Therefore, carbohydrates, especially the simple carbohydrates, raise insulin levels to a greater extent than either proteins or fats do. Insulin promotes glucose uptake by cells, a process necessary for survival. However, it also initiates signal transduction cascades that result in inhibition of lipolysis (fat breakdown), inhibition of fatty acid oxidation, and inhibition of glycogen breakdown by the muscle and liver. At the same time, it stimulates fatty acid and glycogen synthesis (see Table I). These effects explain why elevated insulin levels (hyperinsulinemias) are associated with obesity. Insulin also promotes synthesis of cholesterol, and, therefore, elevated insulin levels are also associated with heart disease.

While insulin is secreted in response to elevated blood sugar, another hormone, glucagon, is secreted in response to low blood sugar. Glucagon acts on some of the same signal cascades as insulin, but with the opposite effect. It results in activation of enzymes that break down fuel while inhibiting the enzymes involved in energy storage.

Therefore, low-carbohydrate diets seek to minimize insulin release while favoring release of glucagon. Where they differ is in the extent to which carbohydrates are restricted, as well as in the amount of protein and fat permitted in the diets. The Atkins plan is the most extreme, calling for a 2-wk induction phase in which carbohydrates are limited to 20 g per day. Americans typically eat 300 g of carbohydrates per day ([9]). This introductory phase thus eliminates not only bread, pasta, and potato products, but many vegetables and all fruits. The diet replaces carbohydrates in this phase with fat and protein. Bacon, eggs, red meat, cream, and butter are permitted on the original Atkins plan. Gradually, carbohydrates are reintroduced into the diet, but at a very controlled level. The South Beach Diet permits carbohydrates from vegetables and sometimes whole fruits, but restricts intake of breads, potatoes, pasta, fruit juices, and sweets. The fat consumed must be largely unsaturated. The Zone Diet aims for meals in which the ratio of protein to carbohydrate is fixed at 2:3 with an estimated daily carbohydrate intake of 170 g ([10]). The Zone Diet also focuses on eicosonoid hormones, the fatty acid-derived substances that control inflammation and related processes. The Montignac method is based on the glycemic index and forbids high-carbohydrate, low-fiber foods that stimulate secretion of insulin. Typical nutrient and caloric comparisons are shown in Table II.

The more carbohydrate-restrictive plans such as the Atkins Diet are called ketogenic because they induce a state of fat burning called ketosis or ketogenesis. When diets are below a threshold level of carbohydrate (estimated at below 65–180 g/day; [11]), glycogen stores begin to be depleted; the decrease in glucose oxidation is compensated by an increase in fat breakdown. In ketosis, fatty acids are not broken down by the usual route, β-oxidation to acetyl coenzyme A (CoA),22 but are instead oxidized to ketones. Some of these ketones enter cells and are used for energy, but others are expelled via the breath, the skin, and the urine without being metabolized.

The role of ketosis in low-carbohydrate diets is a major focus of debate. One question that researchers continue to discuss is whether or not there is something “magical” about ketosis for weight loss. Ketogenic diets generally result in greater rates of weight loss when compared with other diets [12, 13]. Is this difference due to metabolic advantages or merely explained by water loss, appetite suppression, or restrictive food options? These questions are addressed in detail below.


Although low-carbohydrate diets are most closely associated with Dr. Atkins, such regimens far pre-date him. Greek Olympic athletes are reported to have consumed a high-meat, low-vegetable diet to improve their performances [14], and literature from as early as 1825 claims that obesity can be avoided by lowering one's intake of bread and potatoes (Brillat-Savarin, quoted in Ref. 15). However, a pamphlet published from 1864–69, “Letter on Corpulence to the Public” [16] is the most well-known early source for mass popularization of a low-carbohydrate diet. The author was an English funeral director named William Banting. At five-foot-five and 202 pounds (92 kg), Banting had struggled with obesity for decades. He had tried remedies from fasting to Turkish baths to 2 h of rowing in the morning (the latter of which he said gave him “a prodigious appetite”). In the year 1862, he consulted physician Dr. William Harvey who was interested in a new idea that starches convert to fat in the body. He advised his patient to give up sweets, bread, beer, and potatoes, while replacing these items with meat and fish. Within a year of applying this advice, Banting had lost 46 pounds. He was impressed enough with these results that he published his story at his own expense [16]. By the time of the 4th edition, this little pamphlet had sold over 100,000 copies, had been translated into French and German, and had also been published in America. Banting died at age 81, having maintained his weight loss for 19 years. The Banting-Harvey plan, as it came to be called, entered the English language. “To bant” meant “to diet.” Interestingly, the diet generated much of the same controversy then as low-carbohydrate plans do now. The medical establishment is reported to have largely responded to the pamphlet with disapproval [17].

Years later, the low-carbohydrate approach was tested as a treatment for pediatric epilepsy. It had been noted that fasting reduced the frequency of seizures in children. Because fasting has obvious drawbacks for a growing child, physicians tested food combinations and settled on a very high-fat, low-carbohydrate plan to control seizures. It was noted that these children lost weight, even when consuming significant calories in protein and fat. This eating regimen became known as a ketogenic diet [18]. The mechanisms underlying the anti-epileptic properties have yet to be fully elucidated, although theories abound. However, these are beyond the scope of this review.

Noting the weight loss associated with the ketogenic diet is what led a young cardiologist named Dr. Robert Atkins in 1963 to experiment on himself with this eating plan [7, 15]. In 1972, he published a book endorsing a ketogenic diet for weight loss in healthy adults. This book was later reissued and repopularized in the 1990s. Despite the fact that Atkins' approach was opposed by the American Medical Association [19], other plans based on reduction of carbohydrates followed, including the Scarsdale Diet, the Air Force Diet, the Zone Diet, Protein Power, Sugar Busters, and the South Beach Diet, to name a few.


Understanding the role of insulin and the process of ketosis requires some knowledge of metabolism and is a good way to help students retain information about these pathways. I generally distribute the following information throughout class discussions of signal transduction, glycolysis, the citric acid cycle, oxidative phosphorylation, and fatty acid oxidation. Then we put the entire process together. The discussion reinforces many of the concepts learned in these units, provides motivation for learning the metabolic pathways, and engages student interest.

The Centrality of Acetyl CoA

Students must first understand that carbohydrates are broken down largely into glucose; this glucose is oxidized via the glycolytic pathway to pyruvate, which can then be converted to either lactate or acetyl CoA. Using an analogy in which the metabolic pathways are thought of as a subway system, acetyl CoA is a major hub of the system. Acetyl CoA can be made by several different pathways. Glucose oxidation is one; fatty acid oxidation is another. Acetyl CoA also has several potential fates once formed (see Fig. 1). The normal outcome (when glucose is abundant) is for acetyl CoA to condense with oxaloacetate in the mitochondrion and enter the citric acid cycle. Entry of acetyl CoA into the citric acid cycle requires that oxaloacetate be sufficiently abundant for the condensation reaction to proceed at an appreciable rate [20].

Oxaloacetate and Its Role in Gluconeogenesis

Oxaloacetate is a minor hub of the metabolic system and has another potential fate besides combining with acetyl CoA, however. It can be converted to glucose in the liver via gluconeogenesis. Glucose can also be synthesized from muscle, from dietary amino acids, or from the glycerol of fats, but oxaloacetate is a readily available precursor for glucose, because it is stored in the mitochondrion.

Gluconeogenesis is critical to survival because the brain relies heavily on glucose. Human brain cells typically consume 100–125 g (500 kcal) of glucose per day, which amounts to two-thirds of the glucose needed by the body overall [21]. Fatty acids are not usually used by the brain because they tend to be bound to albumin and other serum proteins and are unable to cross the blood-brain barrier [22]. There are some reports of long-chain fatty acids being taken up by the developing brain [23], but in general the brain relies on glucose for energy. Glucose is easily transported into brain cells via the GLUT3 transporter. GLUT3 is usually saturated, because it has a Km of ∼1.6 mM for glucose, and typical plasma levels are about 4.7 mM.

The rate of formation (or depletion) of oxaloacetate determines the overall rate of fatty acid oxidative flux. When gluconeogenesis depletes oxaloacetate so that the rate at which acetyl CoA enters the citric acid cycle is low, acetyl CoA and its precursors build up in the liver. This condition can be triggered by fasting, exercise, or carbohydrate restriction.


As a result of the buildup of acetyl CoA, this metabolite and partially oxidized fatty acids are converted to acetone, acetoacetate (AcAc), and D-β-hydroxybutyrate (HB), the substances known as ketone bodies (see Fig. 2). HB was identified as the principal ketone body in fasting humans [24]. It interconverts with AcAc in a complex reaction catalyzed by the mitochondrial enzyme β-hydroxybutyrate dehydrogenase and requiring the NADH/NAD+ redox couple. AcAc and HB enter into cells via the monocarboxylate transporter. They can then be reconverted into acetyl CoA and used for energy in a process known as ketolysis [25]. The heart is very well-adapted to ketolysis and in fact will use AcAc preferentially over glucose. Acetoacetyl CoA is converted to two molecules of acetyl CoA via the enzyme acetoacetyl CoA thiolase. Acetone is either converted to glucose [26] or excreted without being further metabolized. This ketone is volatile and can be expelled via the breath, skin, or urine. If AcAc and HB are produced faster than they can be used, they also are eliminated via the urine. It is estimated that 10–20% of ketone bodies may be lost in this way [27].

Several enzymes are up-regulated in the metabolic adaptation to ketosis. The most important is the liver mitochondrial 3-hydroxy-3-methylglutaryl-CoA synthase or mHS [28]. Overexpression of the mHS gene in mice results in hyperketogenesis and dramatically lowered levels of serum fatty acids [29]. Loss-of-function mutations in humans result in profound hypoketonemia [30]. A ketogenic diet appears to up-regulate this gene in the brain as well [31].

Ketosis has had a negative connotation because it was initially discovered in the urine of untreated diabetics, especially those experiencing acidosis. In recent years however, scientists have come to realize that ketosis occurs normally at varying levels, particularly in children, who may develop ketosis within hours of a meal, in adults during prolonged exercise, and in pregnant and lactating women [22]. Ketosis spares muscle from being broken down for its glucogenic amino acids. It is in fact a protective mechanism for the brain and for general survival. Cahill and coworkers, in their substantial work with the biochemistry of starvation, suggested that more than 60% of human energy needs can be met by ketolysis [32].

Ketosis is associated with acidosis in diabetes because several of the products can form ketoacids [21]. However, the level of ketones in acidosis is far higher than the levels observed in carbohydrate restriction or in starvation. In normal individuals at rest, HB and AcAc are at concentrations below 0.1 mM. After several days of fasting, levels of AcAc rise to about 1–2 mM; HB concentration rises to 5–8 mM [33]. In diabetic acidosis, blood levels of HB are as high as 25 mM [24].

Fat loss is believed to be faster when humans are in the ketogenic state. The old adage “calories in, calories out,” [34, 35] though often quoted and seemingly valid from a standpoint of the first law of thermodynamics, does not always hold true [36]. Students should understand that metabolic processes can be altered in their efficiency. An excellent example occurs when the proton gradient in the mitochondria is destroyed before ATP synthase can use the potential energy of the gradient to synthesize ATP. Uncoupling proteins (UCPs), molecules such as 2,4-dinitrophenol, and the weight loss supplement usnic acid are examples of molecules able to dissipate the gradient, causing the extra energy to be released as heat.

Ketosis can be thought of as another example of a process that affects metabolic efficiency. When isocaloric diets are compared, weight loss tends to be greater with ketogenic diets [12, 13]. Although there is some debate about the contribution of water to this greater observed weight loss, the preponderance of evidence in both animals and humans shows that ketogenic diets are characterized by increased rates of fat loss [37, 38].

The question as to why ketogenic diets are associated with greater weight reduction is an interesting one, especially because studies of ketone bodies in perfused rat heart have shown that HB is a “superfuel” that actually produces more ATP per mole than that produced through the metabolism of glucose or fatty acids [24]. When added to rat heart, HB increased the ratio of hydraulic work per unit of energy from O2 consumption by 25% [39]. Likewise, a study of epileptic humans using 31P magnetic resonance imaging before and after a ketogenic diet showed increases in fuel molecules, indicating improvement in energy metabolism [40].

Explanations for the Increased Fat Loss Associated with Ketosis

These observations can be explained in several ways. First, from a theoretical standpoint, HB has a higher ΔHcombustion than the glycolytic product pyruvate does (243.6 kcal/mol versus 185.7 kcal/mol). In addition, one can consider that HB is more reduced than pyruvate [24].

Second, from a biochemical standpoint, one could also reason that HB and AcAc have the effect of widening the redox potential in the electron transfer/oxidative phosphorylation pathway that drives synthesis of ATP. Students should understand the basics of this pathway, namely that as redox reactions of metabolism proceed, electrons are harvested and stored in molecules such as NADH (nicotinamide adenine dinucleotide). NADH can deposit its electrons via oxidation at NADH dehydrogenase, or Complex I of the mitochondrial electron transport chain. The electrons are then transferred to the mobile coenzyme Q, which carries them to coenzyme QH2-cytochrome c reductase, or Complex III. Protons are pumped out of the cell at each of the complexes, creating a gradient that then drives the synthesis of ATP by ATP synthase. The number of ATP molecules generated depends on the difference in redox potential between the various complexes.

Ketone bodies are oxidized by the enzyme D-β-hydroxybutyrate dehydrogenase, releasing NADH and altering the ratio [NAD+]/[NADH], thereby reducing the potential at Complex I (–280 to –300 mV). Meanwhile, altered concentrations of metabolites also increase the coenzyme Q couple's potential from –4 to +12 mV. The ultimate effect is to increase the redox span and change the ΔG for electron transfer between Complex I and coenzyme Q from –53 kJ/2 mole e to –60 kJ/2 mol e, making more energy available for ATP synthesis [39].

Third, there is the observation that some of the processes associated with ketosis appear to be less strictly regulated than glycolysis or fatty acid oxidation. This observation arises because an estimated 10–20% of ketone bodies can be excreted without being used for energy [27]. Their production clearly exceeds their use at times.

An additional reason for the observed increase in lipolysis (fat breakdown) observed in low-carbohydrate diets is believed to be the use of fats for gluconeogenesis. When glucose is needed, lipases degrade fats to release glycerol, which can be converted to glucose. In fasting and starvation, approximately one-fifth of de novo glucose synthesis is from glycerol [41, 42]. Westman and coworkers proposed that lipolysis can proceed even when caloric intake exceeds energy expenditure, as long as the fats are needed for gluconeogenesis [43].

Ketosis is sometimes monitored with colorimetric urinalysis tests. “Dip strips,” used by diabetics for years, have been marketed to low-carb dieters. They contain Fe(CN)3, which turns pink/purple upon reaction with acetoacetate. These strips are also marketed as a way to measure fat-burning in general, but in fact they detect only ketosis-mediated fat burning, not general β-oxidation. Moreover, they are limited in their ability to predict plasma levels of ketones, because urinary levels correlate poorly with blood ketone concentrations [24, 44].


Low-carbohydrate diet plans have been a subject of vigorous debate. The subject is surprisingly polarizing. ABC News recently reported that a prime-time special on low-carbohydrate diets generated more mail than any program in its history. One physician-researcher reported that his presentation on the Atkins Diet at a conference generated genuine anger among the academicians in the audience (reported in Ref. 15). A little controversy can create a memorable class experience, however, and can heighten interest for learning.

One source of controversy relates to the role of ketosis. Some researchers claim that the greater short-term success of very low-carbohydrate diets is not due to ketosis, but is primarily due to voluntary caloric restriction. A study comparing nonisoenergetic diets revealed that individuals on low-carbohydrate diets tended to consume one-third fewer calories than individuals on typical low-calorie diets did [45]. Some individuals report decreased appetite when protein is abundant in the diet [46], and ketosis is known to suppress appetite. In fact, ketones are being studied as oral medicines to increase weight loss. Therefore, it may be true that the dramatic effects of low-carbohydrate diets are attributable to fewer calories consumed. Alternatively, a primary mechanism for both ketogenic and nonketogenic (but reduced carbohydrate) diets may simply be the elimination of insulin-mediated swings in blood glucose levels that stimulate hunger [47]. Certainly all these factors, including ketosis, may contribute to the weight loss associated with low-carbohydrate regimens.

While professors without medical credentials should avoid giving medical advice, we should feel free to use the literature to inform. One response to the controversy is to send the students to the literature themselves. An article by Spieth et al. [48] is recommended for its depth and balance, as are an article by Bachman [49] and a review from Westman and coworkers [43]. Simple internet searches can yield interesting results, although the information is often too general for a student of biochemistry. Moreover, students are likely to find alarmist treatises from organizations that promote alternate eating plans for moral or philosophical reasons. The range and fervor of opinions is rather fascinating, as some websites endorse low-carbohydrate eating as the cure to countless maladies, while others warn of its perniciousness.

We have found that students welcome the opportunity to participate in class discussions on this topic. Here we list some of the common questions students have and we respond using the literature.

1. Do Low-carbohydrate Diets Harm the Kidneys?

This is a common fear. Many people associate low-carbohydrate diets with the 400 cal/day high-protein liquid diets that caused some deaths in the 1980s, and the kidney damage observed in body builders who consumed high-protein diets over long periods while simultaneously practicing fluid restriction. Certainly the kidneys work harder and excrete more nitrogen when protein is elevated in the diet; however, the only changes researchers have observed associated with low-carbohydrate diets have been limited to increased glomerular filtration (considered benign) and lower calcium levels [50]. These observations were made in epileptic children who had been on an extreme form of the diet for up to 2 years. (The anti-epileptic form of the diet is higher in fat than that typically used for weight loss and maintains ketosis for a much greater time period). In studies of dieters, researchers have been unable to find evidence of kidney damage in individuals dieting over a period of 6 months or less, whether or not the diets were ketogenic [10, 45]. Increased glomerular filtration rate, increased kidney volume, and increased creatine clearance have been reported, but these changes are not indicative of decreased kidney capacity [49]. Such studies have only involved persons with healthy kidneys, however. It is unknown what effects would be observed in people with compromised kidney function.

2. How Do Low-carbohydrate Diets Compare with the Success of Conventional Low-calorie or Low-fat Regimens?

Dozens of clinical studies have investigated the success of low-carbohydrate and ketogenic eating plans, although most involve a relatively short time period (up to 6 months, which is thought to represent the maximum time most individuals will remain on a diet). Studies have included various genders, races, ages, differing levels of obesity, hyperlipidemic and normal-lipidmic patients, diabetics, and nondiabetics. They largely focus on the initially ketogenic (very low-carbohydrate) plans, and most are conducted on an outpatient basis, relying on individual honesty to document food intake. Some of these results are summarized in Table III. The overriding conclusion is that participants on Atkins-type ketogenic diets experienced substantial weight loss of 2.8–12.0 kg typically (or 6.1–26.4 lbs). Weight loss occurred whether or not the diet was accompanied by increased protein or fat. By comparison, traditional high-carbohydrate, low-fat regimens showed significant weight loss as well. Both methods are clearly effective, but most studies showed greater weight loss in the low-carbohydrate groups (Ref. 43 and references therein). When isocaloric diets are compared, the majority of studies still show greater weight loss in the low-carbohydrate dieters [12, 13, 51, 52]. Nonetheless, while most studies of low-carbohydrate diets showed more rapid rate loss at first, two studies found that the difference after 6 months was not substantial. A study by Lean et al. showed an average weight loss of 6.8 kg (or 15.0 lbs) after 6 months of an initially ketogenic Atkins-type low-carbohydrate diet, but this result was not substantially different from the 5.6 kg (or 12.3 pounds) observed after 6 months of a low-fat diet [51]. Another study by Foster et al. showed that weight loss was greater after 6 months on an Atkins-type plan as compared with a low-fat plan, but after a year, the first group had only lost 4.4 kg, which was not substantially different from the low-fat group's average loss of 2.5 kg [53].

3. Are Low-carbohydrate Diets Bad for the Heart?

The Atkins Diet, especially in its early form, encourages participants to eat fat, including foods high in saturated fat, such as cream and butter, and foods high in cholesterol, such as eggs. The South Beach Diet and the Zone Diet call for limiting fats to the unsaturated variety as much as possible. The idea that bacon, sausage, eggs, and cheese are permitted on the Atkins and similar plans surprises many individuals who have come to associate fats with heart disease. Interestingly, many of the popular low-carbohydrate diet books were authored by cardiologists, and they claim that these eating plans can improve cardiac health. Well-known risk factors for heart disease include high serum triglyceride levels, high total cholesterol and low-density lipoprotein (LDL) cholesterol, and low high-density lipoprotein (HDL) levels. Lesser known risk factors include elevated levels of proteins associated with inflammation such as the high-sensitivity C-reactive protein.

Cardiac disease is multi-factorial, and, with many lifestyle factors other than diet involved, the issue is a complicated one. Results of clinical studies should be interpreted with this fact in mind. A nearly universal result from studies of both ketogenic diets and nonketogenic low-carbohydrate diets is that of lowered plasma triglyceride levels [51, 52, 54]. One study of 41 moderately obese adults following an Atkins-type plan for 6 months reported a mean decrease of as much as 40% in serum triglycerides [55]. However, the role of diet composition in improved triglyceride levels is debatable. Low-fat diets show a similar decrease. Some researchers have reported that nearly any weight loss greater than 5–10% of body mass will result in a decrease in triglyceride levels [56, 57].

Results with LDL-cholesterol levels are more mixed. LDL is the major carrier of cholesterol in the plasma and it has long been known to be atherogenic (associated with plaque development in arteries). At least one study showed unfavorable increases in LDL-cholesterol [58] after an initially ketogenic low-carbohydrate diet after six months, but the majority of studies reported desirable decreases in LDL levels [51, 52] or no change [59].

There is a biochemical justification for the claim that a low-carbohydrate diet will lower total cholesterol, including LDL-cholesterol. Insulin acts on a cAMP-dependent cascade that promotes the activity of 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMG-CoA reductase), the enzyme that catalyzes a rate-limiting step in cholesterol synthesis (see Fig. 3). Reduced insulin levels would be expected to result in less activity of this enzyme and reduced synthesis of cholesterol.

HDL-cholesterol, in contrast to LDL-cholesterol, is not atherogenic. HDLs are the lipoprotein particles that enable cholesterol to be eliminated from blood vessels and broken down by the liver. Higher HDL levels have been found via both clinical and epidemiological studies to be cardioprotective. The literature has reported for years that low-fat, high-carbohydrate diets are associated with unfavorable decreases in HDL levels (for a representative review, see Ref. 60). In general, however, this risk factor is not well-understood. Tests of low-carbohydrate diets over a 6-month period again showed very mixed results in HDL levels. Several studies reported a decrease in HDLs [52], while others showed a desirable increase [55]. (For reviews, see Refs. 45, 57, and 61).

Other risk factors have been investigated as well. In one study, an initially ketogenic Atkins-type low-carbohydrate diet was compared with a low-fat diet in healthy, nonobese, normal-lipidmic women for a 6-month period. There were no significant changes in markers of inflammation such as the C-reactive protein, interleukin-6, or tumor necrosis factor α, nor were there significant changes in the size of the LDL particles [62]. In another study, the high-sensitivity C-reactive protein was measured in women whose diets varied in their glycemic load. A strong positive correlation was found between glycemic load and plasma concentrations of this proinflammatory protein [63].

It is clear that some individuals on low-carbohydrate diets have improved at least some of their cardiac risk factors. While triacylglyceride levels in the plasma are nearly always improved, however, these results do not appear to be dependent on the composition of the diet. Low-fat and low-calorie diets show similar effects. Changes in HDL and LDL levels have thus far been more variable. Although several studies have shown improvement in cardiac profiles, the results have not been universal.

4. Is the Weight Loss Associated with Low-carbohydrate Diets Simply Due to Loss of Water?

The greater initial weight loss seen with low-carbohydrate diets is partly due to water. Breakdown of liver and muscle glycogen results in diuresis (depletion of water). In a typical 70-kg adult, muscle typically stores 400 g of glycogen and the liver typically stores 100 g. Complete mobilization of these stores (representing about 1,600 kcal of energy) can result in a loss of over 2 lbs (about 1 kg). For each gram of glycogen used as energy, twice this mass is lost in water [14]. Ketosis also causes water loss. The kidney filters ketones as anions, increasing distal sodium delivery to the lumen and causing diuresis [14].

The role of water in overall weight loss is controversial. A study compared individuals fed an 800 cal/day ketogenic low-carbohydrate diet with those on an isocaloric mixed diet. Using nitrogen balance calculations, researchers found that while the low-carbohydrate group showed increased weight loss, the difference was entirely due to water. However, both groups lost an average of 6 lbs (2.8 kg) that were not water [12], so both approaches appeared to be effective. One problem with this study, however, is that ketosis was not monitored. The 800 cal/day mixed diet may have induced a certain degree of ketosis in participants. While the initial dramatic decrease in water seen with the low-carbohydrate approach raises the risk of dehydration, the rapid weight loss can also be motivating for a dieter. On the other hand, a drawback is that “cheating” can cause a spike in insulin that can reverse the water loss and cause weight gain [56].

5. How Well Do People “Stick to” These Diets?

According to one study, persons on diets that induce a moderate level of ketosis show attrition rates between 20 and 43%, a similar rate to low-calorie or low-fat eating plans [64]. In another study of overweight hyperlipidemic persons on a very low-carbohydrate diet (<20 g/day followed by slow inclusion of carbohydrates after 2 wk), the low-carbohydrate dieters showed a 25% attrition rate, while low-fat dieters showed a 47% attrition rate [59].

6. What Does the Medical Establishment Say About Low-carbohydrate Diets?

The medical establishment has overall been reluctant to endorse low-carbohydrate diets for the general public, citing the lack of evidence of safety in the long term as compared with the well-tested low-fat alternative diet (for representative publications see Refs. 65 and 66). Although most dieters do not plan to remain on this eating regimen long-term, some physicians are nonetheless opposed on the basis of the lack of long-term studies.


When we began talking about low-carbohydrate diets in class, student body language indicated increased interest (sitting up straight, leaning forward, increased eye contact, asking questions). Students reported having conversations with their friends and families on the topic. A typical comment was “My dad wants me to tell him everything I learn about this.” I also heard statements that I had never heard before, such as “I love metabolism.” While we did not do rigorous assessment of learning before and after using this approach, there was considerable anecdotal evidence that including this highly relevant application of metabolic principles was helpful in motivating students to learn and to provide them with additional familiarity to assist in retention of learning.

Figure Fig. 1..

Summary of several metabolic pathways. Glycolysis and fatty acid oxidation both produce acetyl CoA. Acetyl CoA can enter the citric acid cycle or be degraded via ketosis, among other fates. Concept for drawing inspired from Bhagavan's Medical Biochemistry [1].

Figure Fig. 2..

Formation and use of ketone bodies.

Figure Fig. 3..

Insulin-stimulated synthesis of the precursor to cholesterol.

Table Table I. Some of the biochemical effects of insulin
Glucose transporters Up-regulation
Fatty acid synthaseDephosphorylationActivation
Glycogen synthasesDephosphorylationActivation
Acetyl CoA carboxylasePhosphorylationDeactivation
Glycogen phosphorylasePhosphorylationDeactivation
HMG-CoA reductasePhosphorylationActivation
Table Table II. A modeled comparison (typical diet composition) of two very low-carbohydrate diet plans in ketosis phase with a nonketogenic diet plan and isocaloric American Heart Association 2000 recommendations for daily intakea
 AtkinsbProtein Power b,cZoneAHA
  • a

    a Data adapted from Anderson et al. [10].

  • b

    b During inductive ketogenic phase.

  • c

    c Note that Protein Power can avoid ketosis even in induction phase.

Carbohydrate (g)2237170220
Carbohydrate (%)5%10%40%55%+
Protein (g)14614912028–72
Protein (%)35%35%28%12–18%
Fat (g)104974953
Fat (%)59%53%32%<30%
Fiber (g)41118>25
Calories (kcal)1,6001,6001,6001,600
Table Table III. Clinical studies of low-carbohydrate ketogenic diets compared with low-fat reduced-calorie diets (randomized, controlled trials)
Patients (n)Duration (months)Low-carbohydrate weight lossLow-fat weight lossStudy
30310 kg4 kg[68]
4268.5 kg4.2 kg[69]
13265.8 kg1.9 kg[54]


  1. 1

    Recently, researchers uncovered a transcription factor called the carbohydrate response element-binding protein that activates lipogenic enzyme expression independently of glucose [67].

  2. 2

    The abbreviations used are: CoA, coenzyme A; AcAc, acetoacetate; HB, β-hydroxybutyrate; mSH, 3-hydroxyl-3-methylglutaryl-CoA synthase; LDL, low-density lipoprotein; HDL, high-density lipoprotein; HMG, 3-hydroxy-3-methylglutaryl.