Sugar substitutes

Sugar, in its many guises (dextrose, high-fructose corn syrup, honey, agave nectar, fruit etc.), needs to be minimised on a ketogenic diet, or eliminated as far as practicable. This can be a sticking point for those who are used to, or even addicted to, sweet food. As people and health authorities start to recognise that over-consuming sugar can have detrimental consequences, the market in low-caloric sugar substitutes has grown. These products generally fall into two categories: high-intensity sweeteners (HIS) and sugar-alcohols (SA). There are other ways to categorise sugar substitutes, e.g. natural vs. artificial, nutritive vs non-nutritive however, I don’t find those classifications as precise or helpful.

(1) High-Intensity Sweeteners

To keep the topic under control, I will confine this discussion to just those HIS that are approved by the influential US Food and Drug Administration (FDA). There are eight: saccharin; aspartame; acesulfame potassium; sucralose; neotame; advantame; steviol glycosides and Luo Han Guo (Monk) fruit extract. These are hundreds to thousands of times sweeter than table sugar. With the exception of aspartame, they are non-caloric (i.e. not metabolised) and are eliminated in urine or faeces.

Saccharin is 200 to 700 times sweeter than table sugar. It is the oldest of the HIS, having been discovered and used since 1879. It went out of favour in the 1970s when it was linked to bladder cancer in laboratory rats. FDA approval for saccharin was not withdrawn, however, a warning label was mandated pending further studies. In 2000, the National Institutes for Health (NIH) concluded that the results were specific to rats and did not carry over to the human, and that saccharin was not a potential carcinogen. The requirement for a warning label was removed.

Aspartame is about 200 times sweeter than table sugar. It was discovered in 1965 and consists of two amino acids (components of protein), phenylalanine and aspartate. Both of these are found in many foods (aspartate was first extracted from asparagus, from which it got its name). Phenylalanine is an essential amino acid and we must ingest it. Aspartate is non-essential (our body makes what it needs, but it probably helps to ingest it). Aspartame is metabolised in the body into its components: phenylalanine, aspartate and low-levels of methanol. As a result, like other amino acids, it provides about 4 calories per gram. Not that that matters, given how little is needed for sweetening. The methanol is negligible.

Aspartame was approved for certain applications in 1981, and as a ‘general purpose sweetener’ in 1996. It loses its sweetness when heated and is not commonly used for baking. The FDA reviewed more than 100 studies supporting the safety of aspartame, except for individuals with a rare hereditary disease known as phenylketonuria.

Acesulfame potassium resembles saccharin in taste and level of sweetness. It is often abbreviated to acesulfame K or Ace-K. It was discovered in 1967 and approved as a general sweetener in 2003. It is heat-stable and suitable for baking. The FDA refers to more than 90 safety studies.

Sucralose was discovered in 1976 and it is manufactured from table sugar (sucrose). It is about 600 times sweeter than table sugar and goes under the brand name Splenda. It was approved for use in food as a non-nutritive sweetener in 1999, based on more than 110 safety studies. It is heat-stable to high temperatures.


Neotame is a potent sweetener – at least 7,000 times sweeter than table sugar. A 350ml can of soft drink normally contains about 35g of sugar (~10%) – that could be replaced with 0.006g of neotame. It was developed by scientists at the NutraSweet Company. It is a derivative of aspartame, and made up of the same two amino acids. However, unlike aspartame, the bond between the component amino acids resists breakdown, and neotame is not metabolised, but rather eliminated in the urine and faeces and so is non-caloric.

It was given FDA approval in 2002 (based on over 110 animal and human studies). It is heat-stable at high temperature.

Advantame is the most potent of them all – 20,000 times or more sweeter than table sugar, and with a sugar-like taste. It is another analogue of aspartame, and is heat stable. It is manufactured by Japan’s Ajinomoto Co., the same company that makes MSG. Advantame was approved as recently as 2014. The FDA has this to say: “In determining the safety of advantame, FDA reviewed data from 37 animal and human studies designed to identify possible toxic effects, including effects on the immune system, reproductive and developmental systems, and nervous system. FDA also reviewed pharmacokinetic and carcinogenicity studies, as well as several additional exploratory and screening studies.” 

Steviol glycosides are natural constituents of the leaves of Stevia rebaudiana, a plant native to parts of South America. The glycosides include rebaudioside A (abbreviated Reb A), stevioside or rebaudioside D (Reb D). A product may contain a mixture of these glycosides, and you may see Reb-A on ingredients lists. They are 200 to 400 times sweeter than table sugar. The FDA has not gone through the formal approval process for these sweeteners because it accepts the manufacturer’s claims that they are safe. However, this only applies to these refined glycosides, the FDA does not consider the use of the stevia leaf itself, or crude stevia extracts, to be safe, and their import into the US for use as a sweetener is not permitted.

Luo Han Guo fruit extracts contain mogrosides (related to glycosides). The fruit is native to South China, where it has been traditionally used as a sweetener and herbal remedy. It is also known as monk fruit, Buddha fruit, or by its botanical name Siraitia grosvenorii fruit. Depending on the level of mogrosides, its sweetness ranges from 100-300 that of table sugar. The extract is non-caloric and heat-stable. As with steviol glycosides, the FDA has not formally approved these sweeteners because it accepts the manufacturer’s claims that they are safe.

(2) Sugar Alcohols

The second category of sugar substitutes are the sugar alcohols (SA). These are low-digestible carbohydrates that have a sweetness less than or equal to table sugar. These sugar substitutes are “bulky” and normally used in amounts equal to the amount of sugar they replace. They can be thought of as low-intensity sweeteners.

For most SA, only a small amount is absorbed by the small intestine, and this is either not metabolised, or metabolised in ways that do not need insulin or raise blood glucose significantly and, consequently, they have a low (in some cases zero) glycemic index (GI). The portion that is not absorbed proceeds to the large intestine where it is mostly fermented by gut bacteria into short chain fatty acids. While the FDA has no health concerns for SA, the fermentation may produce abdominal gas and discomfort in some individuals and the FDA requires a warning label for some SA: “Excess consumption may have a laxative effect.” The amount needed to cause symptoms varies greatly based on individual tolerance, and adaptation may develop over time.

Though they are called SA, they are not alcoholic. They do not contain ethanol, which is found in alcoholic beverages. Common examples of SA include: erythritol; isomalt; lactitol; maltitol; sorbitol and xylitol. SA are sometimes called polyols, which represent the ‘P’ in the low FODMAP diet sometimes recommended for irritable bowel syndromes.

SA occur naturally (at low levels) in fruits and vegetables, and can be produced in commercial quantities from starch or glucose (often by controlled fermentation). The SA have a long history of use in baked goods, candy, and other manufactured products. SA do not promote tooth decay, and the FDA allows manufacturers to label their products so. Some, xylitol in particular, inhibit the activity of decay-causing bacteria and are used in chewing gums and toothpaste. Food manufacturers may list the amount in grams per serving of sugar alcohols on the Nutrition Facts Label (under Total Carbohydrate), however, this is voluntary. To know for certain whether SA are in a product, look at the Ingredients List.

In an attempt at brevity, I will confine myself to two SA that are commonly called for in ketogenic recipes (erythritol and xylitol) and a third that is not so common but that may serve another purpose (isomalt).

Erythritol is the exception among SA because it is nearly fully absorbed by the small intestine. However, it is not metabolised while in the body and it is excreted in the urine. It therefore non-caloric and has a GI of 0. It causes minimal fermentation-related effects in the large intestine (because little gets that far). Nevertheless, some individuals report an intolerance to it. Erythritol is ~70% as sweet as table sugar and can be brought up to sugar’s sweetness with a little added HIS. It is stable across temperature, acidic or alkaline environments, and it is not hygroscopic.

Xylitol is a natural carbohydrate found in fibrous vegetables, woody material (xyl means wood in Greek) and corn cobs (which are the main source for industrial manufacture). The human body produces a small amount of xylitol as part of normal metabolism. About half of ingested xylitol is absorbed in the small intestine and metabolised in the liver (without a need for insulin), and the rest can undergo fermentation in the large intestine. Xylitol has the same sweetness as table sugar with about half the calories of sugar and a GI of 7-12. Several studies indicate that daily usage of small amounts of xylitol can improve dental health.

Isomalt is manufactured from table sugar, retaining the glucose and replacing the fructose with equal amounts of the SA sorbitol and mannitol. Like most SA, ingested isomalt is poorly digested and absorbed in the small intestine. Isomalt has about half the sweetness of table sugar and half the calories. It has a GI of 9 and is heat tolerant.

The reason to highlight isomalt is that it can be a substitute for table sugar when the physical properties of sugar are important. For example, isomalt crystallises like table sugar but more slowly, and is often used for sugar decorations and may have application in ice cream. Likewise, it is slightly hygroscopic. The HIS provide only sweetness.

(3) Products containing HIS

These sugar substitutes are high intensity, at least 100 times sweeter than table sugar. This is not a problem (and probably an advantage) if manufacturers add a measured quantity of one of the HIS directly to their product, such as soda drinks. However, when selling HIS in powder form for people to cook with in the normal way (by spoon/cup measures), or to add to their coffee, the sweetness must be scaled back to table sugar levels. A teaspoon of HIS in a coffee would be unbearably sweet. Therefore, the HIS need a ‘filler’. For a sweetener that is 100 times sweeter than sugar, this would require a product to be 99% filler and 1% sweetener. The HIS may not be caloric, or have a GI, but what about the filler?

The most common filler is a carbohydrate known as maltodextrin. These are short chains of glucose molecules, branched or linear (about 5-30 glucose molecules). The shorter the chain-length, the more glucose-like they become. They are classified according to what is known as their Dextrose Equivalent (DE), which is the inverse of their glucose chain length x 100 (so they have DEs of 3-20). Anything higher than 20 (i.e chain-length less than 5) has to be referred to as a syrup. It is quite confusing, I mention it because you may see DE on an Ingredients List label.

Maltodextrins are ubiquitous in manufactured products (not just HIS) because they can be made with a host of properties to suit the product. An extreme example is N-Zorbit M (a tapioca maltodextrin), which is a powder that can absorb oil but remain a powder, and release the oil when water is added – ideal for packet cake mixes.

However, the important property that all maltodextrins share is that they are readily broken down into their component glucose molecules in the acidic environment of the stomach. Consequently, their GI is very high (~100) and comparable to pure glucose (100). For comparison, table sugar has a GI of only 65 because of the fructose component. Maltodextrin is not sugar, so products made with them can legitimately be called sugar free, even if they are 99% maltodextrin and therefore, from a metabolic standpoint, 99% glucose (by weight). I say by weight because maltodextrins can be made to be very light-weight, which will lower glucose by volume (a teaspoon is a volumetric measure). However, whether this is a significant effect will depend on the product and how it is packaged.

Other fillers include dextrose, inulin, cellulose or one of the SA. Dextrose is glucose – it is a different name for the same thing. Inulin and cellulose are low GI dietary fibres. Truvia uses one of the SA – erythritol, meaning it can claim to be non-caloric since erythritol is not metabolised. However, it is marketed as a ‘natural stevia extract’, when in fact it is just erythritol with a touch of added stevia. It is best to look closely at the Ingredients List. The ingredients are ordered from most to least.

If you want to use HIS, it is probably best to use them in liquid form. If diluted, check the solvent.

(4) Consuming HIS and SA

Health authorities, by recommending a limit to daily sugar consumption, have initiated a paradigm shift away from sugar towards sugar substitutes that started around the turn of the present century. We do not currently have the data to predict the consequence of this for individuals or for population health, particularly for certain high-risk individuals such as pregnant and lactating women, diabetics, people with migraine or epilepsy (for example) and, particularly, children. FDA approval assumes that all people are alike and that all HIS are alike in their impact.

And consumers will not necessarily lose weight on these sweeteners, if weight loss is the aim. There is no convincing evidence for that, and a number of studies point to weight gain (likely due to compensatory consumption). US data show that sugar consumption has declined since 2000 as we moved to sugar substitutes, however, obesity has continued to increase. Metabolically, HIS have the potential to confuse appetite/satiety signalling – the brain gets a sweetness cue from the tongue without a follow-up signal from the gastrointestinal tract. Furthermore, there is evidence that some HIS can cross the blood-brain barrier (BBB) where they have the potential to interfere with our main appetite-regulator in the hypothalamus. It is not certain what effects HIS, and more particularly SA, will have on our gut microbiota over time.

A better approach would be to help people become less dependent on the taste of sweet. Taking carbonated drinks as an example, campaign not for lowering sugar but instead campaign against the consumption of the drinks themselves. We are highly-adapted self-regulating organisms that have only recently been made dependent on sugar. There is no biological reason we cannot reverse that, but there are certainly industry and marketing reasons. It is up to the individual.

It is for these reasons that I don’t recommend sugar substitutes, except perhaps in the early transition phase of a ketogenic diet to help with lowering sugar intake. In that case, perhaps the SA have a role, albeit a temporary one. In the longer term, it is better to adapt to a low-sugar diet and, in doing so, become more sensitised to sweet. In this way, it is still possible to enjoy the occasional sweetness.

(5) Where is the sugar?

The sugar industry, said to be the most powerful food lobby group in Washington, has pushed sugar into every niche it can find, including the savoury world. Most sauces (barbecue, ketchup, tomato, Worcestershire, mayonnaise) will contain sugar. Marinated meat will contain sugar (sugar browns, whereas none of the sugar substitutes do). Often, so do brines (including those used in curing bacon). Sugar is often added to canned food (baked beans). Frozen fruit is likely to be dusted with sugar (as a preservative during freeze/thaw). As a rule of thumb, be wary of anything in a can, jar or packet, and mainly eat food you make yourself.

These sugars are hidden, but sometimes they are out in the open – for example, sweet-and-sour sauce.

Sweet-and-sour chicken is ubiquitous in Chinese-style restaurants. This dish was created by a chef from the Chinese province of Hunan (Peng Chang-kuei), who fled to Taiwan (after the fall of the Nationalist government), and then moved to New York in the 1950s to open a restaurant. The original recipe contained no sugar – the Hunanese dislike the combination of sweet and savoury tastes and it doesn’t occur in their cuisine. The sugar was added in NY to appeal to the local preference for a sort of childlike sweetness and familiarity. The original sauce contained tomato paste, soy sauce, vinegar and stock. The Americanised version swapped out the tomato paste for sugar. It became a favourite of Henry Kissinger, who brought Peng and his dish to public notice. Thus, sweet-and-sour sauce is neither traditional, nor is the sweet even necessary. When Peng later opened a restaurant in Taiwan serving his new version, it failed. To Westerners though, it remains one of the fundamentals of Chinese cooking.

(6) Discovery of HIS

The discovery of many HIS is interesting, if somewhat alarming. Usually, it was accidental and down to poor laboratory practice. Laboratory safety practices were a work in progress (only becoming better established in the 1970s) and, unsurprisingly, chemists had a shortened life-expectancy for their times. Commonly, chemists unwittingly got the chemical on their fingers and later noticed residual sweetness when eating, or licking their fingers to turn over a page in a notebook. The sweetener cyclamate was approved by the FDA in 1958, however, approval was withdrawn in 1968 and therefore cyclamate was not covered here (it remains approved in Australia, Europe and Canada, amongst other countries). In this case, the chemist (Michael Sveda) had put his cigarette on the workbench while he was working, and noticed sweetness when he resumed smoking. Perhaps the most remarkable discovery was that of sucralose (as used in Splenda – the top selling sweetener, ahead of Truvia). In 1976, in a laboratory at King’s College London, an Indian student (Shashikant Phadnis) was asked by his supervisor (Leslie Hough) to test a certain chlorinated sugar. Phadnis thought he had been asked to taste it. So he did. It was patented the same year.

Print This Post Print This Post

Further reading:

Lenhart and Chey. A Systematic Review of the Effects of Polyols on Gastrointestinal Health and Irritable Bowel Syndrome
Hofman et al. Nutrition, Health, and Regulatory Aspects of Digestible Maltodextrins
Burke and Small. Physiological mechanisms by which non-nutritive sweeteners may impact body weight and metabolism
Bruyere e al. Review of the nutritional benefits and risks related to intense sweeteners
Yang. Artificial sweeteners and the neurobiology of sugar cravings
Walters. The Sweetener Book
Fushia Dunlop. Revolutionary Chinese Cooking