Think Twice a double dose of reason


Thinking Twice About Attacks on Artificial Sweeteners

by: Ross

I was checking out the “My Athletic Life” blog last week and I read an article titled “The Dish on Sugar and Sweeteners” by Diane Sanfilippo.  While I generally agree with her overall point that sweeteners may have negative effects on health and fat loss, I think there were a number of points made in her post that were problematic and worth thinking twice about.  I thought I would deal with them in sections that correspond with the sections of Sanfilippo’s post.

Section I. “All Sweeteners Can Contribute to Weight Gain or Prohibit Weight Loss”

Sanfilippo’s first section contends that all sweeteners, including non-nutritive sweeteners, can contribute to weight gain or stymie weight loss.  Her argument relies heavily on the idea that cephalic phase insulin response (insulin released by the pancreas in response to the taste of sweet food) causes fat storage, even if the sweet tasting food was non-nutritive.  As she sums it up, “Eat something sweet (regardless of calories) > Insulin is released > Body stores nutrients.”

Sanfilippo encourages her readers to use their independent judgment, rather than “[taking] her word for it,” and, to this end, she supplies three sources (2 of which are peer-reviewed) which she says support her argument.

The problem is that these sources don’t support her argument very well.  The first source[i] (chronologically) is a 1987 paper that examines the effect of acesulfame potassium injection/infusion on peripheral plasma insulin levels in rats.  In other words, this study doesn’t even look at cephalic phase insulin secretion at all—rather, it looks at the effect of artificial sweetener in the blood (not in the mouth) on insulin secretion.  While this might also be an interesting thing to study, the results aren’t too exciting in my opinion.  The researchers achieve increased insulin levels after injection of 100 mg/kg-bw (milligrams per kilogram of body weight) but not at lower levels.  I wasn’t able to find how much acesulfame-K is in a typical diet soft drink, but it has similar sweetness to aspartame, which is usually present in about 180 mg per 12-ounce can of soft drink.  Assuming similar physiology between a human and a rat (which is not always a valid assumption) this study suggests that increasing insulin secretion in the average 70 kg male would require injection of 7g (7000 mg) of acesulfame-K, which is the equivalent of ~39 diet soft drinks.  The infusion data used similarly inflated amounts of artificial sweetener to evoke increased plasma insulin concentrations.  And, perhaps least impressively, the researchers were able to evoke the same responses with equal mass infusions of glucose (sugar).  Given that most diet soft drinks have 27-30g of sugar and 180 mg of acesulfame/aspartame, it doesn’t raise my eyebrows too much to see that 180 mg of acesulfame-K might raise my insulin by as much as 180 mg (0.18g) of sugar.  That's not that much sugar.

There are numerous sources available that use grossly inflated masses of artificial sweeteners to evoke some negative effect in an animal (e.g., rat, mouse) that are then used in attempts to prove that artificial sweeteners are harmful to human beings.  I’m not convinced by these, because dose determines the poison, and few if any of these studies are able to produce negative effects with masses of artificial sweetener that actually mimic that seen in typical human consumption.

There are some sources that suggest that artificial sweeteners can produce a cephalic-phase insulin response, but these are also not particularly impressive to me.  The cephalic phase insulin response shown in these studies has been routinely shown to be brief, lasting only minutes, with a range of increase in plasma insulin between 7-75% in humans and 200% in rats.  Compare this to the average insulin increase from digesting a nutritive meal (500-1200% increase).  Given the body of research I have read, I’m not convinced that the taste of sweeteners contributes to obesity through causing insulin release. [I am open to the idea that it might contribute by other mechanisms, and I think Stephan Guyenet's posts on the food reward hypothesis of obesity to be both interesting and compelling]

Sanfilippo’s second source is a news article from MedScape that reports that an association has been found between aritificial sweeteners and Type-II diabetes mellitus.  Reading the article, however, confirms that this study is only associational and the results can easily be explained by the strong possibility that obese, type-II diabetics are more likely to use artificial sweeteners, perhaps in attempts to lose weight.

Sanfillipo’s third source[ii] is a psychological study examining the role of sweet taste in energy balance.  It’s actually pretty interesting.  The authors hypothesized that sweet-taste indicates caloric content to the brain, and that when sweet taste is decoupled from caloric content, as in foods sweetened with non-nutritive sweeteners, the body responds by increasing caloric intake.  The research was done in rats, which is a weakness when you’re trying to draw a nutritional conclusion for human beings.  Most importantly though, the source is not, as Sanfilippo suggests, an examination of the role of artificial-sweetener-induced insulin-release (particularly CPIR) in weight gain.

Sanfilippo goes on in this section of her post to provide a whole list of symptoms/problems that are allegedly caused by artificial sweeteners.  My thoughts are that, while artificial sweeteners may truly cause these symptoms in some individuals, I am not aware of any evidence suggesting that these effects are widespread (e.g, convulsions/seizures from diet soda).  Sanfilippo gives this list as evidence that everyone, even those who are not overweight, will benefit from removing artificial sweeteners from his/her diet.  To me, this list just suggests that someone experiencing one of the listed symptoms should experiment by removing sweeteners for a period of time (1-6 months), and see if the symptoms return on re-introduction of the sweeteners.  She doesn’t make a very strong case for removal of these agents for people who have healthy weight and no medical problems.

Section II. “Sweeteners from nature are always better to consume than those that come from a lab or factory”

In this section, Sanfilippo posits the argument that natural sweeteners, seemingly in any amount, are better for you than artificial sweeteners, also seemingly in any amount.  She gives the example:  “ I don’t care if it’s a fist full of honey vs. a teaspoon of Splenda – give me the honey!”  My only dispute with this is that I am not familiar with, nor does she provide, any evidence that supports this claim.  The “evidence” provided in the previous section of the post is claimed to support the idea that artificial sweeteners are as bad as natural sweeteners in one or more respects.  Since that claim is not well supported by her evidence, I have a hard time seeing how she can make the jump to claiming, as fact, that artificial sweeteners are worse for your health than natural sweeteners (e.g., cane sugar, maple syrup, etc.)

Another problem I have with this section of the post is the following quote:

“The bottom line:  ANYTHING that comes from nature will ALWAYS be better for you than something that was made in a lab or a factory. Period.”

This is pretty much a literal restatement of what is often referred to as the naturalistic fallacy.  It’s obviously not true.  Streptococcus pyogenes comes from nature; amoxicillin comes from a lab.  Rattlesnake venom comes from nature; rattlesnake anti-venom is made in a lab.  In his post on Paleo 2.0, Kurt Harris has a great Venn diagram demonstrating that foods are not necessarily agents of disease just because they are Neolithic, nor are foods necessarily healthy because they are paleolithic.  Similarly, one cannot conclude that substances are unhealthy solely because they are synthetic, and it is down-right dangerous to conclude that substances are safe/healthy solely because they are natural.

Sections III – V: Minimize Sweeteners Generally; Information/Tips For Doing So

I agree with this part of the post, that sweeteners should be generally limited, and the information is fine as far as I can tell.

What I Do with Artificial Sweeteners?

I only rarely consume anything containing artificial sweeteners.  We’re talking maybe a diet soda every few weeks.  I don’t add artificial sweeteners to anything.  However, I don’t do this because I am aware of convincing scientific data indicting artificial sweeteners as harmful to my health.  I do it almost entirely out of the conviction that eating sweetened food makes me crave sweet food, and that’s a craving I don’t want to have.  I also do take anecdotal evidence that people sometimes have a hard time getting lean or staying lean when they consume artificial sweeteners, and while I may speculate as to why this is the case, I do not present my speculation as fact.

What I Recommend Others Do With Regard to Artificial Sweeteners?

Not much.  I have not seen enough good research, either in animals or in human beings, to conclude that artificial sweeteners, as a matter of fact, are harmful to human health in general.  Everything that has thus far been presented to me as evidence of this has been either merely suggestive or a complete crock.  On the basis of anecdotal evidence, it may be advisable to eliminate these substances from your diet for a period of time (1-6 months) if you are having trouble with your weight or a seemingly causeless medical symptom.  Reintroduce them at the end of that period of time and see if the symptom comes back.  This isn’t perfect because there could be a placebo effect because you will know when you reintroduce the sweeteners (maybe you could get a spouse to prepare your coffee and not tell you when they introduce them back).  With the information I have available to me, that’s the best I can offer.

I do admit that I am a very busy PhD student at the moment, and almost all of my literature research is on the subject of inflammation in cardiovascular disease (the topic of my dissertation).  I have done brief literature searches on sweeteners from time-to-time and haven’t come across anything that strongly indicts artificial sweeteners.  My main point in writing this is that I often see people in the fitness or Paleo “communities” attacking artificial sweeteners with pretty scant evidence.  If, however, you are aware of some good papers that present a strong case against artificial sweeteners, I absolutely welcome you to email them to me or, even better, post the citation(s) in the comments section.


[i] Y Liang et al., “The effect of artificial sweetener on insulin secretion. 1. The effect of acesulfame K on insulin secretion in the rat (studies in vivo),” Hormone and Metabolic Research = Hormon- Und Stoffwechselforschung = Hormones Et Métabolisme 19, no. 6 (June 1987): 233-238.

[ii] Susan E Swithers and Terry L Davidson, “A role for sweet taste: calorie predictive relations in energy regulation by rats,” Behavioral Neuroscience 122, no. 1 (February 2008): 161-173.

Comments (0) Trackbacks (0)

Sorry, the comment form is closed at this time.

Trackbacks are disabled.