Sweet’n Low

I didn’t want to blog about the artificial sweetener study; to be honest, I didn’t even want to read it.  I just wanted to report: 1) how many Diet Cokes are we talking about; and 2) when are you going to die.

Artificial sweeteners induce glucose intolerance by altering the gut microbiota (Suez et al., 2014)

Non-caloric artificial sweeteners (NAS) = saccharin, sucralose, and aspartame. Saccharin worked the best (worst) in the mouse study, so they tested it in humans.  This was the part I found most relevant: seven healthy volunteers (5 men & 2 women, aged 28-36) who did not typically consume a lot of sweeteners were recruited and given 120 mg saccharin three times per day.  360 mg saccharin is ~10 packets of Sweet’n Low.

Exhibit A: 3/7 people tolerated it just fine; the other 4 not so much… or not? The study design, or rather data presentation, is cryptic.  A glucose tolerance test was conducted every day of the experiment.  They combined data from days 1-4 and days 5-7.  This doesn’t make much sense: why not just use day 1 vs. 7?  Artificially inflating statistical power..?

glucose tolerance test
There was only 4 participants in the graph on the left.  How could the data possibly be so tight?  The “Days 1-4” line represents 16 data points (GTT from 4 people, each on days 1, 2, 3, & 4).  Similarly, “Days 5-7” represents 12 data points.  For the Non-responders, “Days 1-4” = 12 data points; “Days 5-7” = 9 data points.

…foul play?  I guarantee the figures above would look MUCH sloppier, and probably lack statistical significance if we were just looking at Day 1 vs. Day 7.



Did something magical happen on day 5 that was fundamentally different from day 4? It appears so, at least for “Responder #3” :s

glycemic response



^^^those are individual data points… that degree of variability pretty much confirms the above speculation about the tight GTT data.

Conclusion of part 1: statistical wizardry produced a compelling GTT figure.  Individual data points confirm statistical wizardry was performed (P<0.05).


Exhibit B. Gut microbes shifted more in responders than non-responders.  I think.

gut microbes days 1 and 7


Fwiw, comparing Day 1 to Day 7 makes more sense.

gut microbes responders non-responders


Conclusion of part 2:  I don’t know.


Exhibit C. Unfortunately, the most compelling data are from more animal studies.  The human data don’t scare me.

next logical step? Fecal transplants: smear human shit on mice, which for some reason I find more gross than smearing mouse shit on mice.  But the data recapitulate the human study quite nicely:

transplanted sample from 1 and 3


above: fecal transplant from a responder on day 7 (after a week of Sweet’n Low) causes insulin resistance relative to a fecal transplant from the same responder prior to Sweet’n Low.  Compelling and consistent.

transplanted sample day 1 and 7

Some of the more marked changes in the microbiome were: huge increases in Bacteroides fragilis, Weissella cibaria, and Candidatus Arthromitus.

Bacteroides fragilis: associated with high protein diet; mixed findings on body weight.


Weissella cibaria: associated with less stinky gas?  Found in some fermented foods.

Candidatus Arthromitus: found in bugs guts (eg, cockroach, termite, etc.)… Arthromitus, arthropod.

My take on this study: diet!?

What differentiates a responder from a non-responder?  Responders appeared to have a different microbial profile at the beginning of the study.  And it appeared to change more than the non-responders, but I have to ask: 1) really? bc hard to tell from the figures; 2) if so, did it change because “responders” responded to the sweeteners by changing their diet? Was it different in the beginning because of different baseline diet? In a study with only 7 participants, this information could’ve & should’ve been comprehensibly assessed, reported, and discussed.  

And grouping days 1-4 and 5-7 seems weird or misleading at best, nefarious at worst.

Mixed findings in PubMed:

Sucrose versus saccharin as an added sweetener in non-insulin-depedent diabetes: short- and medium-term metabolic effects (Cooper et al., 1988)
“No significant treatment effects were observed on fasting concentrations of blood glucose, plasma insulin or serum triglycerides, or on urinary excretion of glucose, sodium or potassium. Following a standard breakfast with either sucrose or saccharin and starch, no differences between meal responses were observed.”

Drinking saccharin increases food intake and preference (Tordoff and Friedman, 1989)

Antihyperglycaemic effect of saccharin in diabetic ob/ob mice (Bailey et al., 1997)

Influence of artificial sweetener on human blood glucose concentration (Skokan et al., 2007)

In any case, 10 packets of Sweet’n Low is probably too many packets of Sweet’n Low… just sayin’


If you like what I do and want to support it, head over to Patreon! Five bucks a month gets you full access to all articles and there are many other options. It’s ad-free and you can cancel if it sucks 🙂

Also, I’m open to suggestions so please don’t hesitate to leave a comment or email me directly at drlagakos@gmail.com.


Affiliate discounts: if you’re still looking for a pair of hot blue blockers, Carbonshade is offering 15% off with the coupon code LAGAKOS and Spectra479 is offering 15% off HERETrueDark is running a pretty big sale HEREIf you have no idea what I’m talking about, read this then this.

20% off some delish stocks and broths from Kettle and Fire HERE

If you want the benefits of  ‘shrooms but don’t like eating them, Real Mushrooms makes great extracts. 10% off with coupon code LAGAKOS.


calories proper

Be Sociable, Share!
  • Colby Vorland had a nice piece summarizing this article too, looking at the other literature in the field http://www.weightymatters.ca/2013/07/guest-post-what-reading-that-artificial.html

  • commonchild2

    Watching some of my golfing buddies put S&L in their iced tea, 10 packets is not that unheard of for a meal :/ It’s not like it is 100 or 1000, which is the non-human comparable dose in other studies. Wish they did more on Splenda, my AS of choice.

    • “Wish they did more on Splenda, my AS of choice.”

      the researchers stacked the deck, by choosing the sweetener that caused the most damage in mice.

      P.S. 10 packets of S&L in their tea!? wow

      • commonchild2

        For three teas over a meal, sure.

  • Gerard Pinzone

    There are quite a large number of people who seem to view the world like a roadrunner cartoon. The coyote chases the roadrunner, runs off a cliff, but doesn’t fall until he realizes his predicament. The idea that the laws of nature are suspended until awareness is raised is absurd, of course. New evidence refines and extends our understanding rather than completely reverse it (assuming the existing knowledge was rigorously confirmed in the first place.)

    If there has been little to no experimental evidence that artificial sweeteners cause weight gain or poor glucose control, that doesn’t magically change one day once new evidence is discovered. If you haven’t been experiencing ill effects, this study really doesn’t impact you in the slightest. Nevertheless, any food’s effects on gut bacteria is a serious concern. I really do hope more research on humans is performed. I’d be interested to know what else could similarly effect gut bacteria in responders. When Bruce Ames developed his method to determine how mutagenic an item can be (thus potentially lead to cancer), he first tested things like foods treated with pesticides. They were indeed mutagenic. Organic food proponents praised him and his discovery. But being a good scientist, he wanted to see what happened when his test was applied to organic foods, too. It turns out they are even MORE mutagenic! (BTW, I’m not implying organic food is carcinogenic. The effects are very small and aren’t worth fretting about.)

    The point is that we shouldn’t just focus on foodstuffs we “know” MUST be bad for us. We also cannot assume other foods which we likely “know” MUST be healthy (usually because they’re “natural”) are automatically absolved from scrutiny.

    • Epic philosophy bomb 🙂

      My only qualm with this particular study, was with the most important part: the human intervention study. And it’s just way too suspect.

      • Gerard Pinzone
        • …everyone there pretty much said the same thing about the human intervention study: only 4/7 people responded…

          my critique was that it is odd (and fishy) to group days 1-4 and 5-7.

          • Gerard Pinzone

            It’s hard to say. The narrative of the mouse study reads so beautifully, but once it gets to the human part, it abruptly shifts gears and leaves you wondering what they were thinking. It’s like the Sopranos finale all over again.

          • “It’s hard to say.”

            disagree! The mouse stuff is quite compelling, but humans are about as far from germ-free mice as you can get! …and I’m confident the raw data (not averages of days 1-4 & 5-7) would sing a different tune

          • Gerard Pinzone

            Look out guys, we’ve got badass over here.


          • Jack Kruse

            Anytime I see a mouse study that jumps to humans I see a PhD looking for funding and not to contribute science to the literature.

    • Dustin Sikstrom

      “We also cannot assume other foods which we likely “know” MUST be healthy (usually because they’re “natural”) are automatically absolved from scrutiny.”

      But everyone knows veggies are the true way to health!

  • Young Chipotle

    I remember asking your thoughts on this paper on twitter when it came out. Good to see you got around to it 😉

    I enjoyed the point on statistical wizardry. Never trusted those statisticians.
    The paper overall wasn’t as meaningful to me as it was to the hippies who claimed to me they were right all along. It mostly pointed towards the need for research on gut microbiota and NAS.

    The presentation of the gut metagenome is Über-sloppy. I was hoping to take a deeper look at the data myself and tried looking for the metagenome data but couldn’t find it. My 4th year metagenomics class could output better visuals using MEGAN5 better than the authors.

    PS Sweet N’ Low tastes like crap. Use Aspartame and Splenda.

    • “The paper overall wasn’t as meaningful to me as it was to the hippies who claimed to me they were right all along.”

      HA! HAhahaha 🙂