Of Mice and Food Additives was going to be the title of this post but I opted against it to keep with my 'Why's that in my food' series; but I thought I was clever so I'm writing it here now. Nature has had a couple papers lately related to nutrition...well, more food additives. I wasn't particularly impressed with either, both how they were presented and how their results were conveyed to the public.
The first paper was on artificial sweeteners, that mostly focused on a couple mouse studies, observing the effect of artificial sweeteners on insulin sensitivity and the microbiome. The first part of the study just dosed mice with 5% sweetener solutions in their water, and noted that they saw impaired insulin sensitivity mostly in the saccharin group; they didn't control/account for body weight changes throughout this period, which is a pretty big confounding factor, as previous research has demonstrated that administration of saccharin leads to weight gain in rats (9). (They mention that kcal intakes aren't *statistically* significantly different but looking in the supplement they appear clinically so). They only saw strong effects for saccharin in this part, and followed up with this rarely used sweetener. The second part of the study used the maximum dose of saccharin, termed the ADI (acceptable daily intake), alongside a high fat diet - this sets us up for a situation of impaired insulin sensitivity in mice, and leaves us wondering how relevant it could be, since the dose is quite high. Given that these data weren't for commonly used artificial sweeteners, it's hard to tell what to make of it, although the effects at this dosage are interesting given the microbiome transplant components of the study. The human saccharin challenge did nothing to control for background diet, and only had 7 people involved, showing that 4 of the 7 had a detrimental effect on blood glucose from the saccharin consumption. This was strengthened by a transfer of 2 samples (from a responder and a non-responder) to germ-free mice, to suggest that it was a real effect. However, this was again giving the ADI of saccharin, an uncommon sweetener that is chemically very distinct from sucralose (Splenda) and aspartame. Hopefully, someone will follow up with a well-designed trial in humans to see if sucralose or aspartame could have this effect, and whether we see responders or non-responders like in this trial of 7 individuals (that could very likely just be random chance). At best, this trial suggested that individuals who excessively consume artificial sweeteners will experience a negative effect mediated by the microbiome, and that the rest of their diet matters significantly. We can't forget that we have many well designed randomized controlled trials that show that artificial sweeteners can be successful additions to weight loss/maintenance regimens (2). RCTs specifically looking at stevia glycosides, sucralose, and saccharin haven't supported the idea that they alter glucose homeostasis (3-6). Despite these relatively lackluster results, the authors conclude: "Our findings suggest that NAS may have directly
contributed to enhancing the exact epidemic that they themselves were
intended to fight." Quite the strong statement with rather limited evidence... This was reflected in the media coverage, which took this and ran, making all sorts of claims about artificial sweeteners safety, even though this study really only focused on one sweetener with little relevance to the American diet - see some of our usual offenders here, here . As a random fun fact, at one point in nutrition history, there was concern that saccharin caused hypoglycemia in individuals (8) - oh how the times change! For any concerned about other aspects of artificial sweetener safety, see the most current reviews (10,11). Keep in mind that artificial sweeteners are chemically diverse molecules and are handled by the body differently - i'd be quite surprised if the hydrolysis of aspartame in the small intestine to 2 amino acids and methanol yields the same results as sweeteners that are not efficiently absorbed/metabolized and pass through to the large intestine. It's far too easy to ignore chemical diversity and just lump these all under 'artificial sweeteners'.
The second study that occurred recently looked at the issue of dietary emulsifiers. The study authors provided wild-type mice and IL10 knockouts with solutions of 2 'emulsifiers', carboxymethylcellulose, and and polysorbate-80. The authors report finding that wild type mice show low grade inflammation and colitis develops in those mice that are genetically prone. The paper in and of itself doesn't strike me as having any really strong methodological problems (from what I can tell) - Ruth Ley is an author on it, she's pretty much the microbiome queen here at Cornell. However, the paper suggests that the 1% solution of these 2 emulsifiers used in water represents what humans would consume, because the FDA allows their use in foods at 1%. But as was noted by Tom Sanders in his comments at sciencemediacentre, these levels are quite high. The ADI for polysorbate-80 is 10mg/kg of body weight as set by the EFSA, yet the amount fed here, given a mouses typical intake of 5mL of water, corresponds to 2500mg/kg (5mL/d x 1g/1000mL = 50mg/d for a 20g mouse, which according to the ADI, should consume, at max, .2mg/d...). In America, the amount of polysorbate 80 varies highly depending on the food product (12). There isn't an ADI for cellulose compounds that i'm aware of. While the amount that humans take in of these isn't tracked closely, the amounts used in this study are highly likely to be extreme levels. It's worth noting that if this mechanism is due to emulsifiers, there are many naturally occurring emulsifiers in foods, such as the lecithin in egg yolks - so, if this does pan out to be true in humans (a relatively BIG if..), this isn't a 'processed' food issue, it's a food issue. Despite this, those with their biases will still use it to fit their perspective:
I always worry when I see nutrition-related papers in non-nutrition journals, because the peer-review tends to be lackluster. These 2 studies show what dosing high levels of food additives can do to microbiomes, and offer little relevance to humans, yet this is not how they are reported on.
I'd like to hope its not, but if the recipe for a Nature paper is to toss high doses of substances and see their effects on the microbiome, I've got a bottle of iron pills in my cabinet at home and a cohort of mice just waiting to have their feces collected...
12. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfCFR/CFRSearch.cfm?fr=172.840
The first paper was on artificial sweeteners, that mostly focused on a couple mouse studies, observing the effect of artificial sweeteners on insulin sensitivity and the microbiome. The first part of the study just dosed mice with 5% sweetener solutions in their water, and noted that they saw impaired insulin sensitivity mostly in the saccharin group; they didn't control/account for body weight changes throughout this period, which is a pretty big confounding factor, as previous research has demonstrated that administration of saccharin leads to weight gain in rats (9). (They mention that kcal intakes aren't *statistically* significantly different but looking in the supplement they appear clinically so). They only saw strong effects for saccharin in this part, and followed up with this rarely used sweetener. The second part of the study used the maximum dose of saccharin, termed the ADI (acceptable daily intake), alongside a high fat diet - this sets us up for a situation of impaired insulin sensitivity in mice, and leaves us wondering how relevant it could be, since the dose is quite high. Given that these data weren't for commonly used artificial sweeteners, it's hard to tell what to make of it, although the effects at this dosage are interesting given the microbiome transplant components of the study. The human saccharin challenge did nothing to control for background diet, and only had 7 people involved, showing that 4 of the 7 had a detrimental effect on blood glucose from the saccharin consumption. This was strengthened by a transfer of 2 samples (from a responder and a non-responder) to germ-free mice, to suggest that it was a real effect. However, this was again giving the ADI of saccharin, an uncommon sweetener that is chemically very distinct from sucralose (Splenda) and aspartame. Hopefully, someone will follow up with a well-designed trial in humans to see if sucralose or aspartame could have this effect, and whether we see responders or non-responders like in this trial of 7 individuals (that could very likely just be random chance). At best, this trial suggested that individuals who excessively consume artificial sweeteners will experience a negative effect mediated by the microbiome, and that the rest of their diet matters significantly. We can't forget that we have many well designed randomized controlled trials that show that artificial sweeteners can be successful additions to weight loss/maintenance regimens (2). RCTs specifically looking at stevia glycosides, sucralose, and saccharin haven't supported the idea that they alter glucose homeostasis (3-6). Despite these relatively lackluster results, the authors conclude: "Our findings suggest that NAS may have directly
contributed to enhancing the exact epidemic that they themselves were
intended to fight." Quite the strong statement with rather limited evidence... This was reflected in the media coverage, which took this and ran, making all sorts of claims about artificial sweeteners safety, even though this study really only focused on one sweetener with little relevance to the American diet - see some of our usual offenders here, here . As a random fun fact, at one point in nutrition history, there was concern that saccharin caused hypoglycemia in individuals (8) - oh how the times change! For any concerned about other aspects of artificial sweetener safety, see the most current reviews (10,11). Keep in mind that artificial sweeteners are chemically diverse molecules and are handled by the body differently - i'd be quite surprised if the hydrolysis of aspartame in the small intestine to 2 amino acids and methanol yields the same results as sweeteners that are not efficiently absorbed/metabolized and pass through to the large intestine. It's far too easy to ignore chemical diversity and just lump these all under 'artificial sweeteners'.The second study that occurred recently looked at the issue of dietary emulsifiers. The study authors provided wild-type mice and IL10 knockouts with solutions of 2 'emulsifiers', carboxymethylcellulose, and and polysorbate-80. The authors report finding that wild type mice show low grade inflammation and colitis develops in those mice that are genetically prone. The paper in and of itself doesn't strike me as having any really strong methodological problems (from what I can tell) - Ruth Ley is an author on it, she's pretty much the microbiome queen here at Cornell. However, the paper suggests that the 1% solution of these 2 emulsifiers used in water represents what humans would consume, because the FDA allows their use in foods at 1%. But as was noted by Tom Sanders in his comments at sciencemediacentre, these levels are quite high. The ADI for polysorbate-80 is 10mg/kg of body weight as set by the EFSA, yet the amount fed here, given a mouses typical intake of 5mL of water, corresponds to 2500mg/kg (5mL/d x 1g/1000mL = 50mg/d for a 20g mouse, which according to the ADI, should consume, at max, .2mg/d...). In America, the amount of polysorbate 80 varies highly depending on the food product (12). There isn't an ADI for cellulose compounds that i'm aware of. While the amount that humans take in of these isn't tracked closely, the amounts used in this study are highly likely to be extreme levels. It's worth noting that if this mechanism is due to emulsifiers, there are many naturally occurring emulsifiers in foods, such as the lecithin in egg yolks - so, if this does pan out to be true in humans (a relatively BIG if..), this isn't a 'processed' food issue, it's a food issue. Despite this, those with their biases will still use it to fit their perspective:
I always worry when I see nutrition-related papers in non-nutrition journals, because the peer-review tends to be lackluster. These 2 studies show what dosing high levels of food additives can do to microbiomes, and offer little relevance to humans, yet this is not how they are reported on.
I'd like to hope its not, but if the recipe for a Nature paper is to toss high doses of substances and see their effects on the microbiome, I've got a bottle of iron pills in my cabinet at home and a cohort of mice just waiting to have their feces collected...
1. http://www.nature.com/nature/journal/v514/n7521/full/nature13793.html
2. http://ajcn.nutrition.org/content/early/2014/06/18/ajcn.113.082826.full.pdf+html
3. http://www.ncbi.nlm.nih.gov/pubmed/18397817?dopt=Abstract
4. http://www.ncbi.nlm.nih.gov/pubmed/18555575?dopt=Abstract
5. http://www.ncbi.nlm.nih.gov/pubmed/14647086?dopt=Abstract
6. http://www.ncbi.nlm.nih.gov/pubmed/2975554?dopt=Abstract
7. http://www.nature.com/nature/journal/v519/n7541/full/nature14232.html
8. http://www.ncbi.nlm.nih.gov/pubmed/4127435
9. http://www.ncbi.nlm.nih.gov/pubmed/2303911
10. http://www.ncbi.nlm.nih.gov/pubmed/23891579
11. http://www.ncbi.nlm.nih.gov/pubmed/2384527312. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfCFR/CFRSearch.cfm?fr=172.840
Comments
Post a Comment