DO SWEETENERS AFFECT THE GUT MICROBIOME? What does the science and evidence tell us? Alexandra Lobach, M.Sc., Ph.D. Manager, Toxicology, Chemistry & Regulatory Affairs Food & Nutrition Health, Environmental & Regulatory Services (HERS) November 9 th, 2017
AGENDA 1) Why the concern? Suez et al. (2014) 2) The gut microbiome Role, function, & dietary modulation 3) Review of the scientific literature: Low/no-calorie sweeteners & the gut microbiome Analysis of the reported nonclinical and clinical studies 4) Molecular and metabolic considerations Different chemical structures lead to different systemic handling 5) Summary & conclusions 2
RECENT HIGH PROFILE ARTICLE CONCLUDED THAT ARTIFICIAL SWEETENERS ALTER THE GUT MICROBIOTA Study objective: To determine the effects of low-calorie sweetener (LCS) consumption (saccharin, aspartame and sucralose) on glucose metabolism and the intestinal microbiota in mice and humans. Study conclusion: Low-calorie sweetener consumption in both mice and humans enhances the risk of glucose intolerance and that these adverse metabolic effects are mediated by modulation of the composition and function of the microbiota. 3
THE GUT MICROBIOME
ROLE AND FUNCTION OF GUT MICROBIOTA Microbiome = trillions of symbiotic microbial cells present in the human body primarily found in the gut Gut microbiota play an important role in the breakdown and metabolism of key dietary components Contribute enzymes that are not encoded by the human genome Extensive metabolic repertoire: hydrolysis, reduction, dehydroxylation, deamination, and ring fission The contributions of host genetics, environmental factors, and diet to shaping the gut microbiota and its functionality remain largely undefined Significant accumulating evidence indicates that diet appears to be the most dominant contributing factor 5
MODULATION OF THE GUT MICROBIOTA BY DIET Evidence indicates that habitual, long term diet, and short term dietary changes all contribute to the overall composition of the gut microbiota Two primary mechanisms are hypothesized: 1. Microbial utilization of dietary substrates Dietary components that are poorly digested in the stomach and small intestine, such as fibre, reach the colon and provide substrates for microbial fermentation 2. Alteration of the microbial environment in the gut Dietary components change the physico-chemical conditions of the gut (e.g., ph, bile salts) - microbial species differ in their tolerance to different environmental conditions Since the diet is considered to be the major contributing factor in determining the composition of the gut microbiota it is therefore critical that studies looking at effects on the microbiome strictly control for dietary intake and changes. 6
REVIEW OF THE SCIENTIFIC DATA NONCLINICAL AND CLINICAL STUDIES INVESTIGATING LOW/NO- CALORIE SWEETENERS AND THE GUT MICROBIOME
ANIMAL STUDIES INVESTIGATING EFFECTS OF LNCS CONSUMPTION ON GUT MICROBIOTA Sweetener Study (details) Microbiome-associated changes Study Confounders Acesulfame potassium Bian et al. 2017 (Mice, 37.5 mg/kg/d, 4 weeks) Uebanso et al. 2017 (Mice, ~15 mg/kg/d, 8 weeks) Aspartame Palmnäs et al. 2014 (Normal rats, 7 mg/kg/d; obese rats, 5 mg/kg/d; 8 weeks Suez et al. 2014 (Mice, 1333 mg/kg/d, 11 weeks) Cyclamate Matsui et al., 1976 (Monkey, 250 mg/kg/d; 30 days) Males: in faecal Bacteroides, Anaerostipes, and Sutterella Females: in faecal Mucispirillum; Lactobacillus, Clostridium, an unassigned Ruminococcaceae genus, and an unassigned Oxalobacteraceae genus No changes in faecal microbiota between groups Normal rats: faecal C. leptum Obese rats: faecal total bacteria, Bifidobacterium spp., Enterobacteriaceae, C. leptum, and Roseburia spp. Food consumption not reported Dose in excess of ADI (2.5x) None to report Food, water intake was not comparable between aspartame and control groups Elevated glycaemic response Dose in excess of ADI (~30x) Food, water consumption not equivalent between groups Unconventional grouping of data; invalid statistical analysis Compared 1 cyclohexylamine producer to 3 controls not consuming cyclamate No change in total faecal bacteria and several microbial populations Small sample size Control animals were different species than test group 8
ANIMAL STUDIES INVESTIGATING EFFECTS OF LNCS CONSUMPTION ON GUT MICROBIOTA (CONT D) Sweetener Study (details) Microbiome-associated changes Study Confounders Saccharin Anderson & Kirkland, 1980 (Rats, 10 14 g/kg/d, 10 days) number of aerobes in caecum caecal anaerobe:aerobe ratio Dose in excess of ADI (~2000x) Suez et al. 2014 (Mice, 3333 mg/kg/d, 11 weeks) Suez et al. 2014 (Mice, 5 mg/kg/d, 5 weeks: fecal transplant germ-free mice) Daly et al. 2014 Daly et al. 2016 (Piglets, 0.015% SUCRAM, 2 weeks) Bian et al. 2017 (Mice, ~27-65 mg/kg/d, 6 months) Elevated glycaemic response faecal abundance Bacteroides, Clostridiales faecal L. reuteri, members of Clostridiales Elevated glycaemic response transferred via faeces to germ-free mice Minor changes (±1.2-fold) in faecal composition in faecal Lactobacillus OTU4228 [authors reported NHDC component of SUCRAM responsible] in faecal Veillonellaceae and Ruminococcaceae in faecal Corynebacterium, Roseburia, and Turicibacter in faecal Ruminococcus, Adlercreutzia, and Dorea Dose in excess of ADI (~650x) Food, water consumption not equivalent between groups Glycaemic responses in donor animals not consistent unclear if all animals were donors, or if only high-response animals were donors Food consumption not reported Saccharin dose (mg/kg) unknown Food consumption not reported Dose appears to be in excess of ADI (at minimum ~5x) 9
ANIMAL STUDIES INVESTIGATING EFFECTS OF LNCS CONSUMPTION ON GUT MICROBIOTA (CONT D) Sweetener Study (details) Microbiome-associated changes Study Confounders Sucralose Suez et al. 2014 (Mice, 1666 mg/kg/d, 11 weeks) Rebaudioside A Harding et al. 2014 (Mice (experimental Crohn s disease), Splenda, dose not reported, 6 weeks) Abou-Donia et al. 2008 (Rats, Splenda sucralose equivalents 1.1 to 11 mg/kg/d, 12 weeks) Bian et al. 2017 (Mice, ~9-22 mg/kg/d, 6 months) Uebanso et al. 2017 (Mice, ~1.5 & 15 mg/kg/d, 8 weeks) Li et al. 2014 (Mice, 5.5 or 139 mg/kg/d, 4 weeks) Elevated glycaemic response Dose in excess of ADI (~100x) Food, water consumption not equivalent between groups in ileal bacteria and Bacteroidetes:Firmicutes ratio in faecal Lactobacillus and E.coli in faecal total anaerobes, bifidobacteria, lactobacilli, and Bacteroides (all doses) Clostridia abundance & total aerobic bacteria (3 highest doses) in faecal Turicibacter, Roseburia, Akkermansia, Clostridiaceae, and Christensenellaceae in faecal Ruminococcus, Streptococcus, Dehalobacterium, and Erysipelotrichaceae in faecal Clostridium IVXa, dose-dependent None to report diversity of faecal lactobacilli (high dose only) No control for sucralose alone Sucralose dose (mg/kg) unknown Food consumption not reported No control for sucralose alone Significant differences in body weight Food consumption not reported Bacterial counts not standardized: faecal weights reported on wet basis Food consumption not reported Unclear if exposure was below ADI for entirety of study High dose in excess of ADI (~10x) Food consumption not reported 10
COMMON ISSUES WITH ANIMAL STUDIES WHEN STUDYING THE GUT MICROBIOTA Lack of isocaloric control groups to account for differences in caloric intake between sweetener and control groups Food consumption and body weight monitoring essential Human dietary relevance is limited because: Utilizing sweetener doses that are significantly greater than the ADI No dose reported at all, often due to lack of food intake monitoring Studies conducted with the commercial product which contains low levels of the sweetener and high levels of a carbohydrate bulking agent (e.g. Maltodextrin) Difficulties translating microbiome findings in animals to humans Example: although the distal gut microbiota of mice and humans harbour the same bacterial phyla, most bacterial genera and species found in mice are not present in humans 11
CLINICAL STUDIES INVESTIGATING EFFECTS OF LNCS CONSUMPTION ON GUT MICROBIOTA Sweetener Study (details) Microbiome-associated changes Study Confounders Acesulfame potassium Frankenfeld et al. 2015 (Humans, consumers & nonconsumers based on daily food records over 4 days) Aspartame Frankenfeld et al. 2015 (same details as above) Saccharin Suez et al. 2014 (Humans, 5 mg/kg/d, 1 week, 7 subjects) Undefined LNCS Suez et al. 2014 (Observational study, 381 nondiabetic individuals, subjects divided for analysis into high- and non-lncs consumers) Bacterial diversity different between consumers & nonconsumers Bacterial diversity different between consumers & nonconsumers Elevated glycaemic responses in 4 subjects, classified as responders Responders microbiome configurations clustered differently Correlations between LNCS consumption and several metabolic syndrome-related clinical parameters Habitual diet not controlled Habitual diet not controlled No control group Habitual diet not controlled Unconventional grouping of data No information on LNCS consumed, amount, or other components of diet Observational study cause and effect cannot be assumed 12
PREVIOUS CLINICAL DATA REPORT LACK OF ASSOCIATION BETWEEN LNCS CONSUMPTION, GLYCAEMIC RESPONSE, & GUT MICROBIOTA Consistent clinical evidence in diabetic and non-diabetic patients reports that LNCS consumption does not alter glycaemic response: A) Saccharin Consumption of 1g/day for 4 weeks by 15 healthy volunteers had no effect on urinary levels of indican, total phenol, or p-cresol (Roberts & Renwick, 1985) (Lawrie & Renwick, 1987) Diabetic patients consuming up to 4.8g/day for 5 months were not observed to have any adverse effects (Neumann, 1926a,b) B) Aspartame No effects on glycaemic response in diabetics reported (Colagiuri et al., 1989; Stern et al., 1976; Nehrling et al., 1985; Okuno et al., 1986) C) Sucralose Consumption of 1 g/day for 12 weeks had no effect on glucose response (Grotz et al., 2003; SCF, 2000; Baird et al., 2000) 13
MOLECULAR AND METABOLIC CONSIDERATIONS
DIFFERENCES IN METABOLISM & LOW EXPOSURE LEVELS LIMIT POTENTIAL TO INTERACT WITH GUT MICROBIOTA Sweetener Molecular Structure ADME Sucrose sweetness equivalence Acesulfame K Not metabolized. Rapidly absorbed. Excreted unchanged in urine. ADI (mg/kg bw/d) Max daily mg intake based on 70kg person 200 x 15 1050 Aspartame Rapidly metabolized to amino acids & methanol that are absorbed in small intestine. 200 x 40 2800 Saccharin Not metabolized. Rapidly absorbed. Excreted unchanged in urine. 400 x 5 350 Sucralose Not metabolized. Poorly absorbed. Excreted unchanged in feces. 600 x 15 1050 15
REBAUDIOSIDE A AND CYLAMATE ARE METABOLIZED BY GUT MICROBIOTA Sweetener Molecular Structure ADME Sucrose sweetness equivalence Cyclamate Small percent of population: unabsorbed fraction can be converted by bacteria in lower GI tract to cyclohexylamine. Excreted in urine. ADI (mg/kg bw/d) Max daily mg intake based on 70kg person ~30 50 x 0 11 770 Rebaudioside A Hydrolyzed by colonic microflora to steviol that is absorbed & excreted in urine. ~300 x 4 280 16
SUMMARY & CONCLUSIONS The studies currently present in the scientific literature provide no evidence that any LNCS alters the gut microbiota in humans at currently permitted human intake levels Consistent clinical evidence continues to support the safety of these food ingredients for their intended uses Numerous clinical trials and long-term animal studies show no effect of permitted LNCS on blood glucose No adverse health effects mediated by gut microflora changes can be assumed based upon the available data Differences in structure and systemic handling means that different sweeteners cannot be grouped together when looking at impacts on the gut microbiome No evidence to suggest that LNCS as a group or individually pose any safety concerns at currently approved levels, which is a viewpoint endorsed by all the major international regulatory authorities 17
THANK YOU! Alexandra Lobach, M.Sc., Ph.D. Manager, Toxicology, Chemistry & Regulatory Affairs Food and Nutrition Group Health, Environmental & Regulatory Services (HERS) Intertek, 2233 Argentia Rd., Suite 201, Mississauga, Ontario L5N 2X7 CANADA +1 905-542-2900 alexandra.lobach@intertek.com www.intertek.com