Debras and colleagues (1) recently assessed an association between consumption of low calorie sweeteners (LCS) and reported cancer cases in a French cohort, the NutriNet-Santé population. These authors previously investigated associations between total sugar intakes and cancer cases in the same cohort, and reported a positive association for higher sugar intakes (2), and sugary beverage intake but no association with beverages containing LCS and overall cancer incidence (3). In this study, they perform a very detailed complete LCS intake assessment and reassess possible associations.
The online-recruited cohort consists mostly of women (78%), average age of 42 years at baseline, with follow up ranging from only 4 to 9 years. Most importantly, over 63% of the population reported consuming no LCS at any time. Age and length of follow-up for each LCS intake category (zero, lower, higher intake) at the time of the assessment were not reported and therefore may vary considerably. Higher LSC consumers were only 40 years of age at baseline. Thus, the appropriateness of this middle-aged, primarily female cohort with a majority being non-consumers, for investigating associations between long-term exposure to LCS and diseases predominately found in adults over the age of 50, or only in men (prostate cancer) seems questionable at best.
Notably, LCS intakes by consumers were very low, even in high consumers. The authors highlight the comprehensiveness of their LCS intake as a unique strength of their study; thus the intakes are unlikely to be underestimates. The average intakes of higher and lower consumers on a mg/day basis were: 7.6 and 79.4 for total LCS; 3.2 and 47.4 for aspartame, 2.7 and 22.4 for acesulfame-K, and; 1.1 and 7.5 for sucralose. Intakes were reported only in mg/day, rather than mg/kg/day, and average body weights were not reported. Using a highly conservative traditional default of 70 kg body weight, average aspartame intake levels were less than 0.05 and 0.7 mg/kg in lower and higher consumers, respectively. Compared to the ADI for aspartame of 40-50 mg/kg/day, these are extremely low consumption levels. With very low LCS consumption levels, the ability to detect an association between LCS intakes and any outcome (positive or negative) is very difficult and confounding factors (i.e. “noise”) in the analysis become more influential. This should be recognized as a major weakness of the study, which is in stark contrast to the authors’ speculation that the low LCS intake levels can be interpreted as an underestimation of the overall risk.

Despite 10-15 fold higher consumptions of LCS in higher consumers, the Hazard Ratio (HR) for all cancers relative to non-consumers was very similar and associations remained in the extremely weak range. For example although reported as statistically significant, the fully adjusted HRs for total LCS for lower and higher consumers were 1.14 and 1.13 for all cancers; and 1.11 and 1.16 for breast cancer (only when pre and postmenopausal cancers were combined; no association was found when considered separately). The lack of change in risk despite a 10-fold change in exposure would typically be interpreted as evidence of no effect, and strongly suggests that observed weak associations of LCS consumption and cancer risk are due to the limitations described by the authors including selection bias, residual confounding risk factors that were more prevalent in the LSC consumers, and reverse causality, despite use of statistical methods to adjust for them.
As recently urged by Mela and colleagues (4), biological fate of LCS (reviewed in 5) must be considered in research and reports on LCS. The authors (and reviewers) either are unaware of or chose to not report on the extensive critical regulatory reviews on the metabolism and safety of LCS, and thus inappropriately cite widely discredited and biologically meaningless studies to support their allegations. For example, the complete digestion of aspartame in the small intestine to the same compounds resulting from digestion of proteins and fruits and vegetables, results in zero levels of intact aspartame in the blood, internal organs and the large intestine (5). This is why in vitro studies where aspartame is added directly to cells or investigations on colonic microflora are considered biologically irrelevant; however the authors cite many of these irrelevant studies and possible mechanisms in their discussion section as support for their conclusions of lack of safety of LCS. Similarly, the very rapid excretion of unchanged acesulfame K, and lack of absorption of sucralose (5) also argue against the biological plausibility of the associations reported by the authors.
The authors acknowledge the lack of support for their findings of an association between LCS consumption and breast or prostate cancer from previous epidemiological studies (6), including their previous study on the same cohort, but attributing their recent observed positive associations to improved intake assessment. The authors cite the study by Bassett et al.(7), in which participants were followed for 19 years, but fail to note that Bassett and colleagues actually conclude that their findings of very weak associations of LCS and cancer risk as not likely to be biologically relevant given the lack of evidence of carcinogenicity of LCS, and thus is probably due to unmeasured confounding or to chance. A comprehensive critical review (8) of aspartame epidemiological studies also did not find support for a positive association of LCS and cancer risk.
In addition, the overwhelming body evidence for the lack of carcinogenic potential of LCS from numerous animal studies conducted using globally recognized quality standards such as Good Laboratory Practices and accepted animal study methodology (9) is described by Debras et al. (1) as “controversial”. This is despite the fact that these studies are considered to be of highest quality and most reliable, by regulatory agencies globally that universally have deemed LCS safe at much higher consumption levels, and by ongoing regulatory reviews of LCS safety(10). In contrast, Debras et al. (1) cite the discredited Ramazzini studies as support for their hypothesis. The many flaws and inappropriate conclusions of Ramazzini studies have been described extensively by toxicology experts in the scientific literature(11, 12, 13 ) and by regulatory agencies such as EFSA, EPA, and FDA(14). Thus disproportionate weight has been given to unreliable studies, which support the authors’ allegations, relative to the large body of evidence to the contrary.
In conclusion, although the study is presented in a well-written paper that appears convincing, the statements by Debras and colleagues (1) regarding safety of LCS and cancer risk are unjustified based on extremely weak associations in a cohort that is primarily middle-aged female non-consumers. There is high probability of residual confounding from the many other risk factors that are more prevalent in consumers, and there is no evidence of increasing risk in the lower to higher consumers despite more than 10-fold differences in consumption levels, an observation that actually provides support for the safety of LCS. The sensational press releases resulting from this study and public scare mongering are without regard for the potential risks from a public health perspective for individuals who would potentially benefit from dietary sugar reductions by using LCS in place of sugar. For example, a National Cancer Institute prospective study (15) reported dramatic reductions in risk of cancer recurrence and overall mortality (adjusted HRs of 0.54 and 0.52, respectively) in colon cancer patients consuming LCS-sweetened beverages compared to patients not consuming LCS. Improved patient outcome with increased LCS-sweetened beverage consumption was attributed completely or in part to substitution for sugar-sweetened alternatives.

1. Debras C, Chazelas E, Srour B, Druesne-Pecollo N, Esseddik Y, Szabo de Edelenyi F, Agaësse C, De Sa A, Lutchia R, Gigandet S, Huybrechts I, Julia C, Kesse-Guyot E, Allès B, Andreeva VA, Galan P, Hercberg S, Deschasaux-Tanguy M, Touvier M. Artificial sweeteners and cancer risk: Results from the NutriNet-Santé population-based cohort study. PLoS Med. 2022 Mar 24;19(3):e1003950. doi: 10.1371/journal.pmed.1003950. PMID: 35324894; PMCID: PMC8946744.

2. Debras C, Chazelas E, Srour B, Kesse-Guyot E, Julia C, Zelek L, Agaësse C, Druesne-Pecollo N, Galan P, Hercberg S, Latino-Martel P, Deschasaux M, Touvier M. Total and added sugar intakes, sugar types, and cancer risk: results from the prospective NutriNet-Santé cohort. Am J Clin Nutr. 2020 Nov 11;112(5):1267-1279. doi: 10.1093/ajcn/nqaa246. PMID: 32936868.

3. Chazelas E, Srour B, Desmetz E, Kesse-Guyot E, Julia C, Deschamps V, Druesne-Pecollo N, Galan P, Hercberg S, Latino-Martel P, Deschasaux M, Touvier M. Sugary drink consumption and risk of cancer: results from NutriNet-Santé prospective cohort. BMJ. 2019 Jul 10;366:l2408. doi: 10.1136/bmj.l2408. PMID: 31292122; PMCID: PMC6614796.

4. Mela DJ, McLaughlin J, Rogers PJ. Perspective: Standards for Research and Reporting on Low-Energy ("Artificial") Sweeteners. Adv Nutr. 2020 May 1;11(3):484-491. doi: 10.1093/advances/nmz137. PMID: 31925418; PMCID: PMC7231577.

5. Magnuson BA, Carakostas MC, Moore NH, Poulos SP, Renwick AG. Biological fate of low-calorie sweeteners. Nutr Rev. 2016 Nov;74(11):670-689. doi: 10.1093/nutrit/nuw032. PMID: 27753624.

6. Hodge AM, Bassett JK, Milne RL, English DR, Giles GG. Consumption of sugar-sweetened and artificially sweetened soft drinks and risk of obesity-related cancers. Public Health Nutr. 2018 Jun;21(9):1618-1626. doi: 10.1017/S1368980017002555. Epub 2018 Feb 21. PMID: 29463332.

7. Bassett JK, Milne RL, English DR, Giles GG, Hodge AM. Consumption of sugar-sweetened and artificially sweetened soft drinks and risk of cancers not related to obesity. Int J Cancer. 2020 Jun 15;146(12):3329-3334. doi: 10.1002/ijc.32772. Epub 2019 Nov 21. PMID: 31693185.

8. Haighton L, Roberts A, Jonaitis T, Lynch B. Evaluation of aspartame cancer epidemiology studies based on quality appraisal criteria. Regul Toxicol Pharmacol. 2019 Apr;103:352-362. doi: 10.1016/j.yrtph.2019.01.033. Epub 2019 Feb 2. PMID: 30716379.

9. Haighton L, Roberts A, Walters B, Lynch B. Systematic review and evaluation of aspartame carcinogenicity bioassays using quality criteria. Regul Toxicol Pharmacol. 2019 Apr;103:332-344. doi: 10.1016/j.yrtph.2018.01.009. Epub 2018 Jan 12. PMID: 29339245.

10. EFSA, 2013. Scientific Opinion on the re-evaluation of aspartame (E 951) as a food additive. The EFSA Journal 2013;11(12):3496, 1-263. DOI: https://doi.org/10.2903/j...

11. Magnuson BA, Burdock GA, Doull J, Kroes RM, Marsh GM, Pariza MW, Spencer PS, Waddell WJ, Walker R, Williams GM. Aspartame: a safety evaluation based on current use levels, regulations, and toxicological and epidemiological studies. Crit Rev Toxicol. 2007;37(8):629-727. doi: 10.1080/10408440701516184. PMID: 17828671.

12. Shibui Y, Fujitani S, Iwata H, Lynch B, Roberts A. Histological analyses of the Ishii (1981) rat carcinogenicity study of aspartame and comparison with the Ramazzini Institute studies. Regul Toxicol Pharmacol. 2019 Mar;102:23-29. doi: 10.1016/j.yrtph.2018.12.010. Epub 2018 Dec 18. PMID: 30572082.

13. Gift JS, Caldwell JC, Jinot J, Evans MV, Cote I, Vandenberg JJ. Scientific considerations for evaluating cancer bioassays conducted by the Ramazzini Institute. Environ Health Perspect. 2013 Nov-Dec;121(11-12):1253-63. doi: 10.1289/ehp.1306661. Epub 2013 Sep 17. PMID: 24045135; PMCID: PMC3852791.

14. EPA, 2019. Update on Ramazzini Institute Data in IRIS Assessments: https://www.epa.gov/iris/...)

15. Guercio BJ, Zhang S, Niedzwiecki D, Li Y, Babic A, Morales-Oyarvide V, Saltz LB, Mayer RJ, Mowat RB, Whittom R, Hantel A, Benson A, Atienza D, Messino M, Kindler H, Venook A, Ogino S, Zoltick ES, Stampfer M, Ng K, Wu K, Willett WC, Giovannucci EL, Meyerhardt JA, Fuchs CS. Associations of artificially sweetened beverage intake with disease recurrence and mortality in stage III colon cancer: Results from CALGB 89803 (Alliance). PLoS One. 2018 Jul 19;13(7):e0199244. doi: 10.1371/journal.pone.0199244. PMID: 30024889; PMCID: PMC6053135.