Artificial sweeteners and cancer risk: Results from the NutriNet-Santé population-based cohort study.

BACKGROUND: The food industry uses artificial sweeteners in a wide range of foods and beverages as alternatives to added sugars, for which deleterious effects on several chronic diseases are now well established. The safety of these food additives is debated, with conflicting findings regarding their role in the aetiology of various diseases. In particular, their carcinogenicity has been suggested by several experimental studies, but robust epidemiological evidence is lacking. Thus, our objective was to investigate the associations between artificial sweetener intakes (total from all dietary sources, and most frequently consumed ones: aspartame [E951], acesulfame-K [E950], and sucralose [E955]) and cancer risk (overall and by site). METHODS AND
FINDINGS: Overall, 102,865 adults from the French population-based cohort NutriNet-Santé (2009-2021) were included (median follow-up time = 7.8 years). Dietary intakes and consumption of sweeteners were obtained by repeated 24-hour dietary records including brand names of industrial products. Associations between sweeteners and cancer incidence were assessed by Cox proportional hazards models, adjusted for age, sex, education, physical activity, smoking, body mass index, height, weight gain during follow-up, diabetes, family history of cancer, number of 24-hour dietary records, and baseline intakes of energy, alcohol, sodium, saturated fatty acids, fibre, sugar, fruit and vegetables, whole-grain foods, and dairy products. Compared to non-consumers, higher consumers of total artificial sweeteners (i.e., above the median exposure in consumers) had higher risk of overall cancer (n = 3,358 cases, hazard ratio [HR] = 1.13 [95% CI 1.03 to 1.25], P-trend = 0.002). In particular, aspartame (HR = 1.15 [95% CI 1.03 to 1.28], P = 0.002) and acesulfame-K (HR = 1.13 [95% CI 1.01 to 1.26], P = 0.007) were associated with increased cancer risk. Higher risks were also observed for breast cancer (n = 979 cases, HR = 1.22 [95% CI 1.01 to 1.48], P = 0.036, for aspartame) and obesity-related cancers (n = 2,023 cases, HR = 1.13 [95% CI 1.00 to 1.28], P = 0.036, for total artificial sweeteners, and HR = 1.15 [95% CI 1.01 to 1.32], P = 0.026, for aspartame). Limitations of this study include potential selection bias, residual confounding, and reverse causality, though sensitivity analyses were performed to address these concerns.
CONCLUSIONS: In this large cohort study, artificial sweeteners (especially aspartame and acesulfame-K), which are used in many food and beverage brands worldwide, were associated with increased cancer risk. These findings provide important and novel insights for the ongoing re-evaluation of food additive sweeteners by the European Food Safety Authority and other health agencies globally. TRIAL REGISTRATION: ClinicalTrials.gov NCT03335644.

Introduction

Given the deleterious health effects of excess sugar intake (e.g., weight gain, cardiometabolic disorders, dental caries), the World Health Organization (WHO) recommends limiting sugar consumption to less than 10% of daily energy intake [1]. However, as liking for sweet taste is widespread globally, the food industry started to use artificial sweeteners as alternatives to reduce added sugar content and corresponding calories while maintaining sweetness. In addition, in order to increase palatability, manufacturers include artificial sweeteners in some food products that do not traditionally contain added sugar (e.g., flavoured potato chips). High-intensity sweeteners (hereafter referred to as ‘artificial sweeteners’) are food additives with high sweetening power yet providing little energy. Aspartame (E951), a well-known artificial sweetener, is found in nearly 1,400 food products on the French market, and more than 6,000 worldwide [2,3]. Its energy value is similar to sugar (4 kcal/g) but its sweetness is 200 times higher [4], meaning a much smaller amount of aspartame is needed for a comparable taste. Other artificial sweeteners are even calorie-free, e.g., acesulfame-K (E950) and sucralose (E955), which are respectively 200 and 600 times sweeter than sucrose [4].

Previous evaluations by health authorities concluded that there was insufficient evidence for risk for the consumption of low- and no-calorie sweeteners under established acceptable daily intakes (ADIs) [4,5]. However, recent epidemiological and experimental studies with conflicting results have reactivated the debate on the safety of these additives. In this context, several health authorities are currently re-evaluating artificial sweeteners, including the European Food Safety Authority (EFSA) [6]. Indeed, while some epidemiological studies did not support the involvement of artificial sweeteners in various health outcomes (e.g., weight loss or weight gain [7-9], glycaemic control [7,8], cardiovascular/kidney diseases [7]), others suggested associations with higher incidence of obesity, hypertension, metabolic syndrome, type 2 diabetes, and cardiovascular events [10].

Regarding cancer, all previous evaluations agreed upon the fact that additional studies, especially in humans, were needed [4]. In particular, experts have urged for a re-evaluation by public health authorities of aspartame’s role in cancer development [11,12], based on previous and recent findings in animal models [11,13], in vitro studies [14,15], and, to a lesser extent, human data [2,16]. Findings about other artificial sweeteners also raise questions regarding their potential role in carcinogenesis based on in vivo studies [13,17]. To our knowledge, no previous prospective cohort has investigated the association of cancer risk with quantitative artificial sweetener intakes from all dietary sources, distinguishing the different types of sweeteners. Indeed, so far, human-derived data have mostly investigated artificial sweetener intakes by using the overall consumption of artificially sweetened beverages (ASBs) as a proxy. A more precise assessment of exposure to artificial sweeteners from a broader range of ultra-processed products (e.g., flavoured yogurts, low-sugar snacks, ready-to-go meals, table-top sweeteners) appears necessary. Besides, since most previous epidemiological studies did not collect data on the brand names of products, data are lacking regarding the specific types of sweeteners consumed by the participants (e.g., aspartame, acesulfame-K, sucralose).

Thus, the objective of our study was to investigate the associations between intakes of artificial sweeteners (total and most consumed ones) and cancer risk (overall and by most frequent cancer sites) in the large-scale population-based NutriNet-Santé cohort, based on detailed dietary data including names/brands of industrial products.

Methods

Study population and data collection

The NutriNet-Santé study is a web-based cohort dedicated to investigating the associations between nutrition and health [18]. Enrolment of participants from the French population was initiated in May 2009 and is still ongoing. The NutriNet-Santé volunteers are adults aged ≥18 years with Internet access recruited through extensive multimedia campaigns. Each participant is followed via questionnaires available and regularly added in their personal account on the study website (https://etude-nutrinet-sante.fr). In particular, detailed information is collected at baseline and every year thereafter through a 5-questionnaire kit, regarding health status (e.g., personal and family history of diseases and drug use), anthropometric data (height, weight) [19,20], physical activity (validated 7-day assessment via the International Physical Activity Questionnaire [IPAQ] [21]), lifestyle and sociodemographic characteristics (e.g., sex, date of birth, educational level, occupation, smoking status, number of children) [22], and diet (see below).

An electronic informed consent is provided by each participant. The NutriNet-Santé study, registered at ClinicalTrials.gov (NCT03335644), is conducted according to the Declaration of Helsinki guidelines and is approved by the Institutional Review Board of the French Institute of Health and Medical Research (Inserm) and the Commission Nationale de l’Informatique et des Libertés (CNIL 908450/909216). All methods have been described in line with the Strengthening the Reporting of Observational Studies in Epidemiology–Nutritional Epidemiology guidelines (S1 STROBE-nut Checklist).

The NutriNet-Santé study was developed to investigate the relationships between multiple dietary exposures and the incidence of chronic diseases, such as cancer. The general protocol of the cohort, written in 2008 before the beginning of the study, is available online [23]. Regarding food additives specifically, the present work is part of a series of pre-specified analyses that are included in a project funded by the European Research Council (https://erc.europa.eu/news-events/magazine/erc-2019-consolidator-grants-examples#ADDITIVES).

Patient involvement statement

The research question developed in this article corresponds to a concern expressed by some participants involved in the NutriNet-Santé cohort, and of the public in general. Participants in the study are thanked in the Acknowledgements. The results of the present study are disseminated to the NutriNet-Santé participants through the cohort website, public seminars, and a press release.

Dietary assessment

Dietary intakes are collected every 6 months by 3 non-consecutive web-based 24-hour dietary records, randomly assigned over 15 days (2 weekdays and 1 weekend day). Participants declare all foods and beverages consumed during main meals and other eating occasions, and they provide information on portion sizes via validated photographs or standard serving containers [24]. Baseline dietary intakes were evaluated by averaging all 24-hour dietary records provided during the first 2 years of follow-up (up to 15 records). Daily intakes of energy, alcohol, and macro- and micronutrients were assessed via the NutriNet-Santé food composition table (providing nutritional composition for about 3,500 items) [25]. Nutrient intakes from composite dishes were estimated according to usual French recipes as defined by nutrition professionals. Dietary energy under-reporters were identified using basal metabolic rate and the Goldberg cut-off method [26], and excluded from the analyses. The detailed methodology for identifying under-reporting is presented in Method A in S1 Appendix. The 24-hour dietary records were validated against an interview by a trained dietitian [27] and against blood and urinary biomarkers [28,29].

Artificial sweetener intakes

Artificial sweetener intakes were assessed through the 24-hour dietary records, in which brands and commercial names of industrial products were routinely collected, enabling us to assess exposure to each food additive. Additive exposure assessment in the NutriNet-Santé cohort has been previously described in detail [30]. Briefly, the presence or absence of each additive in each specific food product consumed was determined using 3 large-scale composition databases: the French food safety agency database Oqali (https://www.oqali.fr/oqali_eng/) [31], Open Food Facts (https://fr-en.openfoodfacts.org/) [3], and Mintel’s Global New Products Database [32]. Dynamic matching was applied, meaning that products were matched date-to-date: The date of consumption of each food or beverage declared by each participant was used to match the product to the closest composition data, thus accounting for potential reformulations. In total, quantitative doses of additives were estimated by approximately 2,700 laboratory assays (either specifically performed by our laboratory for this project—89% of the assays regarding artificial sweeteners—or performed by accredited laboratories upon request of the consumer association UFC–Que Choisir) on different food matrices for the main additive–food vector pairs. Quantitative doses were completed with information for generic food categories provided by EFSA and the Joint FAO/WHO Expert Committee on Food Additives (JECFA) [33]. This methodology allowed us to assess exposure for the following artificial sweeteners: acesulfame-K (European food additive identification number E950), aspartame (E951), cyclamates (E952), saccharin (E954), sucralose (E955), thaumatin (E957), neohesperidine dihydrochalcone (E959), steviol glycosides (E960), and salt of aspartame-acesulfame (E962); the quantities consumed of all these artificial sweeteners were summed to calculate the variable ‘total artificial sweeteners’. Specific analyses were performed for the most represented artificial sweeteners in the cohort: aspartame, acesulfame-K, and sucralose. All other artificial sweeteners were consumed by less than 3.5% of participants.

Cancer case ascertainment

Participants are asked to report all medications/treatment and major health events on the annual health questionnaire, a specific check-up questionnaire every 6 months, or at any time on their NutriNet-Santé account. In order to validate reported incident cancer cases, participants were contacted by a physician of the research team to provide any relevant medical and anatomopathological reports. If necessary, the participant’s physicians and/or hospitals were also contacted to provide the requested information. All cases reported up to 22 January 2021 were investigated. In addition, the data are linked to the medico-administrative databases of the national health insurance system database (SNIIRAM) and the national mortality registry (CépiDc) to ensure completeness of morbidity and mortality information and to limit bias associated with unreported cases. Medical information was obtained for more than 90% of incident cases, and 95% of these were validated; therefore, all incident cases declared were included in the present study, unless they were not validated based on the information provided. Cases were then classified using the International Classification of Diseases–10th Revision [34]. In this study, all first primary cancers (except for basal cell carcinoma) diagnosed between inclusion and 22 January 2021 were considered as cases. Obesity-related cancers are all cancers for which obesity is involved in their aetiology as one of the risk (or protective) factors, as recognised by the World Cancer Research Fund (independently of participant BMI status) [35]: colorectal, stomach, liver, mouth, pharynx, larynx, oesophageal, breast (with opposite associations pre- and postmenopause), ovarian, endometrial, and prostate cancers.

Statistical analysis

Energy under-reporters, as well as those with prevalent cancer at baseline, were excluded. A detailed flowchart is presented in Fig 1.

Fig 1

Flowchart for the selection of the study population: NutriNet-Santé cohort, France, 2009–2021.

Since a substantial proportion of the population were non-consumers of artificial sweeteners, participants were divided into 3 groups: non-consumers, lower consumers, and higher consumers, the latter 2 being separated by the sex-specific median of consumption in the study population. Baseline characteristics were examined across categories of total artificial sweetener intake and were compared using ANOVA tests for continuous variables or χ2 tests for categorical variables. Associations between sweetener intake (all artificial sweeteners, aspartame, acesulfame-K, and sucralose) and cancer risk (overall and by type) were assessed by Cox proportional hazards models with age modelled as the time scale. Specific cancer types considered in this study were breast and prostate (i.e., the most frequent cancer sites in women and men in France [36] and in the cohort) as well as the group of obesity-related cancers. Participants contributed person-time from their inclusion in the cohort until the date of cancer diagnosis, date of last follow-up, date of death, or 22 January 2021, whichever occurred first. Cause-specific hazards were computed so that death and cancer events other than the one studied (for site-specific analyses) occurring during follow-up were handled as competing risks. The Fine and Gray subdistribution hazard model was also tested in sensitivity analysis. Hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated with the non-consumer group as the reference category. P-trend was obtained using the ordinal score for each group (non-consumers: 1; lower consumers: 2; higher consumers: 3). The proportional hazards assumption of the Cox model was confirmed with the rescaled Schoenfeld-type residuals method (Fig A in S1 Appendix). We assessed linearity by comparing the model with the 3 distinct categories of sweetener intake to a model with a linear trend across these categories, using the Akaike information criterion. Consumer (overall) versus non-consumer analyses were also conducted, and this model with 2 categories of exposure was compared to the main model with a formal test for heterogeneity. Missing values for any covariates were handled using the multiple imputation by chained equations (MICE) method [37] (15 imputed datasets) (details in Method B in S1 Appendix). The main analyses were adjusted for the following covariates: sociodemographic characteristics (age [time scale], sex [except for breast and prostate cancer analyses], educational level), lifestyle characteristics (physical activity [IPAQ] [21], smoking status, number of smoked cigarettes in pack-years), anthropometric characteristics (body mass index [BMI], height, percentage weight gain during follow-up), personal and family medical history (prevalent type 1 or type 2 diabetes, family history of cancer), number of 24-hour dietary records, and baseline intakes of energy and food groups/key nutrients for which a direct or indirect role in cancer aetiology has been strongly suggested [35] (alcohol, sodium, saturated fatty acids, fibre, total sugar, fruit and vegetables, whole-grain foods, and dairy products). Breast cancer analyses were additionally adjusted for age at menarche, age at first childbirth, number of biological children, baseline menopausal status, and oral contraceptive use and hormonal treatment for menopause at baseline and during follow-up. Coding for these covariates is indicated in the footnotes to the tables. In analyses specific to 1 artificial sweetener, models were mutually adjusted for the total intake of all other artificial sweeteners. We report minimally adjusted (for age and sex only) and fully adjusted HRs for the associations between artificial sweeteners (total, aspartame, acesulfame-K, and sucralose) and cancer risk (overall, breast, prostate, and obesity-related).

In order to explore the question of which, between sugar and artificial sweeteners, may be more problematic regarding cancer risk, participants were categorised into 6 classes according to their intake levels of artificial sweeteners (non-consumers, lower consumers, and higher consumers) and sugar (. ">38]). Cancer risks were compared 2-by-2 across the 6 categories, and in particular between the categories ‘higher artificial sweetener consumption and sugar intake below the official recommended limit’ and ‘no artificial sweetener consumption and sugar intake exceeding the recommended limit’, with the latter category being the reference category. Further analyses were conducted to investigate the associations of artificial sweetener consumption with premenopausal and postmenopausal breast cancers separately, and menopause-related heterogeneity was assessed using likelihood ratio tests comparing the estimated log-likelihood of a model to that of the same model plus a multiplicative interaction term for menopausal status and the artificial sweetener exposure. Formal interactions between BMI (

A series of sensitivity analyses were performed to assess the robustness of the findings, including restriction of the study population to participants with at least four 24-hour dietary records and exclusion of participants with prevalent diabetes. Models with further adjustments for added sugar intake, sugary beverage consumption, proportion of ultra-processed foods in the diet, being on a weight-loss/calorie-restricted diet, and ‘healthy’ and ‘Western’ dietary patterns (derived by principal component analysis) instead of food groups, and models with artificial sweetener intakes coded as time-dependent exposure variables, were also tested. Some participants with subclinical cancer may get sick and change their dietary habits during the months preceding their diagnosis. Thus, in order to counter this potential reverse causality bias, we performed a sensitivity analysis with follow-up starting at age at entry into the cohort plus 2 years.

The use of Cox models, the 3-category coding for sweetener exposure, and the adjustment for main confounders (sociodemographic, anthropometric, nutritional, and health-related) were pre-specified. The main analyses added following the review process were as follows: linearity check in relation to the trend across categories, interaction tests between artificial sweetener exposure and menopausal status, heterogeneity tests to compare exposure coding strategies, models restricted to non-smoker participants, and analyses exploring multiple exposure to several types of sweeteners. All tests were 2-sided, and P < 0.05 was considered statistically significant. The statistical analysis software SAS, version 9.4, was used for analyses.

Results

Descriptive characteristics

In total, 102,865 participants (78.5% women) were included in the analyses (detailed flowchart shown in Fig 1). Average age at baseline was 42.2 ± 14.5 years. Average number of 24-hour dietary records per participant was 5.6 (SD = 3.0). Artificial sweeteners were consumed by 36.9% of the participants. Table 1 shows baseline characteristics of the study population by categories of total artificial sweetener intake. Compared to non-consumers (unadjusted descriptive comparisons), higher consumers tended to be more often women, younger, smokers, less physically active, more educated, and more likely to have prevalent diabetes. They had lower energy, alcohol, saturated fatty acid, fibre, fruit and vegetables, and whole-grain food intakes and higher intakes of sodium, total sugar, dairy products, sugary foods and drinks, and unsweetened non-alcoholic beverages. The main artificial sweetener was aspartame, contributing to 58% of intakes, followed by acesulfame-K (29%) and sucralose (10%) (Fig 2). These 3 sweeteners were respectively consumed by 28%, 34%, and 14% of the study population. All participants’ intakes of aspartame and acesulfame-K were below the ADIs of 40 mg/kg body weight/day and 9 mg/kg body weight/day, respectively; only 5 participants exceeded the ADI of 15 mg/kg body weight/day for sucralose [5]. Soft drinks with no added sugars, table-top sweeteners, and yogurt/cottage cheese were the main contributors to total artificial sweetener intake, accounting for 53%, 29%, and 8% of intakes, respectively (Fig 3). Table A in S1 Appendix displays the number and percent of participants consuming 1, 2, or 3 of the main artificial sweeteners (aspartame, acesulfame-K, and sucralose). Participants frequently co-consumed several types of artificial sweeteners, but the proportion of those who were consumers of all 3 main artificial sweeteners was low (only 7.1%).

Table 1

Baseline characteristics of the study population, NutriNet-Santé cohort, France, 2009–2021 (n = 102,865).

Characteristic	All participants	Categories of artificial sweetener intake2			P value3
		Non-consumers	Lower consumers	Higher consumers
Number of participants	102,865	64,892 (63.08)	18,986 (18.46)	18,987 (18.46)
Age (years), mean (SD)	42.22 (14.50)	42.82 (14.70)	42.10 (14.54)	40.31 (13.57)	<0.001
Female sex	80,711 (78.46)	49,349 (76.05)	15,681 (82.59)	15,681 (82.59)	<0.001
Height (cm), mean (SD)	166.93 (8.18)	167.24 (8.28)	166.17 (7.94)	166.61 (8.00)	<0.001
Body mass index (kg/m2), mean (SD)	23.69 (4.48)	23.29 (4.17)	23.79 (4.49)	24.96 (5.20)	<0.001
Family history of cancer	39,040 (37.95)	26,643 (37.97)	7,493 (39.46)	6,904 (36.36)	<0.001
Prevalent type 1 diabetes	254 (0.25)	118 (0.18)	43 (0.23)	93 (0.49)	<0.001
Prevalent type 2 diabetes	1,522 (1.48)	676 (1.04)	321 (1.69)	525 (2.76)	<0.001
Educational level					<0.001
Less than high school degree	18,062 (17.42)	11,523 (17.75)	3,263 (17.19)	3,276 (17.25)
2 years or less after high school	17,921 (17.42)	11,269 (17.36)	3,304 (17.40)	3,3348 (17.63)
More than 2 years after high school	66,894 (65.02)	41,109 (64.88)	12,420 (65.41)	12,365 (65.12)
Smoking status					<0.001
Current	17,945 (17.44)	11,188 (17.24)	2,898 (15.26)	3,859 (20.32)
Former	33,030 (32.11)	20,576 (31.70)	6,031 (31.76)	6,423 (33.82)
Never	51,902 (50.45)	33,137 (51.06)	10,058 (52.97)	8,707 (45.85)
Physical activity level4					<0.001
Low	21,443 (20.84)	13,159 (20.28)	4,070 (21.44)	4,214 (22.19)
Moderate	38,152 (37.09)	23,910 (36.84)	7,310 (38.50)	6,932 (36.51)
High	29,023 (28.21)	18,919 (29.15)	5,093 (26.82)	5,011 (26.39)
Number of biological children, mean (SD)	1.28 (1.28)	1.32 (1.31)	1.26 (1.23)	1.18 (1.21)	<0.001
Menopausal or peri-menopausal	28,694 (35.54)	18,019 (36.51)	5,940 (37.88)	4,735 (30.19)	<0.001
Hormonal treatment for menopause5	3,482 (4.31)	2,064 (4.18)	738 (4.71)	680 (4.34)	0.0187
Oral contraception6	22,991 (28.48)	13,052 (26.44)	4,740 (30.23)	5,199 (33.15)	<0.001
Energy intake without alcohol (kcal/day), mean (SD)	1901.69 (471.70)	1913.09 (478.76)	1895.27 (435.87)	1869.15 (480.16)	<0.001
Alcohol intake (g/day), mean (SD)	7.81 (11.88)	8.12 (12.31)	7.65 (11.09)	6.89 (11.05)	<0.001
Saturated fatty acid intake (g/day), mean (SD)	33.21 (12.19)	33.57 (12.34)	33.22 (11.25)	31.95 (12.46)	<0.001
Sodium intake (mg/day), mean (SD)	2719.72 (892.27)	2709.80 (905.87)	2728.75 (826.30)	2744.62 (908.26)	<0.001
Dietary fibre intake (g/day), mean (SD)	19.48 (7.26)	19.82 (7.56)	19.03 (6.32)	18.77 (7.02)	<0.001
Total sugar intake (g/day), mean (SD)	93.47 (33.45)	92.93 (33.85)	95.45 (31.03)	93.35 (34.34)	<0.001
Added sugar intake (g/day), mean (SD)	38.58 (23.92)	38.35 (23.73)	40.12 (22.69)	37.84 (25.66)	<0.001
Percentage of energy from added sugar, mean (SD)	7.95 (4.18)	7.88 (4.15)	8.31 (3.97)	7.85 (4.45)	<0.001
Sugary drink intake (ml/day), mean (SD)	47.94 (107.32)	42.81 (103.77)	55.54 (99.11)	57.90 (124.64)	<0.001
Fruit and vegetable intake (g/day), mean (SD)	405.11 (220.56)	409.05 (223.10)	399.24 (198.46)	397.54 (232.19)	<0.001
Whole-grain food intake (g/day), mean (SD)	34.46 (46.52)	36.01 (49.66)	31.67 (38.98)	31.98 (41.91)	<0.001
Dairy product intake (g/day), mean (SD)	196.48 (148.63)	183.56 (145.11)	202.70 (138.01)	234.40 (163.17)	<0.001
Ultra-processed food intake (percent of the diet in g/day), mean (SD)	17.47 (9.98)	16.04 (9.17)	17.50 (8.73)	22.32 (12.07)	<0.001
Weight-loss diet during the first 2 years of follow-up	17,569 (17.08)	7,747 (11.94)	3,626 (19.10)	6,196 (32.63)	<0.001
Total artificial sweetener intake (mg/day), mean (SD)	16.07 (49.74)	0.00 (0.00)	7.62 (5.05)	79.43 (91.72)	<0.001
Aspartame (E951) intake (mg/day), mean (SD)	9.35 (31.84)	0.00 (0.00)	3.24 (4.06)	47.42 (60.75)	<0.001
Acesulfame-K (E950) intake (mg/day), mean (SD)	4.64 (15.14)	0.00 (0.00)	2.74 (2.86)	22.39 (29.01)	<0.001
Sucralose (E955) intake (mg/day), mean (SD)	1.59 (16.21)	0.00 (0.00)	1.09 (1.98)	7.52 (37.08)	<0.001

1Values are given as number (percentage) unless stated otherwise. 1 kcal = 4.18 kJ = 0.00418 MJ.

2Lower consumers and higher consumers were separated by the sex-specific median among consumers, i.e., 17.44 mg/day in men and 19.00 mg/day in women.

3P values for crude comparison between the 3 categories of sweetener intake by ANOVA or χ2 test as appropriate.

4Available for 88,618 participants, categorised into high, moderate, and low categories according to International Physical Activity Questionnaire guidelines.

5Among menopausal women.

6Among non-menopausal women.

Fig 2

Relative contribution of each specific artificial sweetener to the total intake of artificial sweeteners (percentage), NutriNet-Santé, France, 2009–2021 (n = 102,865).

*Cyclamates (E952), saccharin (E954), thaumatin (E957), neohesperidine dihydrochalcone (E959) steviol glycosides (E960), aspartame-acesulfame salt (E962).

Fig 3

Relative contribution of each food group to the total intake of artificial sweeteners (percentage), NutriNet-Santé, France, 2009–2021 (n = 102,865).

**Artificial sweeteners used as tablets, liquid, or powder, added by the participants in yogurts, hot drinks, etc., or for cooking. ***High-protein food substitutes, sugary foods, cookies, biscuits, cakes, pastries, breakfast cereals, sauces, savoury foods, and ultra-processed fish products.

Associations between intakes of artificial sweeteners and cancer risk

During follow-up (708,905 person-years, median follow-up time = 7.7 years, interquartile range = 4.7–9.4 years), 3,358 incident cancer cases were diagnosed (among which were 982 breast, 403 prostate, and 2,023 obesity-related cancers). Average age at diagnosis was 59.5 ± 12.2 years.

Artificial sweetener intake was positively associated with the risk of overall cancer (HR for higher consumers versus non-consumers = 1.13 [95% CI 1.03 to 1.25], P-trend = 0.002) (Table 2). In particular, higher cancer risks were observed for aspartame (HR = 1.15 [95% CI 1.03 to 1.28], P = 0.002) and acesulfame-K (HR = 1.13 [95% CI 1.01 to 1.26], P = 0.007). Increased risks were observed for breast cancer (HR = 1.22 [95% CI 1.01 to 1.48], P = 0.036, for aspartame) and obesity-related cancers (HR = 1.13 [95% CI 1.00 to 1.28], P = 0.036, for total artificial sweeteners, and HR = 1.15 [95% CI 1.01 to 1.32], P = 0.026, for aspartame). Overall, the same direction of association was observed in pre- and postmenopausal women (Table B in S1 Appendix). Heterogeneity tests showed no difference between pre- and postmenopausal models for total artificial sweeteners, aspartame, and acesulfame-K (P for heterogeneity = 0.440, 0.332, and 0.539, respectively). P for heterogeneity was 0.015 for sucralose, but associations with cancer risk were non-significant in both strata for this food additive, with a low number of consumers per strata. No association was found with prostate cancer (Table 2). Forest plots in Fig B in S1 Appendix present both minimally and fully adjusted associations, showing similar results. Results for competing risk analyses are presented in Result A in S1 Appendix.

Table 2

Association between total artificial sweetener, aspartame, acesulfame-K, and sucralose intakes (mg/day) and cancer risk, NutriNet-Santé cohort, France, 2009–2021 (n = 102,865).

Cancer site	Exposure (mg/day)	Measure	Non-consumers	Lower consumers2	Higher consumers2	P-trend
All cancers	Total artificial sweeteners	Participants/incident cases	64,892/2,013	18,986/744	18,987/601
		HR (95% CI)—minimally adjusted3	1	1.26 (1.16 to 1.37)	1.19 (1.08 to 1.30)	<0.001
		HR (95% CI)—fully adjusted4	1	1.14 (1.05 to 1.25)	1.13 (1.03 to 1.25)	0.002
	Aspartame	Participants/incident cases	74,169/2,309	14,345/572	14,351/477
		HR (95% CI)—minimally adjusted	1	1.21 (1.11 to 1.33)	1.18 (1.07 to 1.31)	<0.001
		HR (95% CI)—fully adjusted	1	1.12 (1.02 to 1.23)	1.15 (1.03 to 1.28)	0.002
	Acesulfame-K	Participants/incident cases	67,662/2,096	17,601/766	17,602/496
		HR (95% CI)—minimally adjusted	1	1.22 (1.12 to 1.33)	1.19 (1.07 to 1.33)	<0.001
		HR (95% CI)—fully adjusted	1	1.12 (1.03 to 1.22)	1.13 (1.01 to 1.26)	0.007
	Sucralose	Participants/incident cases	88,867/2,883	7,005/288	6,993/187
		HR (95% CI)—minimally adjusted	1	1.20 (1.06 to 1.35)	1.00 (0.86 to 1.17)	0.177
		HR (95% CI)—fully adjusted	1	1.03 (0.91 to 1.17)	0.96 (0.82 to 1.12)	0.823
Breast cancer	Total artificial sweeteners	Participants/incident cases	49,349/556	15,681/229	15,681/194
		HR (95% CI)—minimally adjusted	1	1.23 (1.06 to 1.44)	1.16 (0.99 to 1.37)	0.019
		HR (95% CI)—fully adjusted	1	1.11 (0.95 to 1.30)	1.16 (0.97 to 1.38)	0.064
	Aspartame	Participants/incident cases	56,721/647	11,999/176	12,000/156
		HR (95% CI)—minimally adjusted	1	1.17 (0.99 to 1.39)	1.18 (0.98 to 1.42)	0.031
		HR (95% CI)—fully adjusted	1	1.09 (0.92 to 1.29)	1.22 (1.01 to 1.48)	0.036
	Acesulfame-K	Participants/incident cases	51,712/581	14,578/232	14,579/166
		HR (95% CI)—minimally adjusted	1	1.20 (1.03 to 1.40)	1.22 (1.00 to 1.49)	0.014
		HR (95% CI)—fully adjusted	1	1.11 (0.95 to 1.30)	1.17 (0.96 to 1.43)	0.086
	Sucralose	Participants/incident cases	69,189/826	5,772/93	5,750/60
		HR (95% CI)—minimally adjusted	1	1.23 (0.99 to 1.52)	0.99 (0.76 to 1.30)	0.438
		HR (95% CI)—fully adjusted	1	1.04 (0.84 to 1.30)	0.93 (0.71 to 1.22)	0.786
Prostate cancer	Total artificial sweeteners	Participants/incident cases	15,543/282	3,305/63	3,306/58
		HR (95% CI)—minimally adjusted	1	1.02 (0.78 to 1.34)	1.20 (0.90 to 1.59)	0.257
		HR (95% CI)—fully adjusted	1	0.92 (0.70 to 1.22)	1.26 (0.94 to 1.68)	0.274
	Aspartame	Participants/incident cases	17,457/310	2,346/49	2,351/44
		HR (95% CI)—minimally adjusted	1	1.04 (0.77 to 1.41)	1.19 (0.86 to 1.64)	0.324
		HR (95% CI)—fully adjusted	1	0.95 (0.70 to 1.30)	1.28 (0.91 to 1.79)	0.280
	Acesulfame-K	Participants/incident cases	16,108/288	3,023/76	3,023/39
		HR (95% CI)—minimally adjusted	1	1.13 (0.87 to 1.48)	1.25 (0.86 to 1.80)	0.184
		HR (95% CI)—fully adjusted	1	1.06 (0.81 to 1.39)	1.18 (0.82 to 1.71)	0.365
	Sucralose	Participants/incident cases	19,678/365	1,233/25	1,243/13
		HR (95% CI)—minimally adjusted	1	1.02 (0.68 to 1.54)	0.99 (0.57 to 1.74)	0.967
		HR (95% CI)—fully adjusted	1	0.86 (0.57 to 1.30)	1.01 (0.57 to 1.77)	0.699
Obesity-related cancers	Total artificial sweeteners	Participants/incident cases	64,892/1,232	18,986/433	18,987/358
		HR (95% CI)—minimally adjusted	1	1.20 (1.08 to 1.34)	1.17 (1.04 to 1.32)	0.001
		HR (95% CI)—fully adjusted	1	1.08 (0.97 to 1.21)	1.13 (1.00 to 1.28)	0.036
	Aspartame	Participants/Incident cases	74,169/1,401	14,345/337	14,351/285
		HR (95% CI)—minimally adjusted	1	1.17 (1.04 to 1.31)	1.17 (1.03 to 1.33)	0.003
		HR (95% CI)—fully adjusted	1	1.08 (0.96 to 1.22)	1.15 (1.01 to 1.32)	0.026
	Acesulfame-K	Participants/Incident cases	67,662/1,275	17,601/457	17,602/291
		HR (95% CI)—minimally adjusted	1	1.18 (1.05 to 1.31)	1.17 (1.02 to 1.35)	0.004
		HR (95% CI)—fully adjusted	1	1.09 (0.97 to 1.22)	1.13 (0.97 to 1.30)	0.064
	Sucralose	Participants/Incident cases	88,867/1,756	7,005/167	6,993/100
		HR (95% CI)—minimally adjusted	1	1.14 (0.97 to 1.33)	0.90 (0.73 to 1.11)	0.899
		HR (95% CI)—fully adjusted	1	0.98 (0.84 to 1.16)	0.87 (0.71 to 1.07)	0.230

1Median follow-up times for all, breast, prostate, and obesity-related cancers were, respectively, 7.7, 7.6, 8.0, and 7.7 years. Person-years were, respectively, 708,905, 551,803, 157,102, and 708,905.

2The sex-specific cutoffs between higher and lower consumers were 17.44 mg/day in men and 19.00 mg/day in women for total artificial sweeteners, 14.45 mg/day in men and 15.39 mg/day in women for aspartame, 5.06 mg/day in men and 5.50 mg/day in women for acesulfame-K, and 3.46 mg/day in men and 3.43 mg/day in women for sucralose.

3Minimally adjusted models were adjusted for age (time scale) and sex (except for breast and prostate cancer).

4 Fully adjusted multivariable Cox proportional hazards models (main model) were adjusted for age (time scale), sex (except for breast and prostate cancer), BMI (continuous, kg/m2), height (continuous, cm), percentage weight gain during follow-up (continuous), physical activity (categorical International Physical Activity Questionnaire variable: high, moderate, low, missing value), smoking status (categorical: never, former, current), number of smoked cigarettes in pack-years (continuous), educational level (categorical: less than high school degree, ≤2 years after high school degree, >2 years after high school degree), number of 24-hour dietary records (continuous), family history of cancer (categorical: yes, no), prevalent diabetes (categorical: yes, no), energy intake without alcohol (continuous variable: kcal/day), and daily intakes (continuous, g/day) of alcohol, sodium, saturated fatty acids, fibre, sugar, fruit and vegetables, whole-grain foods, and dairy products. Breast cancer models were also adjusted for age at menarche (categorical: <12 years old, ≥12 years old), age at first child (categorical: no child, <30 years, ≥30 years), number of biological children (continuous), baseline menopausal status (categorical: menopausal, non-menopausal), oral contraceptive use at baseline and during follow-up (categorical: yes, no), and hormonal treatment for menopause at baseline and during follow-up (categorical: yes, no). In addition, all models were mutually adjusted for artificial sweetener intake other than the one studied.

Associations in the non-consumer versus consumer analyses (Table C in S1 Appendix) were consistent with those in the non-consumer versus lower and higher consumer analyses, and the heterogeneity tests performed revealed no difference between the two-category model and the three-category model (all P > 0.05). The comparison of the model with 3 categories of intake to the model with a linear trend across categories did not provide evidence of non-linearity (P = 0.107, 0.250, 0.348, and 0.437 for total artificial sweeteners, aspartame, acesulfame-K, and sucralose, respectively, for the overall cancer model). After adjustment for the total dose of artificial sweetener exposure, cancer risk did not differ between participants consuming 1, 2, or 3 different sweeteners (P > 0.05 for all 2-by-2 comparisons).

No interaction was detected for any cancer outcome between artificial sweetener exposures and BMI, nor between the 3 main artificial sweeteners (Table D in S1 Appendix).

We additionally investigated a 6-category composite variable, combining artificial sweetener and sugar intakes, which revealed increased cancer risk associated with both artificial sweetener and sugar intakes (Fig C and Table E in S1 Appendix). In particular, no difference was detected between the categories ‘higher artificial sweetener consumption and sugar intake below the official recommended limit’ and ‘no artificial sweetener consumption and sugar intake exceeding the recommended limit’ (Table F in S1 Appendix). Overall, results remained similar in all sensitivity analyses (Table G in S1 Appendix).

Discussion

Results from this large-scale population-based cohort study suggest a positive association between higher intake of artificial sweeteners (especially aspartame and acesulfame-K) and overall cancer risk. More specifically, aspartame intake was associated with increased breast and obesity-related cancers.

To our knowledge no previous cohort study has directly investigated the association between quantitative artificial sweetener intakes per se—distinguishing the different types of sweeteners, in the whole diet—and cancer risk. However, some proxies have been used. Aspartame intake from ASBs was assessed in the NIH-AARP Diet and Health Study cohort [16], and no association with hematopoietic and brain cancers was revealed. Intakes through table-top sweeteners added by participants was additionally considered in 2 large-scale American cohorts (the Nurses’ Health Study and the Health Professionals Follow-Up Study) [2] and in the Cancer Prevention Study II (CPS-II) Nutrition Cohort [39]. Results from these studies were conflicting; McCullough et al. found no associations with non-Hodgkin lymphoma in the CPS-II cohort [39]. In contrast, Schernhammer et al. [2], who adjusted their model for various additional dietary factors, found increased risks among male participants for non-Hodgkin lymphoma and multiple myeloma. Other studies did not investigate artificial sweeteners but the whole group of ASBs in millilitres or servings per day. A recent systematic review and meta-analysis of sweetened beverages and risk of cancer at different sites [40] stressed the lack of studies on ASBs and cancer risk except for pancreatic cancer, for which they found a positive, although non-significant, association. Likewise, previous analyses of the NutriNet-Santé cohort did not detect an association between ASBs and cancer risk [41], suggesting that measurements of ASBs might be inadequate to accurately characterise the overall dietary exposure to artificial sweeteners. However, ASBs have recently been investigated within the Melbourne Collaborative Cohort Study, revealing a positive association with cancers not related to obesity [42] but not with obesity-related cancers [43]. Evaluating sweetener intake through ASBs might not be sufficient since many other foods are also vectors of artificial sweeteners (e.g., breakfast cereals, yogurts, ice creams, and table-top sweeteners). Several case–control studies have analysed associations between artificial sweeteners or ASBs and different cancer locations, as recently meta-analysed [44-46]. Although these studies bring interesting pieces of evidence, potentially strong reverse causality bias with this type of design limits the interpretability of these studies. It is therefore more appropriate to rely on large-scale prospective studies when available. Several randomized control trials have tested the effect of artificial sweeteners on health parameters such as body weight, BMI, glycaemic control, and eating behaviour [7]. But none, to our knowledge, has considered cancer as a primary or secondary outcome. In a previous publication, we showed that sugar intake was also associated with increased overall (HR for quartile 4 versus quartile 1 = 1.17 [95% CI 1.00 to 1.37], P = 0.02) and breast (HR for quartile 4 versus quartile 1 = 1.51 [95% CI 1.14 to 2.00], P = 0.001) cancers in this cohort [47]. In the present study, the fact that no difference was detected between the categories ‘higher artificial sweetener consumption and sugar intake below the official recommended limit’ and ‘no artificial sweetener consumption and sugar intake exceeding the recommended limit’ suggests that artificial sweeteners and excessive sugar intake may be equally associated with cancer risk.

On the one hand, obesity is a recognised risk factor for many cancers [35]. On the other hand, although it remains unclear, associations between artificial sweetener intake and weight gain have been suggested [8,10,48-51]. Thus, we investigated the associations between artificial sweetener intake and the risk of obesity-related cancers. The positive associations observed suggest that part of this relationship may be driven by overweight-related metabolic disturbances. However, the associations between artificial sweetener intake and cancer risk observed in this study are not entirely explained by weight-gain-related mechanisms, since the models were adjusted for baseline BMI and weight gain during follow-up. Other mechanisms could be involved. Carcinogenicity of artificial sweeteners has long been suspected based on in vitro and in vivo experimental results. Although results from animal studies remain controversial [52-54], some results obtained in rodent models suggested that aspartame was associated with higher risks of different cancers (lymphomas and leukaemias and hepatocellular and alveolar/bronchiolar carcinomas) [11] at dose levels comparable to those to which humans can be exposed. Although these findings have been controversial [55], additional data have been recently published supporting the original findings from the Ramazzini Institute regarding the identification of tumours [56]. This suggests the need for an updated evaluation of aspartame’s carcinogenicity and ADI. Aspartame’s toxicity has also been investigated in several in vitro studies [14,15], the results of which suggested its carcinogenicity [14], potentially through mechanisms related to inflammation, angiogenesis, promotion of DNA damage, and inhibition of apoptosis [15]. More recently, sucralose was shown to increase the risk of malignant tumours and hematopoietic neoplasia in mice [17]. An in vivo study found that acesulfame-K and saccharin elicited even more DNA damage than aspartame [13]. Lastly, Suez et al. revealed findings implicating non-caloric artificial sweeteners (saccharin, sucralose, and aspartame) in the modification of gut microbiota (induction of dysbiosis and glucose intolerance in mice and in healthy humans) [57], which in turn might be involved in the development of some cancers [58].

Beyond its longitudinal design and large sample size, one major strength of our study is its detailed assessment of exposure to artificial sweeteners at the individual level [30]. The repeated 24-hour dietary records allowed us to collect precise information on the consumption of industrial products, including their commercial brands/names. These consumption data were merged with data from several large composition databases and results from thousands of laboratory assays in food matrixes. Also, dynamic matching with the date of consumption was used to select the most appropriate composition data, which reduced potential bias due to reformulations. Thus, total artificial sweetener intake through sources other than just ASBs could be considered.

However, several limitations should be acknowledged. First, caution must be taken in extrapolating the results to the whole adult French population. As generally observed in volunteer-based cohorts, participants were more likely to be women, to have higher educational and socio-professional levels, to have health-conscious behaviours (diet and lifestyle) [59], and to be older on average (while artificial sweetener intake is lower in older individuals). This could contribute to explaining the relatively lower intake levels of artificial sweeteners compared to those described in the literature for national studies (e.g., for aspartame, 0.53/0.40 for women/men in NutriNet-Santé, compared to 0.81/1.08 estimated in simulations for the French population) [4,60]. This suggests that the associations with cancer risk observed among NutriNet-Santé participants may underestimate what would be observed in the general population with a broader range of exposure. In particular, the absence of a relationship between sucralose and cancer risk in this study should be considered with caution since exposure to sucralose was very low compared to the exposures for aspartame and acesulfame-K. However, differences in exposure estimates may also be due to more precise assessment at the individual level in the present study than in simulation studies based on average information for generic product categories. Second, the limited number of cases prevented us from studying associations for other cancer sites (e.g., pancreatic, ovarian, endometrial, kidney, liver, and bladder) than the main ones presented here. Lastly, causal links cannot be established by this unique study; in particular, residual confounding bias cannot be entirely ruled out, although the wide range of adjustment factors accounted for in main and sensitivity analysis models limited this risk. To assess the causal association between cancer incidence and the intake of artificial sweeteners and sugar, genetic markers linked with sweet taste preference (e.g., the rs838133 variant of hepatokine fibroblast growth factor 21 [61]) could be integrated into a polygenic score that could be used as part of a Mendelian randomization study.

This large-scale population-based cohort study suggests associations between artificial sweeteners, especially aspartame and acesulfame-K, and cancer risk, more specifically breast and obesity-related cancers. These results need to be replicated in other large-scale cohorts, and underlying mechanisms clarified by experimental studies. Artificial sweeteners are present in many food and beverage brands worldwide [3] and are consumed by millions of citizens and patients daily. Our findings do not support the use of artificial sweeteners as safe alternatives for sugar in foods or beverages and provide important and novel information to address the controversies about their potential adverse health effects. These results are particularly relevant in the context of the ongoing in-depth re-evaluation of artificial sweeteners by EFSA and other agencies globally.

(DOCX)

Click here for additional data file.

Supplementary material.

Method A: Methodology for identification of energy under-reporting and validation studies for the 24-hour web-based dietary records. Fig A: Proportional hazards assumption testing using rescaled Schoenfeld-type residuals for the association between total artificial sweetener intake and cancer risk, NutriNet-Santé cohort, France, 2009–2021 (n = 102,865). Method B: Multiple imputation by chained equations. Table A: Number of participants in each combination of artificial sweetener consumption for aspartame, acesulfame-K, and sucralose (non-consumers, consumers of 1 type of artificial sweetener, consumers of 2 types, and consumers of 3 types), NutriNet-Santé cohort, France, 2009–2021 (n = 102,865). Table B: Associations between total artificial sweetener, acesulfame-K, aspartame, and sucralose intakes (mg/day) and breast cancer risk in premenopausal and postmenopausal women, NutriNet-Santé cohort, France, 2009–2021 (n = 80,711). Fig B: Forest plots presenting the minimally adjusted and fully adjusted associations between total artificial sweetener, aspartame, acesulfame-K, and sucralose intakes (mg/day) and cancer risk, NutriNet-Santé cohort, France, 2009–2021 (n = 102,865). Result A: Results from competing risk analyses (cause-specific Cox proportional hazards models). Table C: Association between total artificial sweetener, aspartame, acesulfame-K, and sucralose intakes (mg/day) and cancer risk, NutriNet-Santé cohort, France, 2009–2021 (n = 102,865)—consumers versus non-consumers. Table D: Interaction tests for BMI and between the 3 main artificial sweeteners (aspartame, acesulfame-K, and sucralose) for all studied outcomes. Fig C: Cancer risk associated with combined artificial sweetener and sugar intakes, NutriNet-Santé cohort, France, 2009–2021 (n = 102,865). Table E: Overall cancer risk associated with combined artificial sweetener and sugar intakes: 2-by-2 comparisons across categories, NutriNet-Santé cohort, France, 2009–2021 (n = 102,865). Table F: Focus on the comparisons of cancer risk for participants with higher artificial sweetener consumption/lower sugar intake versus participants with no artificial sweetener consumption/higher sugar intake, NutriNet-Santé cohort, France, 2009–2021 (n = 102,865). Table G: Association between artificial sweetener intake (mg/day) and cancer risk, NutriNet-Santé cohort, France, 2009–2021 (n = 102,865)—sensitivity analyses.

(DOCX)

Click here for additional data file.

11 Nov 2021

Dear Dr Debras,

Thank you for submitting your manuscript entitled "Artificial sweeteners and cancer risk in the prospective NutriNet-Santé cohort" for consideration by PLOS Medicine.

Your manuscript has now been evaluated by the PLOS Medicine editorial staff and I am writing to let you know that we would like to send your submission out for external peer review.

However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.

Please re-submit your manuscript within two working days, i.e. by Nov 15 2021 11:59PM.

Login to Editorial Manager here: https://www.editorialmanager.com/pmedicine

Once your full submission is complete, your paper will undergo a series of checks in preparation for peer review. Once your manuscript has passed all checks it will be sent out for review.

Feel free to email us at plosmedicine@plos.org if you have any queries relating to your submission.

Kind regards,

Callam Davidson

Associate Editor

PLOS Medicine

16 Dec 2021

Dear Dr. Debras,

Thank you very much for submitting your manuscript "Artificial sweeteners and cancer risk in the prospective NutriNet-Santé cohort" (PMEDICINE-D-21-04643R1) for consideration at PLOS Medicine.

Your paper was evaluated by an associate editor and discussed among all the editors here. It was also discussed with an academic editor with relevant expertise, and sent to independent reviewers, including a statistical reviewer. The reviews are appended at the bottom of this email and any accompanying reviewer attachments can be seen via the link below:

[LINK]

In light of these reviews, I am afraid that we will not be able to accept the manuscript for publication in the journal in its current form, but we would like to consider a revised version that addresses the reviewers' and editors' comments. Obviously we cannot make any decision about publication until we have seen the revised manuscript and your response, and we plan to seek re-review by one or more of the reviewers.

In revising the manuscript for further consideration, your revisions should address the specific points made by each reviewer and the editors. Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments, the changes you have made in the manuscript, and include either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please submit a clean version of the paper as the main article file; a version with changes marked should be uploaded as a marked up manuscript.

In addition, we request that you upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, please upload your figure files to the PACE digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at PLOSMedicine@plos.org.

We hope to receive your revised manuscript by Jan 06 2022 11:59PM. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

Please use the following link to submit the revised manuscript:

https://www.editorialmanager.com/pmedicine/

Your article can be found in the "Submissions Needing Revision" folder.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

We look forward to receiving your revised manuscript.

Sincerely,

Callam Davidson,

PLOS Medicine

plosmedicine.org

-----------------------------------------------------------

Comments from the academic editor:

‘The dietary assessment method for this study is unique and strong - better than most other cohort studies. However, I have some concerns about potential residual confounding effects. The authors need to carefully address this issue in the revision’

Requests from the editors:

Please revise your title according to PLOS Medicine's style. Your title must be nondeclarative and not a question. It should begin with main concept if possible. "Effect of" should be used only if causality can be inferred, i.e., for an RCT. Please place the study design ("A cohort study") in the subtitle (ie, after a colon).

* Please structure your abstract using the PLOS Medicine headings (Background, Methods and Findings, Conclusions).

* Please combine the Methods and Findings sections into one section, “Methods and findings”.

Abstract Methods and Findings:

* Please include the study design and length of follow up.

* Please include the important dependent variables that are adjusted for in the analyses.

* In the last sentence of the Abstract Methods and Findings section, please describe the main limitation(s) of the study's methodology.

At this stage, we ask that you include a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract. Please see our author guidelines for more information: https://journals.plos.org/plosmedicine/s/revising-your-manuscript#loc-author-summary

Citations should be in square brackets, and preceding punctuation.

Supplementary Figure 1 is central to the understanding of the paper. Please incorporate it into the main paper.

Please remove references to results tables from the methods (lines 177, 192, 205, 215, and 227).

Please ensure that the study is reported according to the STROBE guideline, and include the completed STROBE checklist as Supporting Information. Please add the following statement, or similar, to the Methods: "This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 Checklist)."

The STROBE guideline can be found here: http://www.equator-network.org/reporting-guidelines/strobe/

When completing the checklist, please use section and paragraph numbers, rather than page numbers.

Did your study have a prospective protocol or analysis plan? Please state this (either way) early in the Methods section.

a) If a prospective analysis plan (from your funding proposal, IRB or other ethics committee submission, study protocol, or other planning document written before analyzing the data) was used in designing the study, please include the relevant prospectively written document with your revised manuscript as a Supporting Information file to be published alongside your study, and cite it in the Methods section. A legend for this file should be included at the end of your manuscript.

b) If no such document exists, please make sure that the Methods section transparently describes when analyses were planned, and when/why any data-driven changes to analyses took place.

c) In either case, changes in the analysis-- including those made in response to peer review comments-- should be identified as such in the Methods section of the paper, with rationale.

In Figure 3, please ensure that the y axis is identical for all panels to facilitate comparison.

Throughout, please report P values to 3 decimal places (anything smaller please report as P < 0.001).

Please remove subheadings from your Discussion.

Please begin the Discussion with a short, clear summary of the article's findings.

Please remove the Competing interest statement, Contributorship statement and guarantor, Transparency statement, Copyright/license for publication statement (all PLOS content is published under a CC-BY license), Data sharing statement, Funding statement and statement of independence of researchers from funders, and LARC disclaimer from the main text. If published, all of this information will be captured as metadata based on your responses to the submission form.

Please relocate the Patient involvement statement to the Methods section.

Comments from the reviewers:

Reviewer #1: The Nutri Net-Santé study provides rare data with which to examine the association between artificial sweeteners and health outcomes. It is also a very large data base as required for such work.

Line 29, 'contrasted' seems the wrong word here.

Line 49, remove 'references'.

Line 58, not well written, maybe '…liking for sweet taste is widespread globally…'

Line 65, 360 why do you have 'references' after food? Would 'food products' be correct?

Line 68, 'saccharose' is same as sucrose but I would have thought sucrose was better known so would use this term.

In Supplementary Figure 2 'Schoenfeld' is spelt correctly in the heading but not within the figure.

Reviewer #2: Alex McConnachie, Statistical Review

The paper by Debras et al presents analyses of a large, prospective cohort study, looking at the association between dietary intake of artificial sweeteners and the incidence of cancers. This review considers the statistical parts of the paper.

In short, the statistics in this paper are very good. The background and data sources are well described. The exclusions due to apparent under-reporting are justified. The exclusion of early events to protect against reverse causation is good. Cox models are appropriate, and the key proportional hazards assumption is checked. Models are fully adjusted for a range of potential confounders, and multiple imputation is used to account for missing baseline data. Competing risks models are considered as one of several sensitivity analyses. The results are clearly presented.

The few comments that I do have are fairly minor.

The authors use cubic splines to check linearity. This would be fine, if the main analysis used consumption as a continuous variable, but that is not the case. Each exposure is modelled as a 3-level categorical term (no intake, low intake, high intake), and the significance is tested as a linear trend across categories. It is probably possible for there to exist a linear association between risk and intake as a continuous variable, but a non-linear association between risk and the three categories of intake. There are a few options that I can see: the authors could remove the linearity check from the paper; they could present the HR for each intake as a linear variable (possibly just in the supplements), in which case the linearity check would be relevant; or they could assess linearity in relation to the trend across categories, by comparing the model with 3 categories of intake to a model with a linear trend across categories.

The adjustments in the models are comprehensive, which is good. In similar papers that I have read, it is common to report the unadjusted association, followed by the associations after increasing levels of adjustment for groups of potential confounders. Forest plots can be a good way to present these associations. I would not consider this essential, but it is sometimes interesting to see the size of the unadjusted association, and the extent to which it is confounded by other things, and then which types of factors contribute most to this confounding. Whilst unlikely, it could even be the case that without adjustment, artificial sweetener intake is not associated with outcomes, and the association is only apparent after adjustment.

Line 263-265 states "Analyses conducted among pre- and postmenopausal women distinctively showed that the associations with breast cancers tended to be more pronounced in postmenopausal women." Looking at table S3, I am not convinced. Taking total intakes, the estimated HRs in premenopausal women are 1.21 and 1.24; in postmenopausal women, they are 1.19 and 1.30 - i.e. they are quite similar. The fact that these associations are statistically significant in postmenopausal women only is neither here nor there - I should think that there is no real evidence that the HRs are any different. Any conclusion about different levels of association should be based on statistical interaction tests.

Line 273, the HR for sugar intake is reported as "1." - I reckon this should be about 1.17. Is this in a model that accounts for artificial sweetener intake? This finding is interesting - does it mean that the risk of cancer is roughly the same if someone replaces their sugar intake with artificial sweeteners? How does that affect the conclusion that artificial sweeteners are not a safe alternative to sugar? If the risk of cancer is about the same, but the risk of metabolic conditions is reduced (if that is the case), then maybe they are a safer alternative?

Obviously not for this paper, but are there genetic markers for how much of a "sweet tooth" someone has? Do these factors predict both sugar and artificial sweetener intake? Could a polygenic score be used as part of a Mendelian Randomisation study, to assess the causal association between cancer incidence and the intake of artificial sweeteners (and sugar)?

Reviewer #3: Artificial sweeteners and cancer risk in the prospective NutriNet-Sante cohort

In this paper, the authors examined the association of intake of artificial sweeteners and risk of cancer (both overall and site specific) in approximately 100k participants from France. They find that intake of artificial sweeteners is positively associated with risk of cancer, especially breast cancer. Aspartame and acesulfame-K were each individually associated with increased risk of cancer.

This study is clearly scientifically important. Artificial sweeteners, including aspartame and acesulfame-K, are present in thousands of foods and consumed by millions of people. Any association with cancer risk would be scientifically important. The study also addresses a clear gap in the literature. Few epidemiologic studies have examined the association of artificial sweeteners with risk of cancer, particularly breast cancer. Few cohorts have even meaningfully measured or quantified artificial sweetener intake.

A major strength of the current study is that the authors had previously quantified levels of artificial sweeteners in commonly consumed foods and beverages in France. This allows the present study to link its dietary measures to a food database with detailed measures of artificial sweetener composition, providing a superior and more granular measure of artificial sweetener intake than other cohorts have. The primary limitations are that this is only one study of one population and the sample size is not especially large for cancer epidemiology (which is why only a few specific types of cancer could be examined). Studies in cohorts like EPIC or the AARP or in research consortia often have many hundreds of thousands or millions of participants. With additional studies, it remains possible that these associations will converge toward the null in the aggregate, such as occurs with "Winner's Curse". To the authors' credit, they appropriately acknowledge their study's limitations. My recommendations for revisions are described below.

Specific comments

Abstract background: Artificial sweeteners were originally marketed for weight loss, not for prevention of chronic disease (and excess weight, in of itself, is not a chronic disease). In fact, sodas like Tab, introduced in 1963, pre-date much of the epidemiological evidence on sugar and chronic disease. Consequently, the cause-and-effect sequence described in the first sentence seems somewhat inaccurate.

Abstract background: "findings remain contrasted" should be rewritten as "findings conflict"

Introduction: Some parts of this text need light editing to conform with typical English usage, e.g. "even though these" should be replaced with "that", "expertises" should be "evaluations".

Cancer case ascertainment: Are there any methodological publications documenting the sensitivity and specificity of case ascertainment for the cohort? What percent of cancer cases are ascertained given the relatively recent date of January 22nd, 2021? Could there be a lag in reporting for some cases? If a publication is available, please include a citation here.

Statistical analysis: Please describe the rationale for reverse causation. Is the idea that some participants get sick with subclinical cancer, change their dietary habits, and then are diagnosed? Or is the concern perhaps more accurately described as recall bias?

Statistical analysis: Given that the first two years of cases are excluded, the authors should also state that the first two years of follow-up were also excluded to prevent "immortality bias", since no cancers could occur during this time (i.e. follow-up starts at age at entry into cohort plus two years).

Results, Supplementary Figure 3: The shaded areas do not appear proportionate to the percentages. If they are not proportionate, consider redrawing or presenting as a table.

Results. Associations: Some of the results are hard to read, e.g. breast cancer (HR=1.25 (1.02 to 1.53) P=0.01, HR=1.33 (1.05 to 1.69) P=0.007…respectively). Consider changing use of punctuation to simplify readability.

Results, Associations: "Associations were consistent in the non-consumers vs. consumers (Supplementary Table 1)". This should include a formal test for heterogeneity.

Results, Associations: Was there any consideration given to presenting results for artificially sweetened beverages (ASB)? The results of the current study, at a superficial level, seem to differ from those of the prior BMJ study from this cohort. It would be interesting to know if results differ because of improved measurement in the current study or if measurements of ASB are simply inadequate to characterize artificial sweetener exposure.

Table 1: please clarify what education "no" means. And the other education categories as well. Change "Education level" to "Education level, %". Modify other headings as needed to match.

Table 2: "(mg/d)" is not needed in this table except in the footnote.

Discussion, Mechanistic Plausibility: Reference 12 is a commentary and should not be cited as if it were an independent scientific study. Please also soften the tone of the phrase "recently confirmed in rats". The word "confirmed" implies multiple independent replications, but this is not what was done here. These additional data support, but do not confirm, the original findings.

References: Please update ref. 29 now that it is published.

Reviewer #4: Manuscript: Artificial sweeteners and cancer risk in the prospective NutriNet-Santé cohort

Manuscript # PMEDICINE-D-21-04643R1

General Comments

This manuscript examined the association of total and specific types of artificial sweetener consumption and cancer risk in a large web-based prospective cohort. Higher intake was associated with higher overall cancer, as well as breast and obesity-related cancers. This manuscript is well-written and contributes to the limited evidence on artificial sweeteners and cancer risk. Only a few minor comments are listed below.

Specific Comments

1. Results: Although the authors state that the number of high artificial sweetener consumers who consumed all three main types was low and that no interaction was detected between the three types, there are a notable number of participants who consumed at least 2 different types of sweetener. Food products also frequently contain more than one type of artificial sweetener. Can the authors explain whether they tried to investigate this further? This does not necessarily need to be added to the manuscript, but it would be interesting to understand whether the authors considered this since many individuals consumed more than one type of artificial sweetener.

2. Supplementary Table 2: It is commendable that the authors examined the intake of artificial sweeteners and sugar, as these may or may not be consumed together. Did the authors also compare additional groups, such as "No artificial sweetener and no sugar" and "High artificial sweetener and high sugar," etc. with the two groups included in this table?

3. Discussion, line 292 and 293: It appears that the abbreviation for Cancer Prevention Study-II should be CPS-II.

4. Discussion: It seems that the reason the authors examined the association of artificial sweeteners and obesity-related cancers was because of a potential link between artificial sweeteners and obesity. The manuscript would be strengthened with additional discussion about this.

Any attachments provided with reviews can be seen via the following link:

[LINK]

6 Jan 2022

Submitted filename: AUTHOR RESPONSE LETTER_PlosMed_R1_MT_CD_mt.docx

Click here for additional data file.

21 Jan 2022

Dear Dr. Debras,

Thank you very much for submitting your revised manuscript "Artificial sweeteners and cancer risk: results from the prospective NutriNet-Santé cohort" (PMEDICINE-D-21-04643R2) for consideration at PLOS Medicine.

Your paper was evaluated by an associate editor and discussed among all the editors here. It was also discussed with an academic editor with relevant expertise, and sent back to independent reviewers, including a statistical reviewer. The reviews are appended at the bottom of this email and any accompanying reviewer attachments can be seen via the link below:

[LINK]

In light of these reviews, I am afraid that we will not be able to accept the manuscript for publication in the journal in its current form, but we would like to consider a revised version that addresses the reviewers' and editors' comments. We cannot make any decision about publication until we have seen the revised manuscript and your response, and we plan to seek re-review by one or more of the reviewers.

In revising the manuscript for further consideration, your revisions should address the specific points made by each reviewer and the editors. Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments, the changes you have made in the manuscript, and include either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please submit a clean version of the paper as the main article file; a version with changes marked should be uploaded as a marked up manuscript.

In addition, we request that you upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, please upload your figure files to the PACE digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at PLOSMedicine@plos.org.

We hope to receive your revised manuscript by Feb 11 2022 11:59PM. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

Please use the following link to submit the revised manuscript:

https://www.editorialmanager.com/pmedicine/

Your article can be found in the "Submissions Needing Revision" folder.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

We look forward to receiving your revised manuscript.

Sincerely,

Callam Davidson,

Associate Editor

PLOS Medicine

plosmedicine.org

-----------------------------------------------------------

Requests from the editors:

Please update your title to ‘Artificial sweeteners and cancer risk: a population-based cohort study’ or similar (please do not include the word ‘prospective’).

Lines 49-51: Please shorten the limitations section of your abstract to a single sentence (e.g. ‘Limitations of this study include potential selection bias, residual confounding, and reverse causality, though sensitivity analyses were performed to address these concerns.’ or similar).

Lines 57-8: Please remove the trial registration number/URL from the abstract.

Line 71: Please quantify the key findings mentioned in your Author Summary (with 95% confidence intervals and p-values).

Thank you for including the STROBE-nut checklist – please update the checklist to use section names/paragraph numbers as opposed to page numbers (which are liable to change during the revision process).

As no pre-specified analysis plan is available for this study, the term ‘prospective’ should be avoided throughout; please refer to the study instead as a population-based cohort study. If the authors wish to maintain the ‘prospective’ descriptor, it would be best if a document detailing the pre-specified analysis plan could be provided in the supporting information (the URL provided on lines 145-6 appears to be a news article).

Similar to the above, apologies if I have missed it but I could not locate the general protocol at the provided URL (https://info.etude-nutrinet-sante.fr/siteinfo/), would it be possible to provide a direct link to the protocol?

Please confirm whether SG’s affiliation with Open Food Facts ought to be declared as a COI.

Lines 262/332: Please tabulate this data and present it in the supporting information.

Line 433: Please remove this subheading in the Discussion.

Please show the unadjusted comparisons as well as the adjusted comparisons in Table 2.

Comments from the reviewers:

Reviewer #2: Alex McConnachie, Statistical Review

I thank the authors for their consideration of my original points. I feel that the paper has been improved, though I have one remaining comment.

The data presented in supplementary table 2 do not seem to be in line with the interaction p-values reported. For total artificial sweeteners, the interaction p-value is given as 0.044, suggesting different associations between intake and premenopausal breast cancer compared to postmenopausal breast cancer. However, the hazard ratios do not appear to be very different. For low consumers vs. non-consumers, the HRs are the same (1.09), and for high consumers vs. non-consumers, the difference is small (1.15 vs. 1.20) with confidence intervals that entirely overlap. These HRs and CIs do not, in my opinion, seem consistent with an interaction p-value of 0.044. If anything, it is for sucralose, where the associations trend in opposite directions, where I might anticipate a significant interaction, but this interaction is reported as p=0.152.

Also, I spotted a couple of typos:

In supplementary table 3, for all cancers, sucralose, the HR is reported as 1.10, which is very close to the upper confidence limit. Should this be 1.01?

In supplementary table 4, the HR between [ No AS / Low sugar ] and [ Low AS / Low sugar ] should be less than 1, but is reported as 1.15, the same as the opposite comparison.

Reviewer #3: I am satisfied with the revisions made to this article.

Reviewer #4: The authors have adequately addressed the comments of the Editors and Reviewers.

Minor editing is needed to correct small errors (e.g., Introduction, lines 82-83- could be edited to "…reduce added sugar content and corresponding calories while maintaining sweetness"; Results, page 9, line 326- comma instead of semicolon for 0.250; Discussion, line 357- "CPS-II" cohort, etc.).

Any attachments provided with reviews can be seen via the following link:

[LINK]

26 Jan 2022

Submitted filename: AUTHOR RESPONSE LETTER_PlosMed_R2.docx

Click here for additional data file.

17 Feb 2022

Dear Dr. Debras,

Thank you very much for re-submitting your manuscript "Artificial sweeteners and cancer risk: results from the NutriNet-Santé population-based cohort study" (PMEDICINE-D-21-04643R3) for review by PLOS Medicine.

I have discussed the paper with my colleagues and the academic editor and it was also seen again by the statistical reviewer. I am pleased to say that provided the remaining editorial and production issues are dealt with we are planning to accept the paper for publication in the journal.

The remaining issues that need to be addressed are listed at the end of this email. Any accompanying reviewer attachments can be seen via the link below. Please take these into account before resubmitting your manuscript:

[LINK]

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

In revising the manuscript for further consideration here, please ensure you address the specific points made by each reviewer and the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. If you haven't already, we ask that you provide a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract.

We hope to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We ask every co-author listed on the manuscript to fill in a contributing author statement. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT.

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript.

Please note, when your manuscript is accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you've already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosmedicine@plos.org.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Feb 24 2022 11:59PM.

Sincerely,

Callam Davidson,

Associate Editor

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

Lines 52, 79, and 443: Please update to ‘many food and beverage brands worldwide’.

Line 71: Please add ‘(HR)’.

Figures 2 and 3: The legends for these figures do not explain the asterisks in the figures.

Lines 339-349: Please consider whether some of this content would be better placed in the Discussion (presentation of results from the previous study on lines 339-341 and the interpretation at lines 347-348). Presentation of the data in the Supplementary Materials should however remain in the Results section.

Supplementary Results: I could not find reference to this part of the Supplementary Materials in the manuscript.

Comments from Reviewers:

Reviewer #2: Alex McConnachie, Statistical Review

I thank the authors once again for considering my comments, and I am happy with the modifications they have made. I have no further comments.

Any attachments provided with reviews can be seen via the following link:

[LINK]

22 Feb 2022

Submitted filename: AUTHOR RESPONSE LETTER_PlosMed_R3.docx

Click here for additional data file.

23 Feb 2022

Dear Dr Debras,

On behalf of my colleagues and the Academic Editor, Dr Wei Zheng, I am pleased to inform you that we have agreed to publish your manuscript "Artificial sweeteners and cancer risk: results from the NutriNet-Santé population-based cohort study" (PMEDICINE-D-21-04643R4) in PLOS Medicine.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Once you have received these formatting requests, please note that your manuscript will not be scheduled for publication until you have made the required changes.

In the meantime, please log into Editorial Manager at http://www.editorialmanager.com/pmedicine/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production process.

PRESS

We frequently collaborate with press offices. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximise its impact. If the press office is planning to promote your findings, we would be grateful if they could coordinate with medicinepress@plos.org. If you have not yet opted out of the early version process, we ask that you notify us immediately of any press plans so that we may do so on your behalf.

We also ask that you take this opportunity to read our Embargo Policy regarding the discussion, promotion and media coverage of work that is yet to be published by PLOS. As your manuscript is not yet published, it is bound by the conditions of our Embargo Policy. Please be aware that this policy is in place both to ensure that any press coverage of your article is fully substantiated and to provide a direct link between such coverage and the published work. For full details of our Embargo Policy, please visit http://www.plos.org/about/media-inquiries/embargo-policy/.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Thank you again for submitting to PLOS Medicine. We look forward to publishing your paper.

Sincerely,

Callam Davidson

Associate Editor

PLOS Medicine