This epidemiological study provides, for the first time, exploratory evidence of associations between food coloring additives and cancer incidence. Consumption of total coloring additives was found to be significantly associated with a higher incidence of overall cancer (a rate of 13.3% among high consumers of colorings compared to 12.1% among low consumers), as well as breast cancer and postmenopausal breast cancer. Plain caramel (E150a) was associated with an increased incidence of overall cancer, while beta-carotene (E160a) was associated with an increased incidence of both overall cancer and breast cancer. These findings require confirmation.
Objectives:
To investigate the associations between exposure to food coloring additives and cancer incidence within the NutriNet-Santé cohort.
Study design and methods:
The study included 105,260 participants (78.3% women; mean age 42.0 years [standard deviation 14.5]) from the French NutriNet-Santé cohort who were cancer-free at baseline and followed for over 7 years. Dietary intakes were assessed using repeated 24-hour dietary records (24HDRs) that included industrial product brand names. Cumulative, time-dependent exposure to food additives was evaluated using multiple composition databases and ad hoc laboratory analyses of food matrices. Associations between exposure to food additive colorants (categorized into sex-specific tertiles when the proportion of exposed participants exceeded two-thirds, or into non-exposed/moderately exposed/highly exposed categories based on the sex-specific median otherwise) and cancer incidence were assessed.
Results
A total of 4,226 incident cancer cases (including 508 prostate cancers, 1,208 breast cancers [387 premenopausal and 821 postmenopausal], and 352 colorectal cancers) were identified during the follow-up period. Total food coloring additives were associated with a higher incidence of overall cancer [HR (highest consumers vs. non/low consumers) (95% CI): 1.14 (1.05–1.24)], breast cancer [1.21 (1.03–1.42)], and postmenopausal breast cancer [1.32 (1.09–1.61)]. The absolute risk at age 60 corresponded to an incidence of 13.3%, 5.7%, and 14.9% (among the highest consumers) versus 12.1%, 4.8%, and 12.5% (among non- and low consumers) for overall cancer, breast cancer, and postmenopausal breast cancer, respectively. After adjustment for multiple testing, two colorings stood out for their association with cancer risk: plain caramel (E150a, according to European classification) was associated with an increased incidence of overall cancer [1.15 (1.07–1.25)], and beta-carotene (E160a) with overall cancer [1.16 (1.07–1.25)] and breast cancer [1.41 (1.23–1.62)]. It should be noted that although certain colorings—including ammonium sulfite caramel—exhibit toxicological plausibility linked to the formation of potentially carcinogenic compounds, this epidemiological study did not reveal a significant association between the consumption of this coloring and cancer risk, thereby highlighting the limitations of population-level exposure data. Interaction between coloring additives and ultra-processing The statistical analyses conducted showed no interaction between the consumption of ultra-processed foods and colorings regarding cancer outcomes. These data suggest an effect of coloring additives in their own right, which is not explained by the effect of ultra-processed foods.
Conclusion:
This prospective epidemiological study has identified, for the first time, associations between the consumption of food colorings and cancer in general, breast cancer, and postmenopausal breast cancer. These findings require confirmation. Furthermore, as participants in the NutriNet-Santé cohort are highly educated, health-conscious, and predominantly female, the results cannot be extrapolated to the general French population.
Learn more about the study’s findings.
Orchidali can help you improve the quality of your products.
Français
Food additives and cancer risk in the French NutriNet-Santé cohort
This epidemiological study provides, for the first time, exploratory evidence of associations between food coloring additives and cancer incidence. Consumption of total coloring additives was found to be significantly associated with a higher incidence of overall cancer (a rate of 13.3% among high consumers of colorings compared to 12.1% among low consumers), as well as breast cancer and postmenopausal breast cancer. Plain caramel (E150a) was associated with an increased incidence of overall cancer, while beta-carotene (E160a) was associated with an increased incidence of both overall cancer and breast cancer. These findings require confirmation.
Objectives:
To investigate the associations between exposure to food coloring additives and cancer incidence within the NutriNet-Santé cohort.
Study design and methods:
The study included 105,260 participants (78.3% women; mean age 42.0 years [standard deviation 14.5]) from the French NutriNet-Santé cohort who were cancer-free at baseline and followed for over 7 years. Dietary intakes were assessed using repeated 24-hour dietary records (24HDRs) that included industrial product brand names. Cumulative, time-dependent exposure to food additives was evaluated using multiple composition databases and ad hoc laboratory analyses of food matrices. Associations between exposure to food additive colorants (categorized into sex-specific tertiles when the proportion of exposed participants exceeded two-thirds, or into non-exposed/moderately exposed/highly exposed categories based on the sex-specific median otherwise) and cancer incidence were assessed.
Results
A total of 4,226 incident cancer cases (including 508 prostate cancers, 1,208 breast cancers [387 premenopausal and 821 postmenopausal], and 352 colorectal cancers) were identified during the follow-up period. Total food coloring additives were associated with a higher incidence of overall cancer [HR (highest consumers vs. non/low consumers) (95% CI): 1.14 (1.05–1.24)], breast cancer [1.21 (1.03–1.42)], and postmenopausal breast cancer [1.32 (1.09–1.61)]. The absolute risk at age 60 corresponded to an incidence of 13.3%, 5.7%, and 14.9% (among the highest consumers) versus 12.1%, 4.8%, and 12.5% (among non- and low consumers) for overall cancer, breast cancer, and postmenopausal breast cancer, respectively. After adjustment for multiple testing, two colorings stood out for their association with cancer risk: plain caramel (E150a, according to European classification) was associated with an increased incidence of overall cancer [1.15 (1.07–1.25)], and beta-carotene (E160a) with overall cancer [1.16 (1.07–1.25)] and breast cancer [1.41 (1.23–1.62)]. It should be noted that although certain colorings—including ammonium sulfite caramel—exhibit toxicological plausibility linked to the formation of potentially carcinogenic compounds, this epidemiological study did not reveal a significant association between the consumption of this coloring and cancer risk, thereby highlighting the limitations of population-level exposure data. Interaction between coloring additives and ultra-processing The statistical analyses conducted showed no interaction between the consumption of ultra-processed foods and colorings regarding cancer outcomes. These data suggest an effect of coloring additives in their own right, which is not explained by the effect of ultra-processed foods.
Conclusion:
This prospective epidemiological study has identified, for the first time, associations between the consumption of food colorings and cancer in general, breast cancer, and postmenopausal breast cancer. These findings require confirmation. Furthermore, as participants in the NutriNet-Santé cohort are highly educated, health-conscious, and predominantly female, the results cannot be extrapolated to the general French population.
Learn more about the study’s findings.
Orchidali can help you improve the quality of your products.