Weihrauch and Diehl (2004) reviewed the literature for human epidemiological studies on the health effects from exposure to a range of nonnutritive sweeteners (saccharin, cyclamate, aspartame, acesulfame-K, sucralose, alitame, and neotame) from foods and beverages. They found that assessing the carcinogenic potential of a single sweetener from dietary exposure was not feasible, but that the overall carcinogenic risk from nonnutritive sweeteners was negligible. Renwick (1990) reviewed toxicity reports and magnitude of safety factor for acesulfame-K, aspartame, cyclamate, and saccharin. Acceptable daily intake levels were reviewed and shown to be well below a toxic threshold for acesulfame-K, aspartame, and saccharin. The review reported that cyclamate was banned based on dose-response evidence at the upper dose intake level but the results were not reproducible in subsequent independent studies.
There is only one known large-scale study (Soffritti et al, 2005) to have tested aspartame in an animal model, from 8 weeks of age until natural death, at varying doses (ranging from 0 to 100,000 ppm) that include levels comparable to human exposure through foods and beverages. The authors concluded that administration of aspartame, even at low doses, caused an increase in incidence of malignant tumors of both epithelial and mesenchymal origin.
The data from the study were reviewed by the European Food Safety Authority (EFSA, 2006) and the Committee on Carcinogenicity of Chemicals in Food, Consumer Products, and the Environment (COC) of the UK Food Standards Agency. Both EFSA and the COC concluded that the study was flawed based on a number of errors identified by the panels. For example, the report of increased numbers of lymphomas and leukemias may have been related to the presence of chronic inflammatory disease in the lungs of rats, but the investigators failed to test for mycoplasma (a causative agent) in the animal colony. The reported tumors of the renal pelvis were found to be likely related to the treatment because high doses of chemical irritants are known to cause calcium imbalances in the rat, leading to renal tumors. Neither was a dose-response relationship between aspartame intake and tumor incidence established. Another reported flaw was that all the malignant tumor incidences and all tumor-bearing animals were aggregated for statistical purposes, and the reviewers found that the aggregated data were not sufficient to demonstrate the carcinogenic potential of aspartame.
In a rebuttal statement, Soffritti (2006) responded to the criticism that lymphoma and leukemia tumors were related to underlying lung disease, and that such diseases are common when animals are taken to the point of natural death as they were in his study. The author also pointed out that, if an infection were present in the colony, it would have affected males and females equally. The other criticisms to the study were not addressed