Safety concerns surrounding food-grade titanium dioxide, termed E 171, were raised by findings recently reported by Bettini and coworkers (2017) claiming that oral administration of this ingredient to rats produced lesions in the colon (aberrant crypt foci), and change to the immune system. The study conducted by a team of French government scientists was reviewed by European Food Safety Authority (EFSA), the European Union regulatory agency counterpart to the U.S. Food and Drug Administration, and they determined that the Bettini study had serious methodological flaws. These flaws included administration of E 171 in drinking water despite E 171 being insoluble in water and not present in any liquid food products.
To disperse E 171 in water Bettini and coworkers sonicated E 171 raising questions that the material may have been altered by the sonication procedure. Concerns were also raised that critical controls were omitted in the Bettini study, there was an absence of a statistically significant increase in aberrant crypt foci, and there was a lack of a sufficient number of experimental doses to adequately establish whether E 171 was, in fact, responsible for any of the effects Bettini and coworkers claimed were due to E 171 administration.
To further address the Bettini study, Michigan State University (MSU) investigators at the Center for Research on Ingredient Safety in collaboration with the University of Nebraska Medical Center (UNMC), conducted additional studies correcting for the methodological flaws in the Bettini report. Here E 171 was administered for 7- and 100-days to rats in the diet using multiple doses, which were less than, similar to, and greater than those used by Bettini.
In addition, appropriate experimental controls were included and perhaps most importantly, the study was conducted in a blinded manner such that the investigative team did not know which tissues came from which experimental animals until the analysis of all biological samples was completed to eliminate the possibility of experimental bias.
In this most recent study reported in the journal Food and Chemical Toxicology, MSU and UNMC scientists were unable to demonstrate any adverse effects due to E 171 including aberrant crypt foci or changes in immune parameters reported by Bettini. Moreover, the National Toxicology Program within the National Institute of Environment Health Sciences in the U.S. also conducted a 2-year study in rats that showed no adverse effects by titanium dioxide placed in the diet at doses approximately10-fold higher than the highest dose used by MSU and UNMC researchers.
The EU-based nongovernment organization, Health and Environment Alliance (HEAL), raised several concerns regarding the MSU UNMC study. Specifically, “why did colon samples get obscured, preventing the proper examination of the whole samples (section 3.7, pp. 24-25)? The study mentions an “unfortunate” phenomenon but does not provide a solid explanation.” Although the aforementioned changes were reported for the purpose of thoroughness and complete transparency, evaluation of aberrant crypts was similar between all groups and readily distinguished between DMH-treated and non-DMH-treated groups. Furthermore, there was no impact on histopathologic evaluation.
HEAL also questioned, “why is the increase of early markers of colorectal cancer (visible in table 6, page 56) dismissed in the article (section 3.7, pp. 24-25)”? Table 6 referenced above, in fact, was not dismissed and clearly shows that the increase in aberrant crypt foci is due to dimethylhydrazine (DMH) pretreatment and not E 171 (titanium dioxide). Importantly, Bettini et al. pretreated rats in their study with DMH, to increase the sensitivity of the rats to GI lesions, before treating with E 171 but failed to include the DMH only treatment control. It is noteworthy that DMH is a potent DNA altering chemical termed “genotoxicant” and gives widely varying results in short term studies.
The French government has announced plans for a ban on food products containing E 171 beginning January 2020, which at this time is not supported by other EU countries.