Estimation of Titanium Dioxide Intake by Diet and Stool Assessment among US Healthy Adults

Background Titanium dioxide (TiO2/E171) is used in foods primarily as a whitening agent. Little is known regarding TiO2 exposure in the United States. Objectives To quantify stool TiO2 content among US adults and evaluate its association with estimated intake. Methods Adults participated in phase 1 [three 24-h dietary recalls (DRs) and stool TiO2 measured from 3 matched samples (n = 52)] and/or phase 2 [tailored FFQ and stool TiO2 measured from 3 samples over 3 mo (n = 61)]. TiO2 in foods was estimated from a database, and concentration in 49 additional foods and 339 stool samples were quantified using inductively coupled plasma mass spectrometry. Associations between dietary and stool TiO2 were assessed by log-linear multivariable regression. USDA food groups (n = 49, servings/d) were related to stool TiO2 by stepwise regression. Results TiO2 food content varied by brand. Mean TiO2 intake from three 24-h DRs [0.19 +/- 0.31 mg/(kg body weight center dot d)] was lower than from the FFQ [0.30 +/- 0.21 mg/(kg body weight center dot d)]. Dietary TiO2 was not predictive of stool TiO2, in phase 1 or phase 2, 10<^>(beta) per 10 times higher dietary TiO2: 1.138 [10<^>(95% CI): 0.635, 2.037, P = 0.66] and 0.628 [10<^>(95% CI): 0.206, 1.910, P = 0.41], respectively. Food groups related to stool TiO2 were 1) milk desserts, sauces, and gravies [10<^>(beta) per servings/d: 3.361; 10<^>(95% CI): 0.312, 36.163; P = 0.002] and 2) yeast breads [10<^>(beta): 1.430; 10<^>(95% CI): 0.709, 2.884; P = 0.002] in phase 1 and 1) cream and cream substitutes [10<^>(beta) = 10.925; 10<^>(95% CI): 1.952, 61.137; P = 0.01] and 2) milk and milk drinks [10<^>(beta) = 0.306; 10<^>(95% CI): 0.086, 1.092, P = 0.07] in phase 2. Conclusions Intake of certain foods was associated with higher stool TiO2 content. There is a need for valid estimation of TiO2 intakes via the improvement of a dietary assessment method and a TiO2 food composition database. Future research should assess whether high stool TiO2 content is related to adverse health outcomes.