The association between self-reported acrylamide intake and hemoglobin adducts as biomarkers of exposure.

Background: The validity of epidemiological studies assessing a lifetime cancer risk due to environmental factors, such as nutrition or smoking behavior strongly depends on the validity of the patient's history.

Objectives: This study assesses whether a standardized questionnaire is a valid tool to identify exposure with acrylamide by relating the self-reported food and smoking history with a biomarker, namely hemoglobin-adduct levels of acrylamide.

Methods: Objective parameters of previous exposure, such as hemoglobin-adduct levels of acrylamide and of the smoking-specific acrylonitrile, respectively, were related to self-reported data in 1,008 volunteers of the general population in bivariate analyses and a multiple linear regression analysis using the log-transformed biomarker levels as outcome.

Results: Smoking was significantly associated with adduct levels of acrylamide (p < 0.0001) and had a main contribution to the internal burden with acrylamide. In cigarette smokers a strong correlation between the number of cigarettes smoked daily and the corresponding biomarkers was observed. Focusing on non-smokers (n = 828), a significant but weak correlation was found in bivariate analyses (Spearman rank correlation coefficient: 0.178 (95% CI: 0.089-0.268) in females and 0.168 (95% CI: 0.063-0.273) in males. A multiple linear regression analysis similarly yielded evidence of a significant association between the highest intake category and adduct levels; however, explained variability was very small (R(2) = 0.08).

Conclusion: Self-reported data concerning smoking behavior were highly valid, while self-reported food intake is apparently not as useful for estimating food-related acrylamide exposure.