Abstract 1913: Uses of questionnaire- and biomarker-based measures of pro- and antioxidant exposures in relation to prostate cancer risk

Abstract

Abstract We previously used an oxidative balance score (OBS) as a measure of combined pro- and antioxidant exposure status using either questionnaire or biomarker data. However, questionnaires are subject to recall bias while biomarkers may reflect only recent exposures. In this analysis, we used questionnaire and biomarker data separately and in combination in an effort to better characterize oxidative stress exposures in a study of prostate cancer. Data from a previously conducted population-based case-control study of incident prostate cancer was used for the current analysis. A total of 13 a priori-selected pro- and antioxidant factors were used to calculate the OBS. All continuous variables reflecting pro-oxidant (unsaturated fat and iron) and antioxidant exposures (vitamin C, lycopene, α-carotene, β-carotene, lutein, β-cryptoxanthin, α-tocopherol, and urinary selenium) were divided into low, medium, and high categories using tertile cutoffs among the controls. Participants with low (1st tertile) pro-oxidant exposure were awarded 2 points, those with medium (2nd tertile) exposure received 1 point, and those with high (3rd tertile) exposure received 0 points. For antioxidants, low, medium, and high levels were assigned 0, 1, and 2 points, respectively. A similar scoring approach was used for categorical variables (smoking and use of aspirin and NSAIDs). The overall OBS was then calculated by adding up the points assigned to each participant. The OBS was calculated using three types of data: questionnaire, biomarkers, and a combination of both. In the combined (questionnaire plus biomarker) OBS analysis, 0 points were assigned to those who had low antioxidant exposures for both questionnaire and biomarker measurements, 2 points were assigned for those who had both high antioxidant values, and 1 point for all the other combinations. The association between OBS and prostate cancer was examined using multivariate logistic regression, adjusting for age, race, total energy intake, BMI, plasma cholesterol, and a family history of prostate cancer in a first degree relative. The OBS was divided into three equal intervals, using the lowest interval as reference. In the combined OBS analysis, the adjusted odds ratios (95% CI) for the middle and highest intervals were 0.52 (0.30-0.91) and 0.62 (0.14-2.72), respectively. Similar results were obtained using questionnaire data only. Using only biomarker measurements, the adjusted odds ratios (95% CI) for the middle and highest intervals were 0.56 (0.30-1.08) and 0.19 (0.07-0.49), respectively. These findings provide evidence that the use of questionnaire data instead of and in addition to biomarkers may not necessarily improve the assessment of oxidative stress-related exposures. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 1913. doi:10.1158/1538-7445.AM2011-1913