||5R03CA135661-02 Interpret this number
||University Of Tx Md Anderson Can Ctr
||Diet and Second-Hand Smoke in Lung Cancer Prediction
DESCRIPTION (provided by applicant):
This application is a formative step towards assessing whether diet alone or as a modulator of exposure to second hand smoke (SHS) appreciably influences lung cancer (LC) risk prediction in never smokers. We build upon an epidemiologic risk prediction model for LC, that we recently constructed from an ongoing LC case-control study [Spitz MR et al JNCI 2007]. In the risk model for life-long never smokers (based on 330 Caucasian LC patients and 370 controls), exposure to SHS and family history of cancer were the only significant risk factors for LC. This model was developed using epidemiologic and clinical data and had a discriminatory power of only 59%. Furthermore, this model did not include dietary data. We believe that inclusion of USDA pyramid servings of fruit and vegetables will result in a substantial gain in discriminatory power. We also hypothesize that SHS in the LC risk model may interact with other environmental, behavioral, or lifestyle factors and the joint effects of these factors may substantially alter an individual's LC risk. Therefore, using our on-going case-control study, and the extensive epidemiologic and 135-item food frequency questionnaire database from the parent grant, we propose to evaluate the incremental value of incorporating into the model USDA pyramid servings of fruit and vegetables and the joint effects of fruit and vegetable servings and SHS to improve the discriminatory power of our LC risk model. The specific aims for this project are:
1. To estimate the effect of USDA pyramid servings of fruit and vegetables on LC risk in never smokers- 330 newly diagnosed LC patients and 379 healthy controls. We will also examine whether diet is an indicator for lifestyle (including physical activity, alcohol intake and vitamin mineral supplementation).
2. To examine the joint effects of fruit and vegetable servings and second-hand smoke on LC risk.
3. To construct an expanded LC risk model incorporating the data in aims 1 and 2.
Our long-term goal is to identify the number of USDA pyramid servings of fruit and vegetables that contribute to our LC risk model in never smokers and determine whether specific fruit and vegetable servings modify LC risk from SHS. In order to be clinically useful, the model must use easy to obtain self-reported data derived from an abbreviated dietary questionnaire focusing on specific foods that reduce LC risk from SHS.
Layman's abstract: This project will generate a risk model for lung cancer in never smokers that includes the role of your diet and physical activity. We believe that a lung cancer risk prediction model in never smokers is important to help identify persons at high risk because 10-15 % of lung cancers are diagnosed in never smokers each year.