Skip to Main Content

An official website of the United States government

Dietary predictors of colorectal adenoma in different racial subgroups using three different methods of dietary evaluation

Principal Investigator
Name
Alyson Haslam
Institution
University of Georgia
Position Title
PhD student
Email
About this CDAS Project
Study
PLCO (Learn more about this study)
Project ID
PLCO-131
Initial CDAS Request Approval
Feb 25, 2015
Title
Dietary predictors of colorectal adenoma in different racial subgroups using three different methods of dietary evaluation
Summary
Colorectal adenomas are growths that occur in the lining of the intestines – about 1 in 10-20 people will develop them. These growths often remain benign but can become cancerous if left untreated. Most cancers of the colon and rectum originate from adenomas. Because colorectal cancer is the third most common cancer in men and women in the United States (almost 130,000 people will develop this cancer in 2014; American Cancer Society), preventing adenomas may be a fundamental way to reduce the burden of a common cancer. In addition to the high prevalence of these conditions, a serious concern is the notable racial disparities that occur in colorectal cancer and possibly colorectal adenoma incidence. Approximately 80% of colorectal cancer cases can be attributed to diet, and because colorectal cancer and colorectal adenomas share similar risk factors, it is likely that diet also contributes to the majority of colorectal adenomas. Reasons for the racial disparities in colorectal adenoma incidence are largely unknown, but because previous studies have shown racial differences in individual nutrient intake, diet could be a key way to reduce the disparities that occur with these types of conditions.
Different measurements of diet have been proposed. Few studies, if any, have examined racial differences in overall diet quality in relation to colorectal growths. For the present analyses, three methods for assessing diet quality will be used. Because of the complex association between diet, systemic inflammation, and colorectal cancer, dietary methods that have previously been studied in relation to inflammatory markers will be used in these proposed analyses (Gunter et al., 2006; Schultze et al., 2005). The first two methods are investigator-driven methods (i.e., the investigator determines the dietary scoring) that score the general “healthiness” (e.g. inflammatory nature) of diets, while the third method is a data driven method (i.e., patterns in the data determines the outcome). The first method will use the patented Dietary Inflammatory Index, which is a research-developed method of scoring the inflammatory nature of diet using 45 food items identified as pro- or anti-inflammatory, based on previously published studies (Shivappa et al., 2013). The second method will use the alternative Mediterranean diet index, was developed because of the notable health of the people in the Mediterranean region, and is based upon the consumption of fruits, vegetables, nuts, whole grains, legumes, fish, a high ratio of mono-saturated fats/saturated fats, limited alcohol, and limited meat (Fung et al., 2009). Finally, the third method will use decision tree analysis, which will be used to determine dietary predictors specific for colorectal adenomas and cancer.
Aims

The overarching goal of these analyses is to reduce the racial disparities that occur in colorectal adenomas and colorectal cancer occurrence, through preventable dietary approaches. To accomplish this goal, three specific aims will be employed.
Specific aim 1: To determine whether or not there are racial differences in the consumption of inflammatory diets, as measured by a validated Dietary Inflammatory Index. To determine whether or not an inflammatory diet at baseline predicts the incidence of colorectal growths, and whether or not this association is modified by race. It is hypothesized that blacks have a more inflammatory diet, and that a more inflammatory diet leads to greater colorectal adenoma and cancer occurrence.
Specific aim 2: To determine whether or not there are racial differences in the consumption of score-based diet, namely the Mediterranean diet. To determine whether or not a lower score on the Mediterranean diet at baseline predicts the incidence of colorectal growths, and whether or not this association is modified by race. It is hypothesized that blacks have a less Mediterranean diet, and that a less Mediterranean diet leads to higher colorectal adenoma or cancer incidence.
Specific aim 3: To apply classification trees to data on a large population of men and women in order to determine the specific dietary predictors of colorectal growths in the different racial subgroups, accounting for complex interaction between dietary factors. It is hypothesized that blacks have different predictors of colorectal adenoma and cancer than whites.
For the proposed analyses, only those in the screening arm will be evaluated. For the first two methods (inflammatory diet score and Mediterranean diet score), baseline dietary data will be analyzed and compared between the different races to see if there are differences in the intake of inflammatory diets or diets that are more or less Mediterranean in nature. Follow-up data will then be used to see if the consumption of more inflammatory diets or diets that score lower on the Mediterranean diet index are associated with a higher occurrence of colorectal adenomas or cancer in the racial subgroups. For the third method, the dietary predictors of colorectal growths will be determined using a novel method of decision tree analysis, which uses statistical methods to sort through the complex interactions to identify dietary factors specific for colorectal adenomas. These analyses will be stratified by race to identify dietary predictors specific for each racial subgroup. Analysis using all three methods will adjust for important lifestyle factors such as non-steroidal anti-inflammatory use, physical activity, age, smoking, and socioeconomic status.
It is expected that there will be differences in dietary intake between those of various racial subgroups. It is also expected that these differences contribute, at least in part, to racial disparities in colorectal adenoma/carcinoma incidence. Results of these studies could be used to identify preventative factors for minimizing racial disparities that occur in a disease that affects almost 140,000 people annually and kills about 50,000 annually and is heavily influenced by dietary factors (American Cancer Society, 2014).

Collaborators

Sara Wagner Robb
Mark Ebell
Hanwen Huang
James R Hebert

Related Publications