What is the "Western Pattern?": Refining Factor-Derived Dietary Patterns from the NHANES Food Frequency Questionnaire Using 24-hour Dietary Recall Data

Meal pattern analysis is an emerging technique in nutritional epidemiology useful for identifying eating patterns associated with disease. Two meal patterns commonly found in large epidemiologic studies are the "Western" pattern, characterized by high intakes of red meat, eggs, refined grains, fried foods and added sugars, and the "prudent" or "healthy" pattern, consisting of fruits, vegetables, whole grains, and low fat food items. The Western pattern has become more common worldwide along with the chronic diseases associated with this eating pattern, including obesity, heart disease, type II diabetes mellitus and some diet-related cancers. Factor analysis was used to determine major dietary patterns from 2005- 06 National Health and Nutrition Examination Survey (NHANES) Food Frequency Questionnaire (FFQ) data. Variables were entered into factor analysis with three different grouping schemes. Individuals in the top quintile of each factor were considered to have the greatest adherence to a given pattern. The 39 category analysis provided a balance between simplifying output and retaining predictive value of different food types. Multiple logistic regression models were constructed to determine predictors from demographic and 24-hour dietary recall data, including total energy, food source, food groups represented, number of eating occasions, and breakfast skipping. Men were more likely to follow the Western pattern, and women as well as older Americans were more likely to follow the healthy diet pattern. Mexican Americans and other Hispanics were much less likely, and black Americans much more likely than other ethnicities to follow the Western diet pattern. There was a clear linear relationship showing increased likelihood of following the healthy pattern with each increasing level of education. High income groups were both less likely to follow the healthy diet pattern and less likely to follow the Western pattern when compared to all other income groups. Increasing daily energy had a strong positive association with the Western pattern, and increasing dietary diversity was inversely associated with the Western pattern. Eating out at a restaurant or bar was associated with a lower likelihood of following the healthy pattern. A greater number of eating occasions was generally associated with a lower likelihood of following the Western pattern with the exception of the highest frequency snacking group (9-21 snack items per day). Individuals reporting only 1-2 snack items per day were about half as likely to follow the healthy pattern. Skipping breakfast was associated with an almost 1.5 times greater likelihood of following the healthy diet pattern. These findings are potentially useful for identifying individuals at risk for disease, or targeting health promotion efforts to specific population subgroups. A short form scale based on these results may help to identify strong adherents of the Western dietary pattern in resource-limited settings.Dr.P.H., Community Health and Prevention -- Drexel University, 201