Whole grain consumption and body mass index in adult women: An analysis of NHANES 1999-2000 and the USDA pyramid servings database

Objective: To examine the relationship between whole grain consumption and body mass index (BMI) in a sample of American adult women. Methods: Dietary intake data from the National Health and Nutrition Examination Survey 1999-2000 were linked to the USDA Pyramid Servings Database. Women 19 years of age and older (n = 2,092) were classified into groups based on their average whole grain (WG) intake: 0 servings, more than 0 but less than 1 serving, and >= 1 servings per day. Within these classifications, mean BMI, mean waist circumference and percent overweight/obese (BMI >= 25) were identified as primary dependent variables. Regression and logistic regression analyses were used to assess associations between BMI, waist circumference and percent of the population overweight/obese (BMI >= 25) and WG consumption. Results: Women consuming at least one serving of WG had a significantly lower mean BMI and waist circumference than women with no WG consumption (p < 0.05). Multiple regression analysis showed a significant inverse relationship between BMI and whole grain intake after adjustment for age, energy intake, dietary fiber and alcohol intake (p = 0.004). This effect was mildly attenuated but remained significant after further adjustment for level of physical activity, smoking status, ethnicity and education (p = 0.018). The odds ratio for having a BMI >= 25 was 1.47 (95% CI 1.12-1.94; p for trend 0.013) for women consuming no WG compared to those consuming at least one serving, after adjustment for all covariates. Conclusions: These data support other research suggesting increased WG intake may contribute to a healthy body weight in adult women.