BACKGROUND: Identification of population dietary patterns has been recommended by experts as a key to developing innovative and targeted nutrition interventions and achieving long-term dietary behavior changes for health promotion and chronic disease risk reduction. Essential in this task is the evaluation of methods to accurately identify these unique dietary patterns. OBJECTIVE: To evaluate the validity and test the performance of a method for classifying adult men and women into one of five a priori dietary patterns. DESIGN: The first examination of Framingham Nutrition Studies took place between 1984 and 1987 and included 1,828 women and 1,666 men who completed the Framingham semi-quantitative food frequency questionnaire. Five unique dietary patterns for each sex were identified. Here we used Fisher's discriminant functions to create classification algorithms for identifying the dietary patterns of new individuals. Its validity and performance was evaluated using a variety of statistical tools. RESULTS: The new Framingham Dietary Pattern algorithm classified about 80% of women or men correctly and kappa statistics exceeded 0.70. The cluster-specific sensitivities ranged from 0.70 to 1.00 and specificities were all >0.88. The pooled conditional c statistics were 0.95 and 0.96 for women and men, respectively. Overall, we can be confident that our methodology offers a valid identification of male and female dietary patterns when applied in practice. CONCLUSIONS: The Framingham Dietary Pattern technique is a valid and reliable method for identifying the unique dietary behavior of adults. Our approach can also be used to guide the development and evaluation of other composite dietary quality indices.
BACKGROUND: Identification of population dietary patterns has been recommended by experts as a key to developing innovative and targeted nutrition interventions and achieving long-term dietary behavior changes for health promotion and chronic disease risk reduction. Essential in this task is the evaluation of methods to accurately identify these unique dietary patterns. OBJECTIVE: To evaluate the validity and test the performance of a method for classifying adult men and women into one of five a priori dietary patterns. DESIGN: The first examination of Framingham Nutrition Studies took place between 1984 and 1987 and included 1,828 women and 1,666 men who completed the Framingham semi-quantitative food frequency questionnaire. Five unique dietary patterns for each sex were identified. Here we used Fisher's discriminant functions to create classification algorithms for identifying the dietary patterns of new individuals. Its validity and performance was evaluated using a variety of statistical tools. RESULTS: The new Framingham Dietary Pattern algorithm classified about 80% of women or men correctly and kappa statistics exceeded 0.70. The cluster-specific sensitivities ranged from 0.70 to 1.00 and specificities were all >0.88. The pooled conditional c statistics were 0.95 and 0.96 for women and men, respectively. Overall, we can be confident that our methodology offers a valid identification of male and female dietary patterns when applied in practice. CONCLUSIONS: The Framingham Dietary Pattern technique is a valid and reliable method for identifying the unique dietary behavior of adults. Our approach can also be used to guide the development and evaluation of other composite dietary quality indices.
Authors: P A Quatromoni; D L Copenhafer; S Demissie; R B D'Agostino; C E O'Horo; B-H Nam; B E Millen Journal: J Epidemiol Community Health Date: 2002-05 Impact factor: 3.710
Authors: Barbara E Millen; Paula A Quatromoni; Byung-Ho Nam; Catherine E O'Horo; Joseph F Polak; Philip A Wolf; Ralph B D'Agostino Journal: J Am Diet Assoc Date: 2004-02
Authors: Kim G Jackson; Dave T Clarke; Peter Murray; Julie A Lovegrove; Brendan O'Malley; Anne M Minihane; Christine M Williams Journal: Genes Nutr Date: 2009-09-29 Impact factor: 5.523
Authors: Dolores M Wolongevicz; Lei Zhu; Michael J Pencina; Ruth W Kimokoti; P K Newby; Ralph B D'Agostino; Barbara E Millen Journal: J Obes Date: 2010-01-05
Authors: Daniele B Vinholes; Ione M F Melo; Carlos Alberto Machado; Hilton de Castro Chaves; Flavio D Fuchs; Sandra C Fuchs Journal: ScientificWorldJournal Date: 2012-05-22
Authors: Maria Luz Fernandez; Dele Raheem; Fernando Ramos; Conrado Carrascosa; Ariana Saraiva; António Raposo Journal: Int J Environ Res Public Health Date: 2021-03-10 Impact factor: 3.390