Dietary Intake Among US Adults, 1999-2012

Colin D Rehm, José L Peñalvo, Ashkan Afshin, Dariush Mozaffarian, Colin D Rehm, José L Peñalvo, Ashkan Afshin, Dariush Mozaffarian

Abstract

Importance: Most studies of US dietary trends have evaluated major macronutrients or only a few dietary factors. Understanding trends in summary measures of diet quality for multiple individual foods and nutrients, and the corresponding disparities among population subgroups, is crucial to identify challenges and opportunities to improve dietary intake for all US adults.

Objective: To characterize trends in overall diet quality and multiple dietary components related to major diseases among US adults, including by age, sex, race/ethnicity, education, and income.

Design, setting, and participants: Repeated cross-sectional investigation using 24-hour dietary recalls in nationally representative samples including 33,932 noninstitutionalized US adults aged 20 years or older from 7 National Health and Nutrition Examination Survey (NHANES) cycles (1999-2012). The sample size per cycle ranged from 4237 to 5762.

Exposures: Calendar year and population sociodemographic subgroups.

Main outcomes and measures: Survey-weighted, energy-adjusted mean consumption and proportion meeting targets of the American Heart Association (AHA) 2020 continuous diet scores, AHA score components (primary: total fruits and vegetables, whole grains, fish and shellfish, sugar-sweetened beverages, and sodium; secondary: nuts, seeds, and legumes, processed meat, and saturated fat), and other individual food groups and nutrients.

Results: Several overall dietary improvements were identified (P < .01 for trend for each). The AHA primary diet score (maximum of 50 points) improved from 19.0 to 21.2 (an improvement of 11.6%). The AHA secondary diet score (maximum of 80 points) improved from 35.1 to 38.5 (an improvement of 9.7%). Changes were attributable to increased consumption between 1999-2000 and 2011-2012 of whole grains (0.43 servings/d; 95% CI, 0.34-0.53 servings/d) and nuts or seeds (0.25 servings/d; 95% CI, 0.18-0.34 servings/d) (fish and shellfish intake also increased slightly) and to decreased consumption of sugar-sweetened beverages (0.49 servings/d; 95% CI, 0.28-0.70 servings/d). No significant trend was observed for other score components, including total fruits and vegetables, processed meat, saturated fat, or sodium. The estimated percentage of US adults with poor diets (defined as <40% adherence to the primary AHA diet score components) declined from 55.9% to 45.6%, whereas the percentage with intermediate diets (defined as 40% to 79.9% adherence to the primary AHA diet score components) increased from 43.5% to 52.9%. Other dietary trends included increased consumption of whole fruit (0.15 servings/d; 95% CI, 0.05-0.26 servings/d) and decreased consumption of 100% fruit juice (0.11 servings/d; 95% CI, 0.04-0.18 servings/d). Disparities in diet quality were observed by race/ethnicity, education, and income level; for example, the estimated percentage of non-Hispanic white adults with a poor diet significantly declined (53.9% to 42.8%), whereas similar improvements were not observed for non-Hispanic black or Mexican American adults. There was little evidence of reductions in these disparities and some evidence of worsening by income level.

Conclusions and relevance: In nationally representative US surveys conducted between 1999 and 2012, several improvements in self-reported dietary habits were identified, with additional findings suggesting persistent or worsening disparities based on race/ethnicity, education level, and income level. These findings may inform discussions on emerging successes, areas for greater attention, and corresponding opportunities to improve the diets of individuals living in the United States.