Recent reports suggest persistence of health disparities related to socio-economic position (SEP). To understand if diet may be a contributor to these trends, we examined secular trends in the association of diet and indicators of SEP from 1971–1975 to 1999–2002.

We used data from the National Health and Nutrition Examination Surveys (NHANES) I (1971–1975), II (1976–1980), III (1988–1994) and 1999–2002 to examine the independent associations of poverty income ratio (PIR) and education with diet and biomarkers of diet and disease in 25–74-year-olds (n = 36 600). We used logistic and linear regression methods to adjust for multiple covariates and survey design to examine these associations.

A large PIR differential in the likelihood of reporting a fruit or all five food groups and vitamin C intake, and an education differential in likelihood of obesity and carbohydrate intake, was noted in 1971–1975 but narrowed in 1999–2002 (P < 0.007). The positive association of education with intake of a fruit, vegetable or all five food groups, vitamins A and C, calcium and potassium intake remained unchanged across surveys (P < 0.001). Similarly, the positive association of PIR with the amount of foods and intakes of energy and potassium remained unchanged over three decades (P < 0.001). The education and the PIR differential in energy density, and the PIR differential in the likelihood of obesity, persisted over the period of the four surveys (P < 0.001).

Persistence of unfavourable dietary and biomarker profiles in Americans with low income and education suggests continued need for improvement in the quality of diets of these high-risk groups.