OBJECTIVE: To perform construct validation of the population-based Dietary Inflammatory Index (DII) using dietary data from two different dietary assessments and serum high-sensitivity C-reactive protein (hs-CRP) as the construct validator.

DESIGN: Using data derived from (i) three 24 h dietary recalls (24HR) at baseline and at the end of each subsequent quarter (i.e. up to fifteen over a year) and (ii) a 7 d dietary recall (7DDR) measured at baseline and then quarterly, regression analyses were conducted to test the effect of the DII score on serum hs-CRP as dichotomous (3 mg/l), while controlling for important potential confounders.

SETTING: Existing data from the Seasonal Variation of Blood Cholesterol Study (SEASONS), a longitudinal observational study of healthy participants recruited in Worcester, MA, USA and participants were followed for 1 year.

SUBJECTS: Participants who had at least one hs-CRP measurement over her/his 1-year participation (n 495 for 24HR, n 559 for 7DDR).

RESULTS: Higher DII scores were associated with values of hs-CRP >3 mg/l (OR = 1.08; 95 % CI 1.01, 1.16, P = 0.035 for the 24HR; and OR = 1.10; 95 % CI 1.02, 1.19, P = 0.015 for the 7DDR).

CONCLUSIONS: The population-based DII was associated with interval changes in hs-CRP using both the 24HR and 7DDR. The success of this first-of-a-kind attempt at relating individuals' intakes of inflammation-modulating foods using this refined DII, and the finding that there is virtually no drop-off in predictive capability using a structured questionnaire in comparison to the 24HR standard, sets the stage for use of the DII in a wide variety of other epidemiological and clinical studies.