The comparison of the alternate Mediterranean diet score (aMed) and MedDietScore (MDS) in American samples

摘要

Objectives: The purpose of this study was to assess the relationship between two Mediterranean diet indices and to evaluate the proportion of participants who were ranked into identical tertiles of accordance with the Alternate Mediterranean Diet (aMed) and MedDietScore (MDS) tools when applied to an American samples.;Methods: In this secondary analysis, participants from four samples were pooled into two groups -- one in which respondents completed the web-based VioScreen(TM) (n=200) food frequency questionnaire (FFQ) and the second, those who completed a paper-based Brief Block 2000 FFQ (n=827). Mediterranean diet scores were calculated based on the sex-specific median intakes of nine aMed components as well as meeting target frequency per week of the MDS components. Participants were categorized into a priori tertiles for each score.;Results: Scores for aMed and MDS were moderately correlated in the VioScreen(TM) (rho= 0.546, p<0.001) and Brief Block (rho=0.627, p<0.001) samples. The greatest proportion of participants was classified into Tertile 2 for VioScreen(TM) each scoring paradigm (40% for aMed and 71% for MDS). In the Brief Block sample, 47% of participants were assigned to Tertile 1 for aMed and for MDS 52% into Tertile 2. Only 47% of VioScreen(TM) and 60.3% of Brief Block participants were ranked into identical tertiles for aMed and MDS. Classification agreement between aMed and MDS was fair for VioScreen(TM) (weighted &kgr; = 0.223, p<0.001) and Brief Block samples (weighted &kgr; = 0.384, p<0.001).;Conclusions: Agreement between aMed and MDS was no more than fair for either FFQ, indicating that the scoring paradigms are not interchangeable in measuring accordance to the Mediterranean diet. Further investigation into the effects of FFQ selection on aMed and MDS scoring in addition to factor analysis of the variability between aMed and MDS is warranted in American samples.