Associations between dietary macronutrient intake and plasma lipids demonstrate criterion performance of the Multi-Ethnic Study of Atherosclerosis (MESA) food-frequency questionnaire

摘要: Knowing the quantitative accuracy of dietary assessment tools bolsters validity diet–disease findings, informs design appropriate research questions and allows for calibration risk estimates correcting known error(1). FFQ are most commonly used tool in large epidemiological studies. Validity data from can be assessed utilising a variety study designs. Most often, subset population is selected to complete multiple food records or 24 h recalls over period weeks months as reference comparison with data. However, such measures imperfect, given correlated errors between either diet records(2-4). Opportunities validation via biomarkers limited nutrients and/or items and, some cases, require special sample preparation(5,6). Despite being objective (as opposed self-reported), themselves imperfect markers; thus, non-dietary factors that affect circulating concentrations often need considered interpretation. Consequently, studies(7-9) have taken advantage well-known associations carbohydrate fat plasma lipids (namely, TAG HDL-cholesterol)(10-12). The Multi-Ethnic Study Atherosclerosis (MESA) cohort including men women four different race/ethnic groups. The MESA was based on an originally designed Insulin Resistance (IRAS), which previously studied non-Hispanic whites, African-Americans Hispanics(13,14). Because Chinese-Americans were also included study, modifications made accommodate their unique cuisine. Although FFQ, changed IRAS has not been formally validated through records, it performs well insofar significant various aspects disease factors(15-20) incident type 2 diabetes(21) CVD(22) reported cohort. Since no single method capable characterising all constructs measured by utility any must evaluated basis complementary tests its validity. In context predictive suggested published studies(15-22), further characterisation overall performance collected terms relationships provides another lens understand FFQ. Furthermore, offers opportunity assess race/ethnic-specific Such valuable long-standing concern regarding reliability instruments minorities, African-Americans(23-25). Therefore, we cross-sectional macronutrient intakes fasting HDL full each group represented. Our hypotheses follows: (1) percentage total energy intake would inversely associated but positively TAG; (2) (3) macronutrient–plasma lipid consistent across