Abstract Objectives To evaluate the validity of a single-item diet quality measured as a rapid and cost-effective means of dietary assessment among adults living in Puerto Rico (PR), a population with access to and a cultural preference for specific foods. Methods This cross-sectional study used data from the PR Assessment of Diet, Lifestyle, and Diseases (2015). Self-perception of diet quality was reported (n = 247, age = 35–75 y) for the current day, young adulthood, and childhood (five-point scale: excellent to poor). We used a culturally adapted food frequency questionnaire to measure dietary intake, and computed diet quality (range: 0–110) with the Alternate Healthy Eating Index-2010 (AHEI). We conducted multivariable generalized linear models to assess relationships between current AHEI and perception of diet quality at three-time points. Results We categorized current self-rated diet quality into ‘excellent or very good’ (37%), ‘good’ (35%), or ‘fair or poor’ (28%); the overall mean AHEI was 60.1 +/– 11.0. Sensitivity of single-item measure was 60% and specificity was 34%. Current self-rated diet quality was not significantly associated with AHEI score, although adjusted mean scores for fruit and or long-chain fatty acids were significantly lower among participants reporting fair or poor diet, vs. excellent or very good (β = −0.49, 95% CI = −0.97, −0.014; β = −0.52, 95% CI = −0.98, −0.049). In contrast, mean adjusted AHEI score was significantly higher among those rating their diet quality as fair or poor in young adulthood (36%; β = 6.4, 95% CI = 2.4,10.5), and marginally higher among those reporting fair or poor diet quality in childhood (27%; β = 3.99; 95% CI = −0.01,8.04), vs. those rating their diet as excellent or very good, at respective time points. Conclusions A single-item perceived diet measure is not a valid assessment of current diet quality for adults in PR. Interestingly, perception of low diet quality in earlier life stages is associated with higher quality of the actual current diet. Nutrition researchers may use self-rated diet quality in earlier life stages to estimate current diet quality; the mechanisms driving this relationship warrant further investigation. Funding Sources CG is supported by an NIH/NCI training grant. PRADLAD was funded with anonymous donations and an NIH/NHLBI award.
Read full abstract