Reproducibility and relative validity of a quantitative food frequency questionnaire developed for adults in the coastal area of southeast China.

PURPOSE

To evaluate the reproducibility and validity of a 93-item food frequency questionnaire (FFQ) designed to assess usual dietary habits over the past year among adults in the southeast coastal region of China.

METHODS

Qualified dietary data of 477 subjects (75.26% female, mean age = 59.7 years) were collected via two FFQs over approximately 15 months, alongside four 4-day 24-hour dietary recalls (24 h-DRs) at average intervals of four months. The analysis included 25 general nutrients, 14 particular nutrients and 12 food groups. Wilcoxon signed-rank test, correlation coefficients, cross-classification with weighted kappa statistic and Bland-Altman plots were used for comparison.

RESULTS

For reproducibility, the Spearman, Pearson, and intraclass correlation coefficients between two FFQs ranged from 0.24 to 0.53, 0.21 to 0.53 and 0.19 to 0.53, respectively. After adjusting for energy intake, most coefficients either decreased or remained unchanged. The rates of agreement in classifying subjects in the same or adjacent quartile fell between 62% and 84% for both FFQs. Regarding validity, the crude Spearman, energy-adjusted and de-attenuated Pearson correlation coefficients between FFQs and 24 h-DRs ranged from 0.17 to 0.59, 0.12 to 0.54, and 0.13 to 0.57, respectively. Over 58% of subjects were classified into the same and adjacent categories by both methods. Weighted kappa statistic and Bland-Altman plots demonstrated acceptable concordance.

CONCLUSION

The FFQ developed for adults in the specific coastal region of China, aiming to be used in a population-based cohort study, demonstrates acceptable reproducibility and validity to measure the dietary intakes of certain nutrients and foods.