Using a Patient-Completed Food Frequency Questionnaire to Determine Mediterranean Diet Score in People with Inflammatory Bowel Disease

Purpose: To examine the level of agreement between a patient-completed food frequency questionnaire (FFQ) and assessment of usual intake by a registered dietitian (RD) to score adherence to a Mediterranean diet (MedD) in patients with inflammatory bowel disease (IBD). Methods: Patients with IBD completed a short FFQ and were subsequently interviewed by an RD. A 12-item MedD score (MDS), adapted from the Mediterranean Diet Adherence Screener (MEDAS), was calculated from the FFQ and RD assessments. To determine agreement between individual items, Cohen’s kappa coefficients were calculated. Absolute agreement between assessment methods was quantified using a one-way random intra-class correlation coefficient for a single measure. Results: Forty-six patients with IBD participated. The mean FFQ-MDS was 4.59 (standard deviation [SD] = 1.65), and mean RD-MDS was 4.83 (SD = 1.53). Kappa coefficients for individual MEDAS items ranged from 0.41 to 0.78 (p < 0.01) between the FFQ- and RD-MDS. Most items demonstrated moderate to substantial agreement. The intra-class correlation coefficient for absolute agreement between the summed FFQ-MDS and RD-MDS was 0.71 (95% confidence interval: 0.52–0.83, p < 0.001), indicating moderate reliability. Conclusions: This patient-completed FFQ may be a promising tool in clinical practice and research and would benefit from additional evaluation to validate its use in patients with IBD.