Search in the Abstract Database

Abstracts Search 2019

P768 Development and validation of Processed Foods Questionnaire (PFQ) in Israeli adult inflammatory bowel diseases patients

C. Sarbagili-Shabat*1,2, S. Zelber-Sagi1,3, N. Fliss Isakov1, Y. Ron1, A. Hirsh1, N. Maharshak1,2

1Tel-Aviv Medical Center, IBD Center, Department of Gastroenterology and Liver Diseases, Tel Aviv, Israel, 2Tel-Aviv University, The Sackler Faculty of Medicine, Tel-Aviv, Israel, 3University of Haifa, School of Public Health, Haifa, Israel


Western diet and its food components have been implicated in the pathogenesis of inflammatory bowel diseases (IBD). However, in the absence of a standard quality research that addresses the impact of processed foods on IBD, a validated processed foods questionnaire for IBD patients is essential. The goal of the current study was to develop a Processed Foods Questionnaire (PFQ) and to assess its reliability and validity.


A single-centre validation prospective study. Adult IBD patients, aged 18–65 years, were recruited. Reliability was assessed by comparing the PFQ of one patient at two-time points separated by at least 2 weeks. Validity was assessed by comparing the PFQ to a three to 7 days food diary. Validity was further tested by correlation of PFQ food consumption to urine sodium, which served as a biomarker for processed food consumption. Urine was analysed for sodium level from either the first morning urine or from a 24 h urine collection. Food intake was categorised to: unprocessed, processed and ultra-processed.


Eighty-six IBD patients [57 (66.3%) Crohn’s disease, 29 (34.7%) ulcerative colitis] at a mean age of 33.0 ± 10.7 years, were enrolled. Good test–retest reliability was indicated by intra-class correlation (ICC) of 0.75–0.88 for the different food processing levels. For validity, there were fair to good correlations for different levels of processed food intake between food diaries and the PFQ, ranging between 0.43 to 0.64 (Pearson r, p < 0.001). In addition, Kappa measure of agreement was fair- 0.28-0.4. Mean urine sodium levels were higher in patients with high processed-foods consumption compared with low consumption (104.57 ± 53.26 vs. 78.62 ± 39.08 mmol/l, p = 0.011). Furthermore, consumption of unprocessed foods negatively correlated with urine sodium levels (Pearson r = −0.20, p = 0.034).


The PFQ is a reliable and valid tool for the assessment of processed foods consumption in IBD patients and can be utilised for studying the association between processed food consumption and IBD etiopathogenesis.