Development, validity, and reliability of a food frequency questionnaire for antioxidants in elderly Iranian people

Abstract

Background:The food frequency questionnaire (FFQ), as a way to assess the dietary intake in comparison with other methods, is easier to analyze and takes less time and is less costly. Our aim in this study was to develop and validate an FFQ for estimating the intakes of selected antioxidants in elderly Iranian people.Materials and Methods:A total of 185 elderly people were randomly selected. Three-day food records were completed by the subjects and collected every 2 months and dietary intake levels of zinc, selenium, carotenes and vitamins C and E were estimated. Based on the food records data, geographic location, and age, an FFQ was designed to estimate antioxidant intakes during 1 year. In addition, for controlling energy intake, 2-day food records were also collected with the food frequency questionnaire. To assess the reliability, 40 individuals were asked to complete the FFQ twice with an interval of 3 months in between.Results:Intraclass correlation coefficient (ICC) between the two FFQs for antioxidant C, antioxidant E, carotene, selenium, and zinc were 0.62, 0.47, 0.51, 0.54, and 0.58, respectively (P < 0.05). In addition, Pearson correlation coefficients between the FFQ and the food records, after controlling energy for vitamin C, vitamin E, carotene, selenium, and zinc, were 0.46, 0.48, 0.38, 0.55, and 0.47 respectively (P < 0.001).Conclusion:According to the results of this study, considering the fact that the FFQ was designed for the elderly and the special conditions (patience, memory, etc.) and vulnerability of this age group, the questionnaire is relatively valid and reliable to use.

Food Frequency Questionnaires Developed and Validated in Iran: A Systematic Review

“…Ten of the FFQs were general questionnaires that included food items from all food groups [ 11 , 13 - 21 ], of which 7 had a long format (89-189 food items in the list) and 3 were short (48-80 food items). They were all validated for adults, except for 2 that were specially designed and validated for elderly individuals [ 17 , 18 ].…”

“…Only one FFQ [ 27 ] used an experience-based method by selecting food items based on the opinions of an expert panel regarding the most common food items or foods usually consumed in the study population. Three were developed through a combination of databased and experience-based approaches [ 13 , 18 , 24 ]. In addition, 2 questionnaires were modified versions of the Tehran Lipid and Glucose Study (TLGS) FFQ, in which new food items were added using an experience-based approach [ 17 , 19 ], and 3 questionnaires were translated and modified versions of original questionnaires from other countries [ 20 , 25 , 28 ].…”

“…In most studies, the validity of the developed FFQ was evaluated by assessing the correlation of dietary intake estimates against 24-hour dietary recalls (24hDRs) [ 9 , 11 , 12 , 18 , 19 , 22 , 23 ], DRds [ 14 , 15 , 17 ], or nutrient biomarkers [ 9 , 13 , 20 , 22 , 24 , 27 ] as reference methods ( Table 2 ). The correlation coefficients between measured nutrients from FFQs and reference measures were between 0.07 to 0.82 in the general FFQs and 0.20 to 0.67 in the specialized FFQs.…”

“…All proposed questionnaires appeared valid, except for one [ 14 ]. Six studies reported high correlations (r≥ 0.60) between FFQ estimates and reference measure(s) [ 13 - 15 , 20 , 23 , 25 ], 11 presented moderate correlations (r = 0.40-0.59) [ 11 , 13 - 15 , 18 - 21 , 23 - 25 ], 9 had fair correlations (r=0.30-0.39) [ 14 , 15 , 18 - 21 , 23 - 25 ], and 6 had poor correlations (r< 0.3) [ 9 , 16 , 19 - 21 , 23 ] for some food groups/nutrients derived from FFQs.…”

“…FFQs with a long format had slightly better estimations of nutrient intake. Almost all FFQs used for dietary assessment in the Iranian population were developed and validated for adults, except for one developed to assess calcium intake in children [ 24 ] and 2 developed to assess dietary intake in the elderly [ 18 , 19 ].…”