Development,validity and reproducibility of a dish-based semi-quantitative food frequency questionnaire in Iran

Abstract

BACKGROUND:

The food frequency questionnaire (FFQ) is the method of choice for dietary assessment in epidemiological studies. FFQs focusing on mixed-dishes and simple food items are useful where mixed-dishes are an essential part of food consumption.

OBJECTIVE:

Regarding the fact that the nature of the Iranian diet is mixed-dish, the present study aimed to design and assess the validity and reproducibility of a dish-based semi-quantitative FFQ in the Iranian adult population.

METHODS:

A list of 302 food items was collected from four geographical areas around Iran. The validation study was conducted on 97 healthy adults. The FFQ was introduced at the beginning of the study and 10 months after; two three-day food records were collected during the study. Also, biomarkers including 24-hour urinary potassium and nitrogen, serum retinol, and alpha-tocopherol were measured.

RESULTS:

A 142-food-item FFQ was concluded. The correlation coefficient between the second FFQ and the second three-day food record ranged from 0.225 to 0.323 for macronutrients and 0.128 to 0.476 for micronutrients. The percentile agreements (same or adjacent quartile) between the two methods were more than 60% for all nutrients. The intraclass correlation coefficient between FFQs (except for vitamin E) ranged from 0.363 to 0.578. The correlation coefficient between the second FFQ and the second biomarker assessment was 0.241 for protein.

Keywords

Validity reproducibility food frequency questionnaire FFQ dish-based

1 Introduction

Long-term dietary exposure is conceptual exposure in chronic disease, the food frequency questionnaires (FFQ) is the method of choice to measure this exposure [1].

FFQ is the suitable method for long-term nutritional assessment as inexpensive tool without the need for repeated measurements, which is a common problem in food record and 24-hour dietary recalls [2, 3]. Based on food intakes, the FFQ is used to rank individuals rather than evaluating their absolute intakes [1, 4]. As culture may affect the performance of an FFQ, it is necessary to assess the FFQ validity and reproducibility in the target population [3 , 6].

Validity refers to the accurate measurement of dietary intakes by the FFQ depending on the aim of designing the tool. In FFQ validation studies, FFQ data are compared with dietary reference methods, such as food records, 24-hour recalls, and food diaries [7, 8]. Among various dietary reference methods, food record has the least correlated errors with the FFQ and is considered the ‘gold standard’ method for validation studies [7 , 10]. Biomarkers could be considered as the reference method in FFQ validation studies, because their errors are not correlated with the errors of FFQs [8, 11].

Reproducibility refers to the stability of FFQ data between the two occasions where the FFQ is administered, assuming no considerable dietary change [7].

Mixed dishes have multiple ingredients that are mixed according to the food recipe, and the consumer may not remember each ingredient for assessment purposes [12]. On the other hand, the FFQs that are focused on ingredients may underestimate some micronutrients from the invisible parts of mixed dishes [13]. Several FFQs have been developed and validated for the Iranian population [14 –26] with an emphasis on food items and not mixed dishes, which are highly common in the Iranian diet [12].

The present study aimed to design a new culturally adapted, semi quantitative dish-based FFQ and assess the validity and reproducibility of this FFQ in the Iranian adult population.

2 Method

2.1 Study design

This is a longitudinal and diagnostic test study. The study protocol was approved by the Ethics Committee of Mashhad University of Medical Sciences (No. IR.MUMS.fm.REC.1394.628). The first step was development of a new dish-based semi-quantitative FFQ in 2016. Data collection for assess validity and reproducibility was started in 2018 and continued for 10 months. The FFQ was filled in the first session and in the 10th month of the study. During the study, the participants completed three-day food records in the second and 10th month of the study. In the 4th and 10th month, biochemical samples were obtained from the participants. The study protocol was described comprehensively elsewhere [27]. Table 1 shows the validation study design.

Table 1
Study Design and Timelines of Food Frequency Questionnaire (FFQ) Administration, 3-Day Food Record, and Biochemical Sampling

Month 1 2 3 4 5 6 7 8 9 10

FFQ ^* ^*

3-Day Record ^* ^*

Biochemical Sampling ^* ^*

2.2 Questionnaire design

To develop a new culturally adapted FFQ, the usual food intake was assessed in four geographical dispersed cities of Iran. A comprehensive list of the frequently consumed foods was prepared based on face-to-face interviews with 10 residents of each city, and some items were added to the list based on the comments of experienced local nutritionists. The initial questionnaire including 302 food items (simple items and mixed dishes) was completed by 1,011 adults, including students, employees, and self-employed individuals. The secondary questionnaire items were selected based on the frequency of initial items consumption; the items used by more than 5% of the participants were kept [28], and those used by lower than 1.5% were eliminated. The secondary questionnaire consisted of 160 food items and was provided to three experienced local nutritionists. After assessing the face and content validity, a questionnaire with 150 items was obtained, which was presented to an expert panel of 10 nutritionists. The content validity ratio and content validity index of the items were determined [28], and the final dish-based semi-quantitative FFQ was obtained with 142 food items, including 89 simple items and 53 mixed dishes.

The intake frequencies of food items consist of nine categories ranging from never to more than six times per day. An average portion based on household plates and utensils was defined for every item, and data on the portion sizes were recorded based on three portion size options (less, average, more). The food items were organized in 17 food groups, and a short foods portion size album (supplementary file 1) was added to the beginning of the questionnaire to help recall the portion sizes. The album showed the images of household utensils and some of the food items.

2.3 FFQ Validation study

Prior to completing the questionnaire, the participants were trained on doing so. They completed the questionnaires themselves, and the nutritionist answered their questions during FFQ completion. The data from the FFQ were compared with food records and biochemical measurements to assess validity of the FFQ.

2.3.1 Food records

The participants received verbal instructions on recording their food intake, and a booklet was also provided to each participant, which contained the instructions on food recording, household plates and utensils illustrations, and a short food album. The participants recorded their food intake for three consecutive days in the second and 10th month of the study. Every three-day food record consisted of two weekdays and one weekend.

2.3.2 Biochemical measurements

In the 4th and 10th month of the study, the participants were recruited for laboratory sampling and instructed to fast 12 hours before sampling. Blood samples were obtained during 7 : 00–9 : 00 AM in a sitting position from the antecubital vein and collected in tubes without anticoagulants. The tubes were protected from sunlight by aluminum foil and preserved in a cool box for approximately two hours to allow clot formation. At the next stage, the samples were centrifuged at 3,000 rpm for 10 minutes, and the serum was extracted and transferred to three microtubes per subject. Two microtubes were stored at the temperature of –70°C until analysis, and the third microtube was used for the measurement of the total serum cholesterol and triglycerides (TAG) using an enzymatic method. The serum concentrations of retinol and alpha-tocopherol were measured by high-pressure liquid chromatography method [30, 31].

Also, in the 4th and 10th months, the participants were instructed to collect 24-hour urine samples in two-liter plastic containers. When the containers were delivered to the laboratory, the participants were asked whether urine collection was completed. If more than 50 milliliters of the samples were lost, they were asked to collect the sample again. Total urinary potassium was measured using flame photometry, and the Kjeldahl technique was used to assay the total nitrogen in urine, as well as conversion into urea nitrogen excretion (g/d). Approximately 85% of the urinary nitrogen was considered as urea nitrogen excretion, so that the total protein intake could be calculated using the following formula: 6.25 (urinary nitrogen+2) [32].

2.3.3 Reproducibility assessment

Ten months after the beginning of the study, the participants were invited to the health centers again and trained on completing the questionnaire for the second time to assess the reproducibility of the data.

2.3.4 Food item analysis

The consumption of the food items (simple and mixed) was calculated using the FFQ, and the conversion of the mixed items into their ingredients was performed based on the existing data [33]. After applying the yield factor, the energy and nutrients were calculated based on the USDA Food Composition Database [34]. The daily consumption of each FFQ item was determined by multiplying the consumption frequency by the portion size (g).

2.4 Statistical methods

Data analysis was performed in SPSS version 16.0 (SPSS Inc., Chicago, IL, USA), and the normality of data distribution was assessed using the Shapiro-Wilk test. With the non-normal distribution of data, non-parametric tests were used. In addition, mean and standard deviation (M±SD) were calculated for the energy and nutrients in both FFQs and two three-day food records. The Wilcoxon signed-rank test was also applied to find if the daily nutrient intakes had significant differences between the methods (second FFQ and second three-day food record).

To assess the validity of the questionnaire, the nutrients of the second FFQ and the second three-day food record were compared. Pearson’s or Spearman’s correlation coefficients were estimated between the energy and nutrient intakes of the second FFQ and the second three-day food record, accordingly. The nutrient intake in the second FFQ with the second biomarkers, using correlation coefficients tests were compared, as well. Partial correlation coefficients were adjusted for serum cholesterol and TAG and calculated for serum retinol and alpha-tocopherol concentrations.

The degree of agreement based on quartiles and the nutrient intake of the second FFQ and the second three-day food record were evaluated by examining the classification of the participants using the FFQ and reference method into the same, adjacent, and opposite quartiles.

The reproducibility of data was evaluated using intraclass correlation coefficients (ICC), which were measured for the data of the first and second FFQ.

3 Results

The final 142 items FFQ includes both simple food items 89 and mixed-dishes 53, items are clustered into 17 food groups. The questionnaire is attached in the supplementary file 2.

The mean age of the participants was 42.43±10.36 years (males: 43.00±10.15 years, females: 42.16±10.53 years). The mean BMI of the participants was 26.88±4.27 kg/m² (males: 25.96±3.36 kg/m², females: 27.33±4.61 kg/m²). No significant differences were observed in the age and BMI between the male and female subjects. Table 2 shows some demographic characteristics of the participants.

Table 2
Demographic characteristics of the study population (N = 97)

Demographic feature Labels Number Percent

Gender Male 31 32.0

Female 66 68.0

Educational level Diploma and under Diploma 60 61.85

Under graduate 29 29.89

Post graduate 8 8.24

Job Housewife 47 48.5

Self-employed 26 26.8

Employee 21 21.6

Student 3 3.1

Exercise No exercise 58 61.7

Mild exercise 1 to 3 times a week 23 24.5

Mild exercise 3 to 5 times a week 10 10.6

Mild exercise 6 to 7 times a week 3 3.2

Smoking Status Non-smoker 97 100

Smoker 0 0

Body mass index Less than 18.5 (underweight) 3 3.09

18.5 to 25.0 (normal) 31 31.95

25.0 to 30.0 (over weight) 42 43.29

More than 30.0 (obese) 21 21.64

Demographic feature	Labels	Number	Percent
Gender	Male	31	32.0
	Female	66	68.0
Educational level	Diploma and under Diploma	60	61.85
	Under graduate	29	29.89
	Post graduate	8	8.24
Job	Housewife	47	48.5
	Self-employed	26	26.8
	Employee	21	21.6
	Student	3	3.1
Exercise	No exercise	58	61.7
	Mild exercise 1 to 3 times a week	23	24.5
Mild exercise 3 to 5 times a week	10	10.6
	Mild exercise 6 to 7 times a week	3	3.2
Smoking Status	Non-smoker	97	100
	Smoker	0	0
Body mass index	Less than 18.5 (underweight)	3	3.09
	18.5 to 25.0 (normal)	31	31.95
	25.0 to 30.0 (over weight)	42	43.29
	More than 30.0 (obese)	21	21.64

Table 3 shows the mean intake of nutrients based on the second three-day food record and the FFQs. The second FFQ overestimated nutrient intake compared to the second three-day food records (M±SD). The mean nutrient intake in the second FFQ was significantly higher compared to the second three-day food record in terms of potassium, beta-carotene, folate, trans fatty acids, vitamin A, and vitamin C (P < 0.05).

Table 3

Comparison of Second FFQ and Second 3-Day Food Record/Biomarker Assessment (Percentage of agreement/disagreement in quartiles of second FFQ and second 3-day food record; Comparison of data of first and second FFQs; N = 97)

Energy and Nutrients (unit)	Average Daily Intake						Comparison of 2nd REC and FFQ2	Percentage agreement			Correlation coefficients (CCs) between FFQ2 and 2nd Record/ 2nd Biomarkers	Intraclass Correlation coefficient (ICC) between FFQ1 and FFQ2
	2nd 3-days Record		1st FFQ		2nd FFQ			Between 2nd Record and 2nd FFQ			CC^b	ICC
	Mean	SD	Mean	SD	Mean	SD	p-value^a	Same Quartile (%)	Same or adjacent quartile	Opposite quartile (%)
Energy (Kcal)	1644.03	496.02	1951.56	905.19	1912.71	1319.83	0.659	39.66	80.00	20.00	0.389^*	0.482^*
Protein (g)	63.44	21.41	72.89	40.98	65.59	37.06	0.457	36.66	75.00	25.00	0.291^*	0.570^*
Fat (g)	63.83	23.34	78.73	43.79	76.12	57.00	0.440	26.66	68.32	31.66	0.225	0.578^*
Carbohydrate (g)	220.36	83.20	267.75	136.81	266.28	199.37	0.385	36.66	75.00	25.00	0.323^*	0.480^*
Fiber (g)	20.04	8.95	27.55	17.04	25.90	19.56	0.216	40.00	28.33	31.66	0.251	0.492^*
Calcium (mg)	690.52	333.90	854.90	517.22	770.81	472.55	0.724	31.66	70.00	30.00	0.395	0.546^*
Iron (mg)	11.81	5.19	14.50	7.94	13.31	8.17	0.883	33.33	78.33	21.66	0.284^*	0.486^*
Magnesium (mg)	262.52	157.06	309.38	172.37	278.36	173.03	0.866	43.33	78.33	21.66	0.476^*	0.486^*
Phosphorus (mg)	970.40	393.58	1159.82	630.23	1050.64	590.71	0.702	41.44	81.44	15.00	0.454^*	0.532
Potassium (mg)	2183/79	847.67	3272.83	1949.69	2918.64	1968.40	0.020^*	38.33	76.66	23.33	0.412^*	0.495^*
Zinc (mg)	10.09	10.76	16.90	18.87	14.56	12.65	0.211	38.33	76.66	23.33	0.377	0.456^*
Retinol (μg)	269.45	582.09	343.22	299.80	294.77	275.85	0.233	33.33	68.33	31.66	0.189	0.367^*
Vitamin A (μg)	436.97	160.30	800.89	563.43	674.58	492.43	0.009^*	38.33	73.33	26.66	0.329^*	0.472^*
Betacarotene (μg)	2636.66	1692.51	4875.70	4225.13	4003.60	2972.70	0.006^*	41.66	73.32	26.66	0.348^*	0.477^*
Vitamin E (mg)	13.86	39.78	10.30	5.56	13.71	39.46	0.924	31.66	68.32	31.66	0.237	0.011
Vitamin D (μg)	2.41	12.18	1.07	0.74	1.00	0.77	0.088	30.00	65.00	35.00	0.128	0.450^*
Vitamin C (mg)	98.03	78.56	156.11	111.06	136.42	110.75	0.047^*	31.66	71.66	28.33	0.207	0.416^*
Thiamin (mg)	2.24	7.94	1.27	0.71	1.13	0.69	0.195	40.00	70.00	30.00	0.210	0.567^*
Riboflavin (mg)	1.54	0.74	2.01	1.27	1.77	1.19	0.930	31.66	61.66	38.33	0.317^*	0.649^*
Folate (μg)	236.18	93.23	354.05	197.08	332.12	227.52	0.001^*	26.66	63.32	36.66	0.178	0.514^*
Cholesterol (mg)	215.09	117.98	265.59	162.79	241.40	158.14	0.389	33.33	63.33	36.66	0.167	0.509^*
Trans fatty acids (g)	0.49	0.25	0.92	2.59	0.68	0.60	0.043^*	28.33	70.00	30.00	0.238	0.400^*
Saturated fats (g)	16.57	5.71	10.40	20.09	19.58	13.34	0.791	33.33	68.33	31.66	0.300^*	0.510^*
Monounsaturated fats (g)	23.41	6.82	29.65	14.13	28.80	17.47	0.080	38.33	70.00	30.00	0.433^*	0.380^*
Polyunsaturated fats (g)	12.68	5.24	14.66	6.98	15.32	17.06	0.330	30.00	66.66	33.33	0.179	0.363^*
Biomarkers
Retinol^c	–	–	–	–	–	–	–	–	–	–	–0.068	–
Alpha-tocopherol^c	–	–	–	–	–	–	–	–	–	–	0.059	–
Cholesterol	–	–	–	–	–	–	–	–	–	–	0.057	–
24-hour Urinary Protein	–	–	–	–	–	–	–	–	–	–	0.214	–
24-hour Urinary Potassium	–	–	–	–	–	–	–	–	–	–	–0.105	–

FFQ: food frequency questionnaire; a: p-value of Paired-T test Or Wilcoxone signed rank test between 2nd FFQ and 2nd 3-day food record; b: Pearson’s and Spearman’s CC between the 2nd FFQ and 2nd 3-day food record/2nd biomarkers; c: partial correlation adjusted for serum concentration of cholesterol and TAG; ^*: p-value < 0.05.

The correlation coefficients between the nutrient intake of the second three-day food record and the second FFQ were assessed in case of the normal data using Pearson’s correlation coefficient and Spearman’s correlation coefficient for the non-normal data. The correlation coefficients between energy and macronutrients intake of the second three-day food record and the second FFQ ranged from 0.225 (fat) to 0.389 (energy). The correlation coefficients for micronutrients ranged from 0.128 for vitamin D to 0.476 for magnesium. Table 3 shows correlation coefficients between the second FFQ and the second three-day food record.

The ICC of the FFQs is also shown in Table 3. The ICC between two FFQs ranged from 0.011 to 0.578 for vitamin E and fat. The ICC of protein, fat, calcium, phosphorus, thiamin, riboflavin, folate, cholesterol, and saturated fatty acids was higher than 0.5. Vitamin E was the only micronutrient with non-significant ICC.

The percentile of agreement and disagreement between the second three-day record and the second FFQ quartiles was calculated, and the percentage of the two methods was classified into the same or adjacent percentile ranging from 61.66% for riboflavin to 81.44% for phosphorus with an average of 69.64%. In addition, the complete disagreement of the quartiles ranged from 15.0% for phosphorus to 36.66% for riboflavin and folate.

The correlation coefficient between the second FFQ and the second biomarker assessment of protein was 0.214. The other correlation coefficients of the biomarkers were negligible.

4 Discussion

The present study aimed to design and evaluate the validity and reproducibility of a dish-based semi-quantitative FFQ, which consisted of mixed dishes and simple food items. Iranian mixed dishes are composed of various ingredients, which may not be properly recalled by the consumers. The inclusion of mixed dishes made the questionnaire easier to complete because the contents were similar to the daily intakes of the participants, while the length of the questionnaire also decreased [12].

According to dietary intakes, the ranking of subjects is the main purpose of using FFQs rather than the evaluation of absolute intakes [7 , 36]. Our findings indicated that the FFQ ranked more than 60% of the participants into the correct quartile, which is acceptable and comparable to the previous studies in this regard [1, 37].

In the present study, the correlation coefficients between the second three-day food record and the second FFQ were within the range of 0.207–0.476, which indicated acceptable correlations for most macro and micronutrients. Exceptions were the correlation coefficients of retinol, vitamin D, folate, cholesterol, and polyunsaturated fatty acids, the values of which were lower than 0.2 (poor correlation coefficients). This is consistent with the study by Dehghan [1], which indicated the correlation coefficients to be 0.2–0.47. In addition, Doustmohammadian reported the correlation coefficients of 0.23–0.52, which is in line with the results of the present study [37]. The lower correlation coefficients in the current validation study compared to the studies conducted in western countries [38] may be due to the complexity of the Iranian diet, which consists of traditional, modern, and local foods. This complexity may lead to missing some foods in questionnaire development or the interpretation of food items.

In the current research, overestimation of energy and nutrient consumption was observed in the results of the FFQ compared to the data of the three-day food record, which is consistent with the previous studies in this regard [1, 37].

According to the results of the present study, the ICC of reproducibility (except for vitamin E) was 0.363–0.578, which was considered acceptable. The ICC in a previous study in Iran reported as 0.23–0.76 [37], also in a study in Argentina the ICC of reproducibility study were 0.4–0.64 in the urban and 0.27–0.52 in the rural area [1]. The ICC of our study is in line with the findings of mentioned studies. The ICC in western countries is higher because there are no short-term dietary changes; but nutrition transition in low- and middle-income countries cause lower reproducibility of FFQs [7].

Low correlation coefficients between biomarkers and the second FFQ in our study may be due to the lack of a national food composition table. This makes a systematic error in the calculation of dietary nutrients intake [3].

Furthermore, Cade assessed the method of FFQ administration, reporting that 67% of the questionnaires were self-administered [10]. The mentioned study also indicated that interview-administered questionnaires have better correlation coefficients in validation studies. On the other hand, a systematic review of the FFQ validation studies in Iran showed that only one out of 18 questionnaires was self-administered [39]. In the current research, we used a self-administered FFQ, it seems that an interview is the selected method for the administration of the questionnaire in the illiterate and unmotivated subjects so, administration method could be lead to decrease correlations in the current study.

In the present study, a food record was used as the dietary reference, which corresponds to the emphasis Cade has placed on the fact that food records are the ‘first choice’ of dietary reference methods in FFQ validation studies [10]. In the systematic review of Iranian FFQ validation studies, only three out of 18 food records were used as the dietary reference method [39]. Using food records in participants with limited cooperation and literacy may yield fair correlations [10]. The low accuracy of data collected by food records might reduce the correlation coefficients between the second FFQ and the second three-day food record in our study.

In the general Iranian population, the interview administration of the FFQ and comparison with 24-hour recalls could be the methods of choice for FFQ validation studies despite the associated challenges.

The dish-based semi-quantitative FFQ in Mashhad may be regarded as a new approach to Iranian nutritional epidemiology, which is focused on mixed dishes given the strong correlation between cultural cooking methods and non-communicable diseases. It is hoped that the data obtained by the new dish-based semi-quantitative FFQ become more accurate than formerly used item-based Iranian FFQs in prospective epidemiological studies.

5 Limitation

This study has some limitations. First, although food record is the gold standard method for FFQ validation studies, it seems that the collaboration of the participants is not acceptable in our society despite the training, and strategies should be given to improve their cooperation. Second, lack of a national food composition table may be the reason of low correlation between biochemical markers and dietary intake. Third, the participants of our study took part voluntarily, they may respond more accurately than non-volunteers, and their dietary habits may be different.

6 Conclusion

According to the results, the 142 item semi-quantitative dish-based FFQ had statistically acceptable validity and reproducibility to rank the participants based on their energy and most macro (including: protein, carbohydrate) and micronutrient (including: iron, magnesium, potassium, phosphorus, vitamin A, beta-carotene, riboflavin, saturated, and monounsaturated fatty acids) intakes. So, the semi-quantitative dish-based FFQ could be used in epidemiologic studies.

Footnotes

Acknowledgments

The authors would like to thank the Health Deputy of Mashhad University of Medical Sciences. Also, we thank Mashhad health center staff for their friendly cooperation.

Transparency declaration

The lead author affirms that this manuscript is an honest, accurate, and transparent account of the study being reported. The lead author affirms that no important aspects of the study have been omitted and that any discrepancies from the study as planned have been explained.

Funding

This study received a research grant from Mashhad University of Medical Sciences.

Conflict of interest

None.

Ethical code

Granted by the ethics Committee of Mashhad University of Medical Sciences (No. IR.MUMS.fm.REC.1394.628).

The supplementary material is available in the electronic version of this article: .

References

Dehghan

, del Cerro

, Zhang

, Cuneo

, Linetzky

, Diaz

, et al. Validation of a semi-quantitative Food Frequency Questionnaire for Argentinean adults. PLoS One. 2012;7(5):e37958.

Moghames

, Hammami

, Hwalla

, Yazbeck

, Shoaib

, Nasreddine

, et al. Validity and reliability of a food frequency questionnaire to estimate dietary intake among Lebanese children. Nutr J. 2016;15, 4.

Mirmiran

, Esfahani

, Mehrabi

, Hedayati

, Azizi

. Reliability and relative validity of an FFQ for nutrients in the Tehran lipid and glucose study. Public Health Nutrition. 2010;13(5):654–62.

Dehghan

, Martinez

, Zhang

, Seron

, Lanas

, Islam

, et al. Relative validity of an FFQ to estimate daily food and nutrient intakes for Chilean adults. Public Health Nutrition. 2013;16(10):1782–8.

Pakseresht

, Sharma

. Validation of a culturally appropriate quantitative food frequency questionnaire for Inuvialuit population in the Northwest Territories, Canada. Journal of Human Nutrition and Dietetics. 2010;23, 75–82.

Corsi

, Subramanian

, Chow

, McKee

, Chifamba

, Dagenais

, et al. Prospective Urban Rural Epidemiology (PURE) study: baseline characteristics of the household sample and comparative analyses with national data in 17 countries. American Heart Journal. 2013;166(4):636–46.e4.

Willett

. E. Reproducibility and Validity of Food Frequency Questionnaire. In: Willet W, editor. Nutritional Epidemiology. 40. 3 ed. New York, Oxford: Oxford University Press; 2013. p. 96–145.

Fayet

, Flood

, Petocz

, Samman

. Relative and biomarker-based validity of a food frequency questionnaire that measures the intakes of vitamin B(12), folate, iron, and zinc in young women. Nutrition research (New York, NY). 2011;31(1):14–20.

Lin

, Bromage

, Mostofa

, Allen

, Oken

, Kile

, et al. Validation of a Dish-Based Semiquantitative Food Questionnaire in Rural Bangladesh. Nutrients. 2017;9(1).

10.

Cade

, Thompson

, Burley

, Warm

. Development, validation and utilisation of food-frequency questionnaires - a review. Public Health Nutrition. 2002;5(4):567–87.

11.

Vioque

, Navarrete-Muñoz

E-M

, Gimenez-Monzó

, García-de-la-Hera

, Granado

, Young

, et al. Reproducibility and validity of a food frequency questionnaire among pregnant women in a Mediterranean area. Nutrition Journal. 2013;12(1):1–9.

12.

Keshteli

, Esmaillzadeh

, Rajaie

, Askari

, Feinle-Bisset

, Adibi

. A Dish-based Semi-quantitative Food Frequency Questionnaire for Assessment of Dietary Intakes in Epidemiologic Studies in Iran: Design and Development. International Journal of Preventive Medicine. 2014;5(1):29–36.

13.

Kim

, Kim

, Lee

S-A

, Yoon

, Sasaki

. A study testing the usefulness of a dish-based food-frequency questionnaire developed for epidemiological studies in Korea. British Journal of Nutrition. 2008;101(8):1218–27.

14.

Bijani

, Esmaili

, Ghadimi

, Babazadeh

, Rezaei

, R

, et al. Development and validation of a Semi-quantitative food frequency questionnaire among older people in north of Iran. Caspian Journal of Internal Medicine. 2018;9(1):78–86.

15.

Esfahani

, Asghari

, Mirmiran

, Azizi

. Reproducibility and relative validity of food group intake in a food frequency questionnaire developed for the Tehran Lipid and Glucose Study. Journal of Epidemiology. 2010;20(2):150–8.

16.

Ahmadnezhad

, Asadi

, Miri

, Ferns

, Ghayour-Mobarhan

, Ebrahimi-Mamaghani

. Validation of a Short Semi-Quantitative Food Frequency Questionnaire for Adults: a Pilot study. Journal of Nutritional Sciences and Dietetics. 2017;3(2).

17.

Hadi

, Somi

, Ghavamzadeh

. Validity and reliability of a 40-items Food Frequency Questionnaire to Measure the Consumption of Gluten in Patients with Ulcerative Colitis. Iranian Journal of Nutrition Sciences & Food Technology. 2017;12(1):1–9.

18.

Malekahmadi

, Naeini

, Shab-Bidar

, Feizi

, Djazayery

. Development, validity, and reliability of a food frequency questionnaire for antioxidants in elderly Iranian people. Journal of Research in Medical Sciences: The Official Journal of Isfahan University of Medical Sciences. 2016;21, 14.

19.

Malekshah

, Kimiagar

, Saadatian-Elahi

, Pourshams

, Nouraie

, Goglani

, et al. Validity and reliability of a new food frequency questionnaire compared to 24 h recalls and biochemical measurements: pilot phase of Golestan cohort study of esophageal cancer. European Journal of Clinical Nutrition. 2006;60(8):971–7.

20.

Mohammadifard

, Omidvar

, Houshiarrad

, Neyestani

, Naderi

, Soleymani

. Validity and reproducibility of a food frequency questionnaire for assessment of fruit and vegetable intake in Iranian adults(^*). Journal of Research in Medical Sciences: The Official Journal of Isfahan University of Medical Sciences. 2011;16(10):1286–97.

21.

Mohammadifard

, Sajjadi

, Maghroun

, Alikhasi

, Nilforoushzadeh

, Sarrafzadegan

. Validation of a simplified food frequency questionnaire for the assessment of dietary habits in Iranian adults: Isfahan Healthy Heart Program, Iran. ARYA Atherosclerosis. 2015;11(2):139–46.

22.

Nikniaz

, Tabrizi

, Sadeghi-Bazargani

, Farahbakhsh

, Tahmasebi

, Noroozi

. Reliability and relative validity of short-food frequency questionnaire. British Food Journal. 2017;119(6):1337–48.

23.

Omidvar

, Ghazi-Tabatabie

, Harrison

, Eghtesadi

, Mahboob

, Pourbakht

. Development and validation of a short food-frequency questionnaire for screening women of childbearing age for vitamin A status in northwestern Iran. Food and Nutrition Bulletin. 2002;23(1):73–82.

24.

Pirouzpanah

, Taleban

F-A

, Sabour

, Mehdipour

, Atri

, Farrin

, et al. Validation of food frequency questionnaire to assess folate intake status in breast cancer patients. Razi Journal of Medical Sciences. 2012;18(92):31–41.

25.

Sharifi

, Javadi

, Barikani

. Reliability and validity of short food frequency questionnaire among pregnant females. Biotechnology and Health Sciences. 2016;3(2).

26.

Zeyninejad

, Omidvar

, Neyestani

, Houshiarrad

, Eshraghian

, Stormer

. Development and Validation of a Food Frequency Questionnaire for Assessing Dietary Calcium in Children. Nutrition and Food Sciences Research. 2015;2(1):35–45.

27.

Ayoubi SS, Yaghoubi Z, Amini M, Philippou E, Nematy M, Esmaily H, et al. Development and validation of a dish-based Iranian food frequency questionnaire: a protocol study.

28.

El Sayed Ahmad

, Baroudi

, Shatila

, Nasreddine

, Chokor

FAZ

, Chehab

, et al. Validity and Reproducibility of a Culture-Specific Food Frequency Questionnaire in Lebanon. Nutrients. 2020;12(11):3316.

29.

Lawshe

. A quantitative approach to content validity. Personnel Psychology. 1975;28(4):563–75.

30.

Bystrowska

, Gomolka

, Szczudrawa

, Brandys

, Pawlik

, Milewicz

, et al. Rapid HPLC method for the determination of vitamin A and E and cotinine concentration in human serum in women with CIN and cervical cancer. Ginekol Pol. 2009;80(4):256–62.

31.

Dulińska-Litewka

, Milewicz

, Pawlik

, Brandys

, Szczudrawa

, Gomółka

, et al. Rapid HPLC method for the determination of vitamin A and E and cotinine concentration in human serum in women with CIN and cervical cancer. Ginekologia Polska. 2009;80(4).

32.

Isaksson

. Urinary nitrogen output as a validity test in dietary surveys. Am J Clin Nutr. 1980;33(1):4–5.

33.

Amini

, Esmaillzadeh

, Omidvar

, Abtahi

, Dadkhah-Piraghaj

, Nikooyeh

, et al. Development of a dish-based food frequency questionnaire for Iranian population. Medical Journal of The Islamic Republic of Iran (MJIRI). 2020;34(1):889–93.

34.

USDA National Nutrient Database for Standard Reference, Legacy Release [Internet]. Nutrient Data Laboratory, Beltsville Human Nutrition Research Center, ARS, USDA. 2018 [cited April,2018]. Available from: https://data.nal.usda.gov/dataset/usda-national-nutrient-database-standard-reference-legacy-release.

35.

Dehghan

, Ilow

, Zatonska

, Szuba

, Zhang

, Mente

, et al. Development, reproducibility and validity of the food frequency questionnaire in the Poland arm of the Prospective Urban and Rural Epidemiological (PURE) study. Journal of Human Nutrition and Dietetics. 2012;25(3):225–32.

36.

Prentice

, Huang

. Nutritional epidemiology methods and related statistical challenges and opportunities. Statistical Theory and Related Fields. 2018;2(1):2–10.

37.

Doustmohammadian

, Amini

, Esmaillzadeh

, Omidvar

, Abtahi

, Dadkhah-Piraghaj

, et al. Validity and reliability of a dish-based semi-quantitative food frequency questionnaire for assessment of energy and nutrient intake among Iranian adults. BMC Research Notes. 2020;13(1):1–7.

38.

, Rimm

, Smith-Warner

, Feskanich

, Stampfer

, Ascherio

, et al. Reproducibility and validity of dietary patterns assessed with a food-frequency questionnaire. The American Journal of Clinical Nutrition. 1999;69(2):243–9.

39.

Ayoubi

, Yaghoubi

, Pahlavani

, Philippou

, MalekAhmadi

, Esmaily

, et al. Developed and validated food frequency questionnaires in Iran: A systematic literature review. Journal of Research in Medical Sciences. 2021;26(1):50.

Month	1	2	3	4	5	6	7	8	9	10
FFQ	^*									^*
3-Day Record		^*								^*
Biochemical Sampling				^*						^*

Development,validity and reproducibility of a dish-based semi-quantitative food frequency questionnaire in Iran

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

Keywords

1 Introduction

2 Method

2.1 Study design

Table 1 Study Design and Timelines of Food Frequency Questionnaire (FFQ) Administration, 3-Day Food Record, and Biochemical Sampling Month 1 2 3 4 5 6 7 8 9 10 FFQ * * 3-Day Record * * Biochemical Sampling * *

2.3 FFQ Validation study

2.3.1 Food records

2.3.2 Biochemical measurements

2.3.3 Reproducibility assessment

2.3.4 Food item analysis

2.4 Statistical methods

3 Results

5 Limitation

6 Conclusion

Footnotes

Acknowledgments

Transparency declaration

Funding

Conflict of interest

Ethical code

References

Table 1
Study Design and Timelines of Food Frequency Questionnaire (FFQ) Administration, 3-Day Food Record, and Biochemical Sampling

Month 1 2 3 4 5 6 7 8 9 10

FFQ ^* ^*

3-Day Record ^* ^*

Biochemical Sampling ^* ^*