Abstract
Aim
To translate and adapt the European Prospective Investigation of Cancer (EPIC)-Norfolk food frequency questionnaire (FFQ) for use in Croatia, and to assess the reliability and reproducibility of the Croatian version of the EPIC-Norfolk FFQ.
Methods
Translation and cross-cultural adaptation were performed according to published recommendations. Reliability was assessed in 140 respondents (61 men; age range 8-40 years) divided into three groups: young adults, pregnant women, and children and adolescents. Reproducibility was assessed in the group of young adults (32/61 men), who completed the questionnaire on two occasions three months apart.
Results
The EPIC-Norfolk FFQ showed good reliability (Cronbach alpha = 0.874). Most nutrient intakes showed good reproducibility (intraclass correlation coefficient [ICC] between 0.7 and 0.9). Poor reproducibility was observed for alcohol (ICC = 0.337), and moderate reproducibility was observed for beta-carotene (ICC = 0.692) and total carbohydrates (ICC = 0.698). Nutrient intakes measured by FFQ on two occasions did not significantly differ.
Conclusion
The Croatian version of the EPIC-Norfolk FFQ can be a useful tool for assessing dietary intakes in young people in Croatia and possibly in neighboring countries with similar languages and dietary habits.
Nutritional habits are among the most important lifestyle characteristics related to health. Poor dietary habits are associated with a range of chronic diseases and can contribute to mortality from non-communicable diseases (1).
One of the greatest challenges in nutritional epidemiology is to accurately assess dietary intake (2). Food frequency questionnaires (FFQ) are nutritional assessment tools used to estimate the individual food intake by inquiring about the frequency and quantity of consumption of food items or food groups (3). They offer a low-cost, available, and simple way of recording eating habits, which makes them an instrument of choice in epidemiological studies investigating the links between dietary habits and health or disease risks (4). Besides a structured food list, FFQ usually consist of a food composition database, which compiles the foods commonly consumed in the study population and allows the calculation of nutrient intake (5). The food list should be appropriate for the study group, because dietary habits differ according to ethnicity, economic status, culture, and location (6). It should also precisely specify the portion size to avoid systematic errors in the assessment of nutritional intake (7). One of the easiest FFQs to use in terms of administration and data processing is the European Prospective Investigation of Cancer (EPIC)-Norfolk FFQ. This FFQ is freely available online and provides software for data processing (8). In Croatia, there are available questionnaires for data collection from special populations or assessment of the intake of specific nutrients. However, most of them are constructed to be used by nutritionists and dietitians, who are much more familiar with food composition than are epidemiologists assessing food-related risks. There is no questionnaire in the Croatian language that is easy to administer and that provides software for data processing. Although the EPIC-Norfolk FFQ was designed primarily for the adult population in the UK, the food list is similar to that from the national food consumption survey in Croatia conducted in 2011 and 2012 (9). The EPIC Norfolk FFQ was validated for children and adolescents in Italy (10) and in young adults in Lebanon (11).
The aim of this study was to translate and adapt the EPIC-Norfolk FFQ for use in Croatia, and to assess the reliability and reproducibility of the Croatian version of the FFQ.
Respondents and methods
Study design
The study consisted of two phases: 1) translation and cross-cultural adaptation of the EPIC-Norfolk FFQ and 2) administration of the Croatian version of the questionnaire on two occasions, three months apart, for the reproducibility analysis.
The EPIC-Norfolk food frequency questionnaire
The EPIC-Norfolk FFQ was created in 1988. It was validated and used to assess dietary intake on a large sample in the EPIC study (12). This semiquantitative questionnaire is designed to record the average intake of foods during the previous year. It consists of two parts. Part 1 is a food list consisting of 130 items, which were classified in 10 groups as follows: 1) meat and fish, 2) bread and savory biscuits, 3) cereals, 4) potatoes, rice, and pasta, 5) dairy products and fats, 6) sweets and snacks, 7) soups, sauces, and spreads, 8) drinks, 9) fruits, and 10) vegetables. The portion size was specified for each item. Respondents select the frequency of consumption for each food item from nine frequency categories (from “never or less than once a month” to “six or more times a day”). Part 2 consists of additional questions aimed at obtaining more details about food items from Part 1, particularly breakfast cereals, the type and quantity of milk consumed, the type of fat used in cooking, the amount of visible fat on meat consumed, and supplements used during the previous year. The EPIC-Norfolk FFQ has been validated and widely used (13-15). A copy of the questionnaire and the instructions for completing and coding are available for download from the website of the EPIC study.
Translation and cross-cultural adaptation
The questionnaire was translated and cross-culturally adapted according to WHO instructions (16) and a study by Guillemin et al (17). Two professional translators independently translated the questionnaire from English to Croatian. This process produced two Croatian-language versions of the questionnaire. The versions were compared, and since there was no considerable difference in translation, a unique version was produced by the agreement of the two translators. The questionnaire was then back-translated to English by a third translator, a native English speaker, who was not familiar with the original English version of the questionnaire. Back translation yielded an English version of FFQ, which was not semantically or stylistically different from the original EPIC-Norfolk FFQ.
A focus group of expert nutritionists and dietitians was formed to discuss every item in the questionnaire, the adequacy of the food list for the composition of usual dishes in Croatian cuisine, and the adequacy of portion sizes provided by the questionnaire for the usual nutritional use at different ages. They agreed that portion sizes were adequate for every stage of life because they were small enough for children in Croatia and could be adjusted by the frequency of consumption. For example, if a food item is consumed in a twice-larger size than the portion size provided in the FFQ, the frequency of use can be multiplied by two. Brand names in the English FFQ were replaced with Croatian brand-name products that had the same or very similar nutrient composition. The brand-name products were described in detail to facilitate the use of the questionnaire for those who are unfamiliar with particular brand names. We aimed to make the food list as similar in nutrient composition as possible to the English version to be able to use the data processing software for calculating the nutrient intake. A small focus group (N = 15) was asked to fill in the questionnaire and assess if it was understandable and adequately adapted, and if it included adequate food items in the food list. Their suggestions were accepted (mainly about clarifying the food list by including additional colloquial terms for some foods in the food list), and a final version of FFQ in Croatian was produced (The questionnaire in Croatian is available upon request).
Administration of the Croatin FFQ for reliability and reproducibility analysis
The Croatian questionnaire was administered to 140 participants who volunteered for this study. The need for volunteers was advertised by departmental staff during communication with students and hospital staff with patients. Most of the participants lived in the eastern parts of Croatia: Slavonia, Baranja, and Srijem. To test the reliability of the questionnaire for different populations, we divided the sample into three groups: young adults of participants of both sexes (N = 61); pregnant women (N = 27); and children and adolescents of both sexes (N = 52). The questionnaire was self-administered and paper-based. Children were helped by their parents/caregivers to fill in the questionnaire. We believe that this did not influence answers, since it is usual for children in Croatia to have all their meals at home with parents present. For the purpose of reproducibility analysis, the questionnaire was re-administered three months later to 61 participants from the young adults group. Young adults were selected for reproducibility analysis as they have the most stable nutritional habits among the included groups. The study protocol complied with the latest revision of the Declaration of Helsinki and the national legislation. The study was approved by the Ethics Committee of the University of Osijek, Faculty of Medicine.
Statistical analysis
Descriptive results were expressed as means and standard deviations, or as percentages and frequencies. To test the reliability of the questionnaire, Cronbach's alpha coefficient was used to assess internal consistency, with alpha equal to or greater than 0.70 considered satisfactory (18). To determine the reproducibility of the questionnaire, we checked for agreement between the first and second administration of FFQ (FFQ1 and FFQ2) by using Spearman correlation and intra-class correlation coefficients (ICC) of nutrient intake for all the assessed nutrients (19). The difference in mean nutrient intakes between FFQ1 and FFQ2 was assessed with a paired t test on log transformed data. Bland-Altman plots were used to illustrate the degree of agreement between the intakes of selected nutrients assessed by FFQ1 and FFQ2. The difference between FFQ1 and FFQ2 was plotted against the mean intake of two measurements (FFQ1 and FFQ2). Statistical analysis was performed with SPSS 20.0 (IBM Corp., Armonk, NY, USA). The level of statistical significance was set at P = 0.05.
Results
The final version of the FFQ was very similar to the original EPIC-Norfolk FFQ. Since no difference in nutrient composition was found between the foods listed in the Croatian FFQ compared with the original English FFQ, the nutrient intake was calculated with the FETA software (8). The intakes of 60 nutrients and food types were calculated. The Cronbach alpha for the results of FFQ was 0.874 (95% CI 0.839-0.905).
A total of 140 respondents filled in the FFQ (Table 1). The mean intake of nutrients and food types for the whole sample and for the subgroups is presented in Table 2. The Cronbach alpha values for the subgroups were as follows: young adults - 0.848 (95% CI 0.789-0.898); pregnant women - 0.843 (95% CI 0.745-0.917), children and adolescents - 0.885 (95% CI 0.810-0.941).
Table 1.
General characteristics of respondents included in reliability testing of the translated EPIC-Norfolk Food Frequency Questionnaire
|
|
Sex, n (%) |
Age (years) |
|||
|---|---|---|---|---|---|
| male | female | Min | Max | Mean ± SD | |
|
Young adults
(N = 61) |
32 (52) |
29 (48) |
18 |
27 |
21 ± 2.13 |
|
Pregnant women
(N = 27) |
0 |
27 (100) |
22 |
40 |
30.3 ± 4.07 |
|
Children and adolescents
(N = 52) |
29 (56) |
23 (44) |
8 |
17 |
14.67 ± 2.42 |
| Total (N = 140) | 61 (44) | 79 (56) | 8 | 40 | 21.8 ± 6.05 |
Table 2.
Mean daily intakes of nutrients and food types for the groups of young adults, pregnant women, and children and adolescents
| Nutrient or a food type | Unit | Young adults (N = 61) |
Pregnant women (N = 27) |
Children and adolescents (N = 52) |
Total (N = 140) |
||||
|---|---|---|---|---|---|---|---|---|---|
| mean | SD | mean | SD | mean | SD | mean | SD | ||
|
Energy
|
kcal
|
2375.7 |
992.9 |
1.992.9 |
712.6 |
4948.1 |
1994.8 |
2.792.9 |
1696.9 |
|
Meat and meat products
|
g
|
205.0 |
145.8 |
187.3 |
171.4 |
443.5 |
209.6 |
247.7 |
187.8 |
|
Fish & fish products
|
g
|
39.0 |
33.6 |
29.0 |
24.9 |
62.5 |
31.4 |
44.5 |
42.1 |
|
Milk and milk products
|
g
|
481.2 |
239.8 |
390.1 |
241.7 |
370.2 |
270.8 |
426.2 |
244.9 |
|
Soups & sauces
|
g
|
189.1 |
168.0 |
144.5 |
85.1 |
192.7 |
57.4 |
169.3 |
120.9 |
|
Sugars; preserves and snacks
|
g
|
45.7 |
34.2 |
50.1 |
50.5 |
120.3 |
43.9 |
64.1 |
52.9 |
|
Vegetables
|
g
|
187.4 |
167.6 |
227.6 |
97.7 |
293.1 |
114.7 |
202.8 |
133.4 |
|
Fruit
|
g
|
327.9 |
396.1 |
423.9 |
344.6 |
191.8 |
167.3 |
279.6 |
308.1 |
|
Alcoholic beverages
|
g
|
63.2 |
117.4 |
13.6 |
38.9 |
2.9 |
8.5 |
29.3 |
71.1 |
|
Non-alcoholic beverages
|
g
|
446.7 |
335.3 |
424.9 |
196.3 |
500.1 |
242.0 |
447.6 |
267.9 |
|
Cereals and cereal products
|
g
|
267.2 |
168.1 |
237.9 |
127.5 |
669.9 |
289.0 |
337.1 |
255.5 |
|
Eggs and egg dishes
|
g
|
28.6 |
43.2 |
26.3 |
25.2 |
26.3 |
23.9 |
25.4 |
29.5 |
|
Fats and oils
|
g
|
20.3 |
19.6 |
14.6 |
9.1 |
73.6 |
38.3 |
32.2 |
34.4 |
|
Nuts and seeds
|
g
|
5.0 |
9.8 |
9.2 |
8.1 |
6.9 |
14.2 |
6.3 |
10.1 |
|
Potatoes. rice and pasta
|
g
|
90.6 |
77.5 |
65.4 |
43.2 |
289.5 |
169.8 |
127.5 |
129.3 |
|
Fat - total
|
g
|
96.3 |
44.8 |
81.8 |
33.7 |
247.3 |
102.7 |
125.4 |
90.7 |
|
Monounsaturated fatty acids (total)
|
g
|
34.9 |
17.1 |
29.6 |
13.5 |
91.7 |
38.6 |
46.1 |
34.2 |
|
Polyunsaturated fatty acids (total)
|
g
|
17.7 |
9.7 |
15.2 |
7.5 |
39.9 |
15.5 |
21.6 |
14.7 |
|
Saturated fatty acids (total)
|
g
|
35.1 |
16.2 |
29.8 |
10.9 |
94.1 |
40.2 |
46.7 |
34.8 |
|
Cholesterol
|
mg
|
451.7 |
237.9 |
356.2 |
129.9 |
863.1 |
374.5 |
509.6 |
315.6 |
|
Carbohydrate - total
|
g
|
279.5 |
124.8 |
238.7 |
94.5 |
535.3 |
215.2 |
316.8 |
183.0 |
|
Carbohydrate - sugars (total)
|
g
|
143.5 |
75.9 |
131.1 |
57.8 |
210.8 |
84.3 |
149.8 |
77.2 |
|
Carbohydrate - fructose
|
g
|
26.6 |
19.6 |
26.3 |
12.7 |
31.9 |
12.8 |
25.7 |
15.2 |
|
Carbohydrate - galactose
|
g
|
0.6 |
0.9 |
0.8 |
1.01 |
0.4 |
0.3 |
0.5 |
0.7 |
|
Carbohydrate - glucose
|
g
|
23.5 |
17.0 |
23.9 |
10.4 |
34.7 |
13.1 |
24.8 |
14.3 |
|
Carbohydrate - lactose
|
g
|
24.0 |
11.4 |
19.4 |
10.7 |
23.4 |
14.0 |
22.3 |
11.8 |
|
Carbohydrate - maltose
|
g
|
2.8 |
1.7 |
2.6 |
2.4 |
8.3 |
3.5 |
3.9 |
3.3 |
|
Carbohydrate - starch
|
g
|
131.8 |
70.7 |
103.8 |
51.2 |
319.9 |
137.0 |
162.5 |
118.3 |
|
Carbohydrate - sucrose
|
g
|
60.7 |
38.2 |
55.1 |
37.4 |
103.2 |
43.5 |
66.8 |
42.5 |
|
Englyst fiber – non starch polysaccharides (NSP)
|
g
|
17.1 |
10.4 |
17.7 |
6.5 |
27.7 |
11.6 |
18.4 |
10.3 |
|
Protein
|
g
|
105.6 |
47.2 |
88.9 |
25.6 |
178.1 |
70.9 |
115.5 |
59.2 |
|
Alpha carotene
|
µg
|
403.3 |
343.2 |
660.4 |
363.2 |
572.1 |
390.5 |
469.0 |
360.4 |
|
Beta carotene
|
µg
|
2525.6 |
1686.5 |
3414.3 |
1388.0 |
3.904.2 |
1865.9 |
2.876.8 |
1659.1 |
|
Carotene - total (carotene equivalents)
|
µg
|
3016.7 |
1927.8 |
3971.8 |
1588.2 |
4.552.8 |
2272.0 |
3.378.5 |
1933.7 |
|
Vitamin A - retinol
|
µg
|
1542.1 |
2007.2 |
1032.9 |
1082.4 |
4769.8 |
4496.4 |
2154.5 |
3065.9 |
|
Vitamin A - retinol equivalents
|
µg
|
2048.2 |
2060.4 |
1699.6 |
1226.4 |
5548.2 |
4833.6 |
2724.9 |
3235.1 |
|
Vitamin B2 - riboflavin
|
mg
|
2.4 |
0.9 |
1.9 |
0.7 |
3.6 |
1.7 |
2.5 |
1.3 |
|
Vitamin B1 - thiamin
|
mg
|
1.8 |
0.8 |
1.7 |
0.6 |
3.2 |
1.2 |
2.0 |
1.0 |
|
Vitamin B12 - cobalamin
|
µg
|
9.6 |
6.4 |
7.3 |
4.5 |
22.2 |
16.3 |
11.9 |
11 |
|
Vitamin B6 - pyridoxine
|
mg
|
2.4 |
1.1 |
2.2 |
0.7 |
4.2 |
1.7 |
2.7 |
1.4 |
|
Total folate
|
µg
|
303.2 |
159.7 |
274.1 |
87.3 |
425.9 |
170.1 |
307.1 |
148.7 |
|
Niacin
|
mg
|
28.0 |
13.1 |
24.6 |
6.8 |
43.7 |
16.7 |
29.9 |
14.6 |
|
Vitamin C - ascorbic acid
|
mg
|
137.9 |
124.2 |
162.1 |
89.7 |
173.7 |
72.9 |
138.5 |
94.2 |
|
Vitamin D - ergocalciferol
|
µg
|
3.4 |
1.9 |
2.7 |
1.2 |
7.2 |
3.2 |
4.1 |
2.8 |
|
Vitamin E - alpha tocopherol equivalents
|
mg
|
14.7 |
8.4 |
12.9 |
6.7 |
29.5 |
11.5 |
17.4 |
11.1 |
|
Zinc
|
mg
|
11.4 |
4.8 |
9.6 |
3.0 |
21.5 |
9.4 |
12.9 |
7.3 |
|
Nitrogen
|
g
|
17.0 |
7.6 |
14.4 |
4.1 |
28.9 |
11.6 |
18.7 |
9.7 |
|
Selenium
|
µg
|
79.9 |
34.1 |
69.8 |
25.6 |
138.7 |
56.2 |
89.1 |
47.3 |
|
Phosphorus
|
mg
|
1702.1 |
659.5 |
1455.8 |
415.5 |
2740.4 |
1053.9 |
1833.4 |
859.6 |
|
Magnesium
|
mg
|
322.3 |
149.7 |
311.8 |
88.5 |
507.4 |
196.5 |
347.8 |
163.9 |
|
Manganese
|
mg
|
2.9 |
1.8 |
2.9 |
1.2 |
4.5 |
1.8 |
3.1 |
1.7 |
|
Sodium
|
mg
|
3698 |
1616.5 |
3084.9 |
1208.7 |
7564.1 |
2991.7 |
4268.4 |
2580.9 |
|
Iodine
|
µg
|
184.3 |
72.0 |
152.6 |
59.6 |
256.3 |
106.7 |
189.0 |
89.2 |
|
Potassium
|
mg
|
3893.9 |
1810.8 |
3603.9 |
1031.8 |
6531.3 |
2696.1 |
4241.8 |
2160.1 |
|
Calcium
|
mg
|
1130.9 |
395.2 |
957.2 |
342.1 |
1531.8 |
595.5 |
1146.3 |
478.7 |
|
Chloride
|
mg
|
5451.8 |
2414.8 |
4634.6 |
1792.8 |
10936.2 |
4321.1 |
6260.7 |
3715.8 |
|
Copper
|
mg
|
1.8 |
1.2 |
1.4 |
0.7 |
4.3 |
2.8 |
2.2 |
2 |
| Iron | mg | 11.9 | 5.3 | 10.4 | 3.6 | 23.6 | 10.4 | 13.7 | 8.2 |
The correlations between FFQ1 and FFQ2 in all nutrient intakes were significant. ICC values for most of the nutrient intakes were between 0.75 and 0.9. Moderate reproducibility was found for beta carotene (ICC = 0.684-0.695) and total carbohydrates (ICC = 0.666-0.690). ICC was low (0.337) only for alcohol intake (Table 3) (20). The Bland-Altman plots for selected nutrients: energy, calcium, potassium, sodium, alcohol, and polyunsaturated fatty acids showed a small mean difference between FFQ1 and FFQ2 (Figure 1). The mean differences in nutrient intake between FFQ1 and FFQ2 were not significant (Table 4).
Table 3.
Results of reproducibility analysis: Spearman correlation coefficients and intraclass correlation coefficients (ICC) in two administrations of the food frequency questionnaire in 61 young adults. All the correlations were statistically significant
| Nutrient | Spearman rho | ICC | 95% confidence interval for ICC |
|
|---|---|---|---|---|
| lower bound | upper bound | |||
|
Energy
|
0.840 |
0.861 |
0.829 |
0.879 |
|
Fat (total)
|
0.935 |
0.865 |
0.836 |
0.881 |
|
Monounsaturated fatty acids (total)
|
0.934 |
0.864 |
0.834 |
0.880 |
|
Polyunsaturated fatty acids (total)
|
0.928 |
0.856 |
0.819 |
0.876 |
|
Saturated fatty acids (total)
|
0.916 |
0.872 |
0.849 |
0.885 |
|
Cholesterol
|
0.981 |
0.874 |
0.852 |
0.886 |
|
Carbohydrate (total)
|
0.952 |
0.682 |
0.666 |
0.690 |
|
Carbohydrate - sugars (total)
|
0.948 |
0.884 |
0.871 |
0.891 |
|
Carbohydrate - fructose
|
0.918 |
0.848 |
0.805 |
0.872 |
|
Carbohydrate - galactose
|
0.895 |
0.855 |
0.818 |
0.876 |
|
Carbohydrate - glucose
|
0.907 |
0.852 |
0.812 |
0.874 |
|
Carbohydrate - lactose
|
0.935 |
0.808 |
0.735 |
0.849 |
|
Carbohydrate - maltose
|
0.913 |
0.843 |
0.795 |
0.868 |
|
Carbohydrate - starch
|
0.965 |
0.888 |
0.878 |
0.893 |
|
Carbohydrate - sucrose
|
0.958 |
0.886 |
0.874 |
0.892 |
|
Englyst fiber – non starch polysaccharides
|
0.952 |
0.886 |
0.875 |
0.893 |
|
Alcohol
|
0.922 |
0.337 |
0.036 |
0.581 |
|
Protein
|
0.944 |
0.868 |
0.841 |
0.882 |
|
Alpha carotene
|
0.959 |
0.785 |
0.771 |
0.791 |
|
Beta carotene
|
0.970 |
0.692 |
0.684 |
0.695 |
|
Carotene - total (carotene equivalents)
|
0.971 |
0.791 |
0.783 |
0.795 |
|
Vitamin A - retinol
|
0.918 |
0.790 |
0.704 |
0.839 |
|
Vitamin B2 - riboflavin
|
0.873 |
0.817 |
0.751 |
0.854 |
|
Vitamin B1 - thiamin
|
0.913 |
0.871 |
0.847 |
0.884 |
|
Vitamin B12 - cobalamin
|
0.846 |
0.766 |
0.664 |
0.826 |
|
Vitamin B6 - pyridoxine
|
0.931 |
0.867 |
0.839 |
0.882 |
|
Total folate
|
0.949 |
0.887 |
0.875 |
0.893 |
|
Niacin
|
0.923 |
0.859 |
0.825 |
0.878 |
|
Vitamin C - ascorbic acid
|
0.922 |
0.885 |
0.872 |
0.892 |
|
Vitamin D - ergocalciferol
|
0.936 |
0.858 |
0.823 |
0.877 |
|
Vitamin E - alpha tocopherol equivalents
|
0.942 |
0.875 |
0.855 |
0.886 |
|
Zinc
|
0.927 |
0.854 |
0.815 |
0.875 |
|
Nitrogen
|
0.947 |
0.868 |
0.842 |
0.883 |
|
Selenium
|
0.941 |
0.863 |
0.833 |
0.880 |
|
Phosphorus
|
0.957 |
0.881 |
0.866 |
0.890 |
|
Magnesium
|
0.956 |
0.882 |
0.866 |
0.890 |
|
Manganese
|
0.941 |
0.887 |
0.876 |
0.893 |
|
Sodium
|
0.932 |
0.850 |
0.809 |
0.873 |
|
Iodine
|
0.978 |
0.884 |
0.871 |
0.891 |
|
Potassium
|
0.746 |
0.875 |
0.854 |
0.886 |
|
Calcium
|
0.978 |
0.784 |
0.769 |
0.791 |
|
Chloride
|
0.921 |
0.853 |
0.814 |
0.874 |
|
Copper
|
0.854 |
0.810 |
0.738 |
0.851 |
| Iron | 0.931 | 0.931 | 0.873 | 0.962 |
Figure 1.
Bland-Altman plots show differences between two administrations of the Croatian version of the EPIC-Norfolk Food Frequency Questionnaire (FFQ1 and FFQ2) against their mean value ([FFQ1+FFQ2]/2) for daily intakes of (A) energy (kcal); (B) polyunsaturated fatty acids (PUFA) (g); (C) alcohol (g); (D) calcium (mg); (E) sodium (mg); and (F) potassium (mg). Solid lines represent the mean difference in nutrient intake between two administrations of the FFQ, while dashed lines represent 1.96 standard deviation (SD) of that difference.
Table 4.
Results of the reproducibility analysis: mean daily nutrient intakes estimated by two administrations of the food frequency questionnaire (FFQ1 and FFQ2) in a group of young adults (N = 61)
| Nutrient | Unit | FFQ1 |
FFQ2 |
P* | ||
|---|---|---|---|---|---|---|
| mean | SD | mean | SD | |||
|
Energy
|
kcal
|
2375.7 |
992.9 |
2389.9 |
979.9 |
0.955 |
|
Fat - total
|
g
|
96.3 |
44.8 |
96.4 |
44.8 |
0.939 |
|
Monounsaturated fatty acids (total)
|
g
|
34.9 |
17.1 |
34.8 |
17.0 |
0.897 |
|
Polyunsaturated fatty acids (total)
|
g
|
17.7 |
9.7 |
17.9 |
9.7 |
0.674 |
|
Saturated fatty acids (total)
|
g
|
35.1 |
16.2 |
35.1 |
16.3 |
0.984 |
|
Cholesterol
|
mg
|
451.7 |
237.9 |
454.2 |
239.1 |
0.744 |
|
Carbohydrate - total
|
g
|
279.5 |
124.8 |
281.1 |
123.1 |
0.631 |
|
Carbohydrate - sugars (total)
|
g
|
143.5 |
75.9 |
143.6 |
75.0 |
0.964 |
|
Carbohydrate - fructose
|
g
|
26.6 |
19.6 |
26.3 |
19.2 |
0.742 |
|
Carbohydrate - galactose
|
g
|
0.6 |
0.9 |
0.7 |
0.9 |
0.063 |
|
Carbohydrate - glucose
|
g
|
23.5 |
17.0 |
23.1 |
15.4 |
0.575 |
|
Carbohydrate - lactose
|
g
|
24.0 |
11.4 |
24.5 |
11.9 |
0.467 |
|
Carbohydrate - maltose
|
g
|
2.8 |
1.7 |
2.8 |
1.8 |
0.717 |
|
Carbohydrate - starch
|
g
|
131.8 |
70.7 |
133.1 |
70.4 |
0.387 |
|
Carbohydrate - sucrose
|
g
|
60.7 |
38.2 |
60.5 |
38.1 |
0.804 |
|
Englyst Fiber – non-starch polysaccharides
|
g
|
17.1 |
10.4 |
17.1 |
10.3 |
0.878 |
|
Protein
|
g
|
105.6 |
47.2 |
106.1 |
45.4 |
0.755 |
|
Alcohol
|
g
|
4.4 |
8.3 |
3.7 |
5.9 |
0.116 |
|
Alpha carotene
|
µg
|
403.3 |
343.2 |
393.2 |
339.6 |
0.108 |
|
Beta carotene
|
µg
|
2525.6 |
1686.5 |
2512.7 |
1694.8 |
0.576 |
|
Carotene - total (carotene equivalents)
|
µg
|
3016.7 |
1927.8 |
3009.1 |
1933.0 |
0.785 |
|
Vitamin A - retinol
|
µg
|
1542.1 |
2007.2 |
1505.7 |
2025.4 |
0.805 |
|
Vitamin A - retinol equivalents
|
µg
|
2048.2 |
2060.4 |
2010.4 |
2089.9 |
0.798 |
|
Vitamin B2 - riboflavin
|
mg
|
2.4 |
0.9 |
2.4 |
0.9 |
0.987 |
|
Vitamin B1 - thiamin
|
mg
|
1.8 |
0.8 |
1.8 |
0.8 |
0.200 |
|
Vitamin B12 - cobalamin
|
µg
|
9.6 |
6.4 |
9.5 |
6.4 |
0.794 |
|
Vitamin B6 - pyridoxine
|
mg
|
2.4 |
1.1 |
2.4 |
1.1 |
0.833 |
|
Total folate
|
µg
|
303.2 |
159.7 |
304.7 |
158.3 |
0.670 |
|
Niacin
|
mg
|
28.0 |
13.1 |
28.1 |
12.6 |
0.731 |
|
Vitamin C - ascorbic acid
|
mg
|
137.9 |
124.2 |
137.7 |
123.4 |
0.949 |
|
Vitamin D - ergocalciferol
|
µg
|
3.4 |
1.9 |
3.3 |
1.8 |
0.526 |
|
Vitamin E - alpha tocopherol equivalents
|
mg
|
14.7 |
8.4 |
14.9 |
8.3 |
0.406 |
|
Zinc
|
mg
|
11.4 |
4.8 |
11.3 |
4.7 |
0.838 |
|
Nitrogen
|
g
|
17.0 |
7.6 |
17.1 |
7.3 |
0.749 |
|
Selenium
|
µg
|
79.9 |
34.1 |
80.9 |
32.9 |
0.434 |
|
Phosphorus
|
mg
|
1702.1 |
659.5 |
1713.4 |
648.9 |
0.501 |
|
Magnesium
|
mg
|
322.3 |
149.7 |
324.4 |
146.7 |
0.585 |
|
Manganese
|
mg
|
2.9 |
1.8 |
2.9 |
1.8 |
0.909 |
|
Sodium
|
mg
|
3698 |
1616.5 |
3711 |
1601 |
0.371 |
|
Iodine
|
µg
|
184.3 |
72.0 |
187.3 |
71.5 |
0.061 |
|
Potassium
|
mg
|
3893.9 |
1810.8 |
3893.7 |
1805.3 |
0.476 |
|
Calcium
|
mg
|
1130.9 |
395.2 |
1134.4 |
392.2 |
0.051 |
|
Chloride
|
mg
|
5451.8 |
2414.8 |
5554.2 |
2374.8 |
0.319 |
|
Copper
|
mg
|
1.8 |
1.2 |
1.7 |
1.2 |
0.742 |
| Iron | mg | 11.9 | 5.3 | 11.8 | 5.2 | 0.620 |
*paired t test on log-transformed data.
Discussion
The process of translation and adaptation of the EPIC-Norfolk FFQ yielded a Croatian version of the FFQ, which was very similar to the original questionnaire. The composition of items on the food list was carefully analyzed, with the aim of calculating the usual nutritional intake. The EPIC-Norfolk FFQ has been previously validated and widely used (13-15), therefore, we did not validate it again. The reliability analysis showed that the translated FFQ was a reliable method to assess diet in a self-administered way among children, adolescents, young adults, and pregnant women, with good reproducibility in young adults for almost all nutritional intakes. The group of young adults (N = 61) consisted mostly of college students of both sexes. The second group consisted of pregnant women, who were regarded as a separate group because of specific nutritional needs during pregnancy (21). The third group included children and adolescents. In childhood and adolescence, adequate nutrition is crucial for achieving full growth potential (22). Samuelson et al found that adolescents in Nordic countries had an irregular meal pattern; many of them skipped breakfast and school lunch but had dinner. However, snacking and light meals were very common, contributing to 25%-35% of the daily energy intake (23). This could lead to problems with weight control and consequent health issues.
The reliability of the Croatian FFQ was satisfactory for all groups, with Cronbach alpha values around 0.8-0.9 (18). The EPIC Norfolk FFQ was originally intended for the adult population in the UK and was previously validated for children and adolescents in Italy (10), Lebanon (11), Kazakhstan (24), and Romania (25).
The mean reproducibility was also satisfactory. Reproducibility is defined as a consistency of measurements on more than one administration to the same person at different times (26). ICC values for most of the nutrient intakes were between 0.75 and 0.9, which indicates good reproducibility (27). According to a meta-analysis by Cade et al (7), correlation coefficients between two administrations of a FFQ usually range from 0.5 to 0.7. The higher correlations found in the present study could be explained by our use of the Spearman correlation.
The interval between two administrations of FFQ in this study was three months. The interval between two administrations should be long enough for respondents to forget the answers given in the first administration and not too long so that dietary intakes do not change. Previous studies reported intervals ranging from one week to two years (28). Increasing the length of the interval between two administrations was found to decrease correlation coefficients (7). We believe the interval between administrations in this study was appropriate and long enough for respondents not to remember the answers from the first administration.
According to the meta-analysis by Cui et al (28), an FFQ with a correlation coefficient greater than 0.5 for most nutrients is considered a reliable tool to measure dietary intake. Accordingly, the Croatian version of the EPIC-Norfolk FFQ from this study could be regarded as a reliable questionnaire. A lack of significant differences in nutrient intakes assessed by FFQ1 and FFQ2 presented by Bland-Altman plots confirms this conclusion. The plots represented excellent agreement between FFQ1 and FFQ2, since most of the differences between measurements were very close to zero, and the mean differences were also near zero. Confidence intervals were affected by outliers, and were thus wider than desirable, but the position of the results indicated a small difference between the two administrations of the questionnaire. The reason for the presence of outliers is unknown.
Although in this study the FFQ was culturally adapted for Croatia, it could be used in the neighboring countries with similar language and dietary habits. Balkan region (countries of the former Yugoslavia: Slovenia, Croatia, Bosnia and Herzegovina, Montenegro and Macedonia, Romania, Bulgaria, part of Turkey, Greece, and Albania) (29-31) have similar dietary habits. The traditional food pattern in these countries is similar to the Mediterranean diet, with a distinct influence of western cuisine and food processing industry (29). Although the Croatian language is similar to the languages spoken in neighboring countries (Serbia, Bosnia and Herzegovina, Montenegro) (32), each language developed through time, and some terms in the Croatian language used in this FFQ could differ from terms in other similar languages in the region. However, most people are familiar with linguistic peculiarities of languages spoken in other countries in the region. This makes the possibility for misunderstanding of terms used in this questionnaire negligible. In order to facilitate the use of this questionnaire in other countries in the region it would be appropriate to the clarify food list and eventually include additional colloquial terms for some foods in the food list.
Our study has several strengths. We included the usual dishes from Croatian cuisine, but kept them as similar as possible to the original questionnaire to facilitate result calculation, and international comparisons. We followed the strict protocol to ensure the FFQ was correctly translated to Croatian.
This study also has some limitations. The questionnaire was administered for the second time only to the group of young adults. The main reason was that young adults have the most stable dietary habits among the respondents, as the other two groups were more likely to change eating habits in the test-retest period. Also, it was reported elsewhere that recruiting adolescents can be quite challenging (33), and the sample size of 50 is acceptable for a validation study (7). Finally, since the majority of the respondents were college students and well-educated women, self-administration of this questionnaire to a population with a lower educational level could lead to problems with understanding the questionnaire instructions. The use of this questionnaire in general population would probably require the administration by interviewers instead of self-administration.
In conclusion, this adaptation and reliability study showed that the Croatian FFQ adapted from the EPIC-Norfolk FFQ had good reliability calculated as internal consistency for all included subgroups, and good reproducibility for most nutrients for young adults.
Acknowledgments
Funding This study was a part of the Faculty of Medicine Osijek institutional research projects IP-9-MEFOS-2021 (PI, IJ), IP-10-MEFOS-2021 (PI, AS), IP-02-MEFOS-2022 (PI, AS), and IP-17-MEFOS-2022 (PI, IJ), as well as the European Structural and Investment Funds grant for the Croatian National Scientific Center of Excellence for Personalized Health Care, Josip Juraj Strossmayer University of Osijek (Grant number: KK.01.1.1.01.0010).
Ethical approval granted by the Ethics Committee of the University of Osijek, Faculty of Medicine (approval number for the study of young adults 2158-61-07-12-42, for the study of pregnant women 2158-61-07-21-10, for the study of children and adolescents 2158-61-07-21-06).
Declaration of authorship EDC, ID conceived and designed the study; PŠ, MV, MK, ML, LS acquired the data; EDC, ID, IJ, AS analyzed and interpreted the data; EDC, ID drafted the manuscript; ID, IJ, AS, PŠ, MV, MK, ML, LS critically reviewed the manuscript for important intellectual content; all authors gave approval of the version to be submitted; all authors agree to be accountable for all aspects of the work.
Competing interests All authors have completed the Unified Competing Interest form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organization for the submitted work; no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years; no other relationships or activities that could appear to have influenced the submitted work.
References
- 1. GBD 2017 Diet Collaborators. Health effects of dietary risks in 195 countries, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2019;393:1958–72. doi: 10.1016/S0140-6736(19)30041-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Byers T, Lyle B. The role of epidemiology in determining when evidence is sufficient to support nutrition recommendations Summary statement. Am J Clin Nutr. 1999;69:1365S–7S. doi: 10.1093/ajcn/69.6.1365S. [DOI] [PubMed] [Google Scholar]
- 3. Subar AF. Developing dietary assessment tools. J Am Diet Assoc. 2004;104:769–70. doi: 10.1016/j.jada.2004.02.007. [DOI] [PubMed] [Google Scholar]
- 4. Rodrigo CP, Aranceta J, Salvador G, Varela-Moreiras G. Food frequency Questionnaires. Nutr Hosp. 2015;31:49–56. doi: 10.3305/nh.2015.31.sup3.8751. [DOI] [PubMed] [Google Scholar]
- 5.Thompson FE, Subar AF. Dietary assessment methodology. In: Coulston A, Boushey C (eds). Nutrition in the prevention and treatment of disease, 2nd ed. Academic Press; 2008. pp 3-38. [Google Scholar]
- 6. Shai I, Shahar DR, Vardi H, Fraser D. Selection of food items for inclusion in a newly developed food-frequency questionnaire. Public Health Nutr. 2004;7:745–9. doi: 10.1079/PHN2004599. [DOI] [PubMed] [Google Scholar]
- 7. Cade JE, Burley VJ, Warm DL, Thompson RL, Margetts BM. Food-frequency questionnaire: a review of their design, validation and utilization. Nutr Res Rev. 2004;17:5–22. doi: 10.1079/NRR200370. [DOI] [PubMed] [Google Scholar]
- 8.Mulligan AA, Luben RN, Bhaniani A, Parry-Smith DJ.O’Connor L, Khawaja AP, et alA new tool for converting food frequency questionnaire data into nutrient and food grouP values: FETA research methods and availability. BMJ Open 20144e004503 10.1136/bmjopen-2013-004503 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Sokolić D, Jurković M, Bašić S, Mikec D, Ileš D, Libl Vargović L, et al. Croatian National Food Consumption Survey on Adult population. EFSA Supporting Publications. 2017;14:1297E. [Google Scholar]
- 10. Huybrechts I, Miglio R, Mistura L, Grioni S, Pozzebon I, Odorifero C, et al. Relative validity of an Italian EPIC Food Frequency Questionnaire for dietary factors in children and adolescents. A Rizzoli Orthopedic Institute Study. Nutrients. 2021;13:1245. doi: 10.3390/nu13041245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Khaled K, Hundley V, Bassil M, Bazzi M, Tsofliou F. Validation of the European Prospective Investigation into Cancer (EPIC) FFQ for use among adults in Lebanon. Public Health Nutr. 2021;24:4007–16. doi: 10.1017/S1368980021002123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Day N, Oakes S, Luben R, Khaw KT, Bingham S, Welch A, et al. EPIC-Norfolk: study design and characteristics of the cohort. European Prospective Investigation of Cancer. Br J Cancer. 1999;80:95–103. [PubMed] [Google Scholar]
- 13. Bingham SA, Gill C, Welch A, Cassidy A, Runswick SA, Oakes S, et al. Validation of dietary assessment methods in the UK arm of EPIC using weighed records, and 24-hour urinary nitrogen and potassium and serum vitamin C and carotenoids as biomarkers. Int J Epidemiol. 1997;26:S137–51. doi: 10.1093/ije/26.suppl_1.S137. [DOI] [PubMed] [Google Scholar]
- 14. Bingham SA, Welch AA, McTaggart A, Mulligan AA, Runswick SA, Luben RN, et al. Nutritional methods in the European Prospective Investigation of Cancer in Norfolk. Public Health Nutr. 2001;4:847–58. doi: 10.1079/PHN2000102. [DOI] [PubMed] [Google Scholar]
- 15.McKeown NM, Day NE, Welch AA, Runswick SA, Luben RN, Mulligan AA, et al. Use of biological markers to validate self-reported dietary intake in a random sample of the European Prospective Investigation into Cancer United Kingdom Norfolk cohort. Am J Clin Nutr. 2001;74:188.196. [DOI] [PubMed]
- 16.World Health Organization. Process of translation and adaptation of instruments. Available from: http://www.who.int/substance_abuse/research_tools/translation/en Accessed: June 29, 2022.
- 17. Guillemin F, Bombardier C, Beaton D. Cross-cultural adaptation of health-related quality of life measures: Literature review and proposed guidelines. J Clin Epidemiol. 1993;46:1417–32. doi: 10.1016/0895-4356(93)90142-N. [DOI] [PubMed] [Google Scholar]
- 18.Nunnally JC, Bernstein IH. Psychometric Theory. 3rd Edition New York: McGraw-Hill; 1994. [Google Scholar]
- 19. Rankin G, Stokes M. Reliability of assessment tools in rehabilitation: an illustration of appropriate statistical analyses. Clin Rehabil. 1998;12:187–99. doi: 10.1191/026921598672178340. [DOI] [PubMed] [Google Scholar]
- 20. Liljequist D, Elfving B, Skavberg Roaldsen K. Intraclass correlation - A discussion and demonstration of basic features. PLoS One. 2019;14:e0219854. doi: 10.1371/journal.pone.0219854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Şenol Eren N, Şencan İ, Aksoy H, Koç EM, Kasım İ, Kahveci R, et al. Evaluation of dietary habits during pregnancy. Turk J Obstet Gynecol. 2015;12:89–95. doi: 10.4274/tjod.79923. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Das JK, Salam RA, Thornburg KL, Prentice AM, Campisi S, Lassi ZS, et al. Nutrition in adolescence: physiology, metabolism, and nutritional needs. Ann N Y Acad Sci. 2017;1393:21–33. doi: 10.1111/nyas.13330. [DOI] [PubMed] [Google Scholar]
- 23. Samuelson G. Dietary habits and nutritional status in adolescents over Europe. An overview of current studies in the Nordic countries. Eur J Clin Nutr. 2000;54:S21–8. doi: 10.1038/sj.ejcn.1600980. [DOI] [PubMed] [Google Scholar]
- 24. Tulenova R. Validation of a Kazakh food frequency questionnaire. J Food Process Technol. 2018;9:54. [Google Scholar]
- 25. Gherasim A, Arhire LI, Nita O, Strateanu R, Oprescu AC, Graur M, et al. Can the EPIC Norfolk food frequency questionnaire be applied to the population in Romania? Rev Med Chir Soc Med Nat. 2015;119:856–63. [Google Scholar]
- 26.Willett W, Lenart E. Reproducibility and validity of Food-frequency questionnaires In: Nutritional epidemiology 2. Willett W, editor. Oxford: Oxford University Press; 1998. pp. 101-47. [Google Scholar]
- 27. Koo TK, Li MY. Guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med. 2016;15:155–63. doi: 10.1016/j.jcm.2016.02.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Cui Q, Xia Y, Wu Q, Chang Q, Niu K, Zhao Y. A meta-analysis of the reproducibility of food frequency questionnaires in nutritional epidemiological studies. Int J Behav Nutr Phys Act. 2021;18:12. doi: 10.1186/s12966-020-01078-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Jasar D, Curcic B, Kubelka-Sabit K, Filipovski V, Kukurinov V. Health issues and nutrition in the Balkans. J Hyg Eng Des. 2021;35:96–105. [Google Scholar]
- 30.Krasteva-Blagoeva E, Bogueva D. Chapter 3 - Balkan food cultures and traditions. In: Gostin AI, Bogueva D, Kakurinov V, editors. Nutritional and health aspects of traditional and ethnic foods, nutritional and health aspects of food in the Balkans, Academic Press, 2021. [Google Scholar]
- 31.Vuković AJ, Terzić A. Gastronomy and regional identity: balkan versus national cuisine. In: Peštek A, Kukanja M, Renko S, editors. Gastronomy for Tourism Development. Leeds: Emerald Publishing Limited. 2020. p. 1-25. [Google Scholar]
- 32. Bailyn JF. To what degree are Croatian and Serbian the same language? Evidence from a translation study. Journal of Slavic Linguistics. 2010;18:181–219. doi: 10.1353/jsl.2010.0000. [DOI] [Google Scholar]
- 33. Moreno MA, Waite A, Pumper M, Colburn T, Holm M, Mendoza J. Recruiting adolescent research participants: in person compared to social media approaches. Cyberpsychol Behav Soc Netw. 2017;20:64–7. doi: 10.1089/cyber.2016.0319. [DOI] [PubMed] [Google Scholar]