- Email: [email protected]
Accuracy of reported food intake in a sample of 7–10 year-old children in Serbia
p u b l i c h e a l t h x x x ( 2 0 1 6 ) 1 e6
Available online at www.sciencedirect.com
Public Health journal homepage: www.elsevier.com/puhe
Original Research
Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia a,* S. Sumonja , M. Jevtic b a b
University of Novi Sad, Faculty of Education, Department of Natural Sciences, Serbia University of Novi Sad, Faculty of Medicine, Department of Hygiene, Serbia
article info
abstract
Article history:
Objectives: Children's ability to recall and report dietary intake is affected by age and
Received 21 October 2014
cognitive skills. Dietary intake reporting accuracy in children is associated with age, weight
Received in revised form
status, cognitive, behavioural, social factors and dietary assessment techniques. This study
4 August 2015
analysed accuracy of 7e10 year-old children's reported food intake for one day.
Accepted 15 March 2016
Study design: Validation study.
Available online xxx
Methods: Sample included 94 children aged 7e10 years (median ¼ 9 years) from two
Keywords:
was a combination of 24-h recall and food recognition form. It included recalls for five
Food intake
meals: breakfast at home; snack at home; lunch at home; snack at school and dinner at
Children
home. Parental reports were used as reference information about children's food intake for
Questionnaires
meals obtained at home and observation was used to gain reference information for school
elementary schools in a local community in Serbia. ‘My meals for one day’ questionnaire
meal. Observed and reported amounts were used to calculate omission rate, intrusion rate, corresponding, over-reported and unreported amounts of energy, correspondence rate and inflation ratio. Results: Overall omission rate (37.5%) was higher than overall intrusion rate (36.7%). The same food item (bread) has been the most often correctly reported and omitted food item for breakfast, lunch and dinner. Snack at school had the greatest mean correspondence rate (79.6%) and snack at home the highest mean inflation ratio (90.7%). Conclusion: Most errors in children's recalls were incorrectly reported amounts and not the food items. The questionnaire should be improved to facilitate accurate reports of the amounts. © 2016 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
Introduction An increasing trend of prevalence of overweight and obese children is an important public health problem in Serbia as in
other countries in Europe and USA.1e3 Therefore, validated food intake assessment techniques in child population are important part of research about the determinants of food intake and obesity in children. Food intake assessment in children is a particularly challenging topic for researchers
ka 4, 25000 Sombor, Serbia. Tel.: þ381 638820270; fax: þ381 25416461. * Corresponding author. Podgoric E-mail address: [email protected] (S. Sumonja). http://dx.doi.org/10.1016/j.puhe.2016.03.016 0033-3506/© 2016 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved. M, Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia, Please cite this article in press as: Sumonja S, Jevtic Public Health (2016), http://dx.doi.org/10.1016/j.puhe.2016.03.016
2
p u b l i c h e a l t h x x x ( 2 0 1 6 ) 1 e6
because children's eating habits are variable and influenced by many different factors.4,5 The eating habits of school-aged children are under control of the parents while adolescents practice more out-of-home eating.6,7 Furthermore, children's ability to recall and report dietary intake is age-related and affected primarily by their cognitive skills.8 Previous research shows that children cannot report dietary intake without parental assistance until they reach seven years.7 The period from seven to 11 years presents transitional period in a child's ability to recall and report dietary intake without assistance. That may be one of the reasons why research, including 7e12 year-old children show different results about accuracy of child-reported food intake.9,10 Baranowski and Domel explained that an error in reporting dietary intake may occur during every stage of the process of remembering and recalling dietary information in children.11 Errors in dietary reports may be omitted food items (missing foods), over-reported food items (phantom foods) or incorrectly identifying food items.6 Factors that are associated with children's recall accuracy include age, weight status, cognitive, behavioural, social factors and dietary assessment techniques.8,12e14 The aim of our study was to analyse accuracy of 7e10 year-old children's reported food intake for one day according to parents' reports and observation.
Methods We conducted the study in a local urban community in northern Serbia. Participants were 1st to 4th grade students. In order to obtain a representative sample, we telephoned all five elementary schools in this community to participate in our study and two schools accepted. At each school, two classes from the first, second, third and fourth grades were randomly selected to take part in the study (184 children). After getting child assent and parental written consent total sample included 94 children (response rate 51.1%). Data collection that included completing the questionnaires by children and parents and direct observation occurred between March and May 2014. The questionnaire for children was classroom administered by one researcher in all classes during the last school class. The researcher used the same protocol for classroom questionnaire administration in all classes. Children were instructed by the researcher to recall types and amounts of foods they ate for breakfast, snack before school, lunch at home and snack at school on the day of the data collection. Children were further instructed by the researcher to recall types and amounts of food they ate for dinner the day before data collection. Researcher and teachers were present in the classroom while the children were filling the questionnaire in all classes. The researcher clarified food items that children were not able to recognise and provided no further help to children in completing the questionnaire. ‘My meals for one day’ questionnaire was developed for assessing children's food intake. Questionnaire development was based on a combination of two dietary assessment techniques, 24-h recall and food recognition form.15,16 It consisted of five parts representing five meals: breakfast at home; snack at home; lunch at home; snack at school and dinner at home. Each part included pictures of different food items from each food
group that are usually consumed for these meals and a portion size, which participants used to estimate the amounts of food consumed (e.g. glass, tablespoon, slice, piece). Part of the questionnaire presenting lunch at home was divided into soup, meat, vegetables, bread and pasta, salad, drink and dessert food items. There was an additional option to add food if it was not offered. Parental questionnaire consisted of the same parts as the child's questionnaire except it did not include pictures of food items. Parental questionnaire also included data about their child's height and weight, which were used to calculate children's body mass index (BMI). A consent form, the letter containing instructions for parents and the questionnaire were sent to parents via children on the same day that the children completed their questionnaires. We used direct observation of children eating school meal to obtain reference information about children's food intake for that meal. Observations occurred during a 30-min school meal break. All children from each class were observed whether they ate a school meal or food brought from home. Two researchers simultaneously observed children from one class. Each observer simultaneously observed approximately 12 children and recorded items and amounts that were in a child's plate before and after eating. Amounts were recorded in qualitative terms as amounts eaten per serving. Observers took a training session which was consisted of measuring and observing different amounts of food usually eaten for snack in schools and identifying serving sizes of different food items. In this training session, observers identified food and beverage items and amounts in 10 samples of packed lunches (30 food items) previously measured by the third researcher. Accuracy of reporting food items was 93.3% for the first observer and 94.3% for the second observer. Accuracy of reporting food amounts was 91.6% for the first observer and 92% for the second observer. Food items reported for snack at school were compared to reference information gained by observation of the school meal. Each reported item was classified as a match if it was reported and observed eaten, an intrusion if it was reported, but was not observed eaten and an omission if it was observed eaten, but was not reported. Food items reported for breakfast at home, snack at home, lunch at home and dinner at home were compared to food items reported by parents. Each item was classified as a match if it was reported by both a parent and a child, as an intrusion if it was reported by child but not reported by parent and omission if it was reported by a parent but was not reported by a child. Omission rate and intrusion rate were used to present number (%) of omitted or intruded food items.17 Observed and reported amounts were used to calculate corresponding, overreported and unreported amounts of energy.17 Corresponding, over-reported and unreported amounts of energy were then used to calculate correspondence rate and inflation ratio. Correspondence rate is a measure of reporting accuracy that is sensitive to reporting errors.17 Inflation ratio is a measure of reporting errors.17 Calculations of energy intake were based on a food database available in Serbian language.18 The independent t-test was used to analyse whether there are statistically significant differences in the omission rate, intrusion rate, correspondence rate and inflation ratio between boys
M, Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia, Please cite this article in press as: Sumonja S, Jevtic Public Health (2016), http://dx.doi.org/10.1016/j.puhe.2016.03.016
3
p u b l i c h e a l t h x x x ( 2 0 1 6 ) 1 e6
and girls.19 One-way ANOVA was used to investigate differences between the omission rate, intrusion rate, correspondence rate, and inflation ratio according to meal, grade and BMI.19 After adjustment in the P-values with Bonferroni correction for multiple tests, P-values < 0.017 were considered statistically significant.19
Results The sample characteristics are presented in Table 1. Bread has been the most often correctly reported (matched) food item for breakfast at home (40.4%), lunch at home (38.3%) and dinner at home (30.9%). Bread has also been the most often omitted food item for breakfast (25.5%), lunch (19.3%) and dinner (20.2%) at home. Apples have been the most often correctly reported food item for snack at home (13.8%), but were also the most often intruded (8.6%) and omitted (17.0%) food item for this meal. Cheese pie has been the most often correctly reported food item for snack at school (15.8%) and also the most often intruded (7.9%) food item for this meal. Doughnuts have been the most often omitted (5.3%) food item for snack at school. Milk has been the most often intruded food item for breakfast (5.3%), chocolate for lunch and pizza (7.4%) has been the most often intruded food item for dinner. Omission rates, intrusion rates and reporting errorsensitive variables overall and for each meal are presented in Table 2. Table 2 includes omission rate, intrusion rate, corresponding, over-reported and unreported amounts of energy, correspondence rates and inflation ratios overall and for each meal. Independent samples t-test found that boys (mean ¼ 49.5%) had a significantly higher omission rate than girls (mean ¼ 33.5) (mean difference ¼ 16.0; P ¼ 0.004). Intrusion rate was also significantly higher for boys (mean ¼ 60.8%) than for girls (mean ¼ 37.3%) (mean difference ¼ 23.5; P ¼ 0.017). We did not find any significant differences in omission rate, intrusion rate, correspondence rates and inflation ratio according to age and body mass index.
Discussion The overall omission rate was slightly higher than overall intrusion rate, which may indicate than children in our study
more often omitted to report food items than they intruded food items. This result is consistent with results of the study by Baxter and Thompson, although they used individual interviews to obtain dietary recall.20 Other studies show mixed results about the overall accuracy on the food item level, probably because of differences in methodology.21e23 Hunsberger et al. found 90% accuracy at the food item level, but that study analysed self-reported recalls for one meal.10 Baxter et al. found much lower accuracy and high total inaccuracy in a sample of fourth-grade children, but explored self-reported recalls based on a 24-h recall.23 Baxter et al. also showed that the accuracy of children's food intake recall is less when it is obtained as a part of a 24-h recall rather than as a single meal recall.20 It is expected that accuracy of children's food intake recalls decreases as the number of meals and overall food items increases probably because of information overload.7,11 A large percentage of incorrectly reported food items in our study could partially be explained by information overload.7,11 The result that the same food item (bread, apple, cheese pie) was the most often correctly reported and omitted or intruded food item for the same meal could indicate that most errors in children's recalls for those meals came from incorrectly reported amounts of those food items. This could be further analysed by variables described in Table 2. Reporting error-sensitive variables were used in this study as they were shown to be more sensitive in detecting reporting errors.17 Overall corresponding amount from matches is larger than unreported and over-reported amount from matches, but these variables indicate the extent of errors in reported amounts. Analysing the reporting of error-sensitive variable by meals, it is evident that snack at school was the meal with the greatest mean correspondence rate and lowest intrusion and omission rate. This is an expected result as the retention interval from snack at school and interview was much shorter than the retention interval between interview and other meals.23 We validated snack at school reports against observation which may also affect the mean correspondence rate for this meal. Lunch at home was the meal with the greatest amount that was reported both correctly and incorrectly (mean corresponding amount from matches, mean overreported and unreported amounts from matches, mean over-reported amount from intrusions). Lunch is a meal that usually takes the greatest part in daily energy intake and, consists of many different food items and mixed dishes, so it
Table 1 e Sample characteristics. Total % (n) Gender Boys Girls BMI Underweight Normal weight Overweight Obese Not reported Total
50 (47) 50 (47) 6.4 53.2 21.3 5.3 13.8 100
(6) (50) (20) (5) (13) (94)
1st grade
2nd grade
3rd grade
4th grade
7.4 (7) 6.4 (6)
9.6 (9) 11.7 (11)
13.8 (13) 12.8 (12)
19.1 (18) 19.1 (18)
1.1 11.7 4.2 2.1 2.1 21.3
0.0 18.1 4.2 2.1 2.1 26.5
3.2 (3) 14.9 (14) 12.7 (12) 0.0 (0) 7.4 (7) 38.6 (36)
2.1 8.5 0.0 1.1 2.1 13.8
(2) (8) (0) (1) (2) (13)
(1) (11) (4) (2) (2) (20)
(0) (17) (4) (2) (2) (25)
M, Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia, Please cite this article in press as: Sumonja S, Jevtic Public Health (2016), http://dx.doi.org/10.1016/j.puhe.2016.03.016
4
p u b l i c h e a l t h x x x ( 2 0 1 6 ) 1 e6
Table 2 e Omission rates, intrusion rates, correspondence rates, inflation ratio, corresponding, over-reported and unreported amounts of energy, overall and for each meal. n
Min.
Max.
Mean
SD
c
Omission rate (%) Overall 94 0.00 100.0 94 0.00 100.0 Breakfast at homea 94 0.00 100.0 Snack at homea Lunch at homea 94 0.00 100.0 94 0.00 100.0 Snack at schoolb 94 0.00 100.0 Dinner at homea Intrusion rate (%)d Overall 94 0.00 100.0 94 0.00 100.0 Breakfast at homea 94 0.00 100.0 Snack at homea Lunch at homea 94 0.00 100.0 94 0.00 100.0 Snack at schoolb 94 0.00 100.0 Dinner at homea Correspondence rate (%)e Overall 94 0.00 100.0 94 0.00 100.0 Breakfast at homea 94 0.00 100.0 Snack at homea 94 0.00 100.0 Lunch at homea 94 0.00 100.0 Snack at schoolb 94 0.00 100.0 Dinner at homea Inflation ratio (%)f Overall 94 0.00 226.5 94 0.00 281.5 Breakfast at homea 94 0.00 975.0 Snack at homea 94 0.00 679.8 Lunch at homea Snack at schoolb 94 0.00 346.8 94 0.00 219.9 Dinner at homea Corresponding amount from matches (kcal)g Overall 94 0.00 2938.4 94 0.00 931.0 Breakfast at homea 94 0.00 528.0 Snack at homea 94 0.00 962.0 Lunch at homea 94 0.00 437.0 Snack at schoolb Dinner at homea 94 0.00 902.0 Unreported amount from matches (kcal)h Overall 94 0.00 988.6 94 0.00 723.6 Breakfast at homea 94 0.00 352.0 Snack at homea 94 0.00 681.0 Lunch at homea 94 0.00 154.0 Snack at schoolb Dinner at homea 94 0.00 390.0 Over-reported amount from matches (kcal)i Overall 94 0.00 1011.0 94 0.00 305.0 Breakfast at homea Snack at homea 94 0.00 300.0 94 0.00 812.0 Lunch at homea 94 0.00 195.0 Snack at schoolb Dinner at homea 94 0.00 117.0 Over-reported amount from intrusions (kcal)j Overall 94 0.00 2354.8 94 0.00 1015.8 Breakfast at homea Snack at homea 94 0.00 679.0 94 0.00 1128.0 Lunch at homea 94 0.00 308.0 Snack at schoolb 94 0.00 790.0 Dinner at homea Unreported amount from omissions (kcal)k Overall 94 0.00 3093.4 94 0.00 856.0 Breakfasta 94 0.00 634.0 Morning meala 94 0.00 1193.5 Luncha
Table 2 e (continued ) School mealb Dinnera a b
37.5 43.2 46.4 33.0 14.9 50.1
26.6 34.5 45.4 31.7 17.7 39.9
36.7 30.2 45.5 41.8 28.3 37.9
22.1 34.3 45.1 27.7 30.8 39.9
57.6 57.9 54.1 66.9 79.6 52.1
29.4 36.8 46.1 33.9 38.1 41.4
48.6 39.7 90.8 83.7 37.4 33.0
40.3 62.8 171.7 126.3 64.8 52.8
961.37 223.8 88.6 297.7 176.6 225.3
548.8 181.5 127.9 195.8 132.7 226.5
72.6 26.0 4.7 27.3 9.3 11.3
156.1 92.6 36.8 87.6 35.5 48.7
70.5 9.9 5.2 43.7 22.7 1.7
162.5 41.3 33.3 118.5 47.4 12.9
711.3 121.0 106.4 229.3 53.6 119.3
539.1 181.3 159.3 247.5 79.9 182.2
759.8 188.7 88.2 155.2
653.7 204.3 129.3 206.6
c
d
e
f
g
h
i
j
k
n
Min.
Max.
Mean
SD
94 94
0.00 0.00
675.2 937.0
35.7 207.2
107.0 228.5
With parents reports as reference method. With observation as reference method. Omission rate ¼ (sum of weighted omissions/(sum of weighted omissions þ sum of weighted matches)) 100. Intrusion rate ¼ (sum of weighted intrusions/(sum of weighted intrusions þ sum of weighted matches)) 100. Correspondence rate ¼ (corresponding amount from matches/ reference amount) 100. Inflation ratio ¼ (over-reported amount from matches and intrusions/reference amount) 100. Corresponding amount from matches ¼ reported amount that is equal to the reference amount. Unreported amount from matches ¼ reference amount that exceeds the reported amount. Over-reported amount from matches ¼ reported amount that exceeds reference amount. Over-reported amount from intrusions ¼ entire reported amount for an intrusion. Unreported amount from omissions ¼ entire reference amount for omission.17
could be expected that many errors occur during lunch recall. Chocolate was the most often intruded food item for lunch. This may be an error in parents' and not children's reports because children are more prone to omit rather than overreport less healthful food items.12 Burrows et al. explained that parents e especially mothers e may not be accurate reporters of their children's food intake because they tend to report food that they think their children should consume.24 This may be the reason why parents in our study omitted to report that their children ate chocolate as a dessert for lunch. Inflation ratio was highest for the snack at home. The same food item (apple) was most often correctly and incorrectly reported for snack at home indicating that most errors in the snack at home reports occurred in reporting amounts. Differences in children's and parents' snack at home reports may be explained by the fact that children are not usually with their parents in the morning as parents are at work. Studies show that parents can be reliable reporters of their children's food intake in the home environment, but may not be reliable reporters of children's food intake outside the home or when they are not at home.6 Dinner at home was a meal with the greatest amounts of foods that were unreported. Bread was the most common omitted food item for this meal which means that most children's misreports for dinner were consequences of failing to report bread. Omission rate and intrusion rate were significantly higher for boys than for girls. This means that girls were more accurate reporters than boys. Lyng et al. also found that girls' self-reports were more accurate than boys'.25 In the study conducted by Baxter et al., girls were more accurate reporters of food intake when they were suggested to report meals in forward order.26 These results may be explained by differences between boys' and girls' brain function which results in fewer attention span problems and greater ability to fulfil assignments in girls than in boys.27
M, Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia, Please cite this article in press as: Sumonja S, Jevtic Public Health (2016), http://dx.doi.org/10.1016/j.puhe.2016.03.016
p u b l i c h e a l t h x x x ( 2 0 1 6 ) 1 e6
After adjusting P-values for multiple tests we did not find any significant differences in recall accuracy according to grade and BMI probably due to small sample size or because weight and height were reported by parents. There are several possible limitations of our study. First, we used parents' reports as the reference method to analyse children's recall accuracy of food intake at home. Parents were shown to be reliable reporters of their children's food intake in the home environment.6 When parents report their children's food intake they are also prone to different biases such as the need for social approval or simply memory lapses.7,24,28 Small sample size and lack of random selection of children for observation and reporting may be the reasons why the results cannot be generalised to the whole population of school-aged children in Serbia. Children were aware that they were observed while eating snacks at school so it could affect their recall accuracy. BMI was measured according to parents' reports and not by actual measurement which may affect results that are related to an analysis by BMI. This study analysed recall accuracy based on one day's food intake but the results may be different if it analysed recall accuracy for more days. In spite of these limitations, an important strength of our study is using observation as a generally accepted validation method in assessing children's dietary recall accuracy.8 We also used reporting error-sensitive variables to analyse children's reporting accuracy, which is recommended for studies analysing dietary reporting accuracy.12 This study gave some contribution to the current body of literature concerning dietary assessment methods in children. Giving insight into the differences of children's recall accuracy by each meal we showed that type of meal may also be an important factor affecting children's recall accuracy. This study also gave a contribution to clearing factors that affect children's dietary recall accuracy by analysing food items that are most often incorrectly reported. The questionnaire validated in this study should be improved to facilitate accuracy of reporting amounts. The questionnaire analysed in this study could be used for food intake assessment of 7e11 year-old children in Serbia after a few improvements.
from this study may guide public health practitioners in creating recommendations for dietary assessment as part of obesity prevention programmes.
Author statements Acknowledgements The authors appreciate all children, parents and school teachers who participated in our study.
Ethical approval School boards, parental and child written consent.
Funding None declared.
Competing interests None declared.
references ic K, Vasic M, Grozdanov J, Gudelj-Rakic J, Zivkovi 1. Boric c M, Jadovic Knez N, et al. Health of population of evic Sulovi c
2.
3. 4.
5.
Conclusions and recommendations In summary, the overall correspondence rate was higher than overall inflation ratio, omission and intrusion rates. When children reported a food item correctly they were fairly accurate in reporting food amounts. Intrusions were food items falsely reported eaten in large amounts while omissions were unreported food items eaten in smaller amounts. Snack at school had the greatest mean correspondence rate and snack at home had the highest mean inflation ratio. Overall omission and intrusion rates were significantly higher for boys than for girls. The results of this study may be useful in improving study designs and identifying potential biases in research involving dietary assessment in children. The results of our study may also have some applications to public health practitioners. As validated food intake instruments are very important in the process of monitoring obesity prevention programs, results
5
6. 7.
8.
9.
10.
11.
Serbia 2013. Belgrade, Serbia: Ministry of health of the Republic of Serbia; 2014. Currie C, Zanotti C, Morgan A, Currie D, de Looze M, Roberts C, et al. Health behaviour in school-aged children (HBSC) study: international report from the 2009/2010 survey. Social determinants of health and well-being among young people. Copenhagen: WHO Regional Office for Europe; 2012. Centers for Disease Control and Prevention. Youth risk behavior surveillance. MMWR Surveill Summ 2010;59. SS-5. Skinner JD, Carruth BR, Bounds W, Ziegler PJ. Children's food preferences-a longitudinal analysis. J Am Diet Assoc 2002;102:1638e47. He L, Zhai Y, Engelgau M, Li W, Qian H, Si X, et al. Association of children's eating behaviors with parental education, and teachers' health awareness, attitudes and behaviors: a national school-based survey in China. Eur J Public Health; 2013::1e8. Livingston MBE, Robson PJ. Measurement of dietary intake in children. Proc Nutr Soc 2000;59:279e93. Livingston MBE, Robson PJ, Wallace JMW. Issues in dietary intake assessment of children and adolescents. Br J Nutr 2004;92(Suppl. 2):213e22. Baxter SD. Cognitive processes in children's dietary recalls: insight from methodological studies. Eur J Clin Nutr 2009;63(Suppl. 1):S19e32. Sobo EJ, Rock CL, Neuhouser ML, Maciel TL, NeumarkSztainer D. Caretaker-child interaction during children's 24hour dietary recalls: who contributes what to the recall report? J Am Diet Assoc 2000;100:428e33. Hunsberger M, Pena P, Lissner L, Grafstrӧm L, Vanaelst B, Bӧrnhorst C, et al. Validity of self-reported lunch recalls in Swedish school children aged 6e8 years. Nutr J 2013;12:129. Baranowski T, Domel SB. A cognitive model of children's reporting of food intake. Am J Clin Nutr 1994;59(Suppl. 1):212e7.
M, Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia, Please cite this article in press as: Sumonja S, Jevtic Public Health (2016), http://dx.doi.org/10.1016/j.puhe.2016.03.016
6
p u b l i c h e a l t h x x x ( 2 0 1 6 ) 1 e6
12. Graham FM, Tapper K, Moore L, Murphy S. Cognitive, behavioral and social factors are associated with bias in dietary questionnaire self-reports by schoolchildren aged 9 to 11 years. J Am Diet Assoc 2008;108(11):1865e73. 13. Bӧrnhorst C, Huybrechts I, Ahrens W, Eiben G, Michels N, Pala V, et al. Prevalence and determinants of misreporting among European children in proxy-reported 24 h dietary recalls. Brit J Nutr 2013;109:1257e65. 14. Baxter SD, Hardin JW, Royer JA, Guinn CH, Smith AF. Children's recalls from five dietary reporting validation studies: intrusion in correctly reported and misreported options in school breakfast reports. Appetite 2008;51(3):489e500. 15. California Department of Public Health. Compendium of surveys for fruit and vegetable consumption and physical activity. California, United States: California Department of Public Health; 2011. 16. Edmunds LD, Ziebland S. Development and validation of the Day in the Life Questionnaire (DILQ) as a measure of fruit and vegetable questionnaire in 7e9 year olds. Health Educ Res 2002;17(2):211e20. 17. Baxter SD, Smith AF, Hardin JW, Nichols MD. Conclusions about children's reporting accuracy for energy and macronutrients over multiple interviews depend on the analytic approach for comparing reported information to reference information. J Am Diet Assoc 2007;107(4):585e604. B, Mirosavljev M. Food composition table. In: 18. Novakovic B, Mirosavljev M, editors. Nutrition hygiene. Novi Novakovic Sad, Serbia: University in Novi Sad; 2006. 19. Peacock J, Peacock P. Oxford handbook of medical statistics. Oxford, UK: Oxford University Press; 2011. 20. Baxter SD, Thompson WO. Accuracy by meal component of fourth-graders' school lunch recalls is less when obtained
21.
22.
23.
24.
25.
26.
27. 28.
during a 24-hour recall than as a single meal. Nutr Res 2002;22:679e84. Burrows TL, Martin RJ, Collins CE. A systematic review of the validity of dietary assessment methods in children when compared with the method of doubly labeled water. J Am Diet Assoc 2010;110:1501e10. Baxter SD, Thomson WO, Litaker MS, Frye FHA, Guinn CH. Low accuracy and low consistency of fourth-graders' school breakfast and school lunch recalls. J Am Diet Assoc 2002;102(3):386e95. Baxter SD, Smith AF, Litaker MS, Guinn CH, Nichols MN, Miller PH, et al. Body mass index, sex, interview protocol, and children's accuracy for reporting kilojoules observed eaten at school meals. J Am Diet Assoc 2006;106:1656e66. Burrows TL, Truby H, Morgan PJ, Callister R, Davies PSW, Collins CE. A comparison and validation of child versus parent reporting of children's energy intake using food frequency questionnaires versus food records: who's an accurate reporter? Clin Nutr 2013;32:613e8. Lyng N, Fagt S, Davidsen M, Hoppe C, Holstein B, Teten I. Reporting accuracy of packed lunch consumption among Danish 11-year-olds differ by gender. Food & Nutr Res 2013;57:19621. Baxter SD, Thompson WO, Smith AF, Litaker MS, Yin Z, Frye FH, et al. Reverse versus forward order reporting and the accuracy of fourth-graders' recalls of school breakfast and school lunch. Prev Med 2003;36(5):601e14. Gurian M, Stevens K. With boys and girls in mind. Educ Leadersh 2004;62(3):21e6. McGloin AF, Livingstone MB, Greene LC, Webb SE, Gibson JM, Jebb SA, et al. Energy and fat intake in obese and lean children at varying risk of obesity. Int J Obes Relat Metab Disord 2002;26(2):200e7.
M, Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia, Please cite this article in press as: Sumonja S, Jevtic Public Health (2016), http://dx.doi.org/10.1016/j.puhe.2016.03.016
Available online at www.sciencedirect.com
Public Health journal homepage: www.elsevier.com/puhe
Original Research
Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia a,* S. Sumonja , M. Jevtic b a b
University of Novi Sad, Faculty of Education, Department of Natural Sciences, Serbia University of Novi Sad, Faculty of Medicine, Department of Hygiene, Serbia
article info
abstract
Article history:
Objectives: Children's ability to recall and report dietary intake is affected by age and
Received 21 October 2014
cognitive skills. Dietary intake reporting accuracy in children is associated with age, weight
Received in revised form
status, cognitive, behavioural, social factors and dietary assessment techniques. This study
4 August 2015
analysed accuracy of 7e10 year-old children's reported food intake for one day.
Accepted 15 March 2016
Study design: Validation study.
Available online xxx
Methods: Sample included 94 children aged 7e10 years (median ¼ 9 years) from two
Keywords:
was a combination of 24-h recall and food recognition form. It included recalls for five
Food intake
meals: breakfast at home; snack at home; lunch at home; snack at school and dinner at
Children
home. Parental reports were used as reference information about children's food intake for
Questionnaires
meals obtained at home and observation was used to gain reference information for school
elementary schools in a local community in Serbia. ‘My meals for one day’ questionnaire
meal. Observed and reported amounts were used to calculate omission rate, intrusion rate, corresponding, over-reported and unreported amounts of energy, correspondence rate and inflation ratio. Results: Overall omission rate (37.5%) was higher than overall intrusion rate (36.7%). The same food item (bread) has been the most often correctly reported and omitted food item for breakfast, lunch and dinner. Snack at school had the greatest mean correspondence rate (79.6%) and snack at home the highest mean inflation ratio (90.7%). Conclusion: Most errors in children's recalls were incorrectly reported amounts and not the food items. The questionnaire should be improved to facilitate accurate reports of the amounts. © 2016 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
Introduction An increasing trend of prevalence of overweight and obese children is an important public health problem in Serbia as in
other countries in Europe and USA.1e3 Therefore, validated food intake assessment techniques in child population are important part of research about the determinants of food intake and obesity in children. Food intake assessment in children is a particularly challenging topic for researchers
ka 4, 25000 Sombor, Serbia. Tel.: þ381 638820270; fax: þ381 25416461. * Corresponding author. Podgoric E-mail address: [email protected] (S. Sumonja). http://dx.doi.org/10.1016/j.puhe.2016.03.016 0033-3506/© 2016 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved. M, Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia, Please cite this article in press as: Sumonja S, Jevtic Public Health (2016), http://dx.doi.org/10.1016/j.puhe.2016.03.016
2
p u b l i c h e a l t h x x x ( 2 0 1 6 ) 1 e6
because children's eating habits are variable and influenced by many different factors.4,5 The eating habits of school-aged children are under control of the parents while adolescents practice more out-of-home eating.6,7 Furthermore, children's ability to recall and report dietary intake is age-related and affected primarily by their cognitive skills.8 Previous research shows that children cannot report dietary intake without parental assistance until they reach seven years.7 The period from seven to 11 years presents transitional period in a child's ability to recall and report dietary intake without assistance. That may be one of the reasons why research, including 7e12 year-old children show different results about accuracy of child-reported food intake.9,10 Baranowski and Domel explained that an error in reporting dietary intake may occur during every stage of the process of remembering and recalling dietary information in children.11 Errors in dietary reports may be omitted food items (missing foods), over-reported food items (phantom foods) or incorrectly identifying food items.6 Factors that are associated with children's recall accuracy include age, weight status, cognitive, behavioural, social factors and dietary assessment techniques.8,12e14 The aim of our study was to analyse accuracy of 7e10 year-old children's reported food intake for one day according to parents' reports and observation.
Methods We conducted the study in a local urban community in northern Serbia. Participants were 1st to 4th grade students. In order to obtain a representative sample, we telephoned all five elementary schools in this community to participate in our study and two schools accepted. At each school, two classes from the first, second, third and fourth grades were randomly selected to take part in the study (184 children). After getting child assent and parental written consent total sample included 94 children (response rate 51.1%). Data collection that included completing the questionnaires by children and parents and direct observation occurred between March and May 2014. The questionnaire for children was classroom administered by one researcher in all classes during the last school class. The researcher used the same protocol for classroom questionnaire administration in all classes. Children were instructed by the researcher to recall types and amounts of foods they ate for breakfast, snack before school, lunch at home and snack at school on the day of the data collection. Children were further instructed by the researcher to recall types and amounts of food they ate for dinner the day before data collection. Researcher and teachers were present in the classroom while the children were filling the questionnaire in all classes. The researcher clarified food items that children were not able to recognise and provided no further help to children in completing the questionnaire. ‘My meals for one day’ questionnaire was developed for assessing children's food intake. Questionnaire development was based on a combination of two dietary assessment techniques, 24-h recall and food recognition form.15,16 It consisted of five parts representing five meals: breakfast at home; snack at home; lunch at home; snack at school and dinner at home. Each part included pictures of different food items from each food
group that are usually consumed for these meals and a portion size, which participants used to estimate the amounts of food consumed (e.g. glass, tablespoon, slice, piece). Part of the questionnaire presenting lunch at home was divided into soup, meat, vegetables, bread and pasta, salad, drink and dessert food items. There was an additional option to add food if it was not offered. Parental questionnaire consisted of the same parts as the child's questionnaire except it did not include pictures of food items. Parental questionnaire also included data about their child's height and weight, which were used to calculate children's body mass index (BMI). A consent form, the letter containing instructions for parents and the questionnaire were sent to parents via children on the same day that the children completed their questionnaires. We used direct observation of children eating school meal to obtain reference information about children's food intake for that meal. Observations occurred during a 30-min school meal break. All children from each class were observed whether they ate a school meal or food brought from home. Two researchers simultaneously observed children from one class. Each observer simultaneously observed approximately 12 children and recorded items and amounts that were in a child's plate before and after eating. Amounts were recorded in qualitative terms as amounts eaten per serving. Observers took a training session which was consisted of measuring and observing different amounts of food usually eaten for snack in schools and identifying serving sizes of different food items. In this training session, observers identified food and beverage items and amounts in 10 samples of packed lunches (30 food items) previously measured by the third researcher. Accuracy of reporting food items was 93.3% for the first observer and 94.3% for the second observer. Accuracy of reporting food amounts was 91.6% for the first observer and 92% for the second observer. Food items reported for snack at school were compared to reference information gained by observation of the school meal. Each reported item was classified as a match if it was reported and observed eaten, an intrusion if it was reported, but was not observed eaten and an omission if it was observed eaten, but was not reported. Food items reported for breakfast at home, snack at home, lunch at home and dinner at home were compared to food items reported by parents. Each item was classified as a match if it was reported by both a parent and a child, as an intrusion if it was reported by child but not reported by parent and omission if it was reported by a parent but was not reported by a child. Omission rate and intrusion rate were used to present number (%) of omitted or intruded food items.17 Observed and reported amounts were used to calculate corresponding, overreported and unreported amounts of energy.17 Corresponding, over-reported and unreported amounts of energy were then used to calculate correspondence rate and inflation ratio. Correspondence rate is a measure of reporting accuracy that is sensitive to reporting errors.17 Inflation ratio is a measure of reporting errors.17 Calculations of energy intake were based on a food database available in Serbian language.18 The independent t-test was used to analyse whether there are statistically significant differences in the omission rate, intrusion rate, correspondence rate and inflation ratio between boys
M, Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia, Please cite this article in press as: Sumonja S, Jevtic Public Health (2016), http://dx.doi.org/10.1016/j.puhe.2016.03.016
3
p u b l i c h e a l t h x x x ( 2 0 1 6 ) 1 e6
and girls.19 One-way ANOVA was used to investigate differences between the omission rate, intrusion rate, correspondence rate, and inflation ratio according to meal, grade and BMI.19 After adjustment in the P-values with Bonferroni correction for multiple tests, P-values < 0.017 were considered statistically significant.19
Results The sample characteristics are presented in Table 1. Bread has been the most often correctly reported (matched) food item for breakfast at home (40.4%), lunch at home (38.3%) and dinner at home (30.9%). Bread has also been the most often omitted food item for breakfast (25.5%), lunch (19.3%) and dinner (20.2%) at home. Apples have been the most often correctly reported food item for snack at home (13.8%), but were also the most often intruded (8.6%) and omitted (17.0%) food item for this meal. Cheese pie has been the most often correctly reported food item for snack at school (15.8%) and also the most often intruded (7.9%) food item for this meal. Doughnuts have been the most often omitted (5.3%) food item for snack at school. Milk has been the most often intruded food item for breakfast (5.3%), chocolate for lunch and pizza (7.4%) has been the most often intruded food item for dinner. Omission rates, intrusion rates and reporting errorsensitive variables overall and for each meal are presented in Table 2. Table 2 includes omission rate, intrusion rate, corresponding, over-reported and unreported amounts of energy, correspondence rates and inflation ratios overall and for each meal. Independent samples t-test found that boys (mean ¼ 49.5%) had a significantly higher omission rate than girls (mean ¼ 33.5) (mean difference ¼ 16.0; P ¼ 0.004). Intrusion rate was also significantly higher for boys (mean ¼ 60.8%) than for girls (mean ¼ 37.3%) (mean difference ¼ 23.5; P ¼ 0.017). We did not find any significant differences in omission rate, intrusion rate, correspondence rates and inflation ratio according to age and body mass index.
Discussion The overall omission rate was slightly higher than overall intrusion rate, which may indicate than children in our study
more often omitted to report food items than they intruded food items. This result is consistent with results of the study by Baxter and Thompson, although they used individual interviews to obtain dietary recall.20 Other studies show mixed results about the overall accuracy on the food item level, probably because of differences in methodology.21e23 Hunsberger et al. found 90% accuracy at the food item level, but that study analysed self-reported recalls for one meal.10 Baxter et al. found much lower accuracy and high total inaccuracy in a sample of fourth-grade children, but explored self-reported recalls based on a 24-h recall.23 Baxter et al. also showed that the accuracy of children's food intake recall is less when it is obtained as a part of a 24-h recall rather than as a single meal recall.20 It is expected that accuracy of children's food intake recalls decreases as the number of meals and overall food items increases probably because of information overload.7,11 A large percentage of incorrectly reported food items in our study could partially be explained by information overload.7,11 The result that the same food item (bread, apple, cheese pie) was the most often correctly reported and omitted or intruded food item for the same meal could indicate that most errors in children's recalls for those meals came from incorrectly reported amounts of those food items. This could be further analysed by variables described in Table 2. Reporting error-sensitive variables were used in this study as they were shown to be more sensitive in detecting reporting errors.17 Overall corresponding amount from matches is larger than unreported and over-reported amount from matches, but these variables indicate the extent of errors in reported amounts. Analysing the reporting of error-sensitive variable by meals, it is evident that snack at school was the meal with the greatest mean correspondence rate and lowest intrusion and omission rate. This is an expected result as the retention interval from snack at school and interview was much shorter than the retention interval between interview and other meals.23 We validated snack at school reports against observation which may also affect the mean correspondence rate for this meal. Lunch at home was the meal with the greatest amount that was reported both correctly and incorrectly (mean corresponding amount from matches, mean overreported and unreported amounts from matches, mean over-reported amount from intrusions). Lunch is a meal that usually takes the greatest part in daily energy intake and, consists of many different food items and mixed dishes, so it
Table 1 e Sample characteristics. Total % (n) Gender Boys Girls BMI Underweight Normal weight Overweight Obese Not reported Total
50 (47) 50 (47) 6.4 53.2 21.3 5.3 13.8 100
(6) (50) (20) (5) (13) (94)
1st grade
2nd grade
3rd grade
4th grade
7.4 (7) 6.4 (6)
9.6 (9) 11.7 (11)
13.8 (13) 12.8 (12)
19.1 (18) 19.1 (18)
1.1 11.7 4.2 2.1 2.1 21.3
0.0 18.1 4.2 2.1 2.1 26.5
3.2 (3) 14.9 (14) 12.7 (12) 0.0 (0) 7.4 (7) 38.6 (36)
2.1 8.5 0.0 1.1 2.1 13.8
(2) (8) (0) (1) (2) (13)
(1) (11) (4) (2) (2) (20)
(0) (17) (4) (2) (2) (25)
M, Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia, Please cite this article in press as: Sumonja S, Jevtic Public Health (2016), http://dx.doi.org/10.1016/j.puhe.2016.03.016
4
p u b l i c h e a l t h x x x ( 2 0 1 6 ) 1 e6
Table 2 e Omission rates, intrusion rates, correspondence rates, inflation ratio, corresponding, over-reported and unreported amounts of energy, overall and for each meal. n
Min.
Max.
Mean
SD
c
Omission rate (%) Overall 94 0.00 100.0 94 0.00 100.0 Breakfast at homea 94 0.00 100.0 Snack at homea Lunch at homea 94 0.00 100.0 94 0.00 100.0 Snack at schoolb 94 0.00 100.0 Dinner at homea Intrusion rate (%)d Overall 94 0.00 100.0 94 0.00 100.0 Breakfast at homea 94 0.00 100.0 Snack at homea Lunch at homea 94 0.00 100.0 94 0.00 100.0 Snack at schoolb 94 0.00 100.0 Dinner at homea Correspondence rate (%)e Overall 94 0.00 100.0 94 0.00 100.0 Breakfast at homea 94 0.00 100.0 Snack at homea 94 0.00 100.0 Lunch at homea 94 0.00 100.0 Snack at schoolb 94 0.00 100.0 Dinner at homea Inflation ratio (%)f Overall 94 0.00 226.5 94 0.00 281.5 Breakfast at homea 94 0.00 975.0 Snack at homea 94 0.00 679.8 Lunch at homea Snack at schoolb 94 0.00 346.8 94 0.00 219.9 Dinner at homea Corresponding amount from matches (kcal)g Overall 94 0.00 2938.4 94 0.00 931.0 Breakfast at homea 94 0.00 528.0 Snack at homea 94 0.00 962.0 Lunch at homea 94 0.00 437.0 Snack at schoolb Dinner at homea 94 0.00 902.0 Unreported amount from matches (kcal)h Overall 94 0.00 988.6 94 0.00 723.6 Breakfast at homea 94 0.00 352.0 Snack at homea 94 0.00 681.0 Lunch at homea 94 0.00 154.0 Snack at schoolb Dinner at homea 94 0.00 390.0 Over-reported amount from matches (kcal)i Overall 94 0.00 1011.0 94 0.00 305.0 Breakfast at homea Snack at homea 94 0.00 300.0 94 0.00 812.0 Lunch at homea 94 0.00 195.0 Snack at schoolb Dinner at homea 94 0.00 117.0 Over-reported amount from intrusions (kcal)j Overall 94 0.00 2354.8 94 0.00 1015.8 Breakfast at homea Snack at homea 94 0.00 679.0 94 0.00 1128.0 Lunch at homea 94 0.00 308.0 Snack at schoolb 94 0.00 790.0 Dinner at homea Unreported amount from omissions (kcal)k Overall 94 0.00 3093.4 94 0.00 856.0 Breakfasta 94 0.00 634.0 Morning meala 94 0.00 1193.5 Luncha
Table 2 e (continued ) School mealb Dinnera a b
37.5 43.2 46.4 33.0 14.9 50.1
26.6 34.5 45.4 31.7 17.7 39.9
36.7 30.2 45.5 41.8 28.3 37.9
22.1 34.3 45.1 27.7 30.8 39.9
57.6 57.9 54.1 66.9 79.6 52.1
29.4 36.8 46.1 33.9 38.1 41.4
48.6 39.7 90.8 83.7 37.4 33.0
40.3 62.8 171.7 126.3 64.8 52.8
961.37 223.8 88.6 297.7 176.6 225.3
548.8 181.5 127.9 195.8 132.7 226.5
72.6 26.0 4.7 27.3 9.3 11.3
156.1 92.6 36.8 87.6 35.5 48.7
70.5 9.9 5.2 43.7 22.7 1.7
162.5 41.3 33.3 118.5 47.4 12.9
711.3 121.0 106.4 229.3 53.6 119.3
539.1 181.3 159.3 247.5 79.9 182.2
759.8 188.7 88.2 155.2
653.7 204.3 129.3 206.6
c
d
e
f
g
h
i
j
k
n
Min.
Max.
Mean
SD
94 94
0.00 0.00
675.2 937.0
35.7 207.2
107.0 228.5
With parents reports as reference method. With observation as reference method. Omission rate ¼ (sum of weighted omissions/(sum of weighted omissions þ sum of weighted matches)) 100. Intrusion rate ¼ (sum of weighted intrusions/(sum of weighted intrusions þ sum of weighted matches)) 100. Correspondence rate ¼ (corresponding amount from matches/ reference amount) 100. Inflation ratio ¼ (over-reported amount from matches and intrusions/reference amount) 100. Corresponding amount from matches ¼ reported amount that is equal to the reference amount. Unreported amount from matches ¼ reference amount that exceeds the reported amount. Over-reported amount from matches ¼ reported amount that exceeds reference amount. Over-reported amount from intrusions ¼ entire reported amount for an intrusion. Unreported amount from omissions ¼ entire reference amount for omission.17
could be expected that many errors occur during lunch recall. Chocolate was the most often intruded food item for lunch. This may be an error in parents' and not children's reports because children are more prone to omit rather than overreport less healthful food items.12 Burrows et al. explained that parents e especially mothers e may not be accurate reporters of their children's food intake because they tend to report food that they think their children should consume.24 This may be the reason why parents in our study omitted to report that their children ate chocolate as a dessert for lunch. Inflation ratio was highest for the snack at home. The same food item (apple) was most often correctly and incorrectly reported for snack at home indicating that most errors in the snack at home reports occurred in reporting amounts. Differences in children's and parents' snack at home reports may be explained by the fact that children are not usually with their parents in the morning as parents are at work. Studies show that parents can be reliable reporters of their children's food intake in the home environment, but may not be reliable reporters of children's food intake outside the home or when they are not at home.6 Dinner at home was a meal with the greatest amounts of foods that were unreported. Bread was the most common omitted food item for this meal which means that most children's misreports for dinner were consequences of failing to report bread. Omission rate and intrusion rate were significantly higher for boys than for girls. This means that girls were more accurate reporters than boys. Lyng et al. also found that girls' self-reports were more accurate than boys'.25 In the study conducted by Baxter et al., girls were more accurate reporters of food intake when they were suggested to report meals in forward order.26 These results may be explained by differences between boys' and girls' brain function which results in fewer attention span problems and greater ability to fulfil assignments in girls than in boys.27
M, Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia, Please cite this article in press as: Sumonja S, Jevtic Public Health (2016), http://dx.doi.org/10.1016/j.puhe.2016.03.016
p u b l i c h e a l t h x x x ( 2 0 1 6 ) 1 e6
After adjusting P-values for multiple tests we did not find any significant differences in recall accuracy according to grade and BMI probably due to small sample size or because weight and height were reported by parents. There are several possible limitations of our study. First, we used parents' reports as the reference method to analyse children's recall accuracy of food intake at home. Parents were shown to be reliable reporters of their children's food intake in the home environment.6 When parents report their children's food intake they are also prone to different biases such as the need for social approval or simply memory lapses.7,24,28 Small sample size and lack of random selection of children for observation and reporting may be the reasons why the results cannot be generalised to the whole population of school-aged children in Serbia. Children were aware that they were observed while eating snacks at school so it could affect their recall accuracy. BMI was measured according to parents' reports and not by actual measurement which may affect results that are related to an analysis by BMI. This study analysed recall accuracy based on one day's food intake but the results may be different if it analysed recall accuracy for more days. In spite of these limitations, an important strength of our study is using observation as a generally accepted validation method in assessing children's dietary recall accuracy.8 We also used reporting error-sensitive variables to analyse children's reporting accuracy, which is recommended for studies analysing dietary reporting accuracy.12 This study gave some contribution to the current body of literature concerning dietary assessment methods in children. Giving insight into the differences of children's recall accuracy by each meal we showed that type of meal may also be an important factor affecting children's recall accuracy. This study also gave a contribution to clearing factors that affect children's dietary recall accuracy by analysing food items that are most often incorrectly reported. The questionnaire validated in this study should be improved to facilitate accuracy of reporting amounts. The questionnaire analysed in this study could be used for food intake assessment of 7e11 year-old children in Serbia after a few improvements.
from this study may guide public health practitioners in creating recommendations for dietary assessment as part of obesity prevention programmes.
Author statements Acknowledgements The authors appreciate all children, parents and school teachers who participated in our study.
Ethical approval School boards, parental and child written consent.
Funding None declared.
Competing interests None declared.
references ic K, Vasic M, Grozdanov J, Gudelj-Rakic J, Zivkovi 1. Boric c M, Jadovic Knez N, et al. Health of population of evic Sulovi c
2.
3. 4.
5.
Conclusions and recommendations In summary, the overall correspondence rate was higher than overall inflation ratio, omission and intrusion rates. When children reported a food item correctly they were fairly accurate in reporting food amounts. Intrusions were food items falsely reported eaten in large amounts while omissions were unreported food items eaten in smaller amounts. Snack at school had the greatest mean correspondence rate and snack at home had the highest mean inflation ratio. Overall omission and intrusion rates were significantly higher for boys than for girls. The results of this study may be useful in improving study designs and identifying potential biases in research involving dietary assessment in children. The results of our study may also have some applications to public health practitioners. As validated food intake instruments are very important in the process of monitoring obesity prevention programs, results
5
6. 7.
8.
9.
10.
11.
Serbia 2013. Belgrade, Serbia: Ministry of health of the Republic of Serbia; 2014. Currie C, Zanotti C, Morgan A, Currie D, de Looze M, Roberts C, et al. Health behaviour in school-aged children (HBSC) study: international report from the 2009/2010 survey. Social determinants of health and well-being among young people. Copenhagen: WHO Regional Office for Europe; 2012. Centers for Disease Control and Prevention. Youth risk behavior surveillance. MMWR Surveill Summ 2010;59. SS-5. Skinner JD, Carruth BR, Bounds W, Ziegler PJ. Children's food preferences-a longitudinal analysis. J Am Diet Assoc 2002;102:1638e47. He L, Zhai Y, Engelgau M, Li W, Qian H, Si X, et al. Association of children's eating behaviors with parental education, and teachers' health awareness, attitudes and behaviors: a national school-based survey in China. Eur J Public Health; 2013::1e8. Livingston MBE, Robson PJ. Measurement of dietary intake in children. Proc Nutr Soc 2000;59:279e93. Livingston MBE, Robson PJ, Wallace JMW. Issues in dietary intake assessment of children and adolescents. Br J Nutr 2004;92(Suppl. 2):213e22. Baxter SD. Cognitive processes in children's dietary recalls: insight from methodological studies. Eur J Clin Nutr 2009;63(Suppl. 1):S19e32. Sobo EJ, Rock CL, Neuhouser ML, Maciel TL, NeumarkSztainer D. Caretaker-child interaction during children's 24hour dietary recalls: who contributes what to the recall report? J Am Diet Assoc 2000;100:428e33. Hunsberger M, Pena P, Lissner L, Grafstrӧm L, Vanaelst B, Bӧrnhorst C, et al. Validity of self-reported lunch recalls in Swedish school children aged 6e8 years. Nutr J 2013;12:129. Baranowski T, Domel SB. A cognitive model of children's reporting of food intake. Am J Clin Nutr 1994;59(Suppl. 1):212e7.
M, Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia, Please cite this article in press as: Sumonja S, Jevtic Public Health (2016), http://dx.doi.org/10.1016/j.puhe.2016.03.016
6
p u b l i c h e a l t h x x x ( 2 0 1 6 ) 1 e6
12. Graham FM, Tapper K, Moore L, Murphy S. Cognitive, behavioral and social factors are associated with bias in dietary questionnaire self-reports by schoolchildren aged 9 to 11 years. J Am Diet Assoc 2008;108(11):1865e73. 13. Bӧrnhorst C, Huybrechts I, Ahrens W, Eiben G, Michels N, Pala V, et al. Prevalence and determinants of misreporting among European children in proxy-reported 24 h dietary recalls. Brit J Nutr 2013;109:1257e65. 14. Baxter SD, Hardin JW, Royer JA, Guinn CH, Smith AF. Children's recalls from five dietary reporting validation studies: intrusion in correctly reported and misreported options in school breakfast reports. Appetite 2008;51(3):489e500. 15. California Department of Public Health. Compendium of surveys for fruit and vegetable consumption and physical activity. California, United States: California Department of Public Health; 2011. 16. Edmunds LD, Ziebland S. Development and validation of the Day in the Life Questionnaire (DILQ) as a measure of fruit and vegetable questionnaire in 7e9 year olds. Health Educ Res 2002;17(2):211e20. 17. Baxter SD, Smith AF, Hardin JW, Nichols MD. Conclusions about children's reporting accuracy for energy and macronutrients over multiple interviews depend on the analytic approach for comparing reported information to reference information. J Am Diet Assoc 2007;107(4):585e604. B, Mirosavljev M. Food composition table. In: 18. Novakovic B, Mirosavljev M, editors. Nutrition hygiene. Novi Novakovic Sad, Serbia: University in Novi Sad; 2006. 19. Peacock J, Peacock P. Oxford handbook of medical statistics. Oxford, UK: Oxford University Press; 2011. 20. Baxter SD, Thompson WO. Accuracy by meal component of fourth-graders' school lunch recalls is less when obtained
21.
22.
23.
24.
25.
26.
27. 28.
during a 24-hour recall than as a single meal. Nutr Res 2002;22:679e84. Burrows TL, Martin RJ, Collins CE. A systematic review of the validity of dietary assessment methods in children when compared with the method of doubly labeled water. J Am Diet Assoc 2010;110:1501e10. Baxter SD, Thomson WO, Litaker MS, Frye FHA, Guinn CH. Low accuracy and low consistency of fourth-graders' school breakfast and school lunch recalls. J Am Diet Assoc 2002;102(3):386e95. Baxter SD, Smith AF, Litaker MS, Guinn CH, Nichols MN, Miller PH, et al. Body mass index, sex, interview protocol, and children's accuracy for reporting kilojoules observed eaten at school meals. J Am Diet Assoc 2006;106:1656e66. Burrows TL, Truby H, Morgan PJ, Callister R, Davies PSW, Collins CE. A comparison and validation of child versus parent reporting of children's energy intake using food frequency questionnaires versus food records: who's an accurate reporter? Clin Nutr 2013;32:613e8. Lyng N, Fagt S, Davidsen M, Hoppe C, Holstein B, Teten I. Reporting accuracy of packed lunch consumption among Danish 11-year-olds differ by gender. Food & Nutr Res 2013;57:19621. Baxter SD, Thompson WO, Smith AF, Litaker MS, Yin Z, Frye FH, et al. Reverse versus forward order reporting and the accuracy of fourth-graders' recalls of school breakfast and school lunch. Prev Med 2003;36(5):601e14. Gurian M, Stevens K. With boys and girls in mind. Educ Leadersh 2004;62(3):21e6. McGloin AF, Livingstone MB, Greene LC, Webb SE, Gibson JM, Jebb SA, et al. Energy and fat intake in obese and lean children at varying risk of obesity. Int J Obes Relat Metab Disord 2002;26(2):200e7.
M, Accuracy of reported food intake in a sample of 7e10 year-old children in Serbia, Please cite this article in press as: Sumonja S, Jevtic Public Health (2016), http://dx.doi.org/10.1016/j.puhe.2016.03.016