Journal of Human Nutrition and Dietetics
RESEARCH PAPER Validation of a web-based questionnaire to assess the dietary intake of Brazilian children aged 7–10 years V. F. Davies,1 E. Kupek,1 M. A. de Assis,2 S. Natal,1 P. F. Di Pietro2 & T. Baranowski3 1
Graduate Program in Public Health, Center for Health Sciences at the Federal University of Santa Catarina, Florianopolis, SC, Brazil Graduate Program in Nutrition, Center for Health Sciences at the Federal University of Santa Catarina, Florianopolis, SC, Brazil 3 USDA/ARS Children’s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, TX, USA 2
Keywords intrusions, matches, omissions, online, questionnaire. Correspondence V. F. Davies, Graduate Program in Public Health, Center for Health Sciences at the Federal University of Santa Catarina, Campus Reitor Jo~ao David Ferreira Lima, Rua Delfino Conti, S/N. Bloco H, Cep 88040-370, Florianopolis, SC, Brazil. Tel.: +55 48 3721 9847 Fax: +55 048 3721 9542 E-mail: [email protected]
How to cite this article Davies V.F., Kupek E., De Assis M.A., Natal S., Di Pietro P.F. & Baranowski T. (2015) Validation of a web-based questionnaire to assess the dietary intake of Brazilian children aged 7–10 years. J Hum Nutr Diet. 28 (Suppl. 1), 93–102 doi:10.1111/jhn.12262
Abstract Background: The Food Intake and Physical Activity of School Children (CAAFE) comprises an online questionnaire to self-report diet and physical activity of Brazilian schoolchildren. Background: The present study aimed to assess the validity (matches, omissions and intrusions) and moderating factors of the CAAFE. Methods: Direct observation was made of foods consumed (five public schools) and child self-reporting on the CAAFE. Additional data included school grade, gender, body mass index, completion of food diary, socioeconomic status and access to computer. Data were analysed using regression. Results: In total, 602 children participated in the study [mean (SD) age 9.5 (1.24) years; 53.6% boys]. On average, there were 43% matches, 29% intrusions and 28% omissions. Matches doubled in third grade compared to the second grade (P = 0.004); matches almost tripled for afternoon snack compared to morning snack (P < 0.001); and matches were 69% higher for children with access to a computer at home (P < 0.01). Intrusions decreased by almost one-half in fifth compared to fourth grades (P = 0.004). Omissions declined significantly in the third and fourth grades but increased in the fifth grade. Omissions were 47% lower for children in the highest income and lower among children who completed the food diary. No differences were found for gender or body mass index. Conclusions: Children older than 8 years old, who owned a computer and completed a food diary, performed better in the CAAFE. A high incidence of disagreement was found in relation to the schools and the type of meal. Overall matches (43%), intrusions (29%) and omissions (28%) indicate that further studies are required to improve the validity of the CAAFE.
Introduction Choosing the best method for assessing food consumption includes issues related to characteristics of the individuals investigated (i.e. age group), the objective of the assessment and the available resources (preferably easy to apply and at a low cost), as well as validity and precision (Burrows et al., 2010; Collins et al., 2010). Assessing children’s dietary intake has proved to be an important challenge in nutritional epidemiology. The most commonly used methods for assessing food intake ª 2014 The British Dietetic Association Ltd.
require skills related to memory and attention span, as well as a knowledge of foods and preparations (Livingstone et al., 2004; Ortiz-Andrellucchi et al., 2009; Baxter et al., 2010). Another factor that may influence dietary assessment of children is motivation. Extensive questionnaires that seek to obtain very detailed information are not recommended for children because they can lead to boredom, fatigue and possibly decreased compliance (Magarey et al., 2011; Lu et al., 2012). The use of computers in the assessment of children’s food intake may be a promising alternative to more 93
Web-based dietary questionnaire for children
traditional methods (van Gelder et al., 2010; Illner et al., 2012; Adamson & Baranowski, 2014). The advantages include the use of images and interactive features that attract a child’s attention, as well as allowing researchers to standardise the questionnaires and to conduct an immediate evaluation of the data (Kohlmeier et al., 1997; Ngo et al., 2009; Lu et al., 2012). Several computerised child diet assessment instruments have been developed around the world; for example, SCRAN24, SNAP, Web DASC, CANAA-W and ASA24 – Kids (Biltoft-Jensen et al., 2012; Adamson & Baranowski, 2014; Foster et al., 2014; Moore et al., 2014; Vereecken et al., 2014). The Food Intake and Physical Activity of School Children – CAAFE (Universidade Federal de Santa Catarina, 2013) – comprises an online questionnaire that has been developed to enable the self-reporting of diet and physical activity by Brazilian schoolchildren aged 7–10 years. To create an efficient child-friendly questionnaire, the following steps were initiated: analysis of similar instruments; focus groups with physical education teachers (da Costa et al., 2012) and nutritionists (Davies et al., 2014); meetings of researchers, teachers and experts in web design; analysis of 7-day food and physical activity records; and usability testing (da Costa et al., 2013). In the food consumption section of the questionnaire, the resulting instrument asks the child about their food consumption during the previous 24 h, and quantifies the consumption of priority food groups for the 7–10-year-old age group (e.g. milk and milk products), as well as foods related to obesity (e.g. soft drinks). The present study reports the first validation of the food consumption section of the CAAFE, by identifying the agreement and errors between the children’s answers in the questionnaire and observation of their school meals. It goes beyond other validation studies of online dietary assessment tools by assessing the possible moderating influences of a number of variables, such as socioeconomic status, access to digital technology and the use of a food diary. Materials and methods In the present study, we describe the validation process of the food consumption section of the CAAFE questionnaire. The results for the validity of the physical activity section will be described elsewhere. The validation process included five public schools in the city of Florian opolis, the capital of Santa Catarina state, Brazil. The schools were selected by the Municipal Education Department according to the availability of an adequate computer room (number of computers and working internet connection). Schools were selected (n = 5) in different regions of the city (Central, North, South, and East) to 94
V. F. Davies et al.
represent students from different socioeconomic backgrounds. Six classes from second to fifth grades were selected by the principal of each school (30 classes in total). The classes were selected by the principal based on the class schedule for the week. For example, classes that had already planned field trips or other activities during the research period that could not be rescheduled were not selected to join the study. Children in Brazil who attend second to fifth grade are expected to meet the following age ranges: second grade: 7–8 years old; third grade: 8–9 years old; fourth grade: 9–10 years old; fifth grade: 10–11 years old. In total, 708 students were invited to participate in the study. The children gave their oral consent and the parents provided their written informed consent. No financial reward was offered. The project was approved by the Ethics Committee on Human Research at the Federal University of Santa Catarina (UFSC) (protocol no. 2250/11). Parents completed a questionnaire with sociodemographic information, as well as information regarding the child’s access to a computer at home. For the children who agreed to participate, anthropometric measurements were recorded. The measurements of the children’s body weight and height were taken by trained researchers in the schools themselves, using standard techniques (Lohman et al., 1991). Portable digital scales (with a capacity of 180 kg) were used (Marte LTDA, Belo Horizonte, Brazil). Height was measured using a metal stadiometer (Seca, Hamburg, Germany). The children were weighed and measured barefoot, wearing light clothing. Bofy mass index (BMI) was computed as weight (kg) divided by height squared (m2). In the week before the research, researchers distributed to each student a 7-day-food diary similar to that of the online survey (CAAFE). In the classroom, the children were given instructions on how to complete the diary. The purpose of the diary was to familiarise the children with the images of the foods in the CAAFE, and also as a resource to improve the accuracy of answers when completing the online questionnaire. The children were instructed to consult the diary when completing the online questionnaire. Food diaries included the foods consumed on the day that their food consumption was observed. The CAAFE is a self-reported questionnaire which examines food consumption and physical activity during the previous day (24-h recall). The food consumption section of the CAAFE is divided into six meals (breakfast, morning snack, lunch, afternoon snack, dinner and supper). For each meal, 32 images of food or drinks are presented on the computer screen so that the child can make their own selections. The images of foods and food ª 2014 The British Dietetic Association Ltd.
V. F. Davies et al.
groups were chosen to take into account: the food patterns of children in this age group, the food usually offered on school menus, the suggestions given by focus groups with dietitians (Davies et al., 2014), a 7-day food diary (n = 180 children) and the most consumed foods according to the Brazilian Household Budget Survey (IBGE, 2010a). An avatar in the form of a robot guides the child when completing the questionnaire, explaining the concept of each meal and the time of day at which it is consumed, as well as reinforcing the importance of reporting food consumption from the previous 24 h. The avatar also instructs the child not to click on any food if it was not consumed during the preceding day. To validate the food section of the CAAFE, direct observation was made of the meals in the school environment, including meals that were prepared on the school premises, those bought in the school canteen, or food brought from home. The observers were nutrition students from two large universities in the city where the research was conducted, who, during their classes, were invited by the research coordinator to join the study. To ensure quality, observers took part in both conceptual and practical training sessions. The conceptual session included the presentation of the observation protocol and instructions for completing the worksheet to be used during data collection. The practice phase took place in the different schools at which data were collected. The training consisted of six practice sessions, during which each observer had to observe at least 30 children in total. Agreement on identified foods was 96% between the more experienced observers and the other observers during the training. In addition, because the CAAFE required adjustments before initiating the validation, pilot studies were conducted in three schools. These studies were used as an opportunity to further increase the experience of the observers and thus to improve the quality of the data collected. During data collection, each observer was assigned to observe a maximum of five children, who in turn were identified with name tags and colored ribbons on their arms. The observation took place the day before the computer questionnaire was applied. Some of the observers were in the dining hall and some were in the schoolyard. Observers took note of the types of foods and beverages consumed by children, as well as recording foods that were shared. The day after food consumption was observed, the children were conducted to the school computer room to complete the CAAFE. Prior to that, however, a trained researcher gave demonstration of how to complete the computer questionnaire in each classroom. With the aid of a banner (140 9 105 cm), images of food from the CAAFE were shown, as well as examples of how to answer the questionnaire. Subsequently, the children were ª 2014 The British Dietetic Association Ltd.
Web-based dietary questionnaire for children
instructed to take the previously completed food diaries to the computer room, in case they wished to consult their records. Once in the computer room, researchers were present to assist the children with any questions, and to record the questions raised by the children, when completing the CAAFE. Once the questionnaire was completed, the children were individually interviewed with regard to three questions: the presence of a computer at home, the presence of an Internet connection at home, and the main activity carried out on their home computer. In the present study, matches occurred when foods and/or beverages observed to have been consumed by the children were reported in the CAAFE. Omissions occurred when foods and/or beverages observed to have been consumed by the children were not reported in the CAAFE. Intrusions occurred when foods and/or beverages were reported in the CAAFE but were not observed to have been consumed by the children. Match percentages were calculated: percentage matches = (the number of matches/total food items directly observed at school) 9 100. Intrusion percentages were calculated: percentage intrusions = (total number of intrusions/total food items observed directly in school) 9 100. Omission percentages were calculated: percentage omissions = (the number of omissions/total food items directly observed in school) 9 100. Agreement rates were calculated using Poisson regression for the number of matches, intrusions and omissions. This method of analysis was chosen rather than other standard agreement analyses (i.e. kappa) because Poisson regression allows one to distinguish between two types of mismatch: intrusions and omissions. These are firmly established as distinct types of error, with the former more related to the food environment (e.g. the number of foods offered) and the latter to the difficulties children have to recall the food consumed on the previous day. In addition, rates of agreement calculated by Poisson regression provide the difference between various levels of an independent variable adjusted for all the other independent variables, whereas agreement indexes such as kappa do not (Agresti, 2007). Therefore, regression-type impact measures such as percentage rate difference were used throughout the present study. Poisson regression was used to explain variations in the number of matches, intrusions and omissions with regard to the following independent variables: child gender, school grade, type/location of school, type of meal observed (morning snack, lunch, afternoon snack), children’s BMI (quintiles), parents’ education level, parents’ annual income, access to a computer and Internet at home, number of computers at home, type of activities carried out by children on their computers (research/ 95
V. F. Davies et al.
Web-based dietary questionnaire for children
social media/games), presence of a computer in the child’s room, frequency of computer use, and completion of the food diary. P < 0.05 was considered statistically significant. Analyses were performed using STATA, version 12.0 (StataCorp, College Station, TX, USA). Results This validation study involved 602 children [mean (SD) age 9.5 (1.2) years, 53.6% boys]. Children (n = 106) not included in the final sample were either absent from school during the observation of the meals and/or the application of the CAAFE in the computer room. In total, 16 observers were trained to observe the school meals and participated in the research. Tables 1–3 display percentages of matches, intrusions and omissions in the CAAFE reports according to the characteristics of the study sample. The average percentages of matches, intrusions and omissions were 43%, 29% and 28%, respectively (Table 1). Sociodemographic characteristics of the sample showed that most the children belonged to the second ($3620 to $7236) and third category ($7237 to $10 856) with regard to their parents’
income, which is above the Brazilian national income average (IBGE, 2012). Most parents did not have a college degree, which reflects the Brazilian average education level (IBGE, 2010b). From the third grade onwards, the accuracy (matches) of the CAAFE reports almost doubled (Fig. 1 and Table 4). Intrusions approximately halved in fifth grade compared to fourth grade (Fig. 1 and Table 4). Food omissions were reduced by one-third in third grade compared to second grade, followed by an increase in fifth grade (Fig. 1 and Table 4). Agreement was almost threefold higher when reporting an afternoon snack compared to a morning snack, and 69% higher for children with a computer at home compared to those without. Intrusions were 23% lower for lunch compared to the morning snack reports; 25% lower among children who did not complete the food diary compared to those who did; and 47% fewer for children whose parents had a higher income. No differences were found for gender, BMI quintiles or variables related to digital technology (Table 4). Despite considerable variation in reporting accuracy across schools, only a few reached statistical significance: compared to the reference school, School 5 presented
Table 1 Percentages of matches, intrusions and omissions in the Food Intake and Physical Activity of School Children (CAAFE) reports according to the sociodemographic characteristics of the study sample Variable
Category
n (%)
Matches (%)
Intrusions (%)
Omissions (%)
Child gender
Male Female 2 (mean age 7.8 years) 3 (mean age 8.8 years) 4 (mean age 9.9 years) 5 (mean age 10.9 years) 1 (mean 14.6 kg m–2) 2 (mean 15.9 kg m–2) 3 (mean 17.2 kg m–2) 4 (mean 19.2 kg m–2) 5 (mean 23.9 kg m–2) 1 (≤$3619) 2 ($3620 to $7236) 3 ($7237 to $10 856) 4 ($10 857 to $18 056) 5 ($18 057 to $28 947) 1 (Elementary school not completed) 2 (Elementary school) 3 (High School not completed) 4 (High School) 5 (College not completed) 6 (College) 1 2 3 4 5
279 323 113 140 190 159 119 116 120 115 113 48 250 110 4 23 115 61 73 136 40 52 146 141 100 83 132
42.76 41.83 25.39 47.76 48.90 41.48 43.41 44.90 40.71 43.18 38.77 43.55 40.50 45.02 47.80 61.23 41.96 43.51 39.61 44.52 51.30 47.02 47.03 25.51 28.85 31.80 35.32
30.02 28.60 37.71 26.27 31.57 23.12 28.46 32.73 29.36 28.34 27.99 31.99 28.50 28.00 33.02 21.05 27.27 29.34 29.95 28.04 30.72 29.68 21.62 53.08 37.69 33.47 35.32
27.22 29.57 36.90 25.97 19.54 35.40 28.13 22.37 29.92 28.47 33.24 24.46 31.00 26.98 19.17 17.72 30.76 27.14 30.43 27.44 17.98 23.30 31.35 21.41 33.46 34.73 29.36
School grade
Body mass index (quintiles)
Parents’ annual income (US dollars)
Parents’ education level
School
96
(46.3) (53.6) (18.7) (23.2) (31.5) (26.4) (19.7) (19.2) (19.9) (19.1) (18.7) (7.9) (41.5) (18.2) (7.48) (3.8) (19.1) (10.1) (12.1) (22.9) (6.6) (8.6)
ª 2014 The British Dietetic Association Ltd.
V. F. Davies et al.
Web-based dietary questionnaire for children
Table 2 Percentages of matches, intrusions and omissions in the Food Intake and Physical Activity of School Children (CAAFE) reports by type of meal observed at school and by completion of food diary Variable
Category
n (%)
Matches (%)
Intrusions (%)
Omissions (%)
Meal observed
Morning snack Snack + Lunch Afternoon snack Yes No
213 289 100 442 160
38.83 44.25 43.79 24.13 38.31
32.75 26.07 31.02 51.51 31.84
28.40 29.67 25.19 24.36 31.28
Food diary completed
(35.3) (48.0) (16.6) (73.4) (26.5)
Table 3 Percentages of matches, intrusions and omissions in the Food Intake and Physical Activity of School Children (CAAFE) reports by access to digital technology Variable
Category
n (%)
Matches (%)
Intrusions (%)
Omissions (%)
Computer at home*
No Yes No Yes Research Social media Games 0 1 2 3 No Yes No Sometimes Yes
125 429 174 362 99 312 136 99 275 102 23 396 101 44 165 289
35.56 44.51 37.64 44.66 43.31 39.40 45.45 41.98 45.68 38.90 42.69 43.87 42.19 40.12 42.27 44.38
37.03 27.08 35.16 26.62 28.84 32.36 24.18 31.28 27.07 28.92 27.64 29.25 24.45 29.59 27.09 28.91
27.41 28.41 27.20 28.72 27.85 28.23 30.36 26.73 27.25 32.18 29.67 26.87 33.36 30.29 30.64 26.70
Access to Internet at home* What do the children do when they have access to computer*
Numbers of computer at home†
Computer at children’s bedroom† Children access to computer at home†
(22.5) (77.4) (32.4) (67.5) (16.4) (51.8) (22.5) (19.8) (55.1) (20.4) (4.6) (79.6) (20.3) (8.8) (33.1) (58.0)
*Children’s answer. Parents’ answer.
†
62% lower matching rate and 70% higher intrusion rate; and School 2 presented over twice the rate of intrusions and 38% lower rate of omissions. The children who did not complete the food diary had 33% more omissions than those who did complete the diary (Table 4). The mean number of omissions was high in second grade, decreased considerably in third and fourth grades, only to peak again in fifth grade, when the children consumed more food overall compared to other grades (Fig. 2). Discussion The percentages of matches (43%), intrusions (29%) and omissions (21%) found in the present study are similar to those in previously published studies (Baranowski et al., 2002, 2012; Biltoft-Jensen et al., 2013), although it should be emphasised that those previous studies used different methods to estimate accuracy during their data analyses. Another difference is that the CAAFE study ª 2014 The British Dietetic Association Ltd.
included children from the age of 7 years, whereas the minimum age in the other studies was 8 years (Baranowski et al., 2012; Biltoft-Jensen et al., 2013) and 10 years (Baranowski et al., 2002). By contrast to these other studies, we also gave our students instruction on the use of the CAAFE, and asked them to keep a diary on the day of observation, both of which may have enhanced the accuracy of the younger children’s answers. Children older than 8 years performed better in the CAAFE (Table 1). Other studies also found that the age of the child influences the accuracy of dietary assessment questionnaires (Lanctot et al., 2008; De Assis et al., 2009; Burrows et al., 2010; Baranowski et al., 2012). However, there is no consensus in the literature regarding the age at which a child is able to accurately report their own food consumption. A minimum age of 8 years has been suggested, which is exactly what was found in the present study (Livingstone et al., 2004; Burrows et al., 2010). The unexpected increase in the mean number of omissions 97
V. F. Davies et al.
Web-based dietary questionnaire for children 45
Matches
Intrusions
Omissions
Percentage
40
35
30
25
20 2
3
4
5
School grade
Figure 1 Percentage of matches, intrusions and omissions related to the total of Food Intake and Physical Activity of School Children (CAAFE)-reported food items by school grade.
found in the present study among fifth graders may reflect the larger amount of food consumed in this grade (Fig. 2). The tendency for children to under-report is greater when the number of foods consumed is high (Baranowski et al., 1986). High income was related to a lower number of omissions. B€ ornhorst et al. (2013) showed that income was related to higher frequency of omissions among those with a medium or low income. Lower parental education (Lioret et al., 2011; Rangan et al., 2011) was associated with more omissions. Using a proxy measure (the location of the school) children who attended schools in underprivileged areas omitted more food compared to schools in more affluent areas (Moore et al., 2008). Considering that the socioeconomic variable can be related to literacy levels (Lioret et al., 2011; Rangan et al., 2011), it might be useful in the future to know in advance which children have more difficulty with reading comprehension so that special attention may be paid to these students during the application of the questionnaire. There were decreased omissions among children who completed the food diary. The food diary was used to stimulate information retrieval about the foods consumed. Moore et al. (2005) suggested that habituation or practice (i.e. as a result of keeping a food diary) could improve the children’s performance in dietary assessments. Alternatively, children who completed the diary may have been more motivated to participate in the study. Motivation is an important factor when conducting dietary assessment with children (Lu et al., 2012). Unfortunately, no other published research has been reported with food diaries when completing a computerised recall. 98
Some computer literacy is required to complete the CAAFE and lack of experience with computers could be an extra burden for the children when completing webbased questionnaires (Ngo et al., 2009). Having a computer at home increased children’s matches. No previous studies investigated this influence on the children’s performance in online dietary assessments. This information is important because, in the future, it is intended that the CAAFE will be used all over Brazil, where there may be a considerable variation in the presence of a computer in the home. To alleviate this problem, it may be useful to provide CAAFE training sessions with the children, which could increase the children’s familiarity and confidence with the tool. The results showed differences in relation to the schools (School 5 had the lowest level of agreement; School 2 had the highest intrusions and fewest omissions) and the type of meal (there was better agreement for the afternoon snack, and fewer intrusions for lunch). Such results could have been a result of the differences between the schools with regard to the origin of the food consumed, as well as the children’s behaviour at meal times. In the schools surveyed, the foods consumed by children came from three sources: brought from home, purchased in the school cafeteria, or prepared at the school. When the food was brought from home or purchased in the canteen, children typically consumed it in the school yard when playing with peers and/or sharing it with friends. When the children ate food prepared at school, its consumption was more structured; in other words, the children ate sitting in the cafeteria and there was little trading or sharing of food. Previous studies (Davidson et al., 1986; Poppitt et al., 1998; Robson & Livingstone, 2000; Novotny et al., 2001, Chambers et al., 2000, all cited by Baranowski et al., 1986; Moore et al., 2005) have shown that performing other activities when consuming the food (eating and playing at the same time), or sharing foods, represents a possible barrier to children’s effective recall of their food consumption. Fewer omissions were detected among children who regularly ate in the school canteen, which was attributed to more structured food patterns (sitting down when eating) (Lioret et al., 2011). The results of the present study related to the type of school and type of meal deserve further study, especially relating to the origin of snacks consumed at school. This would help to clarify whether it is necessary to add prompts in the CAAFE to help children remember when they brought food from home, which might improve their answers. Among the strengths of the present study is the method used to validate the answers given by the children in the online questionnaire (observation method). The observers were extensively trained and monitored in ª 2014 The British Dietetic Association Ltd.
V. F. Davies et al.
Web-based dietary questionnaire for children
Table 4 Separate multivariate Poisson regression for matches, intrusions and omissions in the Food Intake and Physical Activity of School Children (CAAFE) reports
Moderating variables
Matches RR (95% CI)†
Child gender Male 1.00‡ Female 1.10 (0.92–1.31) School grade Second 1.00‡ Third 2.02 (1.37–2.98)* Fourth 2.07 (1.34–3.19)* Fifth 2.00 (1.09–3.66)* Type of meal observed at school Morning snack 1.00‡ Snack + Lunch 1.24 (0.95–1.61) Afternoon snack 2.95 (1.34–6.50)* School 1 1.00‡ 2 0.78 (0.58–1.03) 3 0.69 (0.45–1.04) 4 0.78 (0.57–1.07) 5 0.38 (0.17–0.84)* Body mass index (quintiles) 1 1.00‡ 2 1.23 (0.92–1.64) 3 1.08 (0.82–1.42) 4 0.94 (0.70–1.27) 5 0.91 (0.66–1.26) What do the children do when they have access to computer§ Research 1.00‡ Social media 1.14 (0.86–1.52) Games 1.12 (0.84–1.48) Computer at children’s bedroom¶ Yes‡ versus No 1.03 (0.82–1.30) Computer at home§ No‡ versus Yes 1.69 (1.16–2.47)* Internet access§ Yes‡ versus No 0.95 (0.70–1.30) Food diary Yes‡ versus No 1.33 (0.99–1.78) Parents’ income 1‡ 1.00 2 1.05 (0.75–1.47) 3 1.13 (0.77–1.65) 4 1.11 (0.72–1.70) 5 1.47 (0.86–2.54) Parents’ education level 1‡ 1.00 2 0.93 (0.68–1.28) 3 0.94 (0.69–1.27) 4 0.93 (0.71–1.22) 5 0.95 (0.66–1.38) 6 0.97 (0.66–1.43) Number of computers at home¶ 0‡ 1.00 1 0.72 (0.53–0.96)* 2 0.68 (0.49–0.93) 3 0.53 (0.28–1.00)
ª 2014 The British Dietetic Association Ltd.
Intrusions RR (95% CI)†
Omissions RR (95% CI)†
– 0.91 (0.75–1.11)
– 0.97 (0.82–1.15)
– 0.87 (0.62–1.21) 0.95 (0.64–1.41) 0.49 (0.27–0.88)*
– 0.67 (0.48–0.95)* 0.57 (0.38–0.84)* 1.27 (0.76–2.11)
– 0.77 (0.60–0.99)* 0.72 (0.42–1.20)
– 0.98 (0.78–1.24) 0.64 (0.40–1.04)
– 2.19 1.52 1.25 1.70
(1.50–3.15)* (0.98–2.34) (0.82–1.90) (1.02–2.86)*
– 0.62 1.14 1.17 1.29
(0.43–0.89)* (0.80–1.64) (0.86–1.60) (0.83–2.00)
– 0.92 0.72 0.99 0.94
(0.67–1.27) (0.53–0.98)* (0.73–1.34) (0.70–1.27)
– 0.92 1.36 1.12 1.20
(0.67–1.27) (1.02–1.82)* (0.82–1.52) (0.86–1.67)
– 0.90 (0.66–1.21) 0.88 (0.62–1.26)
– 1.03 (0.77–1.38) 1.02 (0.77–1.37)
0.90 (0.69–1.17)
1.07 (0.86–1.34)
0.69 (0.45–1.04)
1.03 (0.72–1.47)
1.01 (0.73–1.39)
1.03 (0.79–1.36)
0.75 (0.60–0.94)*
1.33 (1.00–1.75)*
– 0.88 0.80 0.91 0.89
(0.65–1.19) (0.57–1.12) (0.62–1.34) (0.44–1.77)
– 1.00 1.03 0.76 0.53
(0.77–1.30) (0.76–1.39) (0.48–1.19) (0.33–0.84)*
– 1.25 1.16 1.23 1.31 1.21
(0.92–1.70) (0.84–1.58) (0.92–1.68) (0.89–1.93) (0.85–1.72)
– 0.93 1.05 0.92 0.76 0.99
(0.66–1.30) (0.80–1.38) (0.71–1.18) (0.50–1.16) (0.71–1.39)
– 1.10 (0.74–1.64) 1.05 (0.66–1.67) 1.26 (0.70–2.29)
– 1.17(0.84–1.62) 1.42 (0.99–2.03) 1.46 (0.90–2.35)
99
V. F. Davies et al.
Web-based dietary questionnaire for children Table 4. Continued
Moderating variables
Matches RR (95% CI)†
Intrusions RR (95% CI)†
Omissions RR (95% CI)†
Access to computer¶ Never‡ Sometimes Yes
1.00 1.24 (0.90–1.71) 1.21 (0.90–1.64)
– 0.69 (0.48–0.99)* 0.88 (0.60–1.26)
– 1.10 (0.80–1.51) 0.86 (0.62–1.20)
†
Rate ratio related to the reference group/category. Reference group/category. § Children’s answer. ¶ Parents’ answer. *P < 0.05. CI, confidence interval; RR, relative risk. ‡
Figure 2 Mean number of omissions by school grade and total of the food items consumed.
relation to the quality of the observations made. Another strength was the large number of participants, which enabled comparison of performance indicators across variables in the questionnaire. One limitation of the present study is that the validation was only conducted on the meals consumed at school, which does not allow for the results to be extrapolated for the meals consumed outside of school. However, because that the main objective of CAAFE is to monitor food consumption in the school environment, the research mirrored the real context of anticipated future use of web-based CAAFE questionnaires. Another possible limitation is the use of a convenience sample. Schools with the best access to computers and to the Internet were selected. Had the study been conducted in schools with restricted access to computers and the Internet, the results obtained could have been different. Another limitation is the difficulty of making a direct comparison between the results obtained in the present 100
study and those of other questionnaires for evaluating food consumption as a result of differences in study design, age of children, validation standards and the methods of data analysis. CAAFE usability tests (da Costa et al., 2013) indicated very good performance among children aged 7–10 years, and a reasonably low error rate. Nevertheless, observer notes during the application of the CAAFE revealed points needing further attention, such as children not understanding the questions; children’s curiosity in exploring the tool by clicking on various foods despite not having consumed them; misunderstanding that the CAAFE is a game, which results in competition among peers when completing the questionnaire and fails to focus on dietary assessment accuracy; and not finding the foods consumed among the images on the CAAFE screen. Future studies should include qualitative studies through cognitive and semi-structured interviews involving the target audience, with the aim of investigating the problems highlighted by the observers. Studies about the sources of the food consumed (brought from home or provided by the school), as well as their impact on the accuracy of the responses, would also be worth exploring. Conclusions In conclusion, the results of the present study showed that the CAAFE questionnaire provided a reasonable degree of agreement compared to the observation method. The overall matches (43%), intrusions (29%) and omissions (28%) were to be expected, given the major cognitive demands related to the dietary assessment of children (i.e. memory and attention). In addition, our findings were similar to other studies using computers to evaluate children’s food consumption, with regard to the obtained result that children older than 8 years performed better in the CAAFE. Using a food diary was ª 2014 The British Dietetic Association Ltd.
V. F. Davies et al.
also useful for improving the overall accuracy of answers. Variables that influenced the incidence of disagreement included the meal type and other school related factors, all of which require more in-depth studies. In general, the CAAFE is an attractive tool for children to self-report their diet, although further studies are required to improve its validity. Acknowledgments We gratefully thank the Education Secretary of the Prefeitura Municipal de Florianopolis, the staff of the schools, and the children and parents for their participation in the study.
Conflict of interests, source of funding and authorship The authors declare that there are no conflicts of interest. This research was funded by a grant from the Brazilian Ministry of Health (Departamento de Ci^encia, Tecnologia e Insumos Estrategicos – DECIT), the Brazilian Ministry of Education (Coordenacß~ao de Aperfeicßoamento de Pessoal de NıvelSuperior-CAPES) and the Brazilian Ministry of Science, Technology and Innovation (Conselho Nacional de Desenvolvimento Cientıfico e Tecnol ogico (CNPq). VFD was supported by the CAPES – Brazilian Federal Agency for Support and Evaluation of Graduate Education. VFD participated in its design and coordination of the study and also wrote the first draft of the manuscript. EK, MAAA and PFP were leaders in CAAFE design, coordinated the research and co-wrote the manuscript. NS and TD gave feedback and guided the development of the manuscript. All authors critically reviewed the manuscript and approved the final version submitted for publication.
References Adamson, A.J. & Baranowski, T. (2014) Developing technological solutions for dietary assessment in children and young people. J. Hum. Nutr. Diet. 27(Suppl. 1), 1–4. Agresti, A. (2007) An Introduction to Categorical Data Analysis. Hoboken, NJ: Wiley. Assis, M.A., Benedet, J., Kerpel, R., Vasconcelos, Fde.A., Di Pietro, P.F. & Kupek, E. (2009) Validation of the third version of the Previous Day Food Questionnaire (PDFQ-3) for 6-to-11-years-old schoolchildren. Cad. Saude Publica 25, 1816–1826.
ª 2014 The British Dietetic Association Ltd.
Web-based dietary questionnaire for children
Baranowski, T., Dworkin, R., Henske, J.C., Clearman, D.R., Dunn, K., Nader, P.R. & Hooks, P.C. (1986) The accuracy of children’s self report of diet: Family Health Project. J. Am. Diet. Assoc. 86, 1381–1385. Baranowski, T., Islam, N., Baranowski, J., Cullen, K.W., Myres, D., Marsh, T. & de, M.C. (2002) The food intake recording software system is valid among fourth-grade children. J. Am. Diet. Assoc. 102, 380–385. Baranowski, T., Islam, N., Douglass, D., Dadabhoy, H., Beltran, A., Baranowski, J., Thompson, D., Cullen, K.W. & Subar, A.F. (2012) Food Intake Recording Software System, version 4 (FIRSSt4): a self-completed 24-h dietary recall for children. J. Hum. Nutr. Diet.27(Suppl. 1), 66–71. Baxter, S.D., Hardin, J.W., Guinn, C.H., Royer, J.A., Mackelprang, A.J. & Devlin, C.M. (2010) Children’s body mass index, participation in school meals, and observed energy intake at school meals. Int. J. Behav. Nutr. Phys. Act. 7, 1–8. Biltoft-Jensen, A., Trolle, E., Christensen, T., Islam, N., Andersen, L.F., Egenfeldt-Nielsen, S. & Tetens, I. (2012) WebDASC: a web-based dietary assessment software for 8–11-year old Danish children. J. Hum. Nutr. Diet.27(Suppl. 1), 43–53. Biltoft-Jensen, A., Bysted, A., Trolle, E., Christensen, T., Knuthsen, P., Damsgaard, C.T., Andersen, L.F., Brockhoff, P. & Tetens, I. (2013) Evaluation of Web-based Dietary Assessment Software for Children: comparing reported fruit, juice and vegetable intakes with plasma carotenoid concentration and school lunch observations. Br. J. Nutr. 110, 186–195. B€ ornhorst, C., Huybrechts, I., Ahrens, W., Eiben, G., Michels, N., Pala, V., Molnar, D., Russo, P., Barba, G., Bel-Serrat, S., Moreno, L.A., Papoutsou, S., Veidebaum, T., Loit, H.M., Lissner, L. & Pigeot, I.; IDEFICS Consortium (2013) Prevalence and determinants of misreporting among European children in proxy reported 24 h dietary recalls. Br. J. Nutr. 109, 1257–1265. Burrows, T.L., Martin, R.J. & Collins, C.E. (2010) 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. 110, 1501–1510. Collins, C.E., Watson, J. & Burrows, T. (2010) Measuring dietary intake in children and adolescents in the context of overweight and obesity. Int. J. Obes. (Lond) 34, 1103– 1115. da Costa, F.F., Davies, V.F., Schmoelz, C.P., Kuntz, M.G.F. & de Assis, M.A.A. (2012) Medida da atividade fısica de criancßas: o que professores de educacß~ao fısica t^em a nos dizer? Rev. Bras. Ativ. Fis. e Sa ude. 17, 286–287. da Costa, F.F., Schmoelz, C.P., Davies, V.F., Di Pietro, P.F., Kupek, E. & de Assis, M.A. (2013) Assessment of diet and physical activity of Brazilian schoolchildren: usability testing of a web-based questionnaire. JMIR Res. Protoc. 2, e31. Davies, V.F., Kupek, E., de Assis, M.A., Engel, R., da Costa, F.F., Di Pietro, P.F., Natal, S., Thompson, D. & Baranowski,
101
Web-based dietary questionnaire for children
T. (2014) Qualitative analysis of the contributions of nutritionists to the development of an online instrument for monitoring the food intake of schoolchildren. J. Hum. Nutr. Diet. 31, 1–10. Foster, E., Hawkins, A., Delve, J. & Adamson, A.J. (2014) Reducing the cost of dietary assessment: selfcompleted recall and analysis of nutrition for use with children (SCRAN24). J. Hum. Nutr. Diet. 27 (Suppl. 1), 26–35. van Gelder, M.M., Bretveld, R.W. & Roeleveld, N. (2010) Web-based questionnaires: the future in epidemiology? Am. J. Epidemiol. 172, 1292–1298. Illner, A.K., Freisling, H., Boeing, H., Huybrechts, I., Crispim, S. & Slimani, N. (2012) Review and evaluation of innovative technologies for measuring diet in nutritional epidemiology. Int. J. Epidemiol. 41, 1187–1203. Instituto Brasileiro de Geografia e Estatistica (IBGE). (2010a) Pesquisa de Orcßamentos Familiares 2008/2009 (POF). Available at: http://www.ibge.gov.br/home/presidencia/ noticias/noticia_visualiza.php?id_noticia=1699&id_pagina=1/ (accessed on 10 March 2014). Instituto Brasileiro de Geografia e Estatistica (IBGE). (2010b) Censo Demografico. Available at: http://www.ibge.gov.br/ home/estatistica/populacao/censo2010/resultados_gerais_ amostra/default_resultados_gerais_amostra.shtm (accessed on 23 June 2014). Instituto Brasileiro de Geografia e Estatistica (IBGE). (2012). Available at: http://www.ibge.gov.br/home/estatistica/ populacao/trabalhoerendimento/pnad2012/sintese_ defaultpdf_rendimentos.shtm (accessed on 23 June 2014). Kohlmeier, L., Mendez, M., McDuffie, J. & Miller, M. (1997) Computer assisted self-interviewing: a multimedia approach to dietary assessment. Am. J. Clin. Nutr. 65 (Suppl.), 1275S– 1281S. Lanctot, J.Q., Klesges, R.C., Stockton, M.B. & Klesges, L.M. (2008) Prevalence and characteristics of energy underreporting in African-American girls. Obesity 16, 1407–1412. Lioret, S., Touvier, M., Balin, M., Huybrechts, I., Dubuisson, C., Dufour, A., Bertin, M., Maire, B. & Lafay, L. (2011) Characteristics of energy under-reporting in children and adolescents. Br. J. Nutr. 105, 1671–1680. Livingstone, M.B.E., Robson, P.J. & Wallace, J.M.W. (2004) Issues in dietary intake assessment of children and adolescents. Br. J. Nutr. 92, S213–S222. Lohman, T.G., Roche, A.F. & Martorell, R. (1991) Anthropometric Standardization Reference Manual. Champaign, IL: Human Kinetics.
102
V. F. Davies et al.
Lu, A.S., Baranowski, J., Islam, N. & Baranowski, T. (2012) How to engage children in self administered dietary assessment programmes. J. Hum. Nutr. Diet. 27(Suppl. 1), 5–9. Magarey, A., Watson, J., Golley, R.K., Burrows, T., Sutherland, R., McNaughton, A.S., Denney-Wilson, E., Campbell, K. & Collins, C. (2011) Assessing dietary intake in children and adolescents: considerations and recommendations for obesity research. Int. J. Pediatr. Obes. 6, 2–11. Moore, L., Tapper, K., Dennehy, A. & Cooper, A. (2005) Development and testing of a computerised 24- h recall questionnaire measuring fruit and snack consumption among 9–11 year olds. Eur. J. Clin. Nutr. 59, 809–816. Moore, H.J., Ells, L.J., McLure, S.A., Crooks, S.C., Cumbor, D., Summerbell, C.D. & Batterham, A.M. (2008) The development and evaluation of a novel computer program to assess previous-d ay dietary and physical activity behaviours in school children: the Synchronised Nutrition and Activity ProgramTM (SNAPTM). Br. J. Nutr. 99, 1266–1274. Moore, H.J., Hillier, F.C., Batterham, A.M., Ells, L.J. & Summerbell, C.D. (2014) Technology-based dietary assessment: development of the Synchronised Nutrition and Activity Program (SNAPTM). J. Hum. Nutr. Diet. 27 (Suppl 1), 36–42. Ngo, J., Engelen, A., Molag, M., Roesle, J., Garcia-Segovia, P. & Serra-Majem, L. (2009) A review of the use of information and communication technologies for dietary assessment. Br. J. Nutr. 101, S102–S112. Ortiz-Andrellucchi, A., Sanchez, P.H., Villegas, A.S., Quintana, L.P., Mendez, M. & Serra-Majem, L. (2009) Dietary assessment methods for micronutrient intake in infants, children and adolescents: a systematic review. Br. J. Nutr. 102, S87–S117. Rangan, A.M., Flood, V.M. & Gill, T.P. (2011) Misreporting of energy intake in the 2007 Australian Children’s Survey: identification, characteristics and impact of misreporters. Nutrients 3, 186–199. Universidade Federal de Santa Catarina. (2013) Sistema de monitoramento do consumo alimentar e atividade fısica de escolares. Available at: http://caafe.ufsc.br/ (accessed on 19 March 2013). Vereecken, C., Covents, M., Maes, L. & Moyson, T. (2014) Formative evaluation of the dietary assessment component of Children’s and Adolescents’ Nutrition Assessment and Advice on the Web (CANAA-W). J. Hum. Nutr. Diet. 27(Suppl. 1), 54–65.
ª 2014 The British Dietetic Association Ltd.
Copyright of Journal of Human Nutrition & Dietetics is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
RESEARCH PAPER Validation of a web-based questionnaire to assess the dietary intake of Brazilian children aged 7–10 years V. F. Davies,1 E. Kupek,1 M. A. de Assis,2 S. Natal,1 P. F. Di Pietro2 & T. Baranowski3 1
Graduate Program in Public Health, Center for Health Sciences at the Federal University of Santa Catarina, Florianopolis, SC, Brazil Graduate Program in Nutrition, Center for Health Sciences at the Federal University of Santa Catarina, Florianopolis, SC, Brazil 3 USDA/ARS Children’s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, TX, USA 2
Keywords intrusions, matches, omissions, online, questionnaire. Correspondence V. F. Davies, Graduate Program in Public Health, Center for Health Sciences at the Federal University of Santa Catarina, Campus Reitor Jo~ao David Ferreira Lima, Rua Delfino Conti, S/N. Bloco H, Cep 88040-370, Florianopolis, SC, Brazil. Tel.: +55 48 3721 9847 Fax: +55 048 3721 9542 E-mail: [email protected]
How to cite this article Davies V.F., Kupek E., De Assis M.A., Natal S., Di Pietro P.F. & Baranowski T. (2015) Validation of a web-based questionnaire to assess the dietary intake of Brazilian children aged 7–10 years. J Hum Nutr Diet. 28 (Suppl. 1), 93–102 doi:10.1111/jhn.12262
Abstract Background: The Food Intake and Physical Activity of School Children (CAAFE) comprises an online questionnaire to self-report diet and physical activity of Brazilian schoolchildren. Background: The present study aimed to assess the validity (matches, omissions and intrusions) and moderating factors of the CAAFE. Methods: Direct observation was made of foods consumed (five public schools) and child self-reporting on the CAAFE. Additional data included school grade, gender, body mass index, completion of food diary, socioeconomic status and access to computer. Data were analysed using regression. Results: In total, 602 children participated in the study [mean (SD) age 9.5 (1.24) years; 53.6% boys]. On average, there were 43% matches, 29% intrusions and 28% omissions. Matches doubled in third grade compared to the second grade (P = 0.004); matches almost tripled for afternoon snack compared to morning snack (P < 0.001); and matches were 69% higher for children with access to a computer at home (P < 0.01). Intrusions decreased by almost one-half in fifth compared to fourth grades (P = 0.004). Omissions declined significantly in the third and fourth grades but increased in the fifth grade. Omissions were 47% lower for children in the highest income and lower among children who completed the food diary. No differences were found for gender or body mass index. Conclusions: Children older than 8 years old, who owned a computer and completed a food diary, performed better in the CAAFE. A high incidence of disagreement was found in relation to the schools and the type of meal. Overall matches (43%), intrusions (29%) and omissions (28%) indicate that further studies are required to improve the validity of the CAAFE.
Introduction Choosing the best method for assessing food consumption includes issues related to characteristics of the individuals investigated (i.e. age group), the objective of the assessment and the available resources (preferably easy to apply and at a low cost), as well as validity and precision (Burrows et al., 2010; Collins et al., 2010). Assessing children’s dietary intake has proved to be an important challenge in nutritional epidemiology. The most commonly used methods for assessing food intake ª 2014 The British Dietetic Association Ltd.
require skills related to memory and attention span, as well as a knowledge of foods and preparations (Livingstone et al., 2004; Ortiz-Andrellucchi et al., 2009; Baxter et al., 2010). Another factor that may influence dietary assessment of children is motivation. Extensive questionnaires that seek to obtain very detailed information are not recommended for children because they can lead to boredom, fatigue and possibly decreased compliance (Magarey et al., 2011; Lu et al., 2012). The use of computers in the assessment of children’s food intake may be a promising alternative to more 93
Web-based dietary questionnaire for children
traditional methods (van Gelder et al., 2010; Illner et al., 2012; Adamson & Baranowski, 2014). The advantages include the use of images and interactive features that attract a child’s attention, as well as allowing researchers to standardise the questionnaires and to conduct an immediate evaluation of the data (Kohlmeier et al., 1997; Ngo et al., 2009; Lu et al., 2012). Several computerised child diet assessment instruments have been developed around the world; for example, SCRAN24, SNAP, Web DASC, CANAA-W and ASA24 – Kids (Biltoft-Jensen et al., 2012; Adamson & Baranowski, 2014; Foster et al., 2014; Moore et al., 2014; Vereecken et al., 2014). The Food Intake and Physical Activity of School Children – CAAFE (Universidade Federal de Santa Catarina, 2013) – comprises an online questionnaire that has been developed to enable the self-reporting of diet and physical activity by Brazilian schoolchildren aged 7–10 years. To create an efficient child-friendly questionnaire, the following steps were initiated: analysis of similar instruments; focus groups with physical education teachers (da Costa et al., 2012) and nutritionists (Davies et al., 2014); meetings of researchers, teachers and experts in web design; analysis of 7-day food and physical activity records; and usability testing (da Costa et al., 2013). In the food consumption section of the questionnaire, the resulting instrument asks the child about their food consumption during the previous 24 h, and quantifies the consumption of priority food groups for the 7–10-year-old age group (e.g. milk and milk products), as well as foods related to obesity (e.g. soft drinks). The present study reports the first validation of the food consumption section of the CAAFE, by identifying the agreement and errors between the children’s answers in the questionnaire and observation of their school meals. It goes beyond other validation studies of online dietary assessment tools by assessing the possible moderating influences of a number of variables, such as socioeconomic status, access to digital technology and the use of a food diary. Materials and methods In the present study, we describe the validation process of the food consumption section of the CAAFE questionnaire. The results for the validity of the physical activity section will be described elsewhere. The validation process included five public schools in the city of Florian opolis, the capital of Santa Catarina state, Brazil. The schools were selected by the Municipal Education Department according to the availability of an adequate computer room (number of computers and working internet connection). Schools were selected (n = 5) in different regions of the city (Central, North, South, and East) to 94
V. F. Davies et al.
represent students from different socioeconomic backgrounds. Six classes from second to fifth grades were selected by the principal of each school (30 classes in total). The classes were selected by the principal based on the class schedule for the week. For example, classes that had already planned field trips or other activities during the research period that could not be rescheduled were not selected to join the study. Children in Brazil who attend second to fifth grade are expected to meet the following age ranges: second grade: 7–8 years old; third grade: 8–9 years old; fourth grade: 9–10 years old; fifth grade: 10–11 years old. In total, 708 students were invited to participate in the study. The children gave their oral consent and the parents provided their written informed consent. No financial reward was offered. The project was approved by the Ethics Committee on Human Research at the Federal University of Santa Catarina (UFSC) (protocol no. 2250/11). Parents completed a questionnaire with sociodemographic information, as well as information regarding the child’s access to a computer at home. For the children who agreed to participate, anthropometric measurements were recorded. The measurements of the children’s body weight and height were taken by trained researchers in the schools themselves, using standard techniques (Lohman et al., 1991). Portable digital scales (with a capacity of 180 kg) were used (Marte LTDA, Belo Horizonte, Brazil). Height was measured using a metal stadiometer (Seca, Hamburg, Germany). The children were weighed and measured barefoot, wearing light clothing. Bofy mass index (BMI) was computed as weight (kg) divided by height squared (m2). In the week before the research, researchers distributed to each student a 7-day-food diary similar to that of the online survey (CAAFE). In the classroom, the children were given instructions on how to complete the diary. The purpose of the diary was to familiarise the children with the images of the foods in the CAAFE, and also as a resource to improve the accuracy of answers when completing the online questionnaire. The children were instructed to consult the diary when completing the online questionnaire. Food diaries included the foods consumed on the day that their food consumption was observed. The CAAFE is a self-reported questionnaire which examines food consumption and physical activity during the previous day (24-h recall). The food consumption section of the CAAFE is divided into six meals (breakfast, morning snack, lunch, afternoon snack, dinner and supper). For each meal, 32 images of food or drinks are presented on the computer screen so that the child can make their own selections. The images of foods and food ª 2014 The British Dietetic Association Ltd.
V. F. Davies et al.
groups were chosen to take into account: the food patterns of children in this age group, the food usually offered on school menus, the suggestions given by focus groups with dietitians (Davies et al., 2014), a 7-day food diary (n = 180 children) and the most consumed foods according to the Brazilian Household Budget Survey (IBGE, 2010a). An avatar in the form of a robot guides the child when completing the questionnaire, explaining the concept of each meal and the time of day at which it is consumed, as well as reinforcing the importance of reporting food consumption from the previous 24 h. The avatar also instructs the child not to click on any food if it was not consumed during the preceding day. To validate the food section of the CAAFE, direct observation was made of the meals in the school environment, including meals that were prepared on the school premises, those bought in the school canteen, or food brought from home. The observers were nutrition students from two large universities in the city where the research was conducted, who, during their classes, were invited by the research coordinator to join the study. To ensure quality, observers took part in both conceptual and practical training sessions. The conceptual session included the presentation of the observation protocol and instructions for completing the worksheet to be used during data collection. The practice phase took place in the different schools at which data were collected. The training consisted of six practice sessions, during which each observer had to observe at least 30 children in total. Agreement on identified foods was 96% between the more experienced observers and the other observers during the training. In addition, because the CAAFE required adjustments before initiating the validation, pilot studies were conducted in three schools. These studies were used as an opportunity to further increase the experience of the observers and thus to improve the quality of the data collected. During data collection, each observer was assigned to observe a maximum of five children, who in turn were identified with name tags and colored ribbons on their arms. The observation took place the day before the computer questionnaire was applied. Some of the observers were in the dining hall and some were in the schoolyard. Observers took note of the types of foods and beverages consumed by children, as well as recording foods that were shared. The day after food consumption was observed, the children were conducted to the school computer room to complete the CAAFE. Prior to that, however, a trained researcher gave demonstration of how to complete the computer questionnaire in each classroom. With the aid of a banner (140 9 105 cm), images of food from the CAAFE were shown, as well as examples of how to answer the questionnaire. Subsequently, the children were ª 2014 The British Dietetic Association Ltd.
Web-based dietary questionnaire for children
instructed to take the previously completed food diaries to the computer room, in case they wished to consult their records. Once in the computer room, researchers were present to assist the children with any questions, and to record the questions raised by the children, when completing the CAAFE. Once the questionnaire was completed, the children were individually interviewed with regard to three questions: the presence of a computer at home, the presence of an Internet connection at home, and the main activity carried out on their home computer. In the present study, matches occurred when foods and/or beverages observed to have been consumed by the children were reported in the CAAFE. Omissions occurred when foods and/or beverages observed to have been consumed by the children were not reported in the CAAFE. Intrusions occurred when foods and/or beverages were reported in the CAAFE but were not observed to have been consumed by the children. Match percentages were calculated: percentage matches = (the number of matches/total food items directly observed at school) 9 100. Intrusion percentages were calculated: percentage intrusions = (total number of intrusions/total food items observed directly in school) 9 100. Omission percentages were calculated: percentage omissions = (the number of omissions/total food items directly observed in school) 9 100. Agreement rates were calculated using Poisson regression for the number of matches, intrusions and omissions. This method of analysis was chosen rather than other standard agreement analyses (i.e. kappa) because Poisson regression allows one to distinguish between two types of mismatch: intrusions and omissions. These are firmly established as distinct types of error, with the former more related to the food environment (e.g. the number of foods offered) and the latter to the difficulties children have to recall the food consumed on the previous day. In addition, rates of agreement calculated by Poisson regression provide the difference between various levels of an independent variable adjusted for all the other independent variables, whereas agreement indexes such as kappa do not (Agresti, 2007). Therefore, regression-type impact measures such as percentage rate difference were used throughout the present study. Poisson regression was used to explain variations in the number of matches, intrusions and omissions with regard to the following independent variables: child gender, school grade, type/location of school, type of meal observed (morning snack, lunch, afternoon snack), children’s BMI (quintiles), parents’ education level, parents’ annual income, access to a computer and Internet at home, number of computers at home, type of activities carried out by children on their computers (research/ 95
V. F. Davies et al.
Web-based dietary questionnaire for children
social media/games), presence of a computer in the child’s room, frequency of computer use, and completion of the food diary. P < 0.05 was considered statistically significant. Analyses were performed using STATA, version 12.0 (StataCorp, College Station, TX, USA). Results This validation study involved 602 children [mean (SD) age 9.5 (1.2) years, 53.6% boys]. Children (n = 106) not included in the final sample were either absent from school during the observation of the meals and/or the application of the CAAFE in the computer room. In total, 16 observers were trained to observe the school meals and participated in the research. Tables 1–3 display percentages of matches, intrusions and omissions in the CAAFE reports according to the characteristics of the study sample. The average percentages of matches, intrusions and omissions were 43%, 29% and 28%, respectively (Table 1). Sociodemographic characteristics of the sample showed that most the children belonged to the second ($3620 to $7236) and third category ($7237 to $10 856) with regard to their parents’
income, which is above the Brazilian national income average (IBGE, 2012). Most parents did not have a college degree, which reflects the Brazilian average education level (IBGE, 2010b). From the third grade onwards, the accuracy (matches) of the CAAFE reports almost doubled (Fig. 1 and Table 4). Intrusions approximately halved in fifth grade compared to fourth grade (Fig. 1 and Table 4). Food omissions were reduced by one-third in third grade compared to second grade, followed by an increase in fifth grade (Fig. 1 and Table 4). Agreement was almost threefold higher when reporting an afternoon snack compared to a morning snack, and 69% higher for children with a computer at home compared to those without. Intrusions were 23% lower for lunch compared to the morning snack reports; 25% lower among children who did not complete the food diary compared to those who did; and 47% fewer for children whose parents had a higher income. No differences were found for gender, BMI quintiles or variables related to digital technology (Table 4). Despite considerable variation in reporting accuracy across schools, only a few reached statistical significance: compared to the reference school, School 5 presented
Table 1 Percentages of matches, intrusions and omissions in the Food Intake and Physical Activity of School Children (CAAFE) reports according to the sociodemographic characteristics of the study sample Variable
Category
n (%)
Matches (%)
Intrusions (%)
Omissions (%)
Child gender
Male Female 2 (mean age 7.8 years) 3 (mean age 8.8 years) 4 (mean age 9.9 years) 5 (mean age 10.9 years) 1 (mean 14.6 kg m–2) 2 (mean 15.9 kg m–2) 3 (mean 17.2 kg m–2) 4 (mean 19.2 kg m–2) 5 (mean 23.9 kg m–2) 1 (≤$3619) 2 ($3620 to $7236) 3 ($7237 to $10 856) 4 ($10 857 to $18 056) 5 ($18 057 to $28 947) 1 (Elementary school not completed) 2 (Elementary school) 3 (High School not completed) 4 (High School) 5 (College not completed) 6 (College) 1 2 3 4 5
279 323 113 140 190 159 119 116 120 115 113 48 250 110 4 23 115 61 73 136 40 52 146 141 100 83 132
42.76 41.83 25.39 47.76 48.90 41.48 43.41 44.90 40.71 43.18 38.77 43.55 40.50 45.02 47.80 61.23 41.96 43.51 39.61 44.52 51.30 47.02 47.03 25.51 28.85 31.80 35.32
30.02 28.60 37.71 26.27 31.57 23.12 28.46 32.73 29.36 28.34 27.99 31.99 28.50 28.00 33.02 21.05 27.27 29.34 29.95 28.04 30.72 29.68 21.62 53.08 37.69 33.47 35.32
27.22 29.57 36.90 25.97 19.54 35.40 28.13 22.37 29.92 28.47 33.24 24.46 31.00 26.98 19.17 17.72 30.76 27.14 30.43 27.44 17.98 23.30 31.35 21.41 33.46 34.73 29.36
School grade
Body mass index (quintiles)
Parents’ annual income (US dollars)
Parents’ education level
School
96
(46.3) (53.6) (18.7) (23.2) (31.5) (26.4) (19.7) (19.2) (19.9) (19.1) (18.7) (7.9) (41.5) (18.2) (7.48) (3.8) (19.1) (10.1) (12.1) (22.9) (6.6) (8.6)
ª 2014 The British Dietetic Association Ltd.
V. F. Davies et al.
Web-based dietary questionnaire for children
Table 2 Percentages of matches, intrusions and omissions in the Food Intake and Physical Activity of School Children (CAAFE) reports by type of meal observed at school and by completion of food diary Variable
Category
n (%)
Matches (%)
Intrusions (%)
Omissions (%)
Meal observed
Morning snack Snack + Lunch Afternoon snack Yes No
213 289 100 442 160
38.83 44.25 43.79 24.13 38.31
32.75 26.07 31.02 51.51 31.84
28.40 29.67 25.19 24.36 31.28
Food diary completed
(35.3) (48.0) (16.6) (73.4) (26.5)
Table 3 Percentages of matches, intrusions and omissions in the Food Intake and Physical Activity of School Children (CAAFE) reports by access to digital technology Variable
Category
n (%)
Matches (%)
Intrusions (%)
Omissions (%)
Computer at home*
No Yes No Yes Research Social media Games 0 1 2 3 No Yes No Sometimes Yes
125 429 174 362 99 312 136 99 275 102 23 396 101 44 165 289
35.56 44.51 37.64 44.66 43.31 39.40 45.45 41.98 45.68 38.90 42.69 43.87 42.19 40.12 42.27 44.38
37.03 27.08 35.16 26.62 28.84 32.36 24.18 31.28 27.07 28.92 27.64 29.25 24.45 29.59 27.09 28.91
27.41 28.41 27.20 28.72 27.85 28.23 30.36 26.73 27.25 32.18 29.67 26.87 33.36 30.29 30.64 26.70
Access to Internet at home* What do the children do when they have access to computer*
Numbers of computer at home†
Computer at children’s bedroom† Children access to computer at home†
(22.5) (77.4) (32.4) (67.5) (16.4) (51.8) (22.5) (19.8) (55.1) (20.4) (4.6) (79.6) (20.3) (8.8) (33.1) (58.0)
*Children’s answer. Parents’ answer.
†
62% lower matching rate and 70% higher intrusion rate; and School 2 presented over twice the rate of intrusions and 38% lower rate of omissions. The children who did not complete the food diary had 33% more omissions than those who did complete the diary (Table 4). The mean number of omissions was high in second grade, decreased considerably in third and fourth grades, only to peak again in fifth grade, when the children consumed more food overall compared to other grades (Fig. 2). Discussion The percentages of matches (43%), intrusions (29%) and omissions (21%) found in the present study are similar to those in previously published studies (Baranowski et al., 2002, 2012; Biltoft-Jensen et al., 2013), although it should be emphasised that those previous studies used different methods to estimate accuracy during their data analyses. Another difference is that the CAAFE study ª 2014 The British Dietetic Association Ltd.
included children from the age of 7 years, whereas the minimum age in the other studies was 8 years (Baranowski et al., 2012; Biltoft-Jensen et al., 2013) and 10 years (Baranowski et al., 2002). By contrast to these other studies, we also gave our students instruction on the use of the CAAFE, and asked them to keep a diary on the day of observation, both of which may have enhanced the accuracy of the younger children’s answers. Children older than 8 years performed better in the CAAFE (Table 1). Other studies also found that the age of the child influences the accuracy of dietary assessment questionnaires (Lanctot et al., 2008; De Assis et al., 2009; Burrows et al., 2010; Baranowski et al., 2012). However, there is no consensus in the literature regarding the age at which a child is able to accurately report their own food consumption. A minimum age of 8 years has been suggested, which is exactly what was found in the present study (Livingstone et al., 2004; Burrows et al., 2010). The unexpected increase in the mean number of omissions 97
V. F. Davies et al.
Web-based dietary questionnaire for children 45
Matches
Intrusions
Omissions
Percentage
40
35
30
25
20 2
3
4
5
School grade
Figure 1 Percentage of matches, intrusions and omissions related to the total of Food Intake and Physical Activity of School Children (CAAFE)-reported food items by school grade.
found in the present study among fifth graders may reflect the larger amount of food consumed in this grade (Fig. 2). The tendency for children to under-report is greater when the number of foods consumed is high (Baranowski et al., 1986). High income was related to a lower number of omissions. B€ ornhorst et al. (2013) showed that income was related to higher frequency of omissions among those with a medium or low income. Lower parental education (Lioret et al., 2011; Rangan et al., 2011) was associated with more omissions. Using a proxy measure (the location of the school) children who attended schools in underprivileged areas omitted more food compared to schools in more affluent areas (Moore et al., 2008). Considering that the socioeconomic variable can be related to literacy levels (Lioret et al., 2011; Rangan et al., 2011), it might be useful in the future to know in advance which children have more difficulty with reading comprehension so that special attention may be paid to these students during the application of the questionnaire. There were decreased omissions among children who completed the food diary. The food diary was used to stimulate information retrieval about the foods consumed. Moore et al. (2005) suggested that habituation or practice (i.e. as a result of keeping a food diary) could improve the children’s performance in dietary assessments. Alternatively, children who completed the diary may have been more motivated to participate in the study. Motivation is an important factor when conducting dietary assessment with children (Lu et al., 2012). Unfortunately, no other published research has been reported with food diaries when completing a computerised recall. 98
Some computer literacy is required to complete the CAAFE and lack of experience with computers could be an extra burden for the children when completing webbased questionnaires (Ngo et al., 2009). Having a computer at home increased children’s matches. No previous studies investigated this influence on the children’s performance in online dietary assessments. This information is important because, in the future, it is intended that the CAAFE will be used all over Brazil, where there may be a considerable variation in the presence of a computer in the home. To alleviate this problem, it may be useful to provide CAAFE training sessions with the children, which could increase the children’s familiarity and confidence with the tool. The results showed differences in relation to the schools (School 5 had the lowest level of agreement; School 2 had the highest intrusions and fewest omissions) and the type of meal (there was better agreement for the afternoon snack, and fewer intrusions for lunch). Such results could have been a result of the differences between the schools with regard to the origin of the food consumed, as well as the children’s behaviour at meal times. In the schools surveyed, the foods consumed by children came from three sources: brought from home, purchased in the school cafeteria, or prepared at the school. When the food was brought from home or purchased in the canteen, children typically consumed it in the school yard when playing with peers and/or sharing it with friends. When the children ate food prepared at school, its consumption was more structured; in other words, the children ate sitting in the cafeteria and there was little trading or sharing of food. Previous studies (Davidson et al., 1986; Poppitt et al., 1998; Robson & Livingstone, 2000; Novotny et al., 2001, Chambers et al., 2000, all cited by Baranowski et al., 1986; Moore et al., 2005) have shown that performing other activities when consuming the food (eating and playing at the same time), or sharing foods, represents a possible barrier to children’s effective recall of their food consumption. Fewer omissions were detected among children who regularly ate in the school canteen, which was attributed to more structured food patterns (sitting down when eating) (Lioret et al., 2011). The results of the present study related to the type of school and type of meal deserve further study, especially relating to the origin of snacks consumed at school. This would help to clarify whether it is necessary to add prompts in the CAAFE to help children remember when they brought food from home, which might improve their answers. Among the strengths of the present study is the method used to validate the answers given by the children in the online questionnaire (observation method). The observers were extensively trained and monitored in ª 2014 The British Dietetic Association Ltd.
V. F. Davies et al.
Web-based dietary questionnaire for children
Table 4 Separate multivariate Poisson regression for matches, intrusions and omissions in the Food Intake and Physical Activity of School Children (CAAFE) reports
Moderating variables
Matches RR (95% CI)†
Child gender Male 1.00‡ Female 1.10 (0.92–1.31) School grade Second 1.00‡ Third 2.02 (1.37–2.98)* Fourth 2.07 (1.34–3.19)* Fifth 2.00 (1.09–3.66)* Type of meal observed at school Morning snack 1.00‡ Snack + Lunch 1.24 (0.95–1.61) Afternoon snack 2.95 (1.34–6.50)* School 1 1.00‡ 2 0.78 (0.58–1.03) 3 0.69 (0.45–1.04) 4 0.78 (0.57–1.07) 5 0.38 (0.17–0.84)* Body mass index (quintiles) 1 1.00‡ 2 1.23 (0.92–1.64) 3 1.08 (0.82–1.42) 4 0.94 (0.70–1.27) 5 0.91 (0.66–1.26) What do the children do when they have access to computer§ Research 1.00‡ Social media 1.14 (0.86–1.52) Games 1.12 (0.84–1.48) Computer at children’s bedroom¶ Yes‡ versus No 1.03 (0.82–1.30) Computer at home§ No‡ versus Yes 1.69 (1.16–2.47)* Internet access§ Yes‡ versus No 0.95 (0.70–1.30) Food diary Yes‡ versus No 1.33 (0.99–1.78) Parents’ income 1‡ 1.00 2 1.05 (0.75–1.47) 3 1.13 (0.77–1.65) 4 1.11 (0.72–1.70) 5 1.47 (0.86–2.54) Parents’ education level 1‡ 1.00 2 0.93 (0.68–1.28) 3 0.94 (0.69–1.27) 4 0.93 (0.71–1.22) 5 0.95 (0.66–1.38) 6 0.97 (0.66–1.43) Number of computers at home¶ 0‡ 1.00 1 0.72 (0.53–0.96)* 2 0.68 (0.49–0.93) 3 0.53 (0.28–1.00)
ª 2014 The British Dietetic Association Ltd.
Intrusions RR (95% CI)†
Omissions RR (95% CI)†
– 0.91 (0.75–1.11)
– 0.97 (0.82–1.15)
– 0.87 (0.62–1.21) 0.95 (0.64–1.41) 0.49 (0.27–0.88)*
– 0.67 (0.48–0.95)* 0.57 (0.38–0.84)* 1.27 (0.76–2.11)
– 0.77 (0.60–0.99)* 0.72 (0.42–1.20)
– 0.98 (0.78–1.24) 0.64 (0.40–1.04)
– 2.19 1.52 1.25 1.70
(1.50–3.15)* (0.98–2.34) (0.82–1.90) (1.02–2.86)*
– 0.62 1.14 1.17 1.29
(0.43–0.89)* (0.80–1.64) (0.86–1.60) (0.83–2.00)
– 0.92 0.72 0.99 0.94
(0.67–1.27) (0.53–0.98)* (0.73–1.34) (0.70–1.27)
– 0.92 1.36 1.12 1.20
(0.67–1.27) (1.02–1.82)* (0.82–1.52) (0.86–1.67)
– 0.90 (0.66–1.21) 0.88 (0.62–1.26)
– 1.03 (0.77–1.38) 1.02 (0.77–1.37)
0.90 (0.69–1.17)
1.07 (0.86–1.34)
0.69 (0.45–1.04)
1.03 (0.72–1.47)
1.01 (0.73–1.39)
1.03 (0.79–1.36)
0.75 (0.60–0.94)*
1.33 (1.00–1.75)*
– 0.88 0.80 0.91 0.89
(0.65–1.19) (0.57–1.12) (0.62–1.34) (0.44–1.77)
– 1.00 1.03 0.76 0.53
(0.77–1.30) (0.76–1.39) (0.48–1.19) (0.33–0.84)*
– 1.25 1.16 1.23 1.31 1.21
(0.92–1.70) (0.84–1.58) (0.92–1.68) (0.89–1.93) (0.85–1.72)
– 0.93 1.05 0.92 0.76 0.99
(0.66–1.30) (0.80–1.38) (0.71–1.18) (0.50–1.16) (0.71–1.39)
– 1.10 (0.74–1.64) 1.05 (0.66–1.67) 1.26 (0.70–2.29)
– 1.17(0.84–1.62) 1.42 (0.99–2.03) 1.46 (0.90–2.35)
99
V. F. Davies et al.
Web-based dietary questionnaire for children Table 4. Continued
Moderating variables
Matches RR (95% CI)†
Intrusions RR (95% CI)†
Omissions RR (95% CI)†
Access to computer¶ Never‡ Sometimes Yes
1.00 1.24 (0.90–1.71) 1.21 (0.90–1.64)
– 0.69 (0.48–0.99)* 0.88 (0.60–1.26)
– 1.10 (0.80–1.51) 0.86 (0.62–1.20)
†
Rate ratio related to the reference group/category. Reference group/category. § Children’s answer. ¶ Parents’ answer. *P < 0.05. CI, confidence interval; RR, relative risk. ‡
Figure 2 Mean number of omissions by school grade and total of the food items consumed.
relation to the quality of the observations made. Another strength was the large number of participants, which enabled comparison of performance indicators across variables in the questionnaire. One limitation of the present study is that the validation was only conducted on the meals consumed at school, which does not allow for the results to be extrapolated for the meals consumed outside of school. However, because that the main objective of CAAFE is to monitor food consumption in the school environment, the research mirrored the real context of anticipated future use of web-based CAAFE questionnaires. Another possible limitation is the use of a convenience sample. Schools with the best access to computers and to the Internet were selected. Had the study been conducted in schools with restricted access to computers and the Internet, the results obtained could have been different. Another limitation is the difficulty of making a direct comparison between the results obtained in the present 100
study and those of other questionnaires for evaluating food consumption as a result of differences in study design, age of children, validation standards and the methods of data analysis. CAAFE usability tests (da Costa et al., 2013) indicated very good performance among children aged 7–10 years, and a reasonably low error rate. Nevertheless, observer notes during the application of the CAAFE revealed points needing further attention, such as children not understanding the questions; children’s curiosity in exploring the tool by clicking on various foods despite not having consumed them; misunderstanding that the CAAFE is a game, which results in competition among peers when completing the questionnaire and fails to focus on dietary assessment accuracy; and not finding the foods consumed among the images on the CAAFE screen. Future studies should include qualitative studies through cognitive and semi-structured interviews involving the target audience, with the aim of investigating the problems highlighted by the observers. Studies about the sources of the food consumed (brought from home or provided by the school), as well as their impact on the accuracy of the responses, would also be worth exploring. Conclusions In conclusion, the results of the present study showed that the CAAFE questionnaire provided a reasonable degree of agreement compared to the observation method. The overall matches (43%), intrusions (29%) and omissions (28%) were to be expected, given the major cognitive demands related to the dietary assessment of children (i.e. memory and attention). In addition, our findings were similar to other studies using computers to evaluate children’s food consumption, with regard to the obtained result that children older than 8 years performed better in the CAAFE. Using a food diary was ª 2014 The British Dietetic Association Ltd.
V. F. Davies et al.
also useful for improving the overall accuracy of answers. Variables that influenced the incidence of disagreement included the meal type and other school related factors, all of which require more in-depth studies. In general, the CAAFE is an attractive tool for children to self-report their diet, although further studies are required to improve its validity. Acknowledgments We gratefully thank the Education Secretary of the Prefeitura Municipal de Florianopolis, the staff of the schools, and the children and parents for their participation in the study.
Conflict of interests, source of funding and authorship The authors declare that there are no conflicts of interest. This research was funded by a grant from the Brazilian Ministry of Health (Departamento de Ci^encia, Tecnologia e Insumos Estrategicos – DECIT), the Brazilian Ministry of Education (Coordenacß~ao de Aperfeicßoamento de Pessoal de NıvelSuperior-CAPES) and the Brazilian Ministry of Science, Technology and Innovation (Conselho Nacional de Desenvolvimento Cientıfico e Tecnol ogico (CNPq). VFD was supported by the CAPES – Brazilian Federal Agency for Support and Evaluation of Graduate Education. VFD participated in its design and coordination of the study and also wrote the first draft of the manuscript. EK, MAAA and PFP were leaders in CAAFE design, coordinated the research and co-wrote the manuscript. NS and TD gave feedback and guided the development of the manuscript. All authors critically reviewed the manuscript and approved the final version submitted for publication.
References Adamson, A.J. & Baranowski, T. (2014) Developing technological solutions for dietary assessment in children and young people. J. Hum. Nutr. Diet. 27(Suppl. 1), 1–4. Agresti, A. (2007) An Introduction to Categorical Data Analysis. Hoboken, NJ: Wiley. Assis, M.A., Benedet, J., Kerpel, R., Vasconcelos, Fde.A., Di Pietro, P.F. & Kupek, E. (2009) Validation of the third version of the Previous Day Food Questionnaire (PDFQ-3) for 6-to-11-years-old schoolchildren. Cad. Saude Publica 25, 1816–1826.
ª 2014 The British Dietetic Association Ltd.
Web-based dietary questionnaire for children
Baranowski, T., Dworkin, R., Henske, J.C., Clearman, D.R., Dunn, K., Nader, P.R. & Hooks, P.C. (1986) The accuracy of children’s self report of diet: Family Health Project. J. Am. Diet. Assoc. 86, 1381–1385. Baranowski, T., Islam, N., Baranowski, J., Cullen, K.W., Myres, D., Marsh, T. & de, M.C. (2002) The food intake recording software system is valid among fourth-grade children. J. Am. Diet. Assoc. 102, 380–385. Baranowski, T., Islam, N., Douglass, D., Dadabhoy, H., Beltran, A., Baranowski, J., Thompson, D., Cullen, K.W. & Subar, A.F. (2012) Food Intake Recording Software System, version 4 (FIRSSt4): a self-completed 24-h dietary recall for children. J. Hum. Nutr. Diet.27(Suppl. 1), 66–71. Baxter, S.D., Hardin, J.W., Guinn, C.H., Royer, J.A., Mackelprang, A.J. & Devlin, C.M. (2010) Children’s body mass index, participation in school meals, and observed energy intake at school meals. Int. J. Behav. Nutr. Phys. Act. 7, 1–8. Biltoft-Jensen, A., Trolle, E., Christensen, T., Islam, N., Andersen, L.F., Egenfeldt-Nielsen, S. & Tetens, I. (2012) WebDASC: a web-based dietary assessment software for 8–11-year old Danish children. J. Hum. Nutr. Diet.27(Suppl. 1), 43–53. Biltoft-Jensen, A., Bysted, A., Trolle, E., Christensen, T., Knuthsen, P., Damsgaard, C.T., Andersen, L.F., Brockhoff, P. & Tetens, I. (2013) Evaluation of Web-based Dietary Assessment Software for Children: comparing reported fruit, juice and vegetable intakes with plasma carotenoid concentration and school lunch observations. Br. J. Nutr. 110, 186–195. B€ ornhorst, C., Huybrechts, I., Ahrens, W., Eiben, G., Michels, N., Pala, V., Molnar, D., Russo, P., Barba, G., Bel-Serrat, S., Moreno, L.A., Papoutsou, S., Veidebaum, T., Loit, H.M., Lissner, L. & Pigeot, I.; IDEFICS Consortium (2013) Prevalence and determinants of misreporting among European children in proxy reported 24 h dietary recalls. Br. J. Nutr. 109, 1257–1265. Burrows, T.L., Martin, R.J. & Collins, C.E. (2010) 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. 110, 1501–1510. Collins, C.E., Watson, J. & Burrows, T. (2010) Measuring dietary intake in children and adolescents in the context of overweight and obesity. Int. J. Obes. (Lond) 34, 1103– 1115. da Costa, F.F., Davies, V.F., Schmoelz, C.P., Kuntz, M.G.F. & de Assis, M.A.A. (2012) Medida da atividade fısica de criancßas: o que professores de educacß~ao fısica t^em a nos dizer? Rev. Bras. Ativ. Fis. e Sa ude. 17, 286–287. da Costa, F.F., Schmoelz, C.P., Davies, V.F., Di Pietro, P.F., Kupek, E. & de Assis, M.A. (2013) Assessment of diet and physical activity of Brazilian schoolchildren: usability testing of a web-based questionnaire. JMIR Res. Protoc. 2, e31. Davies, V.F., Kupek, E., de Assis, M.A., Engel, R., da Costa, F.F., Di Pietro, P.F., Natal, S., Thompson, D. & Baranowski,
101
Web-based dietary questionnaire for children
T. (2014) Qualitative analysis of the contributions of nutritionists to the development of an online instrument for monitoring the food intake of schoolchildren. J. Hum. Nutr. Diet. 31, 1–10. Foster, E., Hawkins, A., Delve, J. & Adamson, A.J. (2014) Reducing the cost of dietary assessment: selfcompleted recall and analysis of nutrition for use with children (SCRAN24). J. Hum. Nutr. Diet. 27 (Suppl. 1), 26–35. van Gelder, M.M., Bretveld, R.W. & Roeleveld, N. (2010) Web-based questionnaires: the future in epidemiology? Am. J. Epidemiol. 172, 1292–1298. Illner, A.K., Freisling, H., Boeing, H., Huybrechts, I., Crispim, S. & Slimani, N. (2012) Review and evaluation of innovative technologies for measuring diet in nutritional epidemiology. Int. J. Epidemiol. 41, 1187–1203. Instituto Brasileiro de Geografia e Estatistica (IBGE). (2010a) Pesquisa de Orcßamentos Familiares 2008/2009 (POF). Available at: http://www.ibge.gov.br/home/presidencia/ noticias/noticia_visualiza.php?id_noticia=1699&id_pagina=1/ (accessed on 10 March 2014). Instituto Brasileiro de Geografia e Estatistica (IBGE). (2010b) Censo Demografico. Available at: http://www.ibge.gov.br/ home/estatistica/populacao/censo2010/resultados_gerais_ amostra/default_resultados_gerais_amostra.shtm (accessed on 23 June 2014). Instituto Brasileiro de Geografia e Estatistica (IBGE). (2012). Available at: http://www.ibge.gov.br/home/estatistica/ populacao/trabalhoerendimento/pnad2012/sintese_ defaultpdf_rendimentos.shtm (accessed on 23 June 2014). Kohlmeier, L., Mendez, M., McDuffie, J. & Miller, M. (1997) Computer assisted self-interviewing: a multimedia approach to dietary assessment. Am. J. Clin. Nutr. 65 (Suppl.), 1275S– 1281S. Lanctot, J.Q., Klesges, R.C., Stockton, M.B. & Klesges, L.M. (2008) Prevalence and characteristics of energy underreporting in African-American girls. Obesity 16, 1407–1412. Lioret, S., Touvier, M., Balin, M., Huybrechts, I., Dubuisson, C., Dufour, A., Bertin, M., Maire, B. & Lafay, L. (2011) Characteristics of energy under-reporting in children and adolescents. Br. J. Nutr. 105, 1671–1680. Livingstone, M.B.E., Robson, P.J. & Wallace, J.M.W. (2004) Issues in dietary intake assessment of children and adolescents. Br. J. Nutr. 92, S213–S222. Lohman, T.G., Roche, A.F. & Martorell, R. (1991) Anthropometric Standardization Reference Manual. Champaign, IL: Human Kinetics.
102
V. F. Davies et al.
Lu, A.S., Baranowski, J., Islam, N. & Baranowski, T. (2012) How to engage children in self administered dietary assessment programmes. J. Hum. Nutr. Diet. 27(Suppl. 1), 5–9. Magarey, A., Watson, J., Golley, R.K., Burrows, T., Sutherland, R., McNaughton, A.S., Denney-Wilson, E., Campbell, K. & Collins, C. (2011) Assessing dietary intake in children and adolescents: considerations and recommendations for obesity research. Int. J. Pediatr. Obes. 6, 2–11. Moore, L., Tapper, K., Dennehy, A. & Cooper, A. (2005) Development and testing of a computerised 24- h recall questionnaire measuring fruit and snack consumption among 9–11 year olds. Eur. J. Clin. Nutr. 59, 809–816. Moore, H.J., Ells, L.J., McLure, S.A., Crooks, S.C., Cumbor, D., Summerbell, C.D. & Batterham, A.M. (2008) The development and evaluation of a novel computer program to assess previous-d ay dietary and physical activity behaviours in school children: the Synchronised Nutrition and Activity ProgramTM (SNAPTM). Br. J. Nutr. 99, 1266–1274. Moore, H.J., Hillier, F.C., Batterham, A.M., Ells, L.J. & Summerbell, C.D. (2014) Technology-based dietary assessment: development of the Synchronised Nutrition and Activity Program (SNAPTM). J. Hum. Nutr. Diet. 27 (Suppl 1), 36–42. Ngo, J., Engelen, A., Molag, M., Roesle, J., Garcia-Segovia, P. & Serra-Majem, L. (2009) A review of the use of information and communication technologies for dietary assessment. Br. J. Nutr. 101, S102–S112. Ortiz-Andrellucchi, A., Sanchez, P.H., Villegas, A.S., Quintana, L.P., Mendez, M. & Serra-Majem, L. (2009) Dietary assessment methods for micronutrient intake in infants, children and adolescents: a systematic review. Br. J. Nutr. 102, S87–S117. Rangan, A.M., Flood, V.M. & Gill, T.P. (2011) Misreporting of energy intake in the 2007 Australian Children’s Survey: identification, characteristics and impact of misreporters. Nutrients 3, 186–199. Universidade Federal de Santa Catarina. (2013) Sistema de monitoramento do consumo alimentar e atividade fısica de escolares. Available at: http://caafe.ufsc.br/ (accessed on 19 March 2013). Vereecken, C., Covents, M., Maes, L. & Moyson, T. (2014) Formative evaluation of the dietary assessment component of Children’s and Adolescents’ Nutrition Assessment and Advice on the Web (CANAA-W). J. Hum. Nutr. Diet. 27(Suppl. 1), 54–65.
ª 2014 The British Dietetic Association Ltd.
Copyright of Journal of Human Nutrition & Dietetics is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
