Do children report differently from their parents and from observed data? Cross-sectional data on fruit, water, sugar-sweetened beverages and break-time foods

Insight into children’s consumption habits is important for two main reasons. Firstly, it is widely known that eating and drinking habits can contribute to the development of childhood obesity [1]. Secondly, the consumption habits that we have as children continue into adulthood, when the risk of obesity remains [2]. Therefore, gaining insight into a child’s consumption habit is important. However, for assessing dietary behaviour on the level of pre-specified foods and food groups there seems to be no perfect measurement method [3].

Interventions aimed at changing children’s consumption behaviour are commonly evaluated using information on the habitual consumption behaviour of the child reported by parents [4, 5]. Unfortunately, research has shown that parents may not always be a reliable source of information on the child’s habitual intake of foods and drinks [6, 7]. In addition, social desirability, especially among parents, may lead to over-reporting of the intake of healthy food items and under-reporting of the intake of unhealthy food items [3].

The cognitive ability of children to self-report their intake of foods and food groups is also doubtful [8, 9]. The ability to self-report improves when a child grows older [5], with some suggesting that children should at least be ≥12 years old to report more reliably on their dietary intake [10]. Other research has shown that children from the age of 8 years old may already be reliable sources of information regarding their own food intake over the past 24 h [5, 8, 11, 12]. However, due to their unfamiliarity with concepts such as ‘frequency’ and ‘averaging’, it is debatable whether children of this age respond accurately to food frequency questionnaires when items cover longer periods [5, 8, 11]. Indeed, when Börnhorst and colleagues investigated the nature and magnitude of selective misreporting of food intake of different food items, they found children’s level of under-reporting to be 8.0 %, with over-reporting at 3.4 % [3]. Other studies have shown under-reporting of intake to be significantly higher in obese children [8, 11, 13]. These same studies have also demonstrated that not only does the extent of misreporting increase with age, in contrast to previous mentioned studies [5, 10]. Also, those who under-report on one occasion are likely to under-report on a second occasion, in which case reporter bias cannot be eliminated by repeated measures. Then again, research by Burrows and colleagues suggests that children between the ages 8 and 11 years old may report reliably as compared to either their father or their mother with regard to the child’s dietary intake of specific foods [14].

As can be concluded from the arguments above, all of the available methods for measuring dietary behaviour on the level of pre-specified foods and food groups may have some degree of misreporting and error [8, 11, 15–17], which makes measuring a child’s habitual consumption a challenging aspect of behaviour research. As previous research has indicated, in order for us to choose the most accurate measurement method for assessing children’s eating behaviour, we must improve our understanding of discrepancies between observed and reported behaviour [18–20].

The aim of this study was twofold: (1) to investigate the level of agreement between observed and child-reported break-time food items; and (2) to investigate the level of agreement between children’s reports and those of their parents regarding the children’s overall consumption of fruit, water and sugar-sweetened beverages (SSB).

Our cross-sectional study used data from the population-based ‘Water campaign’ intervention [21]. This controlled trial aims to assess the effects of a combined school- and community-based intervention on children’s SSB consumption. The Medical and Ethical Review Committee of the Erasmus Medical Centre issued a ‘declaration of no objection’ (i.e. formal waver) for this study (reference number MEC-2011-183). Four primary schools located in multi-ethnic, socially deprived neighbourhoods in Rotterdam, the Netherlands, were included in the 'Water campaign' intervention study. This resulted in a total of 1288 children aged 6–13 years (grades 3–8) who were invited to participate. Passive parental consent was obtained. Parents (and children) received an information brochure to notify and inform them about the intervention and study participation. The study was also announced by the school, via the school letter and through the teachers and by flyers which were visible throughout the school. Parents (and children) were free to refuse participation without giving any explanation. They could do so by informing one of the teachers at their school or one of the researchers when present at school. At all times, the researchers could be contacted by a special phone-number or e-mail, for instance to decline participation [21].

The questionnaire items were based on previously widely used questionnaires, mainly used in earlier Dutch studies [22, 23]. We used two questionnaires to assess habitual consumption: one version was completed by children in grades 6–8 and another was completed by the parents of children in grades 3–8. Children filled in their questionnaire at school in the presence of a researcher and their teacher. The parent questionnaire was to be completed at home by the main caregiver of the child, within a period of maximum two months.

To objectively record what children brought to school to consume during break-time we used observation forms. These forms have been frequently applied for these kinds of information gathering by the Youth Health Care in recent years. Observations (unobtrusive) at school for children in grades 3–8 were conducted by trained public health professionals.

For the present study, we used baseline data from children in grades 6–8 only (aged 9–13 years). Pairing of data from the child questionnaires with data from the parent questionnaires or with data from the observations generated a study population of 539 children.

In the parent-child data pairs (n = 275), the children’s mean age was 11.1 years (SD 1.0), 54.5 % were girls, 20.2 % were of Dutch origin, and 26.0 % were overweight or obese. With regard to the caregivers, 88.8 % were female, their mean age was 38.4 years (SD 9.1), and 27.1 % were classified as having a low level of education.

The level of agreement between the observed break-time foods and the children’s self-report was poor to fair: children reported higher quantities on their break-time foods of sandwiches and snacks. There was a poor level of agreement between the reports from parents and children regarding whether children consumed fruit and SSB daily, and a moderate level of agreement for daily water consumption. Despite these differences, on a group level children and parents did report similar average amounts regarding what the child consumes on a day and whether or not he/she drinks water or eats fruit, with the highest ICC for water (good agreement). The reports on average amounts of SSB consumed differed substantially between children and parents however.

Our result of 67 % overall agreement between observed and child-reported break-time foods is similar to that seen in previous studies. For example, in a study by Weber and colleagues, children were able to correctly recall 75 % of the school meal foods that observers had documented that they had on their plate [34]. Subsequently, Baranowski and colleagues found 83 % agreement between observed and recalled 12-h food consumption among children [35]. Despite this agreement, we also found differences, for example between observed and child-reported sandwich intake. The higher reported amounts by children could be explained by the fact that while observers only counted the sandwiches that children brought to school to eat during break-time at 10:00 am, some children may well have reported the number of sandwiches brought to consume during both break-time and lunch. Apart from this, our overall findings indicate that children also report higher amounts of their break-time snacks. A further possible explanation for this lack of agreement is that children find it difficult to estimate the amount of foods and food items, as other studies have reported [8, 9, 11].

Whereas on a group level the reports from parents and children regarding average amounts of fruit and water consumed were similar, those regarding average amounts of SSB consumed were dissimilar. These discrepancies could be due to children being more aware of when and how much fruit and water they consume than how much SSB they consume [8, 9, 11]. Fruit and water may be more straightforward than SSB. Although the definition of SSB was explained to the children and examples of SSB were provided, the children may still not have completely understood what SSB is. After all, it is more difficult to recall your consumption behaviour if you do not fully understand what it is you were supposed to remember, for example when remembering which drinks to include when answering certain questions. As a further explanation for such misreporting, one could consider the argument that healthier food or drink types such as fruit and water are more likely to be over-reported than unhealthy food or drink types such as SSB. This has been addressed by Börnhorst and colleagues in the context of intentional selective misreporting by parents (of 6101 children aged 2–9; mean age 6.1 (SD 1.8)): this study found that foods perceived as unhealthy (such as sugary products) were more likely to be under-reported whereas foods perceived to be more healthy (such as fruit and vegetables) were less likely to be under-reported [3]. So it would seem that there is a tendency towards over-reporting of healthier habits and towards socially desirable answers by parents [36, 37]. This might explain why, in the current study, average amounts of SSB consumed was dissimilar between parent and child reports while average amounts of fruit and water consumed was not. Yet another possible explanation for the discrepancy between child and parent report could be that children could have bought or swapped food items, without their parents knowing. In particular, this could especially be the case for items such as SSB’s. However, the potential over-reporting of healthy foods by parents could mean in our case that water and fruit intake reports of parents are higher than the true intake. Yet, children and parents reports on the child’s intake of fruit and water were similar. The role of social desirability on the answers of parents or children therefore stays obscure. Therefore, further research is required to address whether there are indeed differences between parents’ and children’s reports of healthy and unhealthy consumption in children.

The time between the child and parent reports ranged from one day to two months. Theoretically it is possible that seasonal influences on eating patterns or changes in feeding behaviours could have contributed to the discrepancies between self-reports of parents and children. Because our questions were on daily consumption and average amounts consumed, we assume that if this time delay between reports of parents and children was of influence, it will be of small influence since habit strength in consumption and feeding patterns tends to be high [38].

Others have looked for explanations for the discrepancies seen between parents and children. For example, McPherson and colleagues [8] suggest that one of the reasons that parents and children report different intakes is that they have different perceptions of the child’s food and drink intake. Additionally or alternatively, the discrepancy between parents and children could be explained by reporting bias, with parents reporting socially desirable answers more frequently than children [3].

A possible explanation for the higher reports of snacks as break-time food item (and possible over-reporting of SSB) by children could be due to personal preferences or personal characteristics, for example. Since snacks (and SSB) are more likely to be children’s favourite food (or drink) types, this could have been reflected in their reporting behaviour. In addition, we found statistically significant differences between the children’s reports of obesity-related foods and beverages and that observed or reported by parents, however hereby we contradict the findings found by Bennett and colleagues [36].

Previous research among parents and children reporting on child’s consumption behaviour has found that parental reports are slightly more accurate than children’s reports: 78 % compared with 72 % agreement regarding different food items and food types (children under study between the ages of 6 ang 11 years old) [15]. Given the differences and low levels of agreement that we found, we would recommend combined measurement methods when assessing child’s habitual dietary behaviour on the level of pre-specified foods or food groups. This is also suggested by Eck and colleagues who concluded in their study that the children’s contribution could be of value, given that the combined parent-and-child report was more accurate that the individual parent report (children under study between the ages of 4 and 10 years old) [7]. Reports from multiple sources are therefore recommended over single-source reports, especially for children younger than the age of 8 years old as proposed by Burrows and colleagues [4]. As seen in other public health fields, the question then remains on how to process data from multiple sources [39].

Children not only report higher than observed on amount of break-time foods of sandwiches and snacks, they also report a higher SSB intake than that reported by their parents. However, children and parents have similar estimations of the child’s water and fruit consumption. Since the level of agreement between the observed break-time foods and that reported by children and the agreement of child’s intake between parent and child reports are relatively weak, future studies should focus on improving methods of evaluating children’s consumption behaviour or on ways on how to best use and interpret multiple-source dietary intake data.