PARENTAL PERCEPTIONS
OF THEIR
CHILD’S WEIGHT

Introduction

Obesity is on the increase, so much so that it is now widely referred to as an epidemic of obesity (Hall and Elliman, 2006). Shepherd (2009) defines obesity as excess body fat primarily caused by a shift in the balance between energy intake and energy expenditure over a period of time. The consequences of obesity in childhood can include breathing problems/sleep apnoea, cardiovascular disease, type 2 diabetes, and not to mention the psychological problems associated with low-self esteem and depression (Stewart, 2008). Recent government statistics show that in children aged 2 – 15 years, almost one-third, are classed as overweight with 29% classed as being obese (Information Centre, 2008a). The National Child Measurement Programme (NCMP) found that in 2008, of the children who were measured, one in four aged 4 – 5 year olds and one in three aged 10 – 11 year olds was either overweight or obese.

School nurses are involved in implementing the National Child Measurement Programme (NCMP), which involves weighing and measuring all children in Reception and Year 6. This involves sending the results out to parents along with information on healthy eating and exercise with the hopes that this information will enable parents to make more informed choices and seek help to improve their family’s lifestyle (The Information Centre, 2008b). Having been involved in this process last year, we sent out the results and included details of help that was available to parents of overweight/obese children i.e. cookery classes, one-to-one support, but to my knowledge very few of the parents accessed any of these services. This got me thinking as to why this may be the case, do parents view their child’s weight differently from health professionals and are there any other factors that may influence this.

In accordance with obesity guidelines, published by the National Institute for Health and Clinical Excellence (NICE, 2006) parents should be encouraged, by health professionals, to take main responsibility for lifestyle changes in their overweight/obese child. This is supported by Lindsay, Sussner, Kim and Gortmaker (2006) who argue that interventions aimed at preventing childhood overweight and obesity should involve parents as important forces for change in their children's behaviours. Although I agree with this, we as health professionals need to understand how parents feel and think about obesity and its effects on health in order to help them change their behaviour, which in turn will influence their children.
Therefore the purpose of this literature review is, due to the rising levels of obesity in children, I wish to find out how parents actually perceive their child’s weight. Parahoo (2006) states that a literature review can be carried out to “explore aspects of a phenomenon with the aim of increasing our understanding of it” (p127). Reviewing research material on this subject will broaden my current knowledge but it may also provide me with valuable information that I can relate back to my practice area and may influence the development of new strategies when working with parents.

The sampling frame for the studies included in this literature review was primary research sources within the last 10 years that focused on parental perceptions of their child’s weight. LoBiondo-Wood and Haber (2006) suggest a general time-line of looking back 5 years for evidence-based papers but due to a lack of research articles on this topic I chose to go back 10 years to gain a better insight into my research question. Other inclusion criteria were research studies that: 1) included overweight children in the sample and 2) included children aged between 4 – 9 years in the sample. I chose this age group as it has been documented that parents have more control in shaping their children’s dietary practices and physical activity when they are younger and therefore need to take responsibility for their weight status (NICE, 2006 and Shepherd, 2009).

I used the following electronic databases to assist my search for research articles: Cinahl, PubMed, Scopus and Cochrane. Electronic search terms included parental perceptions/attitudes, children, weight and obesity, I used different groupings with every search to gain a comprehensive list of articles. Undertaking a comprehensive search, albeit time consuming, enabled me to develop my research question so that it was clear and related to my topic of interest. This is supported by Hek, Judd and Moule (2002) who state that systematic searching is fundamental to the review process as it helps to identify previous research undertaken and uncovers all relevant knowledge related to the research question.

From the six studies I read I found three themes, which include how parents view their child’s weight, factors that may influence parents perceptions and parents beliefs about their own child’s obesity.

How Parents View their Child’s Weight

The main objective in studies conducted by Carnell, Edwards, Croker, Boniface, and Wardle (2004) and De La O, Jordan, Ortiz, Moyer-Mileur, Stoddard, Friedrichs, Cox, Carlson, Heap and Mihalopoulos (2009) was to evaluate the accuracy of parents perceptions of their children’s weight status. De La O et al (2009) took this one step further in hypothesising that, “parents of overweight and obese children underestimate their child’s weight status” (p216). A hypothesis is a statement about the relationship between the variables that then predicts an expected outcome, this particular hypothesis is known as directional in that it specifies the expected direction of the relationship between the variables (LoBiondo-Wood and Haber, 2006). It could be argued that this is potentially biased as the researcher has demonstrated a commitment to a particular position in stating an anticipated outcome, however De La O et al (2009) overcome this as they have based their hypothesis on the findings of previous research studies and support it with an extensive literature review.

Non-probability convenience sampling was used for both studies, this meaning that the researcher used the most readily accessible people as subjects in their study (LoBiondo-Wood and Haber, 2006). Neither states the reasoning behind their sampling selection, this type of sampling is very easy to coordinate and is less time consuming than probability sampling, although it is not always representative of the target population (Parahoo, 2006). Both studies took place in large cities and to achieve representativeness they chose large sample sizes and undertook data collection in a range of schools with varying socioeconomic backgrounds, no students or parents were excluded from the study. Due to De La O et al (2009) making a hypothesis it may have been more appropriate to change their sample selection to a probability sample in order to examine relationships between the variables and generalise their findings (Parahoo, 2006). Using a stratified random sample allows the target population to be placed into subgroups according to the variables the researcher believes are important for inclusion, in this instance children’s weight. This would have given the researcher a valid basis for making comparisons among the subgroups and the representativeness of the sample would have been further enhanced (LoBiondo-Wood and Haber, 2006). It could be argued that choosing a large sample size assists with generalising however a poor response rate would impact on this.
Data was collected by the use of postal questionnaires using rating scales, children were weighed and measured and then parents were sent a questionnaire to complete and return. Both studies gained ethical approval and obtained consent from parents to weigh and measure their children. Questionnaires are efficient and collect a large amount of data quickly although they can be difficult to get people to complete and the response rate of postal questionnaires is relatively low (Walsh & Wigens, 2003). This is reflective of both studies as less than half of the sample returned the questionnaires.

De La O et al (2009) state that they used a validated questionnaire but do not state how they achieved this and Carnell et al (2004) only state that they used likert scales to measure parental attitudes, again no mention of the reliability or validity of this method. Likert scales are widely used in research and require respondents to indicate a level of agreement or disagreement to a question (Parahoo, 2006). Rating statements are open to individual interpretation, they may have different meanings to each person therefore could affect the reliability of the response. De La O et al (2009) discuss the possibility that the use of the word ‘extremely overweight’ instead of ‘very overweight’ might have put parents off picking this category. If the questionnaire had undergone a pilot study then this issue may have been brought up earlier and the researcher could have changed the wording prior to starting the study, this also increasing the validity of the data collection tool. A pilot study is a small-scale version of the intended study, which tests the procedures and feasibility of a study, allowing for improvements to be made prior to undergoing the main research (Cormack, 2000).

Carnell et al (2004) found that only 19% (n = 145) of parents with overweight/obese children described their child as overweight and De La O et al (2009) discovered that 86% (n = 91) of parents of obese and overweight children misclassified them as overweight or normal weight, respectively. These results imply that parents of overweight children do underestimate their weight, however both studies acknowledge limitations in that the results could not be generalised due to the poor response rate. Interestingly De La O et al (2009) found that the parents who did not respond to the questionnaires were those with children who were identified, as being overweight, maybe if these parents had responded the results would have differed. This research has proved useful in that it shows that some parents cannot recognise their child as being overweight, however using a quantitative approach prevents a fuller understanding of why parents misperceived their child’s weight, thus highlighting a need for further qualitative research.

Factors Influencing Parental Perceptions

The studies by Baughcum, Chamberlin, Deeks, Powers and Whitaker (2000) and Genovesi, Guissani, Faini, Vigorita, Pieruzzi, Strepparava, Stella and Valsecchi (2005) both identified that mothers underestimated their child’s weight, however also looked further into what factors were associated with this failure. The primary objective for Genovesi et al (2005) was to establish whether maternal education had an impact on the misperception of a child’s body weight whereas Baughcum et al (2000) looked at various other maternal demographic variables such as age, income and education.

Neither of the studies stipulated their research design although it is clear upon reading them that they used a correlational approach. This type of approach is used to investigate relationships between variables, the researchers then describe and try to explain the nature of the relationships that exist they do not examine the cause (Cormack, 2000). Baughcum et al (2000) do state that they used a cross-sectional survey, this type of study collects data on one occasion with the same subjects from varying age groups rather than with the same subjects at different time intervals (LoBiondo-Wood and Haber, 2006). This is relevant to the aims of this study, as opposed to using a longitudinal design, as the researchers were looking for answers now rather than in years to come. This is probably due to the fact that over the past decade childhood obesity has been highlighted as being on the increase therefore action needs to be taken now rather than later to reverse this rate.

Parents and children were recruited, for both studies, when they attended either a clinic or GP practice making this a non-probability sample. It is argued that this type of sample is less likely to represent the target population, however both researchers used inclusion and exclusion criteria detailing who should be included in the sample, thus enhancing the representation of the findings (LoBiondo-Wood and Haber, 2006). I feel this could have been further strengthened if the researchers had undertook the research in other settings such as schools, as people do not regularly attend their GP limiting the sample chosen. Both studies used large, demographically diverse samples and state in their discussion that these represent their target populations. The larger the sample size the more likely you are to represent the target population and generalise your findings (Parahoo, 2006). However, the results may not be applicable to the UK as the research was undertaken in Italy and America where differences exist socially and culturally and obesity figures may differ.
Questionnaires were again used to collect data although response rates were high in these studies compared to Carnell et al (2004) and De La O et al (2009) as they were given to participants on entering the clinic therefore were filled in straight away. Baughcum et al (2000) also used money as an incentive to participate, although I would argue that this might introduce an aspect of bias questioning the respondent’s motivation or commitment to answering the questions truthfully. Parahoo (2006) highlight bias as one of the greatest threats to the reliability and validity of data, therefore to ensure rigor researchers must try to avoid it at all costs. To increase the reliability and validity of the data collection tool Genovesi et al (2005) undertook a pilot study over the course of 2 weeks and test-retest showed a high level of concordance, above 90%. A test re-test is when the questionnaire is administered to the same person on two or more occasions and the results are then compared with each other (Parahoo, 2006).

Baughcum et al (2000) found that low maternal education was the only variable associated with a parent’s failure to perceive their child as being overweight. This is also reflected in Genovesi et al’s (2005) results in that there was a highly significant association between the mother’s level of education and misperception of their child’s weight. Both studies used inferential statistics and the Pearson chi-square test. The chi-square test is commonly used to test whether or not there is a relationship between two variables, in this instance education and weight (Parahoo, 2006). To ascertain the probability that relationships between variables are true, researchers often set a level of confidence such as P