Nutrition and Public Health
(Module code: MMB423283)
Eman Salah Zaky Sayed Ahmed
Presented to the School of Health and Life Science
in partial fulfilment of the requirements for the degree of
MSc Clinical Nutrition and Health
Prevention of Childhood Obesity
While considering the prevention of childhood obesity, several school-based interventions aimed to modify dietary behaviour in primary school children were reviewed and related intervention characteristics to effectiveness.
The literature review was carried on by searching three databases to identify randomised controlled trials (RCTs) of school-based obesity prevention programmes, aimed at children aged 5 to 18 years. The studies were evaluated for primary outcome that measures BMI; core intervention strategies duration and methods; results and findings; follow-up length and measures.
Out of the six interventions, one was found to be effective immediately post intervention, irrespective of its short duration. Studies have shown no differences between intervention and control students for BMI z-scores and body fat. The majority of interventions conducted follow-up, however, they have shown a loss of effectiveness from post-intervention to follow-up.
School-based nutrition education interventions are associated with insignificant improvement in weight outcomes in mixed-weight children and adolescents. No intervention strategy consistently produced benefits. Numerous trials characteristics have been identified. These findings may inform the development of future interventions targeting dietary behaviour in children in the school-based setting.
In the past decade, the prevalence of childhood obesity has increased significantly worldwide raising public health concerns. In spite of the numerous studies highlighting the physiological causes and complications of obesity among children. However, the growing overweight and obesity rates remain the primary childhood health problem in developed nations and, to some degree, in other parts of the world (Ebbeling, Pawlak & Ludwig, 2002).
According to a report produced by the Royal College of Paediatrics and Child Health (2015), the recent trends have shown that nearly one in five 5-year-olds and one in three in 10-year-old children are overweight or obese in the UK. Whereas, almost 17 percent of children aged 2 to15-year-olds were at risk of obesity, based on the Scottish Health Survey 2014 (Obesity Action Scotland, 2017). Latest Food standard agency report, expecting that 40% of the population in Scotland will be obese by 2030 (FSA, 2017).
Although, this epidemic has affected most ethnic groups, yet it has shown a disproportionate way in the distribution according to socioeconomic class. Evidently, extreme obesity is exacerbating particularly in the most deprived areas in the developed countries. In 2014, a survey was conducted in Scotland revealed that 22 percent of children in the most deprived categories were at risk of obesity, compared to 13 percent in the highest income category (Obesity Action Scotland, 2017). Conversely, childhood obesity is most frequent in upper socioeconomic classes of developing nations, probably due to the adoption of Western lifestyle (Ebbeling, Pawlak & Ludwig, 2002).
Obesity is a multifactorial problem with many contributing causes, including dietary behaviour, genetics and environment. Availability of energy dense foods containing concentrated calories from high fat and sugar. Moreover, the rapid decline in levels of physical activity mirrors the rise in obesity thus considered a major contributory factor. (Webster-Gandy, Madden & Holdsworth, 2012).
Evidence support that some individuals are genetically susceptible to the effects of an obesogenic environment more than others. These environmental influences are defined as the sum of surrounding conditions promoting obesity. It includes the socio-cultural and economic factors related to food varying from changes in food type, availability and marketing (Nishtar, Gluckman & Armstrong 2016).
Being obese or overweight is associated with devastating clinical consequences. It increases the risk of developing a range of serious conditions, including type 2 diabetes, hypertension and hyperlipidaemia which are the major contributors to cardiovascular disease. Furthermore, long-term effect on health includes some cancers and premature death (Government, 2016). Besides the physical aspects, obese children tend to suffer from low self-esteem, less social interaction and poorer academic achievement. Also, it is a strong predictor of adult obesity (Webster-Gandy, Madden & Holdsworth, 2012).
The economic burden of childhood obesity considered another critical dimension of the escalating problem. Annually, the estimated spending on medical expenses of conditions related to being overweight or obese in the UK is about £6 billion. Whereas, less than £638 million a year was spent on obesity prevention programmes (RCPCH, 2015).
BMI percentile charts are recommended for routine clinical diagnosis and to identify obesity. It is associated with high specificity and moderate sensitivity (SIGN, 2010).
Children with a BMI above the 98th centile are clinically obese. For epidemiological studies, those above the 95th centile are classed as obese.
Although it's an indicative method, however, it is not precise enough to be used independently to predict or estimate body composition, especially for individuals. Therefore, measuring skin fold thickness and waist circumference is used to predict body fat. Anthropometric measurements can be expressed as Z-scores (Webster-Gandy, Madden & Holdsworth, 2012).
Indorsing healthy habits during childhood influence wellbeing later in life, therefore most efforts to prevent and reduce obesity in children have used either family based or school-based approaches as a first key step towards controlling this condition (Government, 2016). In addition to the strong association between children's health and their cognitive and intellectual capacity, which has led to promoting healthy behaviours through implementing school nutrition education programmes (Langford et al., 2014).
This relationship between health and education forms the basis of the National Institute for Health and Care Excellence (NICE) advises to support primary and secondary schools to adopt a whole school approach oriented towards prevention (Lavis & Robson, 2015). Even though, evidence has been slow to show that school nutrition interventions and physical activity programmes are effective at reducing the prevalence of obesity in children (Rush et al., 2015).
The aim of this review is to determine the quality of the studies on the effectiveness of school-based health awareness programmes, which included nutritional intervention, in preventing obesity among children. A number of randomised control trial studies were selected, where there was a use of students, and the primary outcome measured BMI.
The search for the current literature review was conducted between 14th February and 2nd March 2017.
The inclusion criteria of the eligible studies for this review were, clustered randomised control trials (RCTs) of school-based obesity prevention programmes, aimed at children aged 5 to18 years. Studies were categorised according to the focus on the nutritional intervention, including educational sessions, diet management, direct provision of food and drink and behavioural counselling. Trials included all students' males and females, of all weights. Primary outcome measures childhood obesity defined by anthropometric measures BMI, waist circumference and skin fold thickness.
There were no restrictions placed regarding who delivered the intervention, the duration of the intervention and the engagement of families. Although the selection of papers relied on trials conducted in the UK, America, Brazil and Norway, there were no restrictions on the range of cultures, ethnicities and socioeconomic status of the participants included in the studies.
Excluded, were studies that focused on physical activity only. In addition to, those involved pre-schoolers or the main setting for the intervention was the local community or home. Moreover, trials aimed at treatment, not prevention were not included. No protocol or pilot research articles were included. Only full published articles in the last 6 years from English publications were considered eligible.
Search strategy, articles were identified through a two-stage search:
First, Medline - PubMed and EBSCO (CINAHL) databases were accessed through the Glasgow Caledonian University Library and the University of Strathclyde library:
The following search terms were searched as keywords: Effect AND randomised AND BMI AND School AND Obesity AND prevention
• The search was filtered by Date, from 2011 to 2017
• Studies would refer to findings in humans only
• All controlled trials (interventions), randomised or non-randomised
• English language only articles to be selected
• Peer-reviewed papers to be given priority
The search and filtering process using these criteria resulted in the identification of 16 papers. After applying the inclusion criteria two papers were found suitable for the review.
Applying the same search terms to the Cochrane library database and filtering by trial in addition to considering the inclusion criteria one more article was found out of 81 studies. The full text was not available, thus have been saved and downloaded at a later date.
The second stage of the search and filtering process included systematic reviews. Initially by using Cochrane Database of Systematic Reviews, the same search terms and filtering the results according to Cochrane Reviews. Out of the 125 reviews, The WHO Health Promoting School framework for improving the health and well-being of students and their academic achievement, was selected. Another two relevant articles were hand searched from the reference list.
The last paper found related to the inclusion criteria, was selected after reviewing the articles in a systematic review about prevention of overweight and obesity in children and youth (Peirson et al., 2015).
Using the inclusion and exclusion criteria previously mentioned, studies were assessed by reading the titles and abstracts to identify whether the paper met the pre-determined eligibility criteria. Once selected, a second review of the full text to ensure that only the appropriate studies were included. Suitable studies were saved into a personal database and printed at a later date.
The reference lists of the selected papers were checked and this provided an opportunity to identify additional articles added to the background information used in the review.
The quality of studies was assessed based on the Journal rating measured by Impact factor, which determines the importance of the journal.
According to the web of science and SCOPUS, the original peer-reviewed articles obtained from the following journals:
Journal Impact factor
British Journal of Sports Medicine (BJSM) 6.724
Obesity (Silver Spring, Md.) 4.389.
PLOS ONE 3.234
BMC Public Health 2.209
Journal of School Health 1.547
Data Extracted by using detailed tables found in the appendix.
Key websites used to provide secondary supporting evidence and data included:
• Royal College of Paediatrics and Child Health - www.rcpch.ac.uk
• The Lancet public health - www.thelancet.com
• Obesity Action Scotland- www.obesityactionscotland.org
• Public Health England - www.gov.uk
• SIGN 115 - www.sign.ac.uk
• National Institute for Health and Care Excellence - www.nice.org.uk
• Food Standards Agency Scotland - http://www.foodstandards.gov.scot/
Primary prevention of obesity is critical because of a strong association between overweight and lower academic achievement (Langford et al., 2014). Ideally, dietary education should target all children, however, the emphasis on primary schools of most interventions because the dietary behaviours are forming in this age group (Sharma, 2011).
The purpose of this review was to assess the effectiveness of school-based programmes for preventing childhood obesity that had a nutritional component. Six peer-reviewed research studies that have been published through the last five years were evaluated. Half of the studies were from the United States, while the rest reflects the problem in UK, Norway and the last study from a developing country, Brazil. Given below are the key findings, analysis and discussion.
These studies exclusively measured the BMI of the participants as the primary outcome, after implementing an intervention focused on nutrition combined with physical activity. The age target ranged between 7 and 13 years old students from both genders.
Most of the selected studies included students from Lower socioeconomic class irrespective of their ethnicity. Four of the studies focused their trials on schools with low-income groups. This might be attributed to the increased prevalence of overweight and obesity among children of a low socioeconomic status (Obesity Action Scotland, 2017).
Clustered randomised control trials at the level of schools was used predominantly. Details of randomisation and how clusters had been allocated were provided by descriptive flow charts. Also, the details of the enrolment methods and number of dropouts have been described sufficiently. Moreover, half of the studies reported similarity between intervention and control groups at baseline with respect to age, gender and body composition.
Nevertheless, one study carried out in Rio de Janeiro, which used paired cluster randomised trial design. The sample included 20 classes from 20 schools, randomised according to CONSORT guidelines of RCT, however, it was not applied completely. In addition to the higher prevalence of obesity among the control group which subjected the randomisation process to bias.
Most studies implemented combined nutritional education and physical activity, apart from one study (Cunha et al., 2013), which focused on demonstrating the impact of nutrition education sessions alone on reducing BMI and body fat percentage. Therefore, the primary educator was a trained nutritionist.
On the other trials, the educational sessions were delivered primarily by teachers during classroom lessons, which is a logic choice because they can easily implement the intervention. However, inconsistencies in the lesson delivery and the level of adherence of teachers to the programmes was not assessed, which increase the risk of intervention infidelity.
According to Johnston (2013), teachers received continual assistance by health professionals. This study focused on the teachers and implementation of the message more than the real impact of the programme on students. On the contrary, another study used students as the primary educators to approach their peers during lunchtime (Bogart et al., 2016). The aim was to increase student advocacy. Yet, the data was collected from students by conducting surveys on the 7th graders then again after two years, this subjective method increased the risk of bias.
Family involvement was an essential aspect of all studies. Either they were asked to sign a consent to obtain data, or in an attempt to increase behavioural changes and make an impact on weight gain.
Despite the considerable heterogeneity in intervention strategies, results have been almost consistent across the chosen studies. Five out of the six stated that there were no significant differences in the primary outcome measuring BMI.
Fairclough (2013) on the other hand, reported significant effects on BMI z-scores and waist circumference specifically among girls, obese participants and those from a higher socioeconomic class. His result is in context with a systematic review that focused on prevention of overweight and obesity in children. It concluded that prevention interventions in mixed-weight children showed a small effect in terms of a lowered BMI z-score (Peirson et al., 2015). However, the short-term intervention disabled the researchers to objectively estimate energy intake and macronutrient distribution. Therefore, positive effect on body size was not adequately explained.
In terms of the impact of the program, most studies have shown positive changes in dietary behaviour, food choices and Knowledge. In addition to the favourable effects of increasing physical activity and reducing sedentary behaviour in reducing weight gain and preventing obesity.
One study achieved good results regarding behavioural changes in students' eating habits and reduction of sugar intake (Cunha et al., 2013). While another described the modest effect of environmental modification on dietary fat intake (Williamson et al., 2012).
Studies which have used a holistic environmental approach, (Bogart et al., 2016; Grydeland et al., 2013; Williamson et al., 2012), demonstrated a greater potential to influence adiposity even in a small sector of the participants. These modifications in the environmental cues included food provision, healthy cafeteria food choices and accessible sports equipment with pedometers.
Two studies addressed the identified harms following the school-based diet intervention study, (Johnston et al., 2013) and (Cunha et al., 2013). The first one reported the negative impact on academic achievement and both on body weight. A significant percentage of the students became overweight or obese after the intervention. It indicates a problem either in the approach, the prevention programme itself or in the delivery method. This highlights the reliability of outcome data and the transparency of reporting the results of the research.
There are several common problems were identified among the previous studies. First, the complex multi-level intervention strategy presented in almost all the selected studies formed a challenge in determining the impact of nutritional intervention in isolation from other interventions, in preventing childhood obesity.
Second, the majority of the studies stated that neither students nor investigators were adequately blinded for the condition. However, blinding of participants in such interventions is generally not possible (Langford et al., 2014).
Another limitation, the wide variation in the total number of participants among studies. It ranged from 2165 down to 318 recruited students. The sample size is associated with many factors such as the duration of intervention, trial conditions and the costs. Despite the large sample size in Grydeland (2013) trial, the cluster sizes were unequal which might affect its strength. On the other hand, the only study reported significant effects on BMI z-scores and waist circumference had the smallest sample size, which increased the risk of performance bias.
Regarding Anthropometric data collection and measurements, all studies conducted in 2013 stated that the anthropometric measurements were obtained at baseline by trained staff. They have also considered maturity offset values using sex-specific regression equations. Moreover, Williamson (2012) declared using advanced DXA technology and body impedance in estimating body fat, which is a sensitive outcome measure in childhood obesity. Whereas, less effective measures were used in collecting the same data during follow-up which reflected a weakness.
Additionally, most of the studies relied on surveys, Food recall or Food frequency questionnaire to estimate the usual intake of their participants at baseline. These methods depend on memory, thus there is a possible bias in reporting.
Williamson (2012) used digital photography to measure food selection and intake. The study described incorporating a computer application to estimate food portions. This innovative method provided more reliable and accurate results instead of counting on reporting food intake in dietary assessment. On the contrary, Grydeland (2013) did not provide data about the tools used to measure food intake and physical activity.
Much diversity was found in the duration of the interventions in the school-based settings, varying from five weeks, the shortest intervention (Bogart et al., 2016), up to three academic years (Williamson et al., 2012). Although Fairclough (2013) has shown significant results in the five months trial, it is usually interventions of longer than 6 months considered a minimum duration to detect changes in adiposity indices (Sharma, 2011). In contrast, (Grydeland et al., 2013; Johnston et al., 2013; Williamson et al., 2012) these studies tended to be of longer duration, which inevitably increases the possibility of attrition overtime (Langford et al., 2014).
Post-intervention follow-up ranged from six months to two years were reported in five studies. Most researchers obtained anthropometric measures or conducted surveys at two points after the intervention to assess the adherence of participants and control groups. Conversely, Johnston (2013) provided inadequate data about the follow-up process. While, Bogart (2016), stated that using district records for height and weight outcomes two years post the intervention, was a limitation to the study.
Overall, drop out analyses demonstrated higher rates of participants were lost to follow-up. A possible explanation for this attrition could be resistance to assessment of anthropometrics.
While interventions may be able to produce short term changes in behaviours, a specified period of follow-up is not likely to determine these outcomes. Researchers need to implement follow-up studies in order to determine the long term impact of dietary interventions.
There are no significant results to draw a specific conclusion about the effectiveness of school-based intervention involving nutritional awareness, aimed to prevent childhood obesity. There is limited evidence for the efficacy of these programs to slow the rate of weight gain in children.
No intervention strategy consistently produced benefits. Primary prevention of obesity must include an increase in intensity of the programmes, duration of the intervention and following up with the participants to draw a conclusion about the impact of these programmes.
This might be due to the numerous variations across the implemented programs which make it difficult to isolate any determining features of effectiveness.
Further work is required to test intervention fidelity and the sustained effectiveness of this approach in the medium and long term. Researchers should be encouraged to use validated, objective measures to assess the sample.
In addition, more studies are required to assess the impact of health-promoting in sc
...(download the rest of the essay above)