What Is Known
- Childhood obesity is common in Malta.
- This has been documented via sampling studies, as in all other countries to date.
What Is New
- To our knowledge, this is the first country to measure the entire school-aged children as opposed to a sampling strategy.
- The findings confirm that approximately 40% of these children are overweight or obese, with a social class gradient evidenced increasing body mass index from free state schools, to subsidized church schools to independent, fee-paying schools.
- This exercise is scalable and can be theoretically replicated in large countries using the same modus operandi.
Obesity is a worldwide and chronic disease, a blight of epidemic proportions. The numerous effects of this condition are well documented, with significant morbidity and mortality from cardiovascular, neurological, and oncological sequelae (1).
Obesity often begins in childhood (2), and Malta is by no means unique. Two earlier publications dealing with Maltese children have shown that a quarter to a third of children are overweight or obese (3,4). The first author of the present study won over the Maltese authorities into undertaking a national study in Malta, which would incorporate the measurement of the height and weight of all primary and secondary school children in Malta (and hence their body mass index). The school bags would also be measured for the purposes of a separate study.
Schooling in Malta is provided by free state schools, subsidized Roman Catholic church–run schools, and independent private schools (parents completely fee-paying). These will be respectively referred to as state, church, and independent. The type of schooling that can be afforded by families may be considered an indirect indicator of the family's socioeconomic status because independent schooling is totally funded by families, who must be affluent to bear the school fees. Church schools are partially subsidized (thus more affordable), and are locally perceived by families to be of better academic quality than the completely free state schools. For families with no income to spare on schooling, state schools are the only option; hence, the modicum of social class stratification by school type.
The measurement exercise included the entire cohort of school-aged children in Malta and not just a sample. This constitutes the first occasion wherein a country measured its entire childhood population, as opposed to a sampling methodology. In this manner, follow-up of cohorts within schools would be possible from year to year. In addition, such a high-powered sample would allow deep exploration of health inequalities amongst children, beyond the risk factors that have been identified to date in the local setting.
Ethics and Data Protection
Ethical approval for the study was obtained from the Malta University Research Ethics Committee and the Ethics Committee of the Malta Education Division. Data protection approval was obtained from the Malta National Commissioner for Data Protection. Separate permissions were obtained from the Central Church schools authorities and from the Independent schools. The study was also endorsed by the Maltese Superintendent of Public Health, who went on record stating that the present study is of national public health importance, as per Article 4B of the Malta Public Health Act 2003: “to develop and implement strategies to promote and improve public health.” Nevertheless, the choice of children and parents was respected and they could opt out of the study.
The following were excluded: children who are not able to stand on the scales unaided, children who have medical conditions, which incur the use of plaster or prostheses, signed refusal consent by the legal guardian, children identified as being particularly sensitive about being measured, and those wherein measurement was felt to be inappropriate, for example, those with diagnosed eating or growth disorders. The former included children known to experience anorexia or bulimia or similar conditions. The latter included children with chronic diseases such as achondroplasia or hypochondroplasia or growth hormone deficiency and similar conditions. No school was overrepresented.
Twenty identical scale-stadiometers were purchased (GIMA 27288 PEGASO DIGITAL SCALE) and used to measure children in State and Church Schools. Independent schools purchased their own (identical model) scales. Measurements in all schools were carried out by Physical Education teachers. To reduce interobserver bias, these teachers were trained how to measure height accurately and were also given a handout as an aide-memoire. Furthermore, they were checked once with regard to their measuring methods by physiotherapists. Teachers were also responsible for data inputting. They were aided by physiotherapists coordinated by the Malta Association of Physiotherapists, whereas other administrative staff helped to transport the scales to different schools based on a timetable created by the Education Division. Each school retained a set of scales for approximately 1 week. Using the same instruments helped reduce interobserver bias. The scales were checked and recalibrated before and after the end of each study period, and teachers were actively trained in measurement techniques using standard operating procedures developed for the purpose of the pilot European Health Examination Survey (5).
The spreadsheets used tight data validation techniques in an attempt to minimize errors in data input but errors in data entry could not be excluded (6). Similarly, during data collation, data were checked for glaring errors and corrected where possible, or excluded and marked as absent if this proved impossible to correct.
Data collection was performed for each individual school class using bespoke spreadsheets in Microsoft Excel. These have already been described in detail and will only be outlined here (6). A data entry sheet that auto filled and calculated data (such as decimal age and body mass index [BMI]) and was subject to data validation to minimize errors. This sheet also anonymized the data before sending the sheet to the Maltese Education Division for collation. Data by school were then analyzed by year. Mean age for each class was rounded to the nearest 0.5 BMI unit and this was used to automatically apply BMI cut-offs as per the current World Health Organization (WHO) cut-offs. Key values were overweight (between 85 and 95 percentiles), obese (>95th percentile), the sum of the 2 (overweight and obese), and underweight (less than the 5th percentile) (7). The principal focus of the analysis was to obtain the aggregate of overweight and obese proportions.
As data by school became available, it was rechecked for completeness and input into a master spreadsheet in which additional variables were also added, such as type of school (State, Church, or Independent). This allowed overall analysis by sex for the different types of schools and for the sum of the totals.
Data were also compared with the National Sport School (State school), a coeducational facility for secondary school students with proven sporting talent and a high potential for development in their chosen field of sport. This is achieved in a flexible academic environment that is customized to student sporting demands.
The quadratic equations of Fleiss were used for exact calculation of 95% confidence intervals for ratios (8). Chi square tests and chi square tests for trends were used throughout using the Bio-Med-Stat Excel add-in for contingency tables (9). Throughout, when values for BMI are given, units are in kg/m2.
A P value ≤0.05 was taken to represent a statistically significant result.
The present study included 46,027 children in 145 schools, age ranging from 4.7 to 17 years. A total of 4280 students were not measured (90.7% were measured) comprising 3998 absentees, 199 children who refused to be measured, and 82 students who had a pre-existing ailment that excluded them from the study. The schools included 94 state (students total measured n = 22,539), 39 church (n = 14497), and 12 independent schools (n = 4711), some of which have primary and/or secondary components, and some of which are coeducational whereas others are not. One independent school (an international school) refused to participate in the study (n = 400 students).
Table 1 depicts summary data by WHO cut-off values, by school type and by sex, including for the aggregate of both sex. More than 40% of children in Malta are overweight or obese. The proportion of obese was greater than the proportion of overweight. Levels of overweight and obesity were State>Church>Independent schools. Supplemental Digital Content 1, Table 1, http://links.lww.com/MPG/A806 provides a breakdown of Table 1 by percentage for each school year.
For male primary schools, amalgamated mean BMIs for state, church, and independent schools were 18.2, 17.9, and 17.5, respectively. For girls, the equivalent values were 18.0, 17.9, and 17.3, respectively. For male secondary schools, the equivalent values were 22.3, 22.0, and 21.1, respectively and for girls, the respective values were 22.6, 22.2, and 20.9.
The differences for overweight and obesity between types of schools were statistically significant, with comparisons performed between state and church, and between church and independent, due to the clear gradients in the dataset (Supplemental Digital Content 2, Table 2, http://links.lww.com/MPG/A807). The differences between independent and church schools were greater than that between church and state schools, despite the overall smaller number students in the former group.
Overall, and for both sexes and for the various school types, there was a trend for overweight and obesity to peak for both sexes and for all school types between years 5 and 8, and then decline (Table 1), and almost all of these trends (rises and subsequent falls) were statistically significant (Supplemental Digital Content 3, Table 3, http://links.lww.com/MPG/A808).
Despite these trends, secondary schools showed a higher proportion of overweight and obesity when compared to primary schools for both sexes and for all school types (P ≤ 0.003). Boys had an overall higher proportion of overweight and obese than girls (P < 0.0001).
The least overweight and obese group (independent schools) was compared by sex with the National Sport School data (107 boys and 59 girls at secondary school level). Male and female overweight and obese individuals totaled 8 and 8, and 12 and 10, respectively. These children had significantly lower levels of overweight and obesity as picked up by BMI.
The underweight group was small and there was no significant difference between the 3 cohorts.
Our findings confirm that Malta has a high rate of overweight and obesity among its childhood population. This has been shown by several other studies that compared various countries, not only across the European Union (10), but also worldwide (11). This is of major concern because it is known that obesity tracks into adult life (12), with attendant morbidity and mortality (13,14). The findings also show that the proportion of children who are obese exceeds those that are overweight, contrary to studies in other countries, which show that the overweight proportion exceeds the obese proportion (15). This also goes contrary to extant publications dealing with Malta, which had shown that the proportion of overweight exceeded that of the obese proportion in the samples studied (4,5). The importance of exercise in weight control is also suggested in the lower BMIs in the National Sport School. The peak in overweight and obese in years 5 to 8 and the subsequent slight decline may reflect efforts by various individuals to control weight at/around puberty, along with the effects of puberty itself. The greater proportion of overweight and obese males cannot be explained, but has been noted in several other countries (16). The gradient in type of school (state > church > independent) was previously shown in a cohort of preschool Maltese children, who all undergo mandatory physical examination before school entry. This suggests that factors that predispose to a high BMI are at play before school entry. The nature of the school gradient strongly suggests a social class effect at work, as parental financial outlay to fund education is state < church < independent. Only a small proportion of children were underweight and the effect of overt clinical disease could not be excluded. Furthermore, at approximately <3% across the board, this falls well within the expected distribution, which is defined by WHO as <5th percentile.
The principal limitation in the present study is the unavoidable exclusion of children who were absent within the measurement week when the scales were present at their school. This may have inadvertently led to a degree of exclusion bias as absentees may have had an overrepresentation with higher BMIs (obese or overweight) and who may have been indisposed to attend school due to complications arising from overweight or obesity. Furthermore, the possibility of erroneous data entry could not be entirely excluded. The school that refused to participate is an international school with an almost non-Maltese population. The refusal by the school to participate may also have introduced inescapable bias.
The present study has also provided proof of concept by demonstrating the feasibility of undertaking a relatively inexpensive study of an entire childhood population, as opposed to sampling only, for the precise elaboration of school-age BMI. The modus operandi employed in Malta could relatively easily be up scaled to measure childhood populations of any country since individual classes are measured by individual teachers. Larger countries would merely require a larger logistical office. In Malta, the present study was relatively easily achieved via the whole-hearted cooperation of 2 government ministries (Health and Education), with the direct intervention and help of the politicians in question, and not only that of the civil service. Furthermore, such a study was only possible through the active participation of physical education teachers who were responsible for data collection for the classes that they are normally responsible for teaching. The Malta Association of Physiotherapists was also invaluable in helping to train teachers in measurements, performing actual measurements and helping in overall coordination.
The present study will be further elaborated with the analysis of BMI by locality. An analysis of school bag weight in relation to child weight will also be carried out. The possibility of repeating the present study has been raised, in the same way that a census is redone regularly. Albeit carried out for the first time, since the study ran so smoothly. The possibility of including other measurements (such as skin fold thickness and waist circumference) has also been mooted. This is particularly important since the use of BMI as a measurement tool has its pros and cons. BMI measures weight relative to height rather than adiposity, and cannot distinguish between fat and fat-free mass, which may result in poor estimation of actual degree of overweight or obesity. In contrast, BMI is simple-to-use and a relatively cheap measure, which can be routinely and widely used with reasonable precision. Indeed, the American Academy of Pediatrics and the United States Preventive Services Task Force recommend the use of BMI as a population screening tool for overweight and obesity in children from a young age (17). BMI surveillance also has the backing of the American Heart Association (18). Studies of this type are crucial because the measured outcomes are costly not only to the individual but also to society. Indeed, a previous study based on obesity costings for larger countries (19) had estimated that the cost for Malta for treating the complications of obesity (when the current cohorts of obese children reach adult life) will be around Euros 70,000,000.00 per annum (3).
Because all available and eligible children are measured, the study has the highest possible statistical power. If active measures are undertaken in an attempt to curb the problem of childhood overweight and obesity, a repetition of the present study will enable the detection of even small improvements in the current situation thereby permitting the authorities to focus on and elaborate on efforts that appear to work in this regard, as in an audit process.
The authors would like to thank Hon Evarist Bartolo, Minister for Education and Employment; Hon Christopher Fearne, Minister for Health; Mr Louis Scerri, Assistant Director, Research and Development Department, Education Division; Mr Francis Fabri, Director General, Operations Department, Formerly Director, Research and Development Department, Education Division; Mr Raymond Camilleri, Director, Research and Development Department, Education Division; Members of the Malta Association of Physiotherapists; and all participating physical education teachers and physiotherapists.
1. Llewellyn A, Simmonds M, Owen CG, et al. Childhood obesity as a predictor of morbidity in adulthood: a systematic review and meta-analysis. Obes Rev
2. Freedman DS, Khan LK, Serdula MK, et al. The relation of childhood BMI to adult adiposity: the Bogalusa Heart Study. Pediatrics
3. Grech V, Farrugia Sant’Angelo V. Body mass index estimation in a school-entry aged cohort in Malta. Int J Paediatr Obesity
4. Farrugia Sant’Angelo V, Grech V. Comparison of body mass index of a national cohort of Maltese children over a 3-year interval. Maltese Med J
5. Kuulasmaa K, Tolonen H, Koponen P, et al. An overview of the European Health Examination Survey Pilot Joint Action. Arch Public Health
6. Grech V, Aquilina S, Camilleri E. Depicting the analyses of the first National Maltese Childhood BMI study. Images Paediatr Cardiol
7. World Health Organization (Internet). Geneve: Growth reference 5–19 years; BMI-for-age (5–19 years). http://www.who.int/growthref/who2007_bmi_for_age/en/
.> Accessed October 5, 2016.
8. Fleiss JL. Statistical Methods for Rates and Proportions. 1981; New York: John Wiley and Sons, 2nd ed. 14–15.
9. Slezák P. Microsoft Excel add-in for the statistical analysis of contingency tables. Int J Innovation Educ Res
10. Wijnhoven TM, Van Raaij JM, Spinelli A, et al. WHO European Childhood Obesity Surveillance Initiative 2008: weight, height and body mass index in 6-9-year-old children. Pediatr Obes
11. World Health Organization. Report of the WHO Commission for Ending Childhood Obesity. Geneva: WHO Document Production Services; 2016.
12. Lloyd LJ, Langley-Evans SC, McMullen S. Childhood obesity and risk of the adult metabolic syndrome: a systematic review. Int J Obes (Lond)
13. Deckelbaum RJ, Williams CL. Childhood obesity: the health issue. Obes Res
2001; 9 (suppl 4):239S–243S.
14. Biro FM, Wien M. Childhood obesity and adult morbidities. Am J Clin Nutr
15. Wijnhoven TM, Van Raaij JM, Spinelli A, et al. WHO European Childhood Obesity Surveillance Initiative: body mass index and level of overweight
among 6-9-year-old children from school year 2007/2008 to school year 2009/2010. BMC Public Health
16. Organisation for Economic Co-operation, Development. CO1. 7: Overweight
at Age 15 by Gender. Paris: OECD; 2014.
17. Barton M. US Preventive Services Task Force. Screening for obesity in children and adolescents: US Preventive Services Task Force recommendation statement. Pediatrics
18. American Heart Association. Policy Position Statement on Body Mass Index (BMI) Surveillance and Assessment in Schools. Dallas, TX: American Heart Association; 2008.
19. National Audit Office. Tackling Obesity. London: National Audit Office; 2001.