Background
As a follow-up to the European Ministerial Conference on Counteracting Obesity convened
in 2006 in Turkey, the European Childhood Obesity Surveillance Initiative (COSI) was
launched by the Regional Office for Europe of the World Health Organization (WHO). COSI, a
collaboration between WHO and interested Member States, aims to monitor the magnitude
of overweight and obesity among primary-school children in European countries, to allow
intercountry comparisons and to identify regional differences for informed policy-making. It
collects at regular intervals data on weight and height of primary-school children and on their
nutrition and physical activity behaviours, as well as on school environmental characteristics
supportive to healthy nutrition and physical activity.
Methods
The research described in this PhD thesis is based on the data collected in the first two COSI
rounds by 12 European countries in school year 2007/2008 and by 13 European countries
in school year 2009/2010. Nationally representative samples of children aged 6–9 years
were drawn, whereby a majority of the countries applied a two-stage school-based cluster
sampling approach.
A total of 168 832 children in school year 2007/2008 and 224 920 children in school
year 2009/2010 were included in the anthropometric intercountry data analyses. Children's
weight and height were measured by trained examiners using standardized procedures.
Participating countries were allowed to adhere to their local legal requirements by specified
deviations from standardized procedures, such as in types of clothing worn by the children
during weight and height measurements. For each country, the prevalence of overweight and
obesity, as well as mean Z-scores of anthropometric indices of height, weight and body mass
index (BMI) were computed.
The characteristics included in the analyses on the school environment referred to the
frequency of physical education lessons, the availability of school playgrounds, the possibility
to obtain food items and beverages on the school premises, and the organization of school
initiatives to promote a healthy lifestyle. The school form was usually completed by the school
principal or the teachers involved with the sampled classes. Data from 1831 schools in school
year 2007/2008 and from 2045 schools in school year 2009/2010 were used. For each school,
a school nutrition environment score (range: 0–1) was determined whereby higher scores
correspond to higher support for a healthy school nutrition environment and the mean of the
children's BMI-for-age Z-scores calculated.
Five countries in school year 2007/2008 provided children's data on 13 health-risk
behaviours related to breakfast and food consumption frequency, physical activity, screen
time and sleep duration (n = 15 643). These data were reported by the caregivers alone or
jointly with their child. For each country, the prevalence of the risk behaviours was estimated,
and associations between them and overweight and obesity examined by multilevel logistic
regression analyses.
Results
In both school years, a wide range in overweight and obesity prevalence estimates was
found that differed significantly by country, as well as by European region. In all countries,
the percentage of overweight children was about 20% or more (range: 18–57%), and the
percentage of obese children was 5% or more (range: 5–31%). The findings suggest the
presence of a north–south gradient with the highest overweight and obesity prevalence
estimates found in southern European countries. Furthermore, changes in mean BMI-forage
Z-scores (range: from –0.21 to +0.14) and prevalence of overweight (range: from
–9.0% to +6.2%) from school year 2007/2008 to school year 2009/2010 varied significantly
among countries, whereas a period of two years is considered too short to identify these
developments. The clothes-adjusted overweight prevalence estimates were lower by as much
as 12% than the unadjusted estimates. Monthly BMI-for-age Z-score values within countries
did not show systematic seasonal effects.
Large between-country differences were observed in both school years in the availability
of food items or beverages on the school premises (e.g., fresh fruit could be obtained in
12–95% of schools) and in the organization of initiatives to promote a healthy lifestyle in
the selected classes (range: 42–97%). The provision of physical education lessons and the
availability of school playgrounds were more uniformly present across the countries (range:
76–100%). A large variation was also seen in school nutrition environment scores (range:
0.30–0.93) whereby countries with a low score (<  0.70) graded less than three out of
five characteristics as supportive. High-score countries showed more often than low-score
countries a combined absence of cold drinks containing sugar, sweet snacks and salted snacks
on the school premises.
The prevalence of all 13 health-risk behaviours differed significantly across countries. For
instance, the percentage of children who ate ‘foods like candy bars or chocolate’ > 3 days/
week ranged from 2.2% to 63.4%; this figure ranged from 1.1% to 46.5% for those who
ate ‘foods like potato chips (crisps), corn chips, popcorn or peanuts’ >  3 days/week. The
range for children who did not have breakfast every day was between 4.4% and 32.5%,
and from 4.8% to 35.0% for those who did not play outside ≥  1 hour/day. Not having
breakfast daily and spending screen time ≥ 2 hours/day were clearly positively associated with
obesity. The same was true for eating ‘foods like pizza, French fries, hamburgers, sausages or
meat pies’ > 3 days/week and playing outside < 1 hour/day. While a combination of multiple
less favourable physical activity behaviours was clearly positively associated with obesity, a
combination of the presence of multiple unhealthy eating behaviours did not lead to higher
odds of obesity.
Conclusions
The results found in both COSI school years show that overweight and obesity among 6–9-yearold
children are a serious public health concern, especially in southern European countries,
and show the need for accelerated efforts to prevent excess body weight early in life by all
participating countries. It was possible to detect relevant changes within a period of two years
but to identify clear trends within countries, a longer time interval is necessary. The data on
the school nutrition environment and the children's health-risk behaviours may assist policymakers
in monitoring their national policies targeting school settings and childhood obesity. In
particular, promoting physical activity-related and discouraging sedentary behaviours among
schoolchildren in the context of obesity preventive interventions seem to be essential.