BACKGROUND
It is increasingly recognised that people living in deprived areas not only experience poorer health, but are less physically active and face greater environmental barriers to physical activity (PA) compared with their more affluent counterparts. Disproportionate rises in childhood obesity within lower socio-economic groups might similarly be attributable to their socio-economic environment. The school influences approximately 40% of children’s waking time and, therefore, children’s health-related behaviour at school could also be important in the development of overweight and obesity. Although there has been some research activity in this area, studies using objective PA measurement have often been limited by small numbers and narrow age ranges. This has created a need to further explore school day PA in larger groups of children, with a broader age range, and the potential impact of their socio-economic environment on PA and overweight or obesity (OWOB).
OBJECTIVES
To identify important predictors of PA during the school day and explore the relative contributions of activity levels and the socio-economic environment in relation to OWOB in school children.
METHODS
The study was conducted on a stratified (by Year group and level of deprivation) sample of 32 class groups from primary (n=14) and secondary (n=8) schools across Stoke-on-Trent, UK, between March and April 2006. Actigraph (GT1M) accelerometers were fitted to participating children at the start of the school day and collected at the end, recording minute by minute activity for six hours between 9:15am and 3:15pm. Any children for whom a complete day of accelerometer data was not obtained (e.g. left school early) were excluded from analysis. Total PA was reported as the total number of counts (c) normalised by weight (cn=c/wt x 1000) to account for the positive effect of child’s size on accelerometer output. Time spent in moderate-vigorous intensity activity (MVPA) was defined by cut-off points calculated from the most commonly used algorithm derived elsewhere. Body Mass Index (BMI) was used to classify pupils as normal weight or overweight/obese (using the 85th and 95th percentile, respectively, relative to the British Growth Rate 1990 data set). Children’s home postcodes were used to determine the deprivation level of their area of residence according to the Index of Multiple Deprivation 2004 (based on; income, employment, health, education, housing and services, crime, and living environment). Independent samples t-tests and ANOVA were used to compare group means for total PA. Linear and logistic regression analyses identified correlates of total PA and the likelihood of children being OWOB, respectively.
RESULTS
Out of 851 eligible pupils, assent (and parental consent) was obtained for 827. Data from 716 pupils (374 boys and 342 girls) aged between 7 and 16 years (mean 11.8±2.8) were included in analysis, with exclusions for absence (n=73), technical failure (n=23), and incomplete accelerometer data (n=15). High deprivation levels in Stoke-on-Trent were reflected by almost half of pupils (46.3%) living in areas that fall within the bottom 20% of national rankings for deprivation, and only 13.6% within the top 50%. There was a marked difference in mean MVPA between boys and girls (59.54 vs. 45.28 min/day, p<.001) and a reduction across the Year groups, from Yr 3 to Yr 11 (83.22 vs. 19.39 min/day, p<.001). Linear regression confirmed that increasing age was the strongest negative predictor of total PA (regression coefficient (B)=-.443, p<.001), followed by BMI (B=-.291, p<.001), and the positive influence of school physical education (PE) lessons (B=1.718, p<.001). Female gender (p<.001), weather prohibiting outdoor play (p<.001), and a more deprived area of residence (p=.007) were also significant factors that were negatively associated with total PA. Logistic regression analysis revealed that increasing age, (OR=.788, CI=.728-.853, p<.001), female gender (OR=.788, CI=.468-.966, p=.032), and higher total PA (OR=.672, CI=.608-.742, p<.001), significantly reduced the likelihood of children being overweight/obese. Deprivation level of children’s area of residence on the other hand, was not predictive of OWOB (p=.672). Caution should be observed over the apparent age and gender effects because older children, especially girls, were more likely to be absent or to refuse participation. This resulted in under-representation of older children (particularly girls) in our sample. Despite this, a potential role for PA during the school day in the prevention of childhood overweight/obesity was confirmed.
CONCLUSIONS
School PA appears to be more important in the development of overweight/obesity than the child’s socio-economic environment. School PE lessons and outdoor play make significant contributions towards total PA. These findings suggest that manipulation of the school environment to increase PA should be an important part of strategies to reduce the development of childhood overweight and obesity.
