Risk factors for obesity: further evidence for stronger effects on overweight children and adolescents compared to normal-weight subjects.

Background: We recently showed that in preschoolers risk factors
for overweight show stronger associations with BMI in children with
high BMI values. However, it is unclear whether these findings
might also pertain to adolescents. Methods: We extracted data on
3–10 year-old (n = 7,237) and 11–17 year-old (n = 5,986) children
from a representative cross-sectional German health survey (KiGGS)
conducted between 2003 and 2006 and calculated quantile regression
models for each age group. We used z-scores of children's body mass
index (BMI) as outcome variable and maternal BMI, maternal smoking
in pregnancy, low parental socioeconomic status, exclusive
formula-feeding and high TV viewing time as explanatory variables.
Results: In both age groups, the estimated effects of all risk
factors except formula-feeding on BMI z-score were greatest for
children with the highest BMI z-score. The median BMI z-score of
11–17 year-old children with high TV viewing time, for example, was
0.11 [95% CI: 0.03, 0.19] units higher than the median BMI z-score
of teenage children with low TV viewing time. This risk factor was
associated with an average difference of 0.18 [0.06, 0.30] units at
the 90th percentile of BMI z-score and of 0.20 [0.07, 0.33] units
at the 97th percentile. Conclusions: We confirmed that risk factors
for childhood overweight are associated with greater shifts in the
upper parts of the children's BMI distribution than in the middle
and lower parts. These findings pertain also to teenagers and might
possibly help to explain the secular shift in the upper BMI
percentiles in children and adolescents.