thus, by weight gain during pregnancy.
Pregnancy is a vulnerable period in
women’s life and, in addition, it is temporary.
Successfully tackling the preventable
determinants reduces the
prevalence of childhood overweight by
31.4%. This number is in line with data
of a recent Canadian study where the
preventable proportion of overweight
was estimated to reach 40%.12 In that
study, additional social factors, such as
neighborhood dwelling value, were included
accounting for ∼10% of reduction.
Accordingly, the data of Toschke
et al10 showed a decrease in the
prevalence of overweight by 42.5% by
preventing high TV time, high meal frequency,
and low physical activity. In all
those studies, reducing media time had
the strongest effect. The impact of high
media time is in linewith previous results
of Gortmaker et al,11 showing that 60% of
overweight in children was attributable
to excess TV viewing time; however, that
calculation was unadjusted for other
determinants. In summary, preventive
measures to reduce media time in children
seem to have the largest impact
on childhood overweight. However, it is
tempting to speculate that the impact
of a reduction of media time depends on
the extent of media consumption in the
whole population.
Compared with the international reference
data, the joint AR for all determinants
(model 1) was slightly higher
when overweightwas defined according
to German reference data. However, the
joint ARs of the preventable determinants
were independent of the cutoffs
used (model 2). In that analysis, a short
sleep duration instead of breastfeeding
reached significance. In both analyses,
the ORs of these variables were marginally
or even not significant and, thus,
their impact on reducing overweight
(partial AR: 2.8% for breastfeeding and
2.3% for sleep) was small.
The strengths of our study include the use
of the currently largest German database