51 kg/m2 for ages 6 and 7, respectively. After controlling for country GNI, centre, age and measurement type, there was a statistically significant association between frequent and very frequent fast-food consumption and higher BMI with an apparent dose–response effect (+0.15 and +0.22 kg/m2 for Crizotinib supplier frequent and very frequent fast-food consumption, respectively; table 1). There was no statistically significant interaction between GNI and fast-food consumption in this age group (p=0.06). Table 1 Association between fast-food consumption and BMI (SE) of study participants after adjustment for

GNI When boys and girls were analysed separately, both sexes had a significant positive association between frequent and very frequent fast-food consumption (table 1). Adolescents Figure 3B shows the difference in BMI between adolescents with infrequent fast-food consumption and those with frequent and very frequent fast-food consumption in each centre. In male adolescents, the estimated mean BMIs for those reporting infrequent fast-food consumption were 19.74 and 20.02 kg/m2 for ages 13 and 14, respectively. After controlling for country GNI, centre, age and measurement type, lower BMIs were significantly associated with greater

fast-food consumption (table 1). There was a statistically significant interaction between fast-food consumption and GNI. In low-GNI countries, the frequency of fast-food consumption had no association with BMI (+0.01 and −0.03 kg/m2 in the frequent and very frequent groups, respectively (p=0.89 and 0.56)). In high-GNI countries, lower BMIs were associated with greater fast-food consumption (−0.21 and −0.42 kg/m2 in the frequent

and very frequent groups, respectively (p<0.001 in both groups)). When analyses were restricted to the 20 384 male adolescents who had measured height and weight data, there was a tendency towards a lower BMI with higher fast-food consumption, but this was not statistically significant (−0.03 and −0.11 kg/m2 in the frequent and very frequent groups, respectively (p=0.24)). In female adolescents, the estimated mean BMIs for those reporting infrequent fast-food consumption were 19.56 and 19.98 kg/m2 for ages 13 and AV-951 14, respectively. Lower BMIs were associated with greater fast-food consumption and this was statistically significant. There was a statistically significant interaction between fast-food consumption and GNI. Both low-GNI and high-GNI countries had an association between increasing fast-food consumption and BMI (low-GNI countries −0.10 and −0.11 kg/m2 in the frequent and very frequent groups, respectively (p=0.03 and 0.01), and high-GNI countries −0.24 and −0.22 kg/m2 in the frequent and very frequent groups, respectively (p<0.0001 in both cases)). When analyses were restricted to the 24 127 female adolescents who had measured height and weight data, those who ate fast-food frequently or very frequently had lower BMIs and this was statistically significant (table 1).