Figure 4.

Association between evening daylight and physical activity across study populations, and pooled effect sizes for interactions by sex, age, weight status and maternal education. cpm = counts per minute. Analysis based on 23,188 children from 15 studies in 9 countries, except for the comparison of maternal education which is based on 15,563 children in 11 studies in 8 countries (see Additional file 1: Tables A1 and A2 for details of studies providing maternal education data). On the left, random-effects pooled estimates are presented by country/region, together with 95% confidence intervals. Points to the right of the line indicate that longer evening daylight is associated with increased mean daily cpm, points to the left indicate the reverse. On the right, pooled effect sizes and 95% confidence intervals are shown following tests for interaction, with the adjusted interaction term representing the difference that the interaction variable (e.g. sex) makes to the effect size for evening daylight upon total daily activity measured in cpm. For interaction terms stratified by study population see the Additional file 1: Figures A3 and A4.