Skip to main content
. 2018 Jul 23;126(7):077007. doi: 10.1289/EHP2427

Figure 1.

Four plots A, B, C, and D with interquartile range respectively plotting change in verbal memory scores (y-axis) across brain dose (millijoules per kilogram per day), mobile phone calls (minutes per day), cordless phone calls (minutes per day), data traffic (minutes per day), gaming (minutes per day), text messages (number per day) (x-axis) for the whole sample, sample with operator data, sample 2012–2013, and sample 2014–2015.

Results of linear exposure–response models for change in verbal memory scores (follow-up–baseline): estimates relate to change in memory score for (A) the whole sample per interquartile range (IQR) of exposure of the whole sample; (B) the operator sample per IQR of operator sample; (C) the sample 2012–2013 per IQR of exposure of the whole sample; and (D) the sample 2014–2015 per IQR of exposure of the whole sample. IQRs of the whole sample: brain dose, 953 mJ/kg per day; mobile phone calls, 12.6 min/d; cordless phone calls, 5.1 min/d; data traffic, 55.4 min/d; gaming, 55.7 min/d; and text messages, 40 per day. IQRs of the operator data, brain dose: 341 mJ/kg per day; and mobile phone calls, 1.8 min/d. All models were adjusted for age, gender, baseline score, nationality, school level, physical activity, alcohol, and education of parents and change in height and time between baseline and follow-up investigation. Number of observations for each calculation is indicated below each estimate.