Validity of Self-Reported Mobile Phone Use in the COSMOS Study

Abstract

Background/Aim: Possible health effects of mobile phone use are of considerable public and scientific interest, but previous epidemiological research has been limited by problems associated with (differential) recall bias and cross-sectional health analyses. The COSMOS cohort study, comprising almost 300,000 adult mobile phone users across Europe, has the advantages of a prospective study design and objective mobile phone use data from operator records. We assess validity of self-reported mobile phone use in a sub-population (N=75993) of the COSMOS study, and examine, for the first time, if validity differs according to experience of symptoms during mobile phone use or concerns related to mobile phones. Methods: We used Cohen’s weighted Kappa to assess agreement between self-reported and operator-derived mobile call frequency and duration for a 3-month period. We evaluated sensitivity of both self-reported high (≥ 10 calls/day or ≥ 4 hours/week) and low (≤6 calls/week or <30 minutes/week) mobile phone use, against the gold standard operator data. Results: We observed stronger agreement between self-reported and operator-derived estimates of mobile phone use for call duration compared to call frequency. Self-reported low mobile call frequency and duration demonstrated high sensitivity, however for self-reported high mobile call frequency and duration sensitivity was considerably lower. Self-reported call frequency validity was lower in younger age strata and women. Validity of self-reported call duration was lower in those who reported symptoms during or shortly after mobile phone use, compared to those who reported no symptoms. Conclusions: In contrast to previous validation studies, we observed better agreement overall, but a tendency for greater underestimation than overestimation in self-reporting of mobile phone use. This is most likely due to our use of a pre-defined categorical scale, rather than continuous scale, for self-reporting of mobile phone use.