Which scales measuring problematic internet use are best at predicting anxiety and depression symptoms in medical students?

Abstract

Background. Globally, depression affects one-third of medical students, whereas the prevalence of problematic internet use (PIU) among medical students is approximately five times higher than in the general population [1]. Evidence suggest that PIU contributes to anxiety and depression severity [2]. However, it is unclear which scales measuring PIU are better than others at predicting anxiety and depression severity. Objective. The aim of this study was to evaluate the prognostic value of the Compulsive Internet Use Scale (CIUS), the Problematic Internet Use Questionnaire (PIUQ9) and a single question “whether internet use impairs daily functioning”, to predict depressive and anxiety symptoms in medical students. Methods. Lithuanian medical students participated in two independents online surveys launched between December 2020 and February 2021 (Study 1, N=518 [21.4% men, mean age 23.6 ± 3.0 years]), and between May 2020 and June 2021 (Study 2, N=398 [16.3% men, mean age 23.0 ± 3.4 years]). The participants completed the CIUS [3] and answered question “Impairment to engage in daily activities and social life due to excessive internet use (NO/YES)” (Q_single, Study 1), PIUQ9 (Study 2) [4], the Patient Health Questionnaire (PHQ9) and the Generalized Anxiety Disorder Assessment (GAD7) (both studies) [2]. Block-wise linear regression analyses were performed to estimate the prognostic value of CIUS, PIUQ9 and Q_single on PHQ9 and GAD7 when controlling for gender, age and internet use time. Results. Q_single resulted in increase in the coefficient of determination R2 by 0.175 (p<0.001), R2 of the final model is 0.182, R2 is 0.176 (p<0.001), standardized coefficient beta =0.420 (p<0.001), when predicting PHQ9, and its predictive influence was higher if compared to PIUQ9 (increase in R2 0.110 (p<0.001), R2 0.128, R2 0.120 (p<0.001), beta=0.350 (p<0.001). CIUS also showed a larger influence in predicting PHQ9 increase in the coefficient of determination R2 is 0.247 (p<0.001), R2 of the final model is 0.254 (p<0.001), R2 is 0.249 (p<0.001), beta=0.497 (p<0.001)) if compared to PIUQ9 (Study 2) and Q_single (Study 1). Conclusions. The CIUS performs better than PIUQ9 or single question on function impairment due to PIU in predicting depressive and anxiety symptoms in medical students regardless of age, gender and and internet use time. However, in short medical surveys one question item on functioning difficulties due to PIU could be recommended using as it predicts anxiety and depression symptoms relatively well.