OBJECTIVES: Depression in teenagers has been steadily increasing in the last 10 years, with almost a doubling in incidence, predominantly affecting teenage girls. The increase in depression symptoms has coincided with the ever-increasing use of social networks, starting with Facebook in 2004. This has led to the assumption that social network (SN) use is driving this depression trend, but transversal studies have shown conflicting results. The objective of this study is to address this issue using a novel simulation approach.

METHODS: An open cohort microsimulation model was developed for the French context, simulating depression incidence by year and birth cohorts from 1990 to 2016. Incidence of depression was based on the presence of risk factors including exposure to childhood adversities (parental psychopathology, physical, emotional and sexual abuse, neglect), bullying and SN use. SN use was based on the number of SN used and total duration of use, considering the increase through the years in the number of SN available and the average use duration. Risk factors were applied to a baseline incidence using odds ratio from transversal studies with baseline incidence calibrated on the observed incidence rates in France in teenagers.

RESULTS: The model was correctly fitted to the distribution of childhood adversities observed in France, the evolution of bullying, and the evolution of SN use. After applying these factors to the risk of depression, the model was able to accurately be reproduced the increase in depression incidence in France from a 10% incidence in the 2000s to 20% in 2022. Removing SN use resulted in a stable incidence.

CONCLUSIONS: These results suggest that a causal relationship between SN use in teenagers and the increase in depression seen since the mid-200s is plausible given the strength of the associations between SN use and depression observed in transversal studies.