As the paper points out, this is an analysis of cross-sectional and prospective studies, so while interesting, this evidence tells us nothing about the directio

As the paper points out, this is an analysis of cross-sectional and prospective studies, so while interesting, this evidence tells us nothing about the direction of causation and cannot account for confounding factors. It could easily be that sleep disorders and depression cause people to spend more time online rather than the converse, or it could be that some external factor causes both increased time online as well as mental disorders. To actually figure out if being online causes these disorders, we'd need a randomized controlled trial.