With the explosive rise in the development and popularization of social networking apps, Online Social Networks (OSNs) have now become a cornerstone of everyone's daily lives. Although OSNs tend to broaden the capacity of their users to increase social interactions, they may actually reduce users' real-world person-to-person dialogue. This decrease in the actual interaction can cause many mental health problems like Nomophobia, Phubbing, etc. In this paper, we try to learn about all the efforts made in this field to detect these mental health disorders. We review various literature available regarding the Mental Health Problems caused by Online Social networks and the methods used to detect them using Data Mining. These research works have utilized Regression Techniques, Support Vector Machine, Naive Bayes, etc to predict the outcomes. We further investigate their pre-processing, feature extraction, and classification processes throughout this paper.