DIGITAL PHENOTYPING AND AI-BASED EARLY DETECTION OF DEPRESSION AND COGNITIVE DECLINE AMONG PAKISTANI YOUTH USING SOCIAL MEDIA BEHAVIORAL DATA

Abstract

Depression and cognitive decline are growing mental health concerns among youth, particularly in developing countries where early diagnosis and access to mental healthcare remain limited. This study examined the role of digital phenotyping and artificial intelligence (AI)-based predictive analytics in the early detection of depression and cognitive decline among Pakistani youth using social media behavioral data. A quantitative cross-sectional design was employed, and data were collected from 500 active social media users aged 18–35 years in Pakistan. Social media behavioral indicators, including posting frequency, language and sentiment patterns, social interaction patterns, usage behavior, and engagement activities, were analyzed using Structural Equation Modeling (SEM) and machine learning techniques. The findings indicated that social media behavioral data significantly predicted both depression and cognitive decline. Digital phenotyping was found to mediate these relationships, while AI-based predictive analytics enhanced the accuracy of early detection. The study highlights the potential of integrating digital phenotyping and AI to develop scalable, cost-effective, and culturally relevant mental health screening systems for Pakistani youth. The findings contribute to the growing fields of digital health, computational psychiatry, and AI-enabled mental healthcare.