Ideal-city posts on social media linked to elevated depression risk
Researchers analyze the Sina Weibo Depression Dataset to quantify how idealized sentiment toward urban places relates to depression risk across 88 place types. They build a place-sentiment index (LooM) and use an explainable deep-learning model, finding that depressed individuals express more emotionally mixed language, with surprisingly positive sentiment toward cultural and residential areas associated with higher depression risk. These signals show systemic, city-level patterns that scale with population and vary across Chinese cities, challenging the assumption that positive online sentiment equates to good mental health and highlighting the need for urban design to address residents’ subjective emotional needs.
