Longxiao Li, Yongjun Zhou, Biyu Yang, Zhe Zhang
Algorithmic management has markedly improved operational efficiency in the platform economy, yet its consequences for the occupational health of platform workers remain insufficiently understood. This study therefore examines how efficiency-driven algorithmic pressures produce occupational health burdens among online food delivery (OFD) riders in China. The analysis drew on 10,103 valid rider comments from a major Chinese social media platform, and it combined Latent Dirichlet Allocation topic modeling with co-occurrence network analysis and sentiment intensity assessment. Semi-structured interviews with 32 OFD riders were added to corroborate these computational findings and to reduce reliance on a single data source. Expert review identified five core health burden dimensions: physical exhaustion, social devaluation, disciplinary distress, injury vulnerability, and health-protection deficit. Network analysis showed that these dimensions form a tightly interconnected structure with physical exhaustion as the central hub, and that pressure cascades from operational demands into punitive mechanisms and then into injury risk and social protection deficits. A clear gap emerged between structural prominence and emotional intensity. Riders discussed time pressure most widely and largely accepted it as routine, while punitive mechanisms and health-protection deficits drew the strongest negative responses. The study contributes an integrated text-mining framework for occupational health research. It reframes rider health as a cascading system of mutually reinforcing burdens rather than a set of isolated risks, and it shows that the most central burden is not the one felt most acutely. These insights point to concrete measures for governments, platforms, rider organizations, and industry associations such as portable occupational-injury insurance, algorithmic transparency, fairer timing and rating rules, and accessible grievance channels.