Artificial Intelligence-Driven Assessment and Improvement Strategies for Older Adults' Health Literacy: Integrating Health Management and Mental Health

Background Global population aging presents unprecedented challenges to public health systems, with elderly health literacy emerging as a critical determinant of successful aging. Traditional assessment methods relying on static questionnaires exhibit significant limitations in capturing the dynamic, multifaceted nature of health literacy, particularly regarding mental health dimensions and personalized health management capabilities. These conventional approaches often fail to address urban-rural disparities and lack real-time responsiveness to individual needs. Purpose This study aims to construct a comprehensive artificial intelligence-driven framework for assessing and enhancing elderly health literacy, integrating both physical health management and mental health dimensions. The research seeks to transcend traditional methodological constraints by establishing scientifically rigorous, personalized, and user-friendly intervention mechanisms that address the multidimensional nature of elderly health needs. Methods We developed an integrated assessment system grounded in multidimensional health literacy theory, encompassing health knowledge, practical skills and behavioral attitudes. Leveraging natural language processing, computer vision, and machine learning technologies, we created multimodal intelligent evaluation protocols. Personalized intervention strategies were designed accounting for urban-rural variations, incorporating intelligent guidance systems and hybrid learning platforms. A three-tier ethical risk prevention mechanism was established to govern technical implementation and data management. Results The AI-driven assessment model fundamentally transformed traditional evaluation paradigms, enabling real-time, continuous, and precise measurement of elderly health literacy, including subtle fluctuations in mental health status and health management competencies. The mean difference in the experimental results is 12.5 points, with p less than 0.001. Post-intervention analyses demonstrated statistically significant improvements in total health literacy scores. The integrated online-offline personalized promotion strategy substantially enhanced participant engagement and learning outcomes. Differentiated implementation pathways effectively reduced urban-rural disparities and bridged the digital divide. Conclusions Artificial intelligence significantly advances scientific precision and personalization in elderly health literacy assessment and health management. Incorporating mental health dimensions enriches the comprehensive scope of health literacy interventions, ensuring holistic addressing of multifaceted elderly health needs. However, critical ethical considerations—data privacy protection, algorithmic bias mitigation, and equitable technical accessibility—require ongoing vigilance. This research provides robust theoretical foundations and practical guidance for standardized, ethical AI application in elderly health management and mental well-being promotion, facilitating responsible and inclusive technological integration.