As the aging population continues to grow, ensuring optimal environmental comfort in eldercare facilities has emerged as a critical societal priority. However, existing research on retrofit optimization has predominantly focused on isolated objectives, such as energy or cost reduction, thereby overlooking the complex trade-offs essential for achieving holistic performance in eldercare environments. This study addresses this research gap by proposing a comprehensive, multi-objective optimization framework that simultaneously balances thermal comfort, energy efficiency, and cost-effectiveness. Unlike previous approaches, the proposed methodology integrates architectural design with the evolutionary non-dominated sorting genetic algorithm II, which effectively resolves conflicting objectives and generates Pareto-optimal solutions without the data-intensive demands of conventional machine learning techniques. Focusing on a case study of the Gusu District Shuangta Street Jinfan Community Health Service Center in Suzhou, the research utilizes parametric modeling and the Wallacei plugin in Grasshopper to evaluate over 840,000 retrofit scenarios. The findings reveal tailored strategies – such as the use of Rockwool insulation and low-emissivity glazing – that optimize energy savings while maintaining thermal comfort within budgetary constraints. By bridging architectural innovation with computational optimization, this work not only provides a replicable model for eldercare retrofits but also offers actionable insights for policymakers and practitioners, advocating for interdisciplinary methodologies to meet the evolving demands of aging societies.