In recent decades, environmental pollution has become a pressing global health concern. According to the World Health Organization (WHO), a significant portion of the population is exposed to air pollutant levels exceeding safety guidelines. Cardiovascular diseases (CVDs) — including coronary artery disease, heart attacks, and strokes — are particularly significant health effects of this exposure. In this paper, we investigate the effects of air pollution on cardiovascular health by constructing a dynamic knowledge graph based on extensive biomedical literature. This paper provides a comprehensive exploration of entity identification and relation extraction, leveraging advanced language models. Additionally, we demonstrate how in-context learning with large language models can enhance the accuracy and efficiency of the extraction process. The constructed knowledge graph enables us to analyze the relationships between pollutants and cardiovascular diseases over the years, providing deeper insights into the long-term impact of cumulative exposure, underlying causal mechanisms, vulnerable populations, and the role of emerging contaminants in worsening various cardiac outcomes.