Early disease prediction has become crucial, particularly with the rising incidence of heart disease among young adults. Timely diagnosis allows for effective intervention, helping to manage symptoms and prevent further health complications. This study emphasizes the importance of early heart disease prediction across different age groups and genders to maintain optimal health. We analyze various machine learning algorithms for heart disease prediction using three distinct datasets. Our findings indicate that boosting methods, such as CatBoost, XGBoost, and LightGBM, consistently achieve higher accuracies across all datasets. These methods also perform exceptionally well across most age groups, demonstrating superior accuracy compared to other algorithms. Notably, prediction accuracy is higher for younger adults and declines with increasing age. Additionally, our analysis reveals that accuracies for females are significantly higher than for males when applying various machine learning algorithms.