Electrocardiography (ECG) is a widely used, non-invasive tool for assessing cardiac function, but conventional disease-centric models do not fully capture overall cardiovascular health. This work introduces LeadKAN, a lightweight and explainable Kolmogorov-Arnold Network architecture for ECG age estimation built on LoRKAN layers that combine low-rank bilinear mixing with an RBF-kernelized top. LeadKAN achieves performance comparable to state-of-the-art models while using substantially fewer parameters and multiply-add operations, and its lead-specific encoders support attribution analysis for improved clinical interpretability.