Heart Rate (HR) monitoring in a smart watch is a battery consuming process. Therefore, some smart watches prevent continuous HR sampling during low-level physical activities for improving the power usage of batteries. However, high acceleration events like running require continuous HR sampling for recording the varying HR readings, which consumes a lot of battery power. Therefore, the challenge is to reduce the sampling during running while recording an individual’s HR variation (HRV). Our approach prevents continuous HR sampling for long-running events without missing important HR information. In this context, we analyze the existing real-life HRs and acceleration datasets recorded on humans during long runs. We design an adaptive HR sampling algorithm, A-HeaRing, from the extracted HR variation details. A-HeaRing reduces the battery power consumption of a smart watch by more than 50% while recording a running event. Additionally, we design a data regeneration system for missing HR readings. The regeneration system provides a minimum of 90% accuracy in the HR zone estimation.