Wearable and Implantable Medical Devices (WIMDs) perform critical health monitoring and therapeutic functions. However, current WIMD products lack security safeguards to protect patients from fatal cyber-attacks. In the recent past, electrocardiogram (ECG) signals based security techniques have been widely explored to secure such devices by using two cryptographic primitives, the fuzzy commitment and the fuzzy vault, respectively. Nonetheless, differences, as well as similarities between these two primitives, have not been well investigated, making it difficult to decide which one would be appropriate for a particular setting. In this paper, we perform a critical analysis on both primitives and discuss their merits and drawbacks in the context of ECG-based key distribution. We analyze the critical challenges within each primitive-based key distribution technique, such as binary sequence generation and polynomial computations. Experimental results show that the technique based on the fuzzy commitment has a better false acceptance rate due to the randomness of ECG binary sequences. On the other hand, the fuzzy vault based scheme can achieve an acceptable false reject rate (5%) with less cost to the WIMDs. Future research is suggested to enhance the precision of ECG signal processing, to improve the efficacy of binary sequence generation process as well as to suggest ways to reduce polynomial computations.