Internet enabled body area networks (BANs) will form a core part of future remote health monitoring and Ambient Assisted Living (AAL) technology. In BAN applications, due to the dynamic nature of human activity, the off-body BAN channel can be prone to deep fading caused by body shadowing and multipath fading. Using this knowledge, we present some novel practical adaptive power control protocols based on the channel deviation to simultaneously prolong the lifetime of wearable devices and reduce outage probability. The proposed schemes are both flexible and relatively simple to implement on hardware platforms with constrained resources making them inherently suitable for BAN applications. We present the key algorithm parameters used to dynamically respond to the channel variation. This allows the algorithms to achieve a better energy efficiency and signal reliability in everyday usage scenarios such as those in which a normal person undertakes many different activities (e.g. sitting, walking, standing, etc). We also profile their performance against traditional, optimal and other existing schemes for which it is demonstrated that not only does the outage probability reduce significantly, but the proposed algorithms also save up to 35% average transmit power compared to the competing schemes.