The distance-based location area update mechanism of CDMA 1X EV-DO achieves circular location areas around the cell in which the mobile makes the update. In this paper, we enhance this technique by achieving arbitrary location area shapes tailored to the known mobility patterns of mobiles. We develop a heuristic iterative algorithm that searches for the optimal shape which maximizes the time between location area updates of a mobile, for a given paging cost. We present numerical results that show significant improvement in the time between location area updates of our proposed algorithm over that of the pure distance-based location update scheme, provided there is sufficient bias in the mobility pattern of the mobile. Our heuristic is motivated by formulating the problem as a finite state absorbing Markov chain, with the objective of maximizing the time to absorption of the Markov chain. We investigate and numerically quantify the impact of several other parameters, such as, the total number of cells in the location area and different mean cell residence times across cells.