Massive MIMO enabled ultra-dense heterogeneous networks (UDHNs) have been considered as the indispensable and emerging approach to meet the demand on growing data traffic for next generation networks. Although deployment of large number of antennas causes high circuit power consumption in massive MIMO UDHNs, there is always a tradeoff between energy efficiency (EE) and spectral efficiency (SE). Therefore, an energy efficient and spectral efficient user-cell association will become crucial and challenging in massive MIMO UDHNs. In this paper, we address a user-cell association problem for EE and SE tradeoff. To this end, we formulate a convex multi-objective optimization problem (MOP) and convert it into a single-objective optimization problem (SOP) where a priority is assigned for EE and SE with a weighting factor which means the problem can be adjusted whether priority is on EE or SE. The problem aims to maximize the weighted sum of the EE and SE. As a solution, Lagrange duality analysis is performed and a distributed game theoretical user-cell association (GTUCA) algorithm considering the fairness among users is developed. The results confirm that the proposed algorithm outperforms the baseline algorithm, namely maximum rate-based cell selection in terms of EE and SE, when the weighting factor is set properly.