Strictly speaking, when a body starts to move at a relativistic speed, the entire idea of gravity as a "force" needs to be revisited. However, the error involved in continuing to think of gravity as a force is only about 2:1, so if one isn't too demaning of the model, it's possible to come up with a hand-waving description of the gravitational field of a relativistically moving body (with the understanding that this description will only be approximately correct and not really mathematically rigorous).
With the understanding that the concept isn't totally well defined, the general behavior of the "gravitational field" of a relativistically moving mass will NOT be the same in all directions. It will greatly resemble the eletcric field of a relativistically moving electron. This means that the field will be stronger in the transverse direction than in the direction of motion. Because the field isn't uniform, one has to define the total field strength by some sort of average. Gauss's law for the electric field defines a notion of an average field that is independent of velocity - applying the same notion to the "field" of a moving mass will yield an average field that increases with velocity, so the "average" field of a moving mass is stronger.
Of course the field in a frame co-moving WITH the mass will not be changed! Only the field in a frame moving relative to the mass, or the mass moving relative to the frame, will be changed.