The authors have employed a numerical procedure to analyze the adhesive contact between a soft elastic layer and a rough rigid substrate. The solution of the problem, which belongs to the class of the free boundary problems, is obtained by calculating the Green's function which links the pressure distribution to the normal displacements at the interface. The problem is then formulated in the form of a Fredholm integral equation of the first kind with a logarithmic kernel, and the boundaries of the contact area are calculated by requiring that the energy of the system is stationary. The methodology has been employed to study the adhesive contact between an elastic semi-infinite solid and a randomly rough rigid profile with a self-affine fractal geometry. We show that, even in presence of adhesion, the true contact area still linearly depends on the applied load. The numerical results are then critically compared with the prediction of an extended version of the Persson's contact mechanics theory, able to handle anisotropic surfaces, as 1D interfaces. It is shown that, for any given load, Persson's theory underestimates the contact area of about 50% in comparison with our numerical calculations. We find that this discrepancy is larger than what is found for 2D rough surfaces in case of adhesionless contact. We argue that this increased difference might be explained, at least partially, by considering that Persson's theory is a mean field theory in spirit, so it should work better for 2D rough surfaces rather than for 1D rough surfaces. We also observe, that the predicted value of separation is in very good agreement with our numerical results as well as the exponent of the power spectral density of the contact pressure distribution and of the elastic displacement of the solid. Therefore, we conclude that Persson's theory captures almost exactly the main qualitative behavior of the rough contact phenomena.