The eta-c decay into proton-antiproton cannot be explained by a lowest order perturbative QCD quark scheme. Trying to prove a previous result where diquarks were also considered as nucleon's constituents, the contribution of the spin-flip transition between scalar and vector diquarks inside the nucleon is computed and is shown to be strictly zero. This result excludes the possibility of understanding why this decay is experimentally observed with a branching ratio much greater than those of other charmonium decays into the same final state, CHI-0,1,2-->ppBAR, successfully described by pQCD in terms of quark and diquark components of the protons. A theoretical explanation of this decay rate is then still lacking and it is suggested that pseudoscalar glueballs might play an important role in solving the puzzle. The experimental results are also briefly discussed.