Certain scalar-tensor theories of gravity that generalize Jordan-Fierz-Brans-Dicke theory are known to predict nontrivial phenomenology for neutron stars. In these theories, first proposed by Damour and Esposito-Farèse, the scalar field has a standard kinetic term and couples conformally to the matter fields. The weak equivalence principle is therefore satisfied, but scalar effects may arise in strong-field regimes, e.g., allowing for violations of the strong equivalence principle in neutron stars (“spontaneous scalarization”) or in sufficiently tight binary neutron-star systems (“dynamical/induced scalarization”). The original scalar-tensor theory proposed by Damour and Esposito-Farèse is in tension with Solar System constraints (for couplings that lead to scalarization), if one accounts for cosmological evolution of the scalar field and no mass term is included in the action. We extend here the conformal coupling of that theory, in order to ascertain if, in this way, Solar System tests can be passed, while retaining a nontrivial phenomenology for neutron stars. We find that, even with this generalized conformal coupling, it is impossible to construct a theory that passes both big bang nucleosynthesis and Solar System constraints, while simultaneously allowing for scalarization in isolated/binary neutron stars.