Difficulty closing optimality gap for non-convex MIQCP after parameter change

I am solving a non-convex nonlinear MIQCP using Gurobi.

During a sensitivity analysis, I observed that when I change the value of a certain parameter, the solver stalls with an optimality gap of around 5% and struggles to close the bound.

Interestingly, with the original parameter value (and the same initial point), the model can be solved to optimality in about 20 minutes.

My current hypothesis is that the modified parameter may lead to a degenerate or numerically difficult region of the search space, which slows down bound improvement.

I would like to ask:

1. Are there Gurobi parameters that are particularly helpful for improving bound tightening or gap closing in non-convex MIQCP problems?
2. Are there recommended settings for spatial branch-and-bound performance in such cases?
3. More generally, are there diagnostics in Gurobi that help determine whether the issue is degeneracy, weak relaxations, or numerical instability?

Any suggestions or experiences would be greatly appreciated.