I asked this question on MSE, but I haven't received any comments or responses (also, it has a very low view count), so I thought I would also ask it here.
In this paper, Knight, Pillay, and Steinhorn prove that for any O-minimal structure $\mathfrak{A}$, in which the underlying order types is dense, and if $\mathfrak{B} \equiv \mathfrak{A}$, then $\mathfrak{B}$ is also O-minimal. Is there a counterexample for when the underlying order type is not dense? I haven't been able to construct a counterexample.
Thanks