Let $h$ be a bounded Lebesgue measurable function on $\mathbb{R}$ such that for any finite measure Lebesgue measurable subset $E$ of $\mathbb{R}$,
$$\lim_{n \rightarrow \infty}\int_{E}h(nx)dx = 0$$
My question is can we extend this to general measurable subsets of $\mathbb{R}$ in the following way: the Lebesgue measure on $\mathbb{R}$ is $\sigma$-finite. If $E$ is an arbitrary measurable subset of $\mathbb{R}$, then $E \cap [-t, t]$ is Lebesgue measurable and of finite measure, and this sequence of sets for $t \in \mathbb{N}^+$ increases to $E$, so
$$\lim_{n \rightarrow \infty}\int_{E}h(nx)dx = \lim_{n \rightarrow \infty}\lim_{t \rightarrow \infty}\int_{E \cap [-t, t]}h(nx)dx = \lim_{t \rightarrow \infty}\lim_{n \rightarrow \infty}\int_{E \cap [-t, t]}h(nx)dx = 0$$
I am unsure if the interchange of limits is justified, and if so, why it is appropriate. Comments and explanations welcome.