Suppose we have a Wiener stochastic integral $$\int_{t-h}^t f(r_v) d W_v, \tag{1}$$ It is well known that, by the Ito isometry property, $$\mathbb E \left [\left(\int_{t-h}^t f(r_v) d W_v\right)^2 \bigg | \mathcal F_{t-h}\right] = \int_{t-h}^t \mathbb E[f(r_v)^2|\mathcal F_{t-h}] d v.$$
I am wondering how to extend this result to fourth conditional moments in general $$\mathbb E \left[\left(\int_{t-h}^t f(r_v) d W_v\right)^4 \bigg | \mathcal F_{t-h}\right].$$
There is an answer to a similar question here that implies: $$ \mathbb E \left[\left(\int_{t-h}^t f(r_v) d W_v\right)^4 \bigg| \mathcal F_{t-h}\right] = 3 \left[\int_{t-h}^t \mathbb E[f(r_v)^2 |\mathcal F_{t-h}] d v\right]^2 \tag{2} $$
However, it seems that the answer relies on the fact that $\mathbb E (Y^4) = 3 (\sigma^2)^2$ for any $Y \sim N(0,\sigma^2).$
So my question goes, how can I compute the conditional fourth moment of (1) when $\int_{t-h}^t f(r_v) d W_v$ is not normally distributed? Should I expect result (2)?
An example with $f(x) = x$ would be computing $$\mathbb E \left[ \left(\int_{t-h}^t r_v d W_v\right)^4 \bigg | \mathcal F_{t-h} \right]$$ given, e.g. $$d r_{t} = \mu(r_t) d t + \eta \sqrt{r_t} d W_t.$$
Thank you.
