Does VarianceThreshold of scikit-learn take into account the type of distribution when calculating variance?

Question

For example, the var() method of Pandas Dataframe uses always the square of the standard deviation std()**2 for all the columns of the Dataframe no matter the t