Revisiting the derivation of the fractional diffusion equation