I am quantifying Vit E content in serum samples. Some samples contain white Vit E rich flakes, which turn into a lipid layer after centrifugation. What could I do to homogenize the sample to be able to take a representative aliquot?

To extract the Vit E, I am following an excellent liquid/liquid extraction method, i.e. addition of equal volume of aqueous extraction buffer (2% ascorbic acid, 4% NaCl and 1 mg EDTA per 100 mL), addition of two volumes of ethanol, then protein is precipitated by centrifugation, supernatant is collected and extracted via liquid liquid extraction with hexane.

Is it normal that the lipids come out of solution and form a separate layer in non-fasted serum samples? I feel that this problem must have been encountered and solved before?