yes, I have a 2013 NACRW poster on my page focused on milk, feel free to use it. I needed to homogenize using an automated bead mill, crash using a centrifuge and then further cleanup the sample using supported liquid extraction (SLE+). Recovery for my analytes of interest was >80%,
In the case of milk and cheese, I have used deproteinization with 6% w/v TCA and a second extraction step of the supernatant with hexane (for fat removal), It will probably work for yogurt, too.
You can cleanup the protein with crashing solvent such as methanol, acetonitrile with and without acid(formic or acetic), or use TCA or mineral acid. This will eliminate protein. You can do methylene chloride or hexane extraction to get rid of fat. However, you may need to get rid of phospholipids which may cause matrix suppression in your sample or foul up your column. The most important thing is you need to find out if your analyte will bind with component in milk and cause poor recovery. Case and point, I spike glyphosate and extract it with all that and get poor recovery. I found out that it may bind to cationic substance in milk and I need to break it down before extraction. It is very depending on the nature of your analytes. First I thought milk is easy, but it is very a complicate matrix.