These seem to be the best references, although directly collecting buccal cells is probably preferable to collecting saliva: 

http://onlinelibrary.wiley.com/doi/10.1002/ajmg.b.32278/pdf

http://www.tandfonline.com/doi/full/10.4161/epi.24362#.VNiQ1vnF_To

We used a protocol for saliva with an RNAase to stop the enzymatic breakdown when collecting saliva.protease inhibitor cocktail (Calbiochem; 1:100 dilution) within ten minutes of collection, prior to freezing at -80°C.but we were looking at interleukins. 

Thank you both very much for your answers. Since we might have trouble freezing the samples in a controlled way during travel, collecting buccal cells seems like the best option. Good luck with your research.

Additional  considerations regarding epigenetics: To my understanding, there can be tissue specific changes in epigenetic markers although I have not seen a good study comparing epigenomic markers from a simple histologic point of very. Do you know of a reference? Therefore, the question remains, "Are the epigenomic changes simply related to the buccal cheek cells or peripheral blood cells versus an end organ susceptibility or related disease process?" 

One criticism that I have received for use of buccal cheek cells in an epigenomic protocol was, "How do the epigenomic changes relate to temporal and age related changes during growth and development?"

Then,  when  said that we would use site specific cells, the criticism was that the problem that interfered with growth and development already happened and may not  be observed at a later time at a specific tissue site. My question is, "When and where does imprinting occur?"

If you are interested in a collaboration on this topic, I could really use the help. I am very interested in conducting epigenomic studies related to disease susceptibility and a special area related to birth defects.

Are violence and stress epigenomic pertubations?