Right, here we go then. The question at hand is whether the last sip of a soda is mostly (e.g. 90%) saliva from the 'backwash'. See below for the shocking answer. (I could tell you now, but I did the damn experiment, you're gonna at least have to page down to the bottom of the report.)
You all remember, of course, the pioneering work in this field by Dr. Andy Walton. He concluded that there was no detectable saliva in the last sip, based on the traditional "swish" and "hold-it-up-to-the-light" tests. In an attempt to quantify his results, I have used a sensitive protein detection assay to quantify the amount of saliva in the last sip.
To summarize the experiment, I drank a mug full of water (and Mountain Dew, but see below for the bad news on that), sampling the water before the first sip and when there was approximately 5 ml left. I also took a sample of saliva as a control. I then assayed the protein concentrations in each sample, and used that to calculate the percentage of saliva in the final sip.
The mug was a Peet's Coffee travel mug; I filled it in order to give as much chance for saliva addition as possible. The protein assay was the bicinchoninic acid (BCA) protocol, manufactured by Pierce Chemicals. The protocol I used is accurate for protein concentrations of less than 50 micrograms/ml (ug/ml) to about 200 ug/ml. Saliva was donated by a geeky but nevertheless startlingly handsome researcher, as I was the only one in the lab.
I made a series of dilutions of the saliva, from 1/2 to 1/200 in water. The before sample (tap water) and the after sample ('backwash') I used straight. I also made a series of protein concentration standards using the bovine serum albumin supplied with the kit. Here are the results (OD562 = optical density at 562 nanometers).
|
| | ||
| concentration | OD562 | ||
| 50 ug/ml | 0.18 | ||
| 100 ug/ml | 0.34 | ||
| 150 ug/ml | 0.47 | ||
| 200 ug/ml | 0.68 | ||
| Dilution | OD562 | ||
| 1/2 | (off scale) | ||
| 1/10 | 0.54 | ||
| 1/25 | 0.30 | ||
| 1/50 | 0.21 | ||
| 1/100 | 0.16 | ||
| 1/200 | 0.08 | ||
So the protein concentration in my saliva is about 1.6 - 1.7 mg/ml. This is more or less linear until the 1/200 ratio, so the assay is measuring accurately down to about 15 - 20 ug/ml. In other words, I can detect saliva if it's 1% or more of the sample; below that the assay is not quantitative. (I may be detecting things other than protein in this, but of course for the purposes of the experiment that doesn't matter.)
| Samples | OD562 |
| Before | 0.05 |
| After | 0.05 |
Conclusion (Cheers from the audience) The level of 'backwash' saliva is less than 1% of the final sip.
I made no particular effort to drink either carefully or sloppily, but obviously, someone with a more primitive drinking style may have different results.
I tried this with Mountain Dew as well, but foolishly forgot that this particular assay is screwed up by sugar, so those results were useless (everything was off the scale). I should have tried it with a diet drink, but I detest them so I won't. Also, the four hour incubation that was keeping me here and bored is over.
Ian "daredevil" York