Someone just emailed me a link to an article showing that dogs can be trained to detect colon cancer in breath and stool samples, even in its’ early stages. He asked if this could be applied to sleep apnea. I though long and hard about this. There are so many metabolites in urine, stools, sweat, saliva, and even cerumen from untreated sleep apnea that could possibly be detected by dogs, but not any one of these substances would be specific, like cancer cells.
However, there’s an even easier way: Look at the person’s facial features. Look for narrow and/or recessed jaws, flat cheekbones, narrow nasal width, mouth breathing, or forward head lean. Inside the mouth, the classic findings would be multiple missing teeth, dental crowding, a high arched high palate, narrow and crowded dental arches, tongue scalloping, and a relatively large tongue that prevents you from seeing the uvula. Tongue scalloping alone has been found to predict apneas in 89%, oxygen desaturation in 89%, and abnormal AHI is 67%.
If you combine these exam findings which can take just a few minutes to observe, along with simple screening questionnaires such as the Epworth Sleepiness Scale or the STOP-BANG, you could probably screen for and pick out from 1/2 to 2/3 of all sleep apnea patients. Currently, it’s estimated that only about 10% of sleep apnea is diagnosed. Notice I didn’t mention being overweight, being male, or even the fact that you have to snore.
These days, I play an informal game with myself whenever I see patients. Just by looking at the medical history and external facial appearances, I try to guess who has sleep apnea versus who doesn’t. It’s scary, but I’m close to being about 95 to 98% accurate.
I know this may sound dangerously like facial profiling, but since I’m able to help many of these people breathe better and sleep better, I’ll continue to engage in this practice. Try it on yourself and then on your family members, and then on your friends. With some practice, you can be just a good as me.