Peter Attia· MD
so to come back to your question I love the way that you think about that problem and I agree with it very much that out sums disease is probably going to be a heterogeneous I think the pathology already points to that and its complexity the way that sleep plugs in to that heterogeneous disorder that we call Alzheimer's is also going to be heterogeneous in nature different sleep features will be predictive of different aspects of the amyloid pathology cascade and were we just submitted a paper which I'm gonna blow now in terms of media sort of outlet it hasn't been accepted yet but we just tried to ask the question across someone's lifespan if you bucket them into 10-year decades and you ask what was the quantity and quality of their sleep from their 30s and 40s 40s and 50s 50s and 60s so on and so forth and then at the end of their life you stick them in a brain scanner and you map how much amyloid is in the brain and how much towers in the brain which is what we can do here at the sleep center and then you essentially can in a retrospective manner say was sleep at certain times of life more critical in predicting how much amyloid you will have late in life so in other words are the certain decades during an adult's lifespan worse sleep is even more important for determining your how can you do that retrospectively so let's say you had this cohort of patients maybe at best you could through questionnaires infer duration but how would you be able to infer anything about depth or quality so depth we can't do but quality we can do so we've got a set of standardized questioners that have been essentially validated against gold standard what we call polysomnography which essentially is sleep laboratory grade recording with electrodes and what we find is that those questioners have good predictive validity relative to objective gold standard measures so it's a hop skip and a jump but you can say look these subjective questioners seemed to go pretty well hand-in-hand with sleep quantity and quality let's take those questioners and then let's ask of those patients what they were doing in terms of their sleep quantity and quality in those different bucket decades of their life and it has all of the problems that you can imagine because this seems like a slightly better version of a food frequency questionnaire which your peers would be relying on to do the same question so it's reliant on an individual you know it's got problems with history cuz you know the further you go back does that signal of accuracy degrade how big are the hazard ratios because I guess this is where one of those things where if sleep could account for 25% of this disease or if it if it increased their risk by say 25% would be a better way to say that it strikes me as unlikely that you could see that signal because that's not a big enough signal with all the noise you've just described but if you were talking about a hazard ratio of 6 or 7 or you know it's so funny today one of my kids asked me about smoking like why do people smoke and this I couldn't resist going down the Bradford Hill pathway of unfortunately he's four and a half but he got a probably about a 12-minute lecture on epidemiology and the hazard ratio of small cell lung cancer and smoking but the punchline in my story to him Peter attea ladies and gentlemen PETA I was trying to explain to him what it would mean to have a 14 X increase in risk which was basically what was observed in the first physician study of smoking and you know the point there is