One of our audience with professional expertise comments: I don’t know what to make of this article. I agree that it does have some importance in at least indicating the possibility of increased symptoms with time. The problem I have is that there are a lot of percentages given, since I have not looked at the article itself but just your review. I don’t know if there were more sophisticated statistical treatments of the data. Usually you mention the type of test used so I am assuming descriptive stats are it. Such a treatment of the data can be of interest but pretty meaningless given the variability in time. The one thing that I find of some concern was the fact that the outcome could not be predicted by age, BMI, interval of time between testing, upper airway anatomy, blood gas tensions or respiratory function. If there is in fact an exacerbation of the disorder over time then one would expect time interval between testing to be correlated with an index of severity. The variability in test retest time in this study should lend itself to such an analysis. A simple linear regression might be expected. All in all I think the N is just too small to make much sense with such variability and the large number of variables. However, given all of this, there is the fact that 30 out of 55 patients got worse as measured by a 25% increase in AHI. And morbidity was pretty high. Do you know what the morbidity rate for people in this age group is for similar conditions to the ones cited in the study? Or rates with CPAP? I am 52, I worry about 57 given the figures in this study. I hate to see these sorts of numbers thrown around without reference to what is expected in the normal population, for all I know this may represent an improvement over normal expectations. If you have 55 overweight smokers how many would you expect to have some sort of heart problem over the course of 2 years at age 57. Beats me! So I don’t know what all this really means. I guess you could look all that stuff up somewhere, I am pretty sure insurance companies have a good idea of the numbers. This is, on the surface, sort of an alarming study though isn’t it?

My Thoughts About These Points: Yes, it is alarming; though its findings presuppose absence of treatment, which seems likely to alter course favorably. To answer some of your questions, the study report is not heavy with statistical analyses, just looking at the significance of before & after differences, and possible predictors of these changes. I agree the relatively small number sets limits on how much they can say about predictors, for example the interval between testing which is so variable. One would like to see patients grouped by time interval and compared to gauge how fast the disease progresses. If there really were no correlation of severity with time, I would have to rethink the idea that worsening was an intrinsic characteristic of the illness, and wonder whether the patients came for evaluation at a time point when some sort of abrupt, short-term drop-off in functioning was in the works. Actually, a couple of years is not a long time for a chronic disease. Many of the most virulent diseases use five-year outcomes as their measure. The authors do not provide, and I cannot say, what the normative rate of complications for this age group without sleep apnea would be, just that the frequencies impressed me as large. I know there are many issues in comparing a patient group like this with normative population data. I usually try to get a copy of my summary to the authors for their comments; in this case I will make a special point of it, including any comments that come in soon. Readers should be aware that, to the authors, the audience comment sections are the most interesting. They want to hear what patients think about their work, and how they relate it to personal experiences. So the more audience contributions we get, the more author contributions we are likely to draw. Make your contribution now! Be sure to mention the article number or name in your e-mail.