To the Editor:
I appreciate the critique from Drs. Rupp and Balkin. Our findings are indeed counterintuitive, and open debate will hopefully lead to clearer understanding.
We used the Positive Affect Negative Affect Scale because it covers a range of emotional states. As noted in our paper, it has been validated.1,2 It has not been widely used in sleep research, but we were also interested in anger3 and other emotions.
We compared injured people not only to an uninjured control group, but also to themselves 24 hours before. Granted, recall bias is still likely, but the very similar findings with 2 separate control groups enhance the credibility of the results.
We all think that sleepiness and sleep deprivation are strongly associated with injury risk, but look again at the case-control studies we reviewed in the introduction to our paper. We identified 9, 3 of which examined sleepiness or recent hours of sleep in adults. Two found that acute sleepiness (dichotomizing the Stanford Sleepiness Scale) was associated with injury risk, but 1 found that yawning was associated with a significantly reduced risk of injury. The third found an odds ratio of 0.63; although not statistically significant, it's <1. Previous studies’ findings have not been as consistent as we would like them to be. Perhaps it's all due to recall or reporting bias, but perhaps other factors are involved.
It is difficult to study sleepiness and injury. Randomized controlled trials are possible only in a laboratory. Real world studies will have to be observational and retrospective for the foreseeable future. Our study, like the previous case-control studies, has very real weaknesses. We hope that sleep and injury researchers will challenge our study's findings by doing better ones.
For our patients and for our own behavior (eg, while driving), Alan Bavley, writing for the Kansas City Star (February 17, 2007), had good advice when he reported on this paper: “Daniel Vinson wants you to play it safe and ignore what you are about to read.”