Fatigue Resistance Assessed in Five Tasks for a Single Session of Sleep Deprivation

http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA493497&Location=U2&doc=GetTRDoc.pdf
Chaiken, Scott R., Harville, Donald L., Harrison, Richard, Fischer, Joe, Fisher, Dion, Whitmore, Jeff
Date October 2008

AFRL-RH-BR-TR-2008-0067

To assess whether individuals (n=89) could be put on a trait dimension of fatigue resistance (c.f., Von Dongen, Maislin, & Dinges, 2004), we observed performance on cognitive tasks in a single 48-hour sustained-wake protocol. Individual differences in task performance were largest late in the protocol. Next we developed methods for classifying a participant as fatigue resistant or susceptible, as part of a larger project investigating genetic factors in fatigue-resistance. We considered a rule based on percent-change decrement with fatigue and another rule based on residuals of task performance predicted by (presumably non-genetic) sleep behaviors, which were shown to bias raw percent-change classifications. Classifications based on ranking residuals were less confounded by sleep behaviors than similar classifications based on percent change. Finally, we assessed the SAFTE fatigue model (Hursh, Redmond, Johnson, Thorne, Belenky, Balkin, Storm, Mil! ler, Eddy, 2004) on sustained performance across the different tasks. While this model is a simulation theory of physiological fatigue, and as such task-independent, we discuss relatively simple ways to put theory and task performances on the same quantitative scale to assess model adequacy.