Fatigue and processing speed predict subsequent fatigue override, evidence from a coordinated analysis of six EMA studies

Abstract In daily life, people often ignore feelings of fatigue to meet cognitive demands. This ‘fatigue override’ is the ability to maintain performance despite feeling tired. In both clinical and non-clinical samples, we examined whether better performance (i.e., faster processing speed) or high fatigue levels in preceding hours predicted decreased override capacity. We conducted coordinated analysis of six intensive longitudinal studies involving 1088 participants from clinical, non-clinical, and mixed samples in the United States. Included studies were ones where, at each measurement occasion, fatigue was measured with ecological momentary assessments and processing speed with the Symbol Search task in people’s real-world environments. For each study, coordination models were used to examine both concurrent associations between fatigue and processing speed and whether current states predicted subsequent fatigue override, and the overall effect across studies was computed with individual participant data coordinated analysis. Experiencing greater fatigue than average was associated with slightly poorer performance on the processing speed test (overall r = -0.068, 95% CI -0.127 to -0.008, p = .027). Being more fatigued than usual and performing faster than one’s typical processing speed were each associated with stronger subsequent coupling between fatigue and speed, indicating a decreased capacity to override fatigue in the following hours (overall B = 0.047, 95% CI 0.002 to 0.091, and B = 0.044, 95% CI 0.005 to 0.083, respectively). Clinical samples demonstrated lower capacity to override fatigue. Results suggest that both elevated fatigue and unusually high cognitive performance may signal reduced capacity to sustain performance later.