Attentional bias for sleep-related stimuli and the transition from acute to persistent insomnia: results from some recent experiments

Attentional bias develops when disproportionate information processing resources are allocated to exemplars of a given class, compared with otherwise equivalent stimuli. Biases for sleep-relevant stimuli have been proposed in conceptual models of insomnia [1, 2]. Our aim was to investigate such processing biases in persistent primary insomnia (PPI) relative to acute insomnia (AI), delayed sleep phase syndrome (DSPS: anendogenous circadian disorder), and good sleep (GS). In this report we summarize data from recent experiments using three different paradigms –the emotional Stroop task, the flicker paradigm for inducing change blindness and the dot probe task. Experimental studies have been completed involving almost 300 participants. Adults with AI and PPI, associated with cancer (n = 33; mean age = 47 years) were selected for the Stroopstudies. The flicker task was completed by 192 campus-recruited adults (mean age = 32 years), subsequently grouped into poor (PS), moderate(MS) and GS using Pittsburgh Sleep Quality Index (PSQI) scores. The dot probe task was completed by university students meeting diagnostic criteria for PPI, DSPS and GS (n = 63). In the Stroop task, as expected, both AI and PPI demonstrated attentional bias for cancer-related word cues, but there was also a significant group · word type interaction [F(2.62) = 1.7, P < 0.001]. Calculation of interference index scores illustrated that this effect was due to PPI exhibiting interference (delayed reaction time) for sleep words. The flicker paradigm explores attentional bias using visual scenes rather than textual words. Analyses of change detection latency scores (number of flickers to detection of one changed feature of scene) revealed only the change type (sleep-relevant/non-relevant) · sleep quality (PS/MS/GS) interaction were significant [F(2.180) = 7.12,P < 0.01). Regression analyses confirmed that PSQI scores explained variance in change detection scores [R2 = 10.6 (P < 0.05); b = –0.351(P < 0.05)]. In the dot probe task, visual attention is measured by detection latency for a probe appearing in the spatial location of either of two words (above/below fixation), immediately after the display has terminated. Trials providing data of interest are those in which one of the words has emotional salience (sleep-related). Our results from this trial are more equivocal. We found no between group or interaction main effects across PPI DSPS and GS groups (all P > 0.05). Shift in attentional focus from (transient) life events to sleep state function may help to explain emergence of PPI and attentional bias may be a factor maintaining dysfunctional sleep preoccupation.