Fig. 5: Association of chronotype with higher-order cognition (working memory and attention).

Behavioral and electrophysiological data were analyzed using mixed-factorial (daytime × chronotype) ANOVAs (see Table 3). a ECs had significantly higher “Hits” and percentage of “Hits” in the morning vs evening (t = 2.53, P = 0.023), whereas LCs showed the reverse pattern of results (t = 2.36, P = 0.032) in the 3-back working memory task. Both groups showed the same pattern for the d prime (t_ECs = 2.55, P = 0.022; t_LCs = 2.49, P = 0.025) showing a significantly higher accuracy percentage and d prime value at their circadian-preferred time. b RT was not significantly different across time of day in the groups (t_ECs = 1.13, P = 0.275; t_LCs = 0.18, P = 0.858). c ECs displayed a larger P300 component in the morning vs evening (t = 3.62, P = 0.003) and LCs showed a larger P300 in the evening (t = 2.27, P = 0.038) at electrode Pz. d Both, ECs (t_morning = 1.91, P = 0.074; t_evening = 5.88, P < 0.001) and LCs (t_morning = 4.32, P = 0.001; t_evening = 5.60, P < 0.001) displayed a stronger Stroop interference effect (RT_incongruent − RT_congruent) at their circadian non-preferred time. RT of the congruent, incongruent and overall trials were significantly slower at the circadian non-preferred time (ECs: t_con = 1.66, P = 0.117; t_incon = 2.63, P = 0.019; t_overall = 3.34, P = 0.004; LCs: t_con = 4.13, P = 0.001; t_incon = 5.26, P < 0.001; t_overall = 5.17, P < 0.001). e, f The N200 amplitude at electrode Fz was calculated in the Stroop stage for both, congruent and incongruent trials. It was larger for both groups at their circadian-preferred times, but the difference was significant only in ECs (t_con = 4.81, P = 0.001; t_incon = 2.65, P = 0.018). However, chorotypes had a significant effect on the “morning vs evening” N200 amplitude difference (see “Results”). For the N450 component details see supplementary Fig. 4f. g, h In the AX-CPT, both groups showed enhanced sustained attention, as indexed by higher accuracy (ECs: t = 2.64, P = 0.018; LCs: t = 2.62, P = 0.019) at their circadian-preferred times. The respective RT difference was significant in LCs only (t = 4.22, P = 0.001). i The P300 difference between morning and evening was significantly larger only for LCs (t = 2.63, P = 0.019). In a, b, d, g, h, the horizontal bar shows the median, the + shows the mean, the upper and lower boundaries show the 25th and 75th percentiles, respectively, and the whiskers show 5–95 percentile. All pairwise comparisons in panels a–d are calculated using Student’s t test (P < 0.05, two-tailed). n = 32 (16 per group). Data are presented as mean values ± SEM. ECs early chronotypes, LCs late chronotypes, eve evening, mo morning, RT  reaction time, AX-CPT   AX-continuous performance test, ERP   event-related potential, µV microvolt, P3   P300 component, N2   N200, ns nonsignificant. Asterisks [*] represent statistically significant differences.