Polymorphisms of ADORA2A modulate psychomotor vigilance and the effects of caffeine on neurobehavioural performance and sleep EEG after sleep deprivation

Abstract

Background and purpose: Prolonged wakefulness impairs sustained vigilant attention, measured with the psychomotor vigilance task (PVT), and induces a compensatory increase in sleep intensity in recovery sleep, quantified by slow-wave activity (SWA) in the sleep electroencephalogram (EEG). These effects of sleep deprivation are counteracted by the adenosine receptor antagonist caffeine, implying involvement of the adenosine neuromodulator/receptor system. To examine a role for adenosine A(2A) receptors, we investigated whether variation of the A(2A) receptor gene (ADORA2A) modified effects of caffeine on PVT and SWA after sleep deprivation.

Experimental approach: A haplotype analysis of eight single-nucleotide polymorphisms of ADORA2A was performed in 82 volunteers. In 45 young men carrying five different allele combinations, we investigated the effects of prolonged waking and 2 × 200 mg caffeine or 2 × 100 mg modafinil on psychomotor vigilance, sleepiness, and the waking and sleep EEG.

Key results: Throughout extended wakefulness, the carriers of haplotype HT4 performed faster on the PVT than carriers of non-HT4 haplotype alleles. In haplotype HT4, caffeine failed to counteract the waking-induced impairment of PVT performance and the rebound of SWA in recovery sleep. However, caffeine was effective in non-HT4 allele carriers, and modafinil reduced the consequences of prolonged waking, independently of ADORA2A haplotype.

Conclusions and implications: Common genetic variation of ADORA2A is an important determinant of psychomotor vigilance in rested and sleep-deprived state. It also modulates individual responses to caffeine after sleep deprivation. These findings demonstrate a role for adenosine A(2A) receptors in the effects of prolonged wakefulness on vigilant attention and the sleep EEG.

Figure 1

Genetic variation of ADORA2A determines inter-individual difference in psychomotor vigilance during prolonged wakefulness. Mean values (+ SEM) in carriers of HT4 (n = 14) and non-HT4 haplotype alleles of ADORA2A are plotted. HT1: n = 21, HT2: n = 16, HT3: n = 15, HT5: n = 13. A majority of subjects were heterozygous HT allele carriers (see Supporting Information Table S1). Ticks on the x-axes were rounded to the previous hour. Starting 30 min after wake-up from the baseline night, a 10 min PVT was administered at 3 h intervals during 40 h prolonged wakefulness. The 90th percentile of reaction times on the PVT per session was expressed as speed (1/reaction time). Two-way mixed-model anova with the between-subjects factor ‘haplotype’ (HT1–HT5) and the within-subjects factor ‘session’ (1–14) and the co-variate ‘age’ revealed that individuals with haplotype HT4 performed faster than non-HT4 allele carriers throughout sleep deprivation (‘haplotype’: F_4,1033= 15.41, P < 0.0001; ‘session’: F_13,1033= 22.03, P < 0.0001; ‘haplotype’בsession’: F_52,1034= 0.14, P > 0.9).

Figure 2

Stable difference in psychomotor vigilance between carriers of HT4 (n = 14) and non-HT4 haplotype (n = 31) alleles of ADORA2A. The 90th percentile of reaction times (expressed as speed, 1/reaction time) on the PVT in test sessions at 3, 6 and 9 h awake (prior to first placebo/stimulant administration) was averaged. Means + SEM are shown. Individuals with HT4 haplotype performed consistently faster than non-HT4 allele carriers in placebo and stimulant conditions, occurring in random order 1 week apart (anova: ‘haplotype’: F_1,43= 9.3, P < 0.004; ‘condition’: F_1,43= 0.19, P > 0.6; ‘haplotype’×’condition’: F_1,43= 0.13, P > 0.7). P-values refer to unpaired two-tailed t-tests.

Figure 3

Genetic variation of ADORA2A differently modulates the effects of caffeine and modafinil on impaired psychomotor vigilance after sleep loss. The PVT was administered at 3 h intervals throughout prolonged wakefulness, beginning 30 min after wake-up from the baseline nights. Mean values (±SEM) of the 90th percentile of reaction times (expressed as speed, 1/reaction time) in carriers of HT4 (left panels) and non-HT4 haplotype (right panels) alleles of ADORA2A are plotted. Ticks on the x-axes were rounded to the previous hour. Study participants received 2 × placebo, and 2 × 200 mg caffeine or 2 × 100 mg modafinil during two 40 h periods of prolonged wakefulness, separated by 1 week. (A and B) Caffeine improved PVT response speed after sleep loss in non-HT4 haplotype carriers of ADORA2A only (anova: ‘haplotype’: F_1,21.2= 7.91, P < 0.02; ‘session’: F_13,205= 13.43, P < 0.0001, ‘treatment’: F_1,68= 8.45, P < 0.005; ‘haplotype’בtreatment’בsession’: F_26,219= 2.1, P < 0.003). (C and D) Modafinil improved PVT response speed after sleep loss independently of ADORA2A haplotype (‘session’: F_13,198= 10.24, P < 0.0001; ‘treatment’: F_1,64.9= 12.45, P = 0.0008; ‘haplotype’בsession’: F_13,198= 1.87, P < 0.04; ‘haplotype’בtreatment’בsession’: F_26,205.= 1.35, P > 0.12). *P < 0.05 (stimulant vs. placebo, paired two-tailed t-tests).

Figure 4

Caffeine attenuates the rebound of EEG delta oscillations (0.75–4.5 Hz) after sleep loss in subjects with non-HT4 haplotype of ADORA2A only. Bars represent the mean increase (+SEM) in slow-wave activity in the first NREM sleep episode (stages 1–4) of the recovery night, expressed as a percentage of the corresponding value in the baseline night. Individuals with HT4 (n = 5) and non-HT4 haplotype (n = 15) alleles received 2 × placebo and 2 × 200 mg caffeine during two 40 h periods of prolonged wakefulness, separated by one week. Caffeine reduced the rebound in delta activity when compared with placebo (anova: ‘treatment’: F_1,18= 5.59, P < 0.03; ‘haplotype’: F_1,18= 1.13, P = 0.3; ‘treatment’×’haplotype’: F_1,18= 2.23, P = 0.15). P-value refers to paired two-tailed t-test.