Concentration–effect relationships of plasma caffeine on EEG delta power and cardiac autonomic activity during human sleep

Abstract

SummaryAcute caffeine intake affects brain and cardiovascular physiology, yet the concentration–effect relationships on the electroencephalogram and cardiac autonomic activity during sleep are poorly understood. To tackle this question, we simultaneously quantified the plasma caffeine concentration with ultra‐high‐performance liquid chromatography, as well as the electroencephalogram, heart rate and high‐frequency (0.15–0.4 Hz) spectral power in heart rate variability, representing parasympathetic activity, with standard polysomnography during undisturbed human sleep. Twenty‐one healthy young men in randomized, double‐blind, crossover fashion, ingested 160 mg caffeine or placebo in a delayed, pulsatile‐release caffeine formula at their habitual bedtime, and initiated a 4‐hr sleep opportunity 4.5 hr later. The mean caffeine levels during sleep exhibited high individual variability between 0.2 and 18.4 μmol L−1. Across the first two non‐rapid‐eye‐movement (NREM)–rapid‐eye‐movement sleep cycles, electroencephalogram delta (0.75–2.5 Hz) activity and heart rate were reliably modulated by waking and sleep states. Caffeine dose‐dependently reduced delta activity and heart rate, and increased high‐frequency heart rate variability in NREM sleep when compared with placebo. The average reduction in heart rate equalled 3.24 ± 0.77 beats per minute. Non‐linear statistical models suggest that caffeine levels above ~7.4 μmol L−1 decreased electroencephalogram delta activity, whereas concentrations above ~4.3 μmol L−1 and ~ 4.9 μmol L−1, respectively, reduced heart rate and increased high‐frequency heart rate variability. These findings provide quantitative concentration–effect relationships of caffeine, electroencephalogram delta power and cardiac autonomic activity, and suggest increased parasympathetic activity during sleep after intake of caffeine.