Diurnal variations of interleukin-6 plasma levels are confounded by blood drawing procedures

Abstract

Recent findings suggest that inflammatory cytokines are involved in sleep regulation. In part, this idea is based on studies showing that systemic levels of interleukin-6 (IL-6) are affected by sleep and sleep deprivation. However, intravenous (IV) catheters used for repetitive blood sampling were reported to increase local IL-6 production, which might confound sleep-dependent or circadian changes in the plasma concentrations of this cytokine. To further examine the effects of blood drawing procedures on IL-6 plasma levels, 12 healthy young male subjects participated in a 24-h cross-over study protocol involving sleep and sleep deprivation. Blood was collected half-hourly through an IV line and one additional sample was taken by a simple needle stick from the contralateral arm in parallel to the last sample from the catheter. Difficulties in blood sampling, the plasma levels of IL-6, cortisol and subjective sleepiness were quantified. In samples from the IV line there was a linear increase in IL-6 levels in both conditions, whereas the amount of IL-6 detected in the needle stick sample at the end did not differ from baseline. IL-6 levels were significantly higher in samples rated as difficult and those difficulties were more frequent during sleep compared to nocturnal wakefulness. IL-6 levels did not correlate to variations in sleepiness or cortisol levels. We conclude that variations in IL-6 plasma levels measured in samples from an IV catheter are caused, at least in part, by changes in local cytokine production rather than by physiological changes in circulating IL-6 levels.

Introduction

Cytokines have been suggested to act as physiologic sleep regulatory substances. Animal studies have shown that interleukin (IL)-1 and tumor necrosis factor (TNF)-α fulfil many of the criteria that have been proposed to qualify a substance as a sleep regulatory one (Krueger and Obál, 1997). In humans, little is known about the role of inflammatory cytokines in the sleep–wake processes. With respect to IL-1 and TNF-α, the few studies performed so far are not consistent, showing either sleep–wake-related changes of the in vivo or in vitro production of IL-1 and TNF-α (Moldofsky et al., 1986, Hohagen et al., 1993, Young et al., 1995, Entzian et al., 1996, Petrovsky et al., 1998), or documenting no substantial variations (Dinges et al., 1995, Born et al., 1997). However, there is somewhat more consistent evidence that in humans peripheral IL-6 concentrations are related to sleep–wake processes. For example, circulating levels of IL-6 have been found to vary across the nycthemeron with highest levels occurring during sleep (Gudewill et al., 1992, Bauer et al., 1994, Sothern et al., 1995, Vgontzas et al., 1999), and recently, Redwine et al. (2000) have shown that sleep onset was associated with an increase of IL-6 levels, independent of whether sleep started at the habitual time or about 4 h later. However, sleep deprivation has also been reported to be associated with increased IL-6 levels: Vgontzas et al. (1999) reported that after one night of sleep deprivation the mean daytime IL-6 levels were higher than after a normal night sleep the day before. Recently, Shearer et al. (2001) found that after 4 days of total sleep deprivation average daytime IL-6 levels were higher than after 4 days of partial sleep deprivation.

These diurnal and sleep–wake associated variations indicate that IL-6 might play a role in the regulation of sleep. In these studies, an IV catheter for blood sampling was used. Such catheters, however, have been shown to influence local IL-6 production. When circulating IL-6 levels were analyzed in plasma samples obtained from an indwelling venous catheter, they increased about 100% or even more after sampling periods of 10 h or longer. In contrast, no increase was observed when IL-6 concentrations were analyzed in samples obtained after the same time interval, but by a simple needle stick at the contralateral arm (Seiler et al., 1994, Gudmundsson et al., 1997, Haack et al., 2000). These studies suggest that circulating levels of IL-6 measured in samples obtained by an IV catheter reflect local, rather than systemic increases in IL-6 levels. This local increase of IL-6 levels might reflect subtle inflammatory processes at the site of the catheter. The levels of other cytokines and soluble cytokine receptors, such as TNF-α and the soluble TNF receptors p55 and p75, are only slightly influenced by an intravenous catheter (Haack et al., 2000). This suggests that IL-6 might also be sensitive to other confounding factors: For example, the blood drawing process and related difficulties, e.g. frequent and/or strong aspiration of blood, may influence IL-6 levels. In general, blood drawing from an indwelling catheter is more difficult when subjects are asleep, because during sleep arm positions are often inept for blood drawing. These difficulties might induce subtle mechanical damage to the vessel wall or to individual cells, resulting in cytokine release. Therefore, blood samples taken during sleep from an indwelling catheter might yield higher IL-6 levels than those obtained during wakefulness. This may confound findings of differences of IL-6 levels between sleep and wakefulness.

The purpose of the present study was to examine in detail whether diurnal and/or sleep–wake dependent variations of IL-6 levels are influenced by the blood drawing process. Based on previous studies, we expected a continuous increase of IL-6 levels across 24 h and an additional sleep-related increase. Specifically, we hypothezised that due to catheter-induced alterations, IL-6 levels would not reach baseline values again after 24 h of blood sampling through an IV catheter, and that difficulties in blood drawing are more frequent during sleep as compared to wakefulness and go along with increased IL-6 levels.