Response to letter to the editor “Anesthetics and clock genes expression in critically ill patients”

We thank Dr. Ben-Hamouda for his interest in our recent paper describing abnormal circadian gene expression rhythms in critically ill patients.(1) He raised thought-provoking points related to the role of light exposure and anesthetic medications as potential mediators of the gene expression disruptions we observed. We appreciate the opportunity to provide our perspectives on his comments.

With respect to light exposure, Dr. Ben-Hamouda pointed to prior work in which we reported that critically ill patients receive abnormally low daytime light exposure, and proposed that the disruption observed in clock gene expression rhythms in critically ill patients could results from lack of light.(2) While inadequate daytime light may contribute to rhythm disruption over a long time period and light therapy may represent an opportunity for therapeutic intervention, elimination of normal daytime light does not immediately desynchronize central or peripheral rhythms such as we observed in critically ill patients. In fact, the control samples in our study were obtained from healthy patients in a constant routine that included dim (<20 lux) daytime lighting. Given that the critically ill patients were sampled beginning within the first day of hospitalization as well, the light exposure profiles of both groups were similar.

Dr. Ben-Hamouda also draws our attention to interesting animal studies in which several sedative anesthetics commonly used in intensive care units have been shown to reduce clock gene expression in rats and other rodents. Medications could certainly affect gene expression directly or indirectly by affecting sleep-wake state. Our recent study of melatonin rhythms did not identify an independent effect of sedatives on melatonin secretion, so we had not analyzed for an effect of sedatives in these data.(3) Prompted by Dr. Ben-Hamouda’s insights, we explored our data with further analyses and found that the amplitude of NR1D1 and CRY2 were diminished in patients exposed to sedatives by univariate analyses and after adjustment for Glasgow Coma Scale scores in linear regression models (sedation exposure for NR1D1 β 0.29, p=0.011 and for CRY2 β 0.12, p=0.019). Absolute error for TImeSignature was higher in sedative exposed patients as well (mean 5.9 versus 4.3 hours), but the difference was not significant by univariate testing or in an adjusted model. With the caveats that these results are from a set of multiple, exploratory hypothesis tests not adjusted for type 1 error risk and residual confounding may be present, in the context of similar findings in animal data, they at least support the merit of reassessing these potential associations in a larger sample of human subjects.

Identifying risk factors for disrupted gene expression rhythms is an important goal for future research. Some factors, such as sedative exposure, may be modifiable. Other factors that are not directly involved in pathologic disruption, like light exposure, could still be exploited for circadian rhythm restoration. Gene expression rhythms are essential for cellular and physiologic functions, and persistently disrupted peripheral clocks may impede rehabilitation and full recovery from acute illness.