Abstract
Circadian rhythms are known to be entrained to the 24 hour day through interactions between endogenous mechanisms and a number exogenous factors that include the light-dark cycle and social interaction. Intensive care patients are thought to be prone to disturbances in their circadian rhythms due to isolation from environmental changes and sensory deprivation as a result of sedation and sleep disturbance. The main aim of this thesis was to determine whether circadian rhythms in critically ill patients are desynchronised from the environment in comparison with those observed in healthy individuals, using urinary aMT6s excretion, plasma melatonin and core body temperature as markers of the endogenous clock. Patients in the Intensive Care Unit (ICU) at the Royal Free Hospital, London, who were stable, inotrope-free, not undergoing haemodiafiltration and with an ICU stay of more than three days were studied. Cosinor analysis of retrospective core body temperature data taken from nursing charts showed that a statistically significant rhythm with a period of 24h was present in 49.5% of the patient days studied. The position of acrophase showed abnormalities within individuals and between individuals. Prospective core body temperature studies confirmed the findings of abnormal circadian rhythmicity but suggested that the rhythms were influenced by fever associated with infection. Urinary 6-sulphatoxymelatonin (aMT6s) excretion data detected rhythmic activity and allowed for the identification of four patterns of excretion rate over 24h: normal rhythmicity, phase shifted rhythms, abnormal with frequent fluctuations and abnormal but arrhythmic. Urinary aMT6s excretion rate measured on two or three occasions at weekly intervals identified changes associated with the normalisation of circadian rhythmicity, namely an increase in the amplitude in some patients and the development of significant diurnal changes from grossly abnormal patterns of urinary aMT6s excretion in others. This trend was seen in patients that were improving clinically, but could not be associated with disease severity as determined by APACHE II scoring. Studies on the circadian rhythms of urine volume and creatinine clearance suggested that renal function in the absence of acute renal failure was still impaired or possibly influenced by treatment medication and could modify urinary aMT6s excretion. Measurements of plasma melatonin identified circadian rhythmicity in 7 out of 9 patients of which 5 were significant by cosinor analysis. These 5 patients had plasma melatonin rhythms that were either normal or phase shifted. The circadian rhythms of plasma cortisol, plasma prolactin and plasma TSH in most ICU patients were also abnormal and showed poor correlations with plasma melatonin. Overall this thesis provides evidence of abnormal circadian rhythmicity in critically ill patients and identifies this group of patients as one that might benefit from light or melatonin treatment to help resynchronise rhythms.