Circadian and Sleep-Related Endocrine Rhythms in Schizophrenia Eve Van Cauter, PhD; Paul Linkowski, MD, PhD; Myriam Kerkhofs, MS; Philippe Hubain, MD; Mireille L'Hermite-Bal\l=e'\riaux,MS; Raoul Leclercq, MD; Michelle Brasseur, MD; Georges Copinschi, MD, PhD; Julien Mendlewicz, MD, PhD
\s=b\ Plasma levels of prolactin, growth hormone, corticotropin, and cortisol were measured at 15-minute intervals for 24 hours in nine unmedicated male schizophrenic patients and in nine agematched normal male subjects. Each study was preceded by 3 days of habituation to the laboratory environment. Sleep was polygraphically recorded. The circadian and pulsatile variations present in each hormonal profile were quantitatively characterized with the use of computer algorithms specifically designed for analyses of hormonal fluctuations. The major abnormality of neuroendocrine release that was observed in the schizophrenic patients was an almost threefold enhancement of the sleeprelated increase in the prolactin level, associated with an intensified frequency of nocturnal prolactin pulses. This increased stimulatory effect of sleep on prolactin secretion was evident immediately after sleep onset. The normal inhibition of cortisol secretion during early sleep was absent in schizophrenic patients. The major sleep abnormalities were a prolonged sleep latency and a reduction in total rapid eye movement stage sleep. During wakefulness, prolactin and cortisol levels were normal. The 24-hour profile of growth hormone was unaltered in schizophrenic patients, and a sleep-onset growth hormone pulse was observed in all patients. No abnormalities were noted in the levels or temporal organization of corticotropin secretion. Both the amplitude and the timing of the cortisol rhythm were normal. We conclude that, in schizophrenic men, pituitary-adrenal function and circadian time-keeping are normal but prolactin secretion is hyperresponsive to the physiologic stimulus of sleep onset. Schizophrenia thus appears to be characterized by a subset of neuroendocrine disturbances distinct from that observed in major endogenous depression. (Arch Gen Psychiatry. 1991 ;48:348-356)
Accepted for publication July 20,1990. From the Department of Medicine, University of Chicago (III) (Dr Van Cauter), and the School of Medicine, Universit\l=e'\Libre de Bruxelles, Belgium (Drs Linkowski, Hubain, Leclercq, Brasseur, Copinschi, and Mendlewicz and
Mss L'Hermite-Bal\l=e'\riauxand Kerkhofs). Reprint requests to the Department of Medicine, Box 138, University of Chicago, 5841S Maryland Ave, Chicago, IL 60637 (Dr Van Cauter).
Mostphrenia pituitary
studies on neuroendocrine abnormalities in schizo¬ have involved analyses of the hypothalamicaxis responses to pharmacologie challenges.1,2 The studies that have involved investigation of the physiologic variations of basal pituitary secretion in unmedicated schizo¬ phrenic subjects3"8 either have considered only part of the 24-hour cycle or were based on sampling every 2 to 9 hours, regimens that fail to delineate the profound diurnal and pulsa¬ tile fluctuations that are known to modulate endocrine release in each of the hypothalamic-pituitary axes. In contrast, in major depressive illness, the temporal profile of pituitary and pituitary-dependent hormone release under basal conditions has been characterized in detail.9"20 These studies have shown that continuously elevated levels of plasma cortisol character¬ ize a subset of acutely depressed subjects.9,13"16 Increased day¬ time growth hormone (GH) secretion,12 decreased sleep-re¬ lated GH release,1221 and abnormal patterns of nocturnal pro¬ lactin secretion10,19 have also been observed in these patients. Furthermore, some ofthese chronobiologic studies have dem¬ onstrated that a number of endocrine events timed by the central circadian pacemaker occur earlier in acutely ill pa¬ tients with major endogenous depression compared with nor¬ mal age-matched control subjects.10,12,13,16,19 Considered with the frequent finding of abnormally short rapid eye movement (REM) latencies in depressed subjects, these studies have provided the basis of the "phase-advance hypothesis of de¬ pression," which postulates that a disorder of circadian time¬ keeping may be part of the pathophysiologic nature of affec¬ tive illness.22"24 However, a series of findings in schizophrenic patients have questioned the specificity of abnormalities of corticotro¬ pin secretion and alterations in circadian time-keeping for depression. Indeed, several studies have documented, in some schizophrenic patients, the existence of neuroendocrine and sleep abnormalities similar to those found in depression. In particular, abnormal responses to the dexamethasone sup¬ pression test (DST)25 have been observed in some schizo¬ phrenic patients, suggesting that hypercortisolism may not be a specific marker for depressive illness. Findings regard¬ ing the occurrence of pathologically short REM sleep laten-
Downloaded From: http://archpsyc.jamanetwork.com/ by a DALHOUSIE UNIVERSITY-DAL-11762 User on 06/24/2015
Table 1 .—Clinical Characteristics of Schizophrenic Patients* Duration of Wash-out REM
Subject/Age, y 1/24
Latency, min
116
Diagnosis and
32
BPRS Score 68
Subtype
19
Age at
-"->
16 h
23 h
Onset of Illness, y
\s=b\ Plasma levels of prolactin, growth hormone, corticotropin, and cortisol were measured at 15-minute intervals for 24 hours in nine unmedicated male schizophrenic patients and in nine agematched normal male subjects. Each study was preceded by 3 days of habituation to the laboratory environment. Sleep was polygraphically recorded. The circadian and pulsatile variations present in each hormonal profile were quantitatively characterized with the use of computer algorithms specifically designed for analyses of hormonal fluctuations. The major abnormality of neuroendocrine release that was observed in the schizophrenic patients was an almost threefold enhancement of the sleeprelated increase in the prolactin level, associated with an intensified frequency of nocturnal prolactin pulses. This increased stimulatory effect of sleep on prolactin secretion was evident immediately after sleep onset. The normal inhibition of cortisol secretion during early sleep was absent in schizophrenic patients. The major sleep abnormalities were a prolonged sleep latency and a reduction in total rapid eye movement stage sleep. During wakefulness, prolactin and cortisol levels were normal. The 24-hour profile of growth hormone was unaltered in schizophrenic patients, and a sleep-onset growth hormone pulse was observed in all patients. No abnormalities were noted in the levels or temporal organization of corticotropin secretion. Both the amplitude and the timing of the cortisol rhythm were normal. We conclude that, in schizophrenic men, pituitary-adrenal function and circadian time-keeping are normal but prolactin secretion is hyperresponsive to the physiologic stimulus of sleep onset. Schizophrenia thus appears to be characterized by a subset of neuroendocrine disturbances distinct from that observed in major endogenous depression. (Arch Gen Psychiatry. 1991 ;48:348-356)
Accepted for publication July 20,1990. From the Department of Medicine, University of Chicago (III) (Dr Van Cauter), and the School of Medicine, Universit\l=e'\Libre de Bruxelles, Belgium (Drs Linkowski, Hubain, Leclercq, Brasseur, Copinschi, and Mendlewicz and
Mss L'Hermite-Bal\l=e'\riauxand Kerkhofs). Reprint requests to the Department of Medicine, Box 138, University of Chicago, 5841S Maryland Ave, Chicago, IL 60637 (Dr Van Cauter).
Mostphrenia pituitary
studies on neuroendocrine abnormalities in schizo¬ have involved analyses of the hypothalamicaxis responses to pharmacologie challenges.1,2 The studies that have involved investigation of the physiologic variations of basal pituitary secretion in unmedicated schizo¬ phrenic subjects3"8 either have considered only part of the 24-hour cycle or were based on sampling every 2 to 9 hours, regimens that fail to delineate the profound diurnal and pulsa¬ tile fluctuations that are known to modulate endocrine release in each of the hypothalamic-pituitary axes. In contrast, in major depressive illness, the temporal profile of pituitary and pituitary-dependent hormone release under basal conditions has been characterized in detail.9"20 These studies have shown that continuously elevated levels of plasma cortisol character¬ ize a subset of acutely depressed subjects.9,13"16 Increased day¬ time growth hormone (GH) secretion,12 decreased sleep-re¬ lated GH release,1221 and abnormal patterns of nocturnal pro¬ lactin secretion10,19 have also been observed in these patients. Furthermore, some ofthese chronobiologic studies have dem¬ onstrated that a number of endocrine events timed by the central circadian pacemaker occur earlier in acutely ill pa¬ tients with major endogenous depression compared with nor¬ mal age-matched control subjects.10,12,13,16,19 Considered with the frequent finding of abnormally short rapid eye movement (REM) latencies in depressed subjects, these studies have provided the basis of the "phase-advance hypothesis of de¬ pression," which postulates that a disorder of circadian time¬ keeping may be part of the pathophysiologic nature of affec¬ tive illness.22"24 However, a series of findings in schizophrenic patients have questioned the specificity of abnormalities of corticotro¬ pin secretion and alterations in circadian time-keeping for depression. Indeed, several studies have documented, in some schizophrenic patients, the existence of neuroendocrine and sleep abnormalities similar to those found in depression. In particular, abnormal responses to the dexamethasone sup¬ pression test (DST)25 have been observed in some schizo¬ phrenic patients, suggesting that hypercortisolism may not be a specific marker for depressive illness. Findings regard¬ ing the occurrence of pathologically short REM sleep laten-
Downloaded From: http://archpsyc.jamanetwork.com/ by a DALHOUSIE UNIVERSITY-DAL-11762 User on 06/24/2015
Table 1 .—Clinical Characteristics of Schizophrenic Patients* Duration of Wash-out REM
Subject/Age, y 1/24
Latency, min
116
Diagnosis and
32
BPRS Score 68
Subtype
19
Age at
-"->
16 h
23 h
Onset of Illness, y
