- Email: [email protected]
Cerebral glucose utilization during stage 2 sleep in man
Brain Research, 571 (1992) 149-153 Elsevier
149
BRES 25013
Cerebral glucose utilization during stage 2 sleep in man Pierre Maquet, Dominique Dive, Eric Salmon, Bernard Sadzot, Gianni Franco, Robert Poirrier and Georges Franck Department of Neurology, University of Lidge, CHU Sart Tilman, Liege (Belgium) and Cyclotron Research Center, University of Liege, Sart Tilman, Liege (Belgium) (Accepted 15 October 1991)
Key words: Positron emission tomography; [lSF]Fluorodeoxyglucose; Cerebral glucose metabolism; Sleep/wake cycle; Stage 2 sleep; Slow wave sleep; Rapid eye movement sleep
Using [lSF]fluorodeoxyglucosemethod and positron emission tomography, we performed paired determinations of the cerebral glucose utilization at one week intervals during sleep and wakefulness, in 12 young normal subjects. During 6 of 28 sleep runs, a stable stage 2 SWS was observed that fulfilled the steady-state conditions of the model. The cerebral glucose utilization during stage 2 SWS was lower than during wakefulness, but the variation did not significantly differ from zero (mean variation: -11.5 _+ 25.57%, P ffi 0.28). The analysis of 89 regions of interest showed that glucose metabolism differed significantly from that observed at wake in 6 brain regions, among them both thalamic nuclei. We conclude that the brain energy metabolism is not homogeneous throughout all the stages of non-REMS but decreases from stage 2 SWS to deep SWS; we suggest that a low thalamic glucose metabolism is a metabolic feature common to both stage 2 and deep SWS, reflecting the inhibitory processes observed in the thalamus during these stages of sleep. Stage 2 SWS might protect the stability of sleep by insulating the subject from the environment and might be a prerequisite to the full development of other phases of sleep, especially deep SWS. Usually considered as a transitional stage of sleep, stage 2 slow wave sleep takes however nearly half of the total sleep time of the young adult 1°. Despite the temporal importance of this stage of sleep, the cerebral glucose metabolism has not .yet been selectively measured during stage 2 SWS. In the framework of an experiment measuring the cerebral glucose utilization in man during deep slow wave (SWS) and R E M phases of sleep using positron emission tomography (PET) and [lSF]fluorodeoxyglucose (FDG) method s, we had the opportunity to determine the CMRGIu of 6 subjects who maintained a stable stage 2 SWS during the uptake period of FDG. These results are presented here. Materials and methods were extensively described in a previous article s. Twelve healthy subjects selected as good sleepers aged from 21 to 33 years old (mean: 23.8 + 3.0) were examined by P E T using F D G method for a total of 28 man-nights (40 PET studies, taking waking studies into account). For each volunteer, besides a PET study during diurnal wakefulness, 2 or 3 nocturnal sleeps were investigated at one week interval. Sleep cerebral glucose metabolism of each subject was compared to his own waking cerebral metabolic rates, acting as control, In the evening of PET studies, the subjects fell asleep
spontaneously under continuous polygraphic monitoring. Five to 10 mCi of F D G were injected intravenously when the sleep stage under study remained steady for a few minutes. Arterial blood samples were drawn during the whole procedure. The subjects were woken up 45-50 rain after the F D G injection, then quickly placed on the table of the tomograph (Neuro-ECAT, E G & G Ortec), transverse resolution: 12.4 mm FWHM, axial resolution: 15 m m FWHM). Nine to 15 scan planes were acquired parallel to and 20-78 mm above the orbito-meatal plane. Scanning lasted about 20 min. Attenuation correction was calculated using a uniform absorption coefficient (0.088 cm -~) and an operator-positioned ellipse to define the edge of the scalp. Cerebral metabolic rates (CMRGIu) were calculated using the equation and the values of rate and lumped constants of Phelps et al.6 for normal awake humans. Hypnograms of the uptake periods were established using the Rechtschaffen and Kales criteria s. As previously described 5, the influence of a particular stage of sleep on the final value of CMRGIu is estimated by an 'index of stability', namely the ratio of the amount of plasma F D G available during this stage weighted by the amount of F D G transported by the plasma during the
Abbreviations: CMRGIu, cerebral metabolic rate of glucose; deep SWS, stage 3 and 4 SWS; EEG, eleetroancephalographic; FDG, fluorodeoxyglucose; LRD, left/right difference; MI, metabolic index; non-REMS: stage 2, 3 and 4 SWS; PET, positron emission tomography; REMS, rapid eye movement sleep; ROI, region of interest; SWS, slow wave sleep. Correspondence: O. Franek, Department of Neurology, CHU Sart Tilman (B35), B-4000 Liege, Belgium.
150 8 3
4
9
19
10
11 i ~
1
12
i3
17 18
20
21
92
27
28
23 !
,24
29
A 46
32
30 31
33
47
36
B 56
.54
52
53
5.5
57
68
, 69
58
: 59
50
,o
.
.
.
,2
.
,4,5
°o
72
6, 65
63 E
73
77
78
79
~
6~
70 71
76 :i",:i
...... °,
. . . . . .
,3 D
C
......
~
80
..
e= 03
8~F
1 86 F
:
~
~ , g7
G
Fig. 1. The 7 planes of references and the 89 preselected regions of interest used in this study. Several regions of interest were lumped in areas of interest as follows: Medial frontal 8,9,17,18,30,31,52,53,70,71 Lateral frontal 19,20,32,33,54,55,72,73 Orbital frontal 10,11 Inferior frontal 21,22,34,35,56,57,74,75 Anterior temporal 1,2 Middle temporal 12,13,40,41,60,61,78,79 Cortical regions Superior temporal 25,26,38,39,58,59 Parietal 86,87 Lateral occipital 42,43,62,63,80,81 Medial occipital 44,45,64,65,82,83 Insula 23,24,36,37 Mean grey matter Sensorimotor cortex 76,77,84,85 Hippocampus 3,4,14,15
Subcortical regions
whole F D G uptake period. We considered that the steady-state conditions of the model were met if IS >~ 0.75. Regional cerebral metabolic rates were obtained from 89 elliptical regions of interest (ROIs; see Fig. 1). Several regions of interest appear on different planes. For easier presentation of the data, they were lumped together in larger areas (areas of interest) spreading over a
Anterior striatum Caudate nuclei Lenticular nuclei Thalami Brainstem Cerebellum
27,28 46,47,66,67 48,49 50,51,68,69 7,16,29 5,6
large part of a lobe. The C M R G I u of an area of interest is the average of the metabolic rates of its constituent regions of interest. The modifications of cerebral glucose utilization during sleep were calculated by: A C M R G I u = (sleep C M R G l u - wake CMRGlu)/wake C M R G I u x 100. The regional distribution of cerebral glucose metabolism was estimated using regional metabolic indices (MI):
151
TABLE I Cerebral metabolic rates (CMRGIu) and standard deviation (SD) during stage 2 sleep and during wakefulness of the 6 selected runs in the major areas of interest
Mean variations of metabolic rates (ACMRGIu) between them and their standard deviations (SD). For comparison, ACMRGlu observed between stage 3 and 4 SWS and wakefulness in 5 other subjects appear in the two last colums (ref. 5). Areas of interest
During stage 2 sleep (mg/lO0 g.min)
During wakefulness (mg/lO0 g.min)
(st. 2 SWS vs wake)
(st. 3 & 4 SWS vs wake)
CMRGIu
SD
CMRGlu
SD
ACMRGIu1 (%)
SD (%)
ACMRGIu 1 (%)
SD (%)
Medial frontal Lateral frontal Orbito frontal Inferior frontal Anterior temporal Middle temporal Superior temporal Lateral occipital Medial occipital Insula Hippocampus
5.51 4.84 4.91 5.40 4.59 4.73 5.43 4.11 5.63 4.88 3.94
1.86 1.64 1.86 1.79 1.55 1.62 1.73 1.45 1.83 1.57 1.34
6.05 5.70 5.74 6.24 4.66 5.15 6.12 4.26 6.22 5.45 4.15
1.17 1.03 1.24 1.05 1.35 1.02 0.96 0.70 1.31 0.82 1.23
-9.64 -15.41 -18.2 -14.29 -8.12 -9.13 -12.12 -4.71 -9.65 -11.47 -4.76
26.76 26.83 18.84 24.64 21.94 25.49 24.41 29.72 24.76 23.76 24.74
-42.832 --46.032 -43.662 -43.812 -39.002 -44.392 -42.732 -39.702 -43.232 -42.142 -35.212
14.26 13.67 13.65 14.84 4.23 14.47 13.85 19.38 17.40 12.44 0.94
Mean cortical regions
5.00
1.66
5.56
1.00
-11.06
25.73
-43.832
14.49
Anterior striatum Caudate nucleus Lenticular nucleus Thalamus Brainstem Cerebellum
5.59 5.59 5.93 5.04 3.97 3.97
1.74 1.82 1.84 1.71 1.32 1.23
6.23 6.23 6.75 6.34 4.36 4.38
0.84 1.06 0.99 1.15 0.98 1.32
-11.05 -10.42 -12.66 -21.53 -10.00 -13.24
24.76 27.89 24.42 23.55 22.83 26.03
-37.962 -36.092 -41.852 -48.952 -42.462 -48.822
12.54 5.94 12.25 14.75 15.79 7.29
Mean subcortical regions
4.96
1.58
5.70
1.00
-13.82
23.34
-43.302
11.78
Mean (grey matter)
4.99
1.65
5.58
0.99
-11.55
25.27
-43.802
14.10
1 The variation is negative if the sleep CMRGlu is lower than wake CMRGIu. 2 P < 0.05.
MI = regional C M R G l u / m e a n cortical C M R G l u . Metabolic indices, being by definition regional values, were not lumped in larger areas. The intra-individual variations of MI were evaluated for each subject and for each particular region by (MIs - MIw)/(MIs + MIw) x 200 for the sleep population where MI s and MIw are the metabolic indices of the sleep and wake studies respectively. The left/right differences ( L R D ) were obtained by: L R D = (left C M R G I u - fight CMRGlu)/(left C M R G I u + right C M R G l u ) x 200. Paired two-tailed Student's t-test was performed between sleep and wake values of the metabolic rates and the metabolic indices. One-sample two-tailed Student's t-test was used to determine the significance from zero of the L R D . Paired two-tailed Student's t-test was performed between sleep and wake values of the L R D . The correction of Bonferroni was performed for each Student's t-test. In 6 runs, out of the 28 nocturnal P E T scans that we performed, these periods of stage 2 SWS were long
enough to fulfill our criterion of stability. In these cases, stage 2 SWS was always observed as a transition phase from other stages of sleep. In 2 cases, stage 2 SWS appeared after a deep SWS (stability indices: 0.95 and 0.87). In the other 4 cases it appeared after a period of R E M S (stability index: 0.84, 0.86, 0.92 and 0.97). In no case was the stage 2 SWS observed after wakefulness when the subject was falling asleep. Table I shows the mean C M R G I u observed during wakefulness and during the stage 2 SWS in the 6 aforementioned subjects, as well as the average variation of CMRGlu. O n the average, a trend toward a decrease in C M R G I u (-11.55%) was observed during stage 2 SWS. This variation was not significant, however (P = 0.28). No significant asymmetry was observed during stage 2 SWS (average asymmetry: 0.68 + 1.33%, P = 0.27). The M I did not change significantly during stage 2 SWS in all but 6 regions of interest (Table II) in which MI was significantly increased or decreased. Especially, the MI of both thalami was decreased. Because of the
152 TABLE II Metabolic indices (MI) and standard deviation (SD) during stage 2 sleep and during wakefulness in the 7 ROls where their variation (AMI; %) was significant (P < 0.05)
For comparison, AMI observed between stage 3 and 4 SWS and wakefulness in 5 other subjects appear in the two last cotums (ref. 5). ROls 2
Left lateral frontal Left lateral frontal Right medial occipital Right lateral occipital
Left thalamus Right thalamus
During st. 2 sleep
During wakefulness
AM11
SD
MI
SD
MI
SD
(192)
0.90
0.09
0.98
0.08
-8.85
7.71
(72)
1.01
0.90
0.99
0.07
4.11
0.50
(65)
1.12
0.98
1.05
0.11
6.50
5.07
(81)
0.85
0.77
0.74
0.63
8.35
1.38
(50)
0.97
0.011
1.12
0.52
-14.86
13.09
(51)
1.00
0.10
1.12
0.08
-12.27
11.47
ROls
AMI 1
SD
Right lateral frontal
(33)
-7.49
Left lateral occipital Left lateral occipital
(42)
11.79 1.60
(62)
8.10 2.72
(28)
10.78 3.09
(50)
-9.94
1.71
(51)
-12.55
2.91
Anterior striatum Left thalamus Right thalamus
2.80
1 The variation is negative when the sleep MI is lower than wake MI. 2 ROI number as it appears in Fig. 1.
great number of tests performed (89), no significant variation was observed when the correction of Bonferroni was applied. The CMRGIu during non-REMS was shown to be lower than during wakefulness in cat 7, monkey 4 and in man 1'5. Presuming that, in man, the cerebral glucose metabolism might differ from light to deep SWS, we investigated separately these two types of non-REM sleep. Our results show indeed that cerebral glucose metabolism is not homogeneous throughout all the stages of non-REMS. We showed previouslys that CMRGIu was dramatically decreased during the deep phases of nonR E M sleep (about 40% decrease; see Table I). In contrast, during stage 2 SWS, the CMRGIu was not found to be significantly different from that observed when awake. Thus, deep SWS, but not stage 2 SWS, appears to be linked to a unique status of cerebral energetic metabolism, drastically different from that observed when awake, However, if the rate of cerebral glucose utili7ation during stage 2 SWS is close to that of wakefulness, the distribution of the cerebral glucose metabolism during stage 2 SWS shares some similarity with that observed in deep SWS. In particular, a low metabolism in both thalamic nuclei (as reflected by the significant decrease of their MI) was observed both during stage 2 SWS (present results) and during deep SWS 5. Because of the number
of t-tests performed, the significant variation of the MI in thalamic nuclei as in 4 other ROIs might be due to type I error. However, the low thalamic glucose metabolism might also reflect the dysfacilitation and inhibition processes within thalamic structures during slow sleep 9. These phenomena induce the outbreak of long-lasting inhibitory potentials in thalamic neurons and conseqnently, a potent blockade of the ascending afferent volleys 2. Stage 2 SWS thus appears as a state during which ascending afferences are dampened and cortical activation is decreased. Stage 2 SWS might protect the stability of sleep by insulating the subject from the environment and might be a prerequisite to the full development of other phases of sleep, especially deep SWS. The level of cerebral glucose metabolism during stage 2 SWS is similar to the awake level: only deep SWS seems to be linked with a energy metabolism definitely different from that of wakefulness. Future research will determine if during deep SWS, some as yet unknown metabolic processes are taking place, that might be related to brain tissue restitution 3'5.
This research was supported by a grant from the Belgian National Fund for Medical Research (F.R.S.M. no. 3.4508.83) and from the Reine Elisabeth Medical Foundation. P.M. is Senior Research Assistant at the National Fund of Scientific Research (Belgium).
153 1 Buchsbaum, M.S., Giilin, J.C., WU, J., Hazlett, E., Sicotte, N., Dupont, R.M. and Bunney, W.E., Regional cerebral glucose metabolic rate in human sleep assessed by positron emission tomography, Life Sci., 45 (1989) 1349-1356. 2 Glen, L.L. and Steriade, M., Discharge rate and excitability of cortically projecting intralaminar thalamic neurons during waking and sleep states, J. Neurosci., 2 (1982) 1387-1404. 3 Home, J.A., Why We Sleep, Oxford University Press, Oxford, 1988, 319 pp. 4 Kennedy, C., Gillin, J.C., Mendelson, W., Soda, S., Miyaoka, M., Ito, M., Nakamura, R.K., Storch, F.I., Pettigrew, K., Mishkin, M. and Sokoloff, L., Local cerebral glucose utilization in non-rapid eye movement sleep, Nature, 297 (1982) 325-327. 5 Maquet, P., Dive, D., Salmon, E., Sadzot, B., Franco, G., Poirrier, R., yon Frenckel, R. and Franck, G., Cerebral glucose utilization during sleep-wake cycle in man determined by positron emission tomography and [lSF]2-fluoro-2-deoxy-D-
glucose method, Brain Research, 513 (1990) 136-143. 6 Phelps, M.E., Huang, S.C., Hoffman, E.J., Selin, C., Sokoloff, L. and Kuhl, D.E., Tomographic measurement of local cerebral glucose metabofic rate in humans with [lSF]2-fluoro-2-deoxy-Dglucose: validation of method, Ann. Neurol., 6 (1979) 371-388. 7 Ramm, P. and Frost, B.J., Cerebral and local cerebral metabolism in the eat during slow wave and REM sleep, Brain Research, 365 (1986) 112-124. 8 Rechtschaffen, A. and Kales, A.A., A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects, U.S. Department of Health, Education andWelfare, Bethesda, MD, 1968. 9 Stedade, M. and Hobson, J.A., Neuronal activity during the sleep-waking cycle, Prog. Neurobiol., 6 (1976) 155-376. 10 Webb, W.B. and Agnew, H.W., Sleep characteristics of long and short sleepers, Science, 168 (1970)145-147.
149
BRES 25013
Cerebral glucose utilization during stage 2 sleep in man Pierre Maquet, Dominique Dive, Eric Salmon, Bernard Sadzot, Gianni Franco, Robert Poirrier and Georges Franck Department of Neurology, University of Lidge, CHU Sart Tilman, Liege (Belgium) and Cyclotron Research Center, University of Liege, Sart Tilman, Liege (Belgium) (Accepted 15 October 1991)
Key words: Positron emission tomography; [lSF]Fluorodeoxyglucose; Cerebral glucose metabolism; Sleep/wake cycle; Stage 2 sleep; Slow wave sleep; Rapid eye movement sleep
Using [lSF]fluorodeoxyglucosemethod and positron emission tomography, we performed paired determinations of the cerebral glucose utilization at one week intervals during sleep and wakefulness, in 12 young normal subjects. During 6 of 28 sleep runs, a stable stage 2 SWS was observed that fulfilled the steady-state conditions of the model. The cerebral glucose utilization during stage 2 SWS was lower than during wakefulness, but the variation did not significantly differ from zero (mean variation: -11.5 _+ 25.57%, P ffi 0.28). The analysis of 89 regions of interest showed that glucose metabolism differed significantly from that observed at wake in 6 brain regions, among them both thalamic nuclei. We conclude that the brain energy metabolism is not homogeneous throughout all the stages of non-REMS but decreases from stage 2 SWS to deep SWS; we suggest that a low thalamic glucose metabolism is a metabolic feature common to both stage 2 and deep SWS, reflecting the inhibitory processes observed in the thalamus during these stages of sleep. Stage 2 SWS might protect the stability of sleep by insulating the subject from the environment and might be a prerequisite to the full development of other phases of sleep, especially deep SWS. Usually considered as a transitional stage of sleep, stage 2 slow wave sleep takes however nearly half of the total sleep time of the young adult 1°. Despite the temporal importance of this stage of sleep, the cerebral glucose metabolism has not .yet been selectively measured during stage 2 SWS. In the framework of an experiment measuring the cerebral glucose utilization in man during deep slow wave (SWS) and R E M phases of sleep using positron emission tomography (PET) and [lSF]fluorodeoxyglucose (FDG) method s, we had the opportunity to determine the CMRGIu of 6 subjects who maintained a stable stage 2 SWS during the uptake period of FDG. These results are presented here. Materials and methods were extensively described in a previous article s. Twelve healthy subjects selected as good sleepers aged from 21 to 33 years old (mean: 23.8 + 3.0) were examined by P E T using F D G method for a total of 28 man-nights (40 PET studies, taking waking studies into account). For each volunteer, besides a PET study during diurnal wakefulness, 2 or 3 nocturnal sleeps were investigated at one week interval. Sleep cerebral glucose metabolism of each subject was compared to his own waking cerebral metabolic rates, acting as control, In the evening of PET studies, the subjects fell asleep
spontaneously under continuous polygraphic monitoring. Five to 10 mCi of F D G were injected intravenously when the sleep stage under study remained steady for a few minutes. Arterial blood samples were drawn during the whole procedure. The subjects were woken up 45-50 rain after the F D G injection, then quickly placed on the table of the tomograph (Neuro-ECAT, E G & G Ortec), transverse resolution: 12.4 mm FWHM, axial resolution: 15 m m FWHM). Nine to 15 scan planes were acquired parallel to and 20-78 mm above the orbito-meatal plane. Scanning lasted about 20 min. Attenuation correction was calculated using a uniform absorption coefficient (0.088 cm -~) and an operator-positioned ellipse to define the edge of the scalp. Cerebral metabolic rates (CMRGIu) were calculated using the equation and the values of rate and lumped constants of Phelps et al.6 for normal awake humans. Hypnograms of the uptake periods were established using the Rechtschaffen and Kales criteria s. As previously described 5, the influence of a particular stage of sleep on the final value of CMRGIu is estimated by an 'index of stability', namely the ratio of the amount of plasma F D G available during this stage weighted by the amount of F D G transported by the plasma during the
Abbreviations: CMRGIu, cerebral metabolic rate of glucose; deep SWS, stage 3 and 4 SWS; EEG, eleetroancephalographic; FDG, fluorodeoxyglucose; LRD, left/right difference; MI, metabolic index; non-REMS: stage 2, 3 and 4 SWS; PET, positron emission tomography; REMS, rapid eye movement sleep; ROI, region of interest; SWS, slow wave sleep. Correspondence: O. Franek, Department of Neurology, CHU Sart Tilman (B35), B-4000 Liege, Belgium.
150 8 3
4
9
19
10
11 i ~
1
12
i3
17 18
20
21
92
27
28
23 !
,24
29
A 46
32
30 31
33
47
36
B 56
.54
52
53
5.5
57
68
, 69
58
: 59
50
,o
.
.
.
,2
.
,4,5
°o
72
6, 65
63 E
73
77
78
79
~
6~
70 71
76 :i",:i
...... °,
. . . . . .
,3 D
C
......
~
80
..
e= 03
8~F
1 86 F
:
~
~ , g7
G
Fig. 1. The 7 planes of references and the 89 preselected regions of interest used in this study. Several regions of interest were lumped in areas of interest as follows: Medial frontal 8,9,17,18,30,31,52,53,70,71 Lateral frontal 19,20,32,33,54,55,72,73 Orbital frontal 10,11 Inferior frontal 21,22,34,35,56,57,74,75 Anterior temporal 1,2 Middle temporal 12,13,40,41,60,61,78,79 Cortical regions Superior temporal 25,26,38,39,58,59 Parietal 86,87 Lateral occipital 42,43,62,63,80,81 Medial occipital 44,45,64,65,82,83 Insula 23,24,36,37 Mean grey matter Sensorimotor cortex 76,77,84,85 Hippocampus 3,4,14,15
Subcortical regions
whole F D G uptake period. We considered that the steady-state conditions of the model were met if IS >~ 0.75. Regional cerebral metabolic rates were obtained from 89 elliptical regions of interest (ROIs; see Fig. 1). Several regions of interest appear on different planes. For easier presentation of the data, they were lumped together in larger areas (areas of interest) spreading over a
Anterior striatum Caudate nuclei Lenticular nuclei Thalami Brainstem Cerebellum
27,28 46,47,66,67 48,49 50,51,68,69 7,16,29 5,6
large part of a lobe. The C M R G I u of an area of interest is the average of the metabolic rates of its constituent regions of interest. The modifications of cerebral glucose utilization during sleep were calculated by: A C M R G I u = (sleep C M R G l u - wake CMRGlu)/wake C M R G I u x 100. The regional distribution of cerebral glucose metabolism was estimated using regional metabolic indices (MI):
151
TABLE I Cerebral metabolic rates (CMRGIu) and standard deviation (SD) during stage 2 sleep and during wakefulness of the 6 selected runs in the major areas of interest
Mean variations of metabolic rates (ACMRGIu) between them and their standard deviations (SD). For comparison, ACMRGlu observed between stage 3 and 4 SWS and wakefulness in 5 other subjects appear in the two last colums (ref. 5). Areas of interest
During stage 2 sleep (mg/lO0 g.min)
During wakefulness (mg/lO0 g.min)
(st. 2 SWS vs wake)
(st. 3 & 4 SWS vs wake)
CMRGIu
SD
CMRGlu
SD
ACMRGIu1 (%)
SD (%)
ACMRGIu 1 (%)
SD (%)
Medial frontal Lateral frontal Orbito frontal Inferior frontal Anterior temporal Middle temporal Superior temporal Lateral occipital Medial occipital Insula Hippocampus
5.51 4.84 4.91 5.40 4.59 4.73 5.43 4.11 5.63 4.88 3.94
1.86 1.64 1.86 1.79 1.55 1.62 1.73 1.45 1.83 1.57 1.34
6.05 5.70 5.74 6.24 4.66 5.15 6.12 4.26 6.22 5.45 4.15
1.17 1.03 1.24 1.05 1.35 1.02 0.96 0.70 1.31 0.82 1.23
-9.64 -15.41 -18.2 -14.29 -8.12 -9.13 -12.12 -4.71 -9.65 -11.47 -4.76
26.76 26.83 18.84 24.64 21.94 25.49 24.41 29.72 24.76 23.76 24.74
-42.832 --46.032 -43.662 -43.812 -39.002 -44.392 -42.732 -39.702 -43.232 -42.142 -35.212
14.26 13.67 13.65 14.84 4.23 14.47 13.85 19.38 17.40 12.44 0.94
Mean cortical regions
5.00
1.66
5.56
1.00
-11.06
25.73
-43.832
14.49
Anterior striatum Caudate nucleus Lenticular nucleus Thalamus Brainstem Cerebellum
5.59 5.59 5.93 5.04 3.97 3.97
1.74 1.82 1.84 1.71 1.32 1.23
6.23 6.23 6.75 6.34 4.36 4.38
0.84 1.06 0.99 1.15 0.98 1.32
-11.05 -10.42 -12.66 -21.53 -10.00 -13.24
24.76 27.89 24.42 23.55 22.83 26.03
-37.962 -36.092 -41.852 -48.952 -42.462 -48.822
12.54 5.94 12.25 14.75 15.79 7.29
Mean subcortical regions
4.96
1.58
5.70
1.00
-13.82
23.34
-43.302
11.78
Mean (grey matter)
4.99
1.65
5.58
0.99
-11.55
25.27
-43.802
14.10
1 The variation is negative if the sleep CMRGlu is lower than wake CMRGIu. 2 P < 0.05.
MI = regional C M R G l u / m e a n cortical C M R G l u . Metabolic indices, being by definition regional values, were not lumped in larger areas. The intra-individual variations of MI were evaluated for each subject and for each particular region by (MIs - MIw)/(MIs + MIw) x 200 for the sleep population where MI s and MIw are the metabolic indices of the sleep and wake studies respectively. The left/right differences ( L R D ) were obtained by: L R D = (left C M R G I u - fight CMRGlu)/(left C M R G I u + right C M R G l u ) x 200. Paired two-tailed Student's t-test was performed between sleep and wake values of the metabolic rates and the metabolic indices. One-sample two-tailed Student's t-test was used to determine the significance from zero of the L R D . Paired two-tailed Student's t-test was performed between sleep and wake values of the L R D . The correction of Bonferroni was performed for each Student's t-test. In 6 runs, out of the 28 nocturnal P E T scans that we performed, these periods of stage 2 SWS were long
enough to fulfill our criterion of stability. In these cases, stage 2 SWS was always observed as a transition phase from other stages of sleep. In 2 cases, stage 2 SWS appeared after a deep SWS (stability indices: 0.95 and 0.87). In the other 4 cases it appeared after a period of R E M S (stability index: 0.84, 0.86, 0.92 and 0.97). In no case was the stage 2 SWS observed after wakefulness when the subject was falling asleep. Table I shows the mean C M R G I u observed during wakefulness and during the stage 2 SWS in the 6 aforementioned subjects, as well as the average variation of CMRGlu. O n the average, a trend toward a decrease in C M R G I u (-11.55%) was observed during stage 2 SWS. This variation was not significant, however (P = 0.28). No significant asymmetry was observed during stage 2 SWS (average asymmetry: 0.68 + 1.33%, P = 0.27). The M I did not change significantly during stage 2 SWS in all but 6 regions of interest (Table II) in which MI was significantly increased or decreased. Especially, the MI of both thalami was decreased. Because of the
152 TABLE II Metabolic indices (MI) and standard deviation (SD) during stage 2 sleep and during wakefulness in the 7 ROls where their variation (AMI; %) was significant (P < 0.05)
For comparison, AMI observed between stage 3 and 4 SWS and wakefulness in 5 other subjects appear in the two last cotums (ref. 5). ROls 2
Left lateral frontal Left lateral frontal Right medial occipital Right lateral occipital
Left thalamus Right thalamus
During st. 2 sleep
During wakefulness
AM11
SD
MI
SD
MI
SD
(192)
0.90
0.09
0.98
0.08
-8.85
7.71
(72)
1.01
0.90
0.99
0.07
4.11
0.50
(65)
1.12
0.98
1.05
0.11
6.50
5.07
(81)
0.85
0.77
0.74
0.63
8.35
1.38
(50)
0.97
0.011
1.12
0.52
-14.86
13.09
(51)
1.00
0.10
1.12
0.08
-12.27
11.47
ROls
AMI 1
SD
Right lateral frontal
(33)
-7.49
Left lateral occipital Left lateral occipital
(42)
11.79 1.60
(62)
8.10 2.72
(28)
10.78 3.09
(50)
-9.94
1.71
(51)
-12.55
2.91
Anterior striatum Left thalamus Right thalamus
2.80
1 The variation is negative when the sleep MI is lower than wake MI. 2 ROI number as it appears in Fig. 1.
great number of tests performed (89), no significant variation was observed when the correction of Bonferroni was applied. The CMRGIu during non-REMS was shown to be lower than during wakefulness in cat 7, monkey 4 and in man 1'5. Presuming that, in man, the cerebral glucose metabolism might differ from light to deep SWS, we investigated separately these two types of non-REM sleep. Our results show indeed that cerebral glucose metabolism is not homogeneous throughout all the stages of non-REMS. We showed previouslys that CMRGIu was dramatically decreased during the deep phases of nonR E M sleep (about 40% decrease; see Table I). In contrast, during stage 2 SWS, the CMRGIu was not found to be significantly different from that observed when awake. Thus, deep SWS, but not stage 2 SWS, appears to be linked to a unique status of cerebral energetic metabolism, drastically different from that observed when awake, However, if the rate of cerebral glucose utili7ation during stage 2 SWS is close to that of wakefulness, the distribution of the cerebral glucose metabolism during stage 2 SWS shares some similarity with that observed in deep SWS. In particular, a low metabolism in both thalamic nuclei (as reflected by the significant decrease of their MI) was observed both during stage 2 SWS (present results) and during deep SWS 5. Because of the number
of t-tests performed, the significant variation of the MI in thalamic nuclei as in 4 other ROIs might be due to type I error. However, the low thalamic glucose metabolism might also reflect the dysfacilitation and inhibition processes within thalamic structures during slow sleep 9. These phenomena induce the outbreak of long-lasting inhibitory potentials in thalamic neurons and conseqnently, a potent blockade of the ascending afferent volleys 2. Stage 2 SWS thus appears as a state during which ascending afferences are dampened and cortical activation is decreased. Stage 2 SWS might protect the stability of sleep by insulating the subject from the environment and might be a prerequisite to the full development of other phases of sleep, especially deep SWS. The level of cerebral glucose metabolism during stage 2 SWS is similar to the awake level: only deep SWS seems to be linked with a energy metabolism definitely different from that of wakefulness. Future research will determine if during deep SWS, some as yet unknown metabolic processes are taking place, that might be related to brain tissue restitution 3'5.
This research was supported by a grant from the Belgian National Fund for Medical Research (F.R.S.M. no. 3.4508.83) and from the Reine Elisabeth Medical Foundation. P.M. is Senior Research Assistant at the National Fund of Scientific Research (Belgium).
153 1 Buchsbaum, M.S., Giilin, J.C., WU, J., Hazlett, E., Sicotte, N., Dupont, R.M. and Bunney, W.E., Regional cerebral glucose metabolic rate in human sleep assessed by positron emission tomography, Life Sci., 45 (1989) 1349-1356. 2 Glen, L.L. and Steriade, M., Discharge rate and excitability of cortically projecting intralaminar thalamic neurons during waking and sleep states, J. Neurosci., 2 (1982) 1387-1404. 3 Home, J.A., Why We Sleep, Oxford University Press, Oxford, 1988, 319 pp. 4 Kennedy, C., Gillin, J.C., Mendelson, W., Soda, S., Miyaoka, M., Ito, M., Nakamura, R.K., Storch, F.I., Pettigrew, K., Mishkin, M. and Sokoloff, L., Local cerebral glucose utilization in non-rapid eye movement sleep, Nature, 297 (1982) 325-327. 5 Maquet, P., Dive, D., Salmon, E., Sadzot, B., Franco, G., Poirrier, R., yon Frenckel, R. and Franck, G., Cerebral glucose utilization during sleep-wake cycle in man determined by positron emission tomography and [lSF]2-fluoro-2-deoxy-D-
glucose method, Brain Research, 513 (1990) 136-143. 6 Phelps, M.E., Huang, S.C., Hoffman, E.J., Selin, C., Sokoloff, L. and Kuhl, D.E., Tomographic measurement of local cerebral glucose metabofic rate in humans with [lSF]2-fluoro-2-deoxy-Dglucose: validation of method, Ann. Neurol., 6 (1979) 371-388. 7 Ramm, P. and Frost, B.J., Cerebral and local cerebral metabolism in the eat during slow wave and REM sleep, Brain Research, 365 (1986) 112-124. 8 Rechtschaffen, A. and Kales, A.A., A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects, U.S. Department of Health, Education andWelfare, Bethesda, MD, 1968. 9 Stedade, M. and Hobson, J.A., Neuronal activity during the sleep-waking cycle, Prog. Neurobiol., 6 (1976) 155-376. 10 Webb, W.B. and Agnew, H.W., Sleep characteristics of long and short sleepers, Science, 168 (1970)145-147.