Brain Research 1004 (2004) 45 – 51

www.elsevier.com/locate/brainres

Research report
Sleep deprivation induced by the modified multiple platform technique:
quantification of sleep loss and recovery
Ricardo Borges Machado, Débora C. Hipólide *, Ana Amélia Benedito-Silva, Sergio Tufik
Psychobiology Department, Universidade Federal de São Paulo (UNIFESP), Rua Napoleão de Barros 925, São Paulo, SP, 04024-002, Brazil
Accepted 14 January 2004

Abstract

Vigilance status was continually monitored in socially stable groups of rats exposed to the modified multiple platform (MMP) technique
for sleep deprivation. For comparison, sleep parameters were also monitored in socially isolated rats deprived of sleep by the single platform
(SP) method. In all cases, sleep was continuously recorded during baseline, during 96 h of sleep deprivation and during 4 days of recovery.
Both multiple- and single-platform techniques completely abolished paradoxical sleep (PS) during the deprivation period, but also resulted in
significant decreases in slow wave sleep (SWS) ( 31% and 37%, respectively). Unexpectedly, animals on large platforms, which are
normally intended as controls, also showed significant reductions in PS and SWS, and these effects were more pronounced in rats deprived in
groups than in animals deprived in isolation. Another control preparation, rats placed on wire-mesh grids in the deprivation tank, also showed
PS reduction ( 39%) but no loss of SWS during the 4 test days. Paradoxical sleep rebound was observed in the first 24 h in all groups,
except for grid controls. Overall, no significant differences were found between single- and multiple-platform procedures during the 4 days of
deprivation. However, sleep rebound was more pronounced in MMP-deprived rats than in SP-deprived rats. Sleep loss in both control groups
may reflect residual effect of stress that remain in the platform technique. These findings indicate that the MMP technique is effective in
inducing PS deprivation (PSD). However, the fact that SWS is also affected may have implications for conclusions on paradoxical sleep
function based upon paradoxical sleep deprivation.
D 2004 Elsevier B.V. All rights reserved.

Theme: Neural basis of behavior

Topic: Biological rhythms and sleep

Keywords: Sleep deprivation; Paradoxical sleep; Slow wave sleep; Platform technique; Recording

1. Introduction platform into the water and awaken. Because platform

Sleep loss or fragmentation is a feature of a number of
sleep disorders and other pathological conditions in humans.
Historically, considerable research interest has focused on
the selective effects of paradoxical sleep (PS) loss. Exper-
imentally, most of the available techniques for PS depriva-
tion (PSD) take advantage of the atonia that occurs during
this phase of sleep, following the basic technique introduced
by Jouvet in 1964 for cats [11], and later adapted for rats by
Cohen and Dement [6]. In these procedures, animals are
placed on top of a small platform that is surrounded by
water. As animals enter the paradoxical phase of sleep, they
lose postural tone (atonia) and partially or fully slip from the
* Corresponding author. Tel.: +55-11-5539-0155; fax: +55-11-5572-
5092.
E-mail address: hipolide@psicobio.epm.br (D.C. Hipólide).
techniques do not require complex instrumentation, this
basic method has been widely used in PSD studies.
It is well known, however, that PSD experiments are
affected by a number of potentially confounding variables,
including stress associated with isolation, movement restric-
tion, wetness and muscle fatigue [5,9,14,27]. In addition, a
significant loss of slow wave sleep (SWS) may also occur
[10,19]. In order to reduce movement restriction associate
with the original platform technique, Van Hulzen and
Coenen [33] introduced a multiple platform procedure where
one rat is deprived in a large tank containing a number of
platforms. However, significant indices of stress were still
seen with this procedure [5]. The technique was subsequent-
ly modified by placing several rats in large tanks containing
a number of platforms, with the aim of removing social
isolation and restraint factors [20]. However, Suchecki et al.
[27] demonstrated that rats subjected to this procedure

0006-8993/$ - see front matter D 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.brainres.2004.01.019
46 R.B. Machado et al. / Brain Research 1004 (2004) 45–51
showed augmented ACTH and corticosterone responses and
that this response was attributed to social instability. More
recently, it has been shown that animals raised in groups and
sleep-deprived along with their peers in a large tank water
exhibit attenuated stress indices compared to animals that
were sleep-deprived as part of a socially unstable group [26].
A further modification consisted of exposing controls to the
same apparatus except that a wire mesh was placed over the
platforms to prevent any contact with the water. The simul-
taneous deprivation of a socially stable group and the wire-
mesh control group define what has been referred to as the
modified multiple platform (MMP) technique.
Although both single- [19] and multiple-platform [34]
techniques have been electrophysiologically evaluated,
there has been no direct assessment of changes in vigilance
state in rats subjected to the modified multiple platform
procedure. The aim of the present study was therefore to
quantify the nature and extent of sleep changes induced by
the modified multiple platform technique and proposed
control procedures, by continuous recording during 4 days
of sleep deprivation and recovery. For comparative purpo-
ses, we also evaluated sleep changes in rats deprived of
sleep by the conventional single platform (SP) technique.
2. Materials and methods
2.1. Animals
Male Wistar rats (300 – 400 g) from our colony were used
at present study. The experiments were approved by the
Ethical Research Committee of the Universidade Federal de
São Paulo and followed international guidelines for the care
of research animals. Rats were housed and maintained under
a 12-h light– dark cycle with lights on at 7:00 AM. Room
temperature was kept at 20 F 2 jC and animals had free
access to water and food at all times. Animals defined as
‘‘socially stable’’ were housed in groups of five. Isolated
animals were individually housed in hanging wire-mesh
cages.
Under ketamine – diazepam anesthesia, rats were imp-
lanted with two ipsilateral stainless steel screws for electro-
encephalogram (EEG) monitoring, one overlaying the right
lateral fronto-parietal area and the left medial parietal cortex
[30]. One additional pair of nickel – chromium fine wire
electrodes was implanted in the dorsal neck muscle for
electromyogram (EMG) recording. The electrodes were sol-
dered to a connector, which was fixed to the animal cranium
with acrylic dental cement. After surgery, penicillin and
diclofenac were administered and 15 days were allowed for
recovery.
2.2. Sleep recording
Electrophysiological signals were recorded on a digital
polygraph (Neurofax QP 223 A nNihon Kohden, Tokyo,
Japan) at a sampling rate of 200 Hz. Recordings were
visually scored on 30-s epochs according to classical
sleep-waking criteria as reflecting either (i) the waking state
(W), which is characterized by low-voltage, fast EEG
cortical and high EMG activity; (ii) slow wave sleep
(SWS), characterized by continuous high amplitude, slow
EEG activity and low EMG activity; or (iii) paradoxical
(PS), characterized by high EEG cortical activity and
absence of EMG activity [30]. Each 30-s epoch was defined
by the wave pattern (W, SWS or PS) which occupied more
than 50% of the epoch. When artifacts or noise did not allow
characterization of an epoch, assessments were made based
on immediately surrounding epochs.
Animals were habituated to the recording system for 3
days before a 24-h baseline recording was performed. Base-
line recordings were performed in group cages (N = 5) for
animals subjected to the MMP method (Experiment 1) or in
individual home cages (Experiment 2). Following baseline
recording, animals were adapted to the sleep deprivation
procedure for 30 min on 3 consecutive days. Sleep parameters
were obtained in minutes and then converted to percentage of
total recording time (usually 23 h) for each day.
2.2.1. Experiment 1: modified multiple platform method
Rats were tested in socially stable groups of five in
separate water tanks (123  44  44 cm) containing either
18 round platforms of 6.5-cm-diameter (small platforms) or
18 14-cm-diameter platforms (large platforms). The ‘‘large
platform’’ group is often used as a control in sleep depri-
vation experiments. An additional control group was placed
in a third tank on a stainless steel wire mesh (2.3-mm
openings), which allowed rats to lie down without touching
the water. This is referred to as the grid control group. The
tanks were filled with water to a level 1 cm below the
surface of the platforms or the grid. Continuous electro-
physiological monitoring was performed in one rat in each
group of five. After 96 h of sleep deprivation, rats were
returned to the home cage, where recording continued for 4
additional days. Ten deprivation runs were conducted, with
different animals, to achieve a final N of 10 recorded
animals per group.
2.2.2. Experiment 2: single platform method
The single platform procedure consisted of placing an
animal on a single small (6.5-cm diameter) platform or a
large (14-cm diameter) platform in an enclosed tank
(23  23  35 cm). As in the group procedure, the platform
was surrounded by water to about 1 cm below platform
surface. After 96 h of sleep deprivation, rats were returned
to the home cage, where recording continued for 4 addi-
tional days. Twelve animals were tested in each group.
2.3. Statistical analyses
In all experiments, sleep parameters were obtained in
minutes and then converted to percentage of total recording
 R.B. Machado et al. / Brain Research 1004 (2004) 45–51 47

Table 1
Sleep parameters in rats subjected to the modified multiple platform procedurea
B D1 D2 D3 D4 R1 R2 R3 R4
SP
W 45.2 F 6.4 67.5 F 9.6* 70.3 F 7.7* 71.3 F 8.4* 72.5 F 9.2* 35.5 F 7.8* 48.0 F 9.4 47.2 F 7.7 49.7 F 4.4
SWS 43.8 F 5.3 32.5 F 9.6* 29.6 F 7.7* 28.3 F 8.4* 27.5 F 9.2* 36.1 F 6.1* 38.0 F 7.1 40.3 F 6.3 39.0 F 4.2
PS 11.0 F 3.3 0.0 F 0.0* 0.0 F 0.0* 0.0 F 0.0* 0.0 F 0.0* 28.4 F 3.3* 14.0 F 4.3 12.4 F 2.6 11.3 F 2.3

LP
W 50.2 F 6.1 69.2 F 5.2* 66.9 F 7.6* 66.2 F 8.0* 71.4 F 6.2* 41.0 F 10.7* 50.0 F 9.5 50.7 F 9.8 52.7 F 5.5
SWS 41.2 F 4.4 29.7 F 4.6* 31.2 F 6.9* 31.1 F 5.7* 27.0 F 5.5* 39.2 F 7.1 40.7 F 7.3 40.1 F 7.4 38.0 F 3.4
PS 8.6 F 2.6 1.2 F 1.3* 1.9 F 1.3* 2.7 F 2.9* 1.5 F 1.3* 19.7 F 5.4* 9.3 F 3.4 9.2 F 3.4 9.3 F 2.9

Grid controls
W 48.1 F 5.7 56.7 F 7.6 54.4 F 4.5 58.8 F 6.4* 57.4 F 4.2* 50.2 F 9.4 55.0 F 6.4 55.4 F 5.3 57.6 F 9.3
SWS 41.3 F 5.0 37.5 F 5.9 38.8 F 3.5 35.2 F 5.5 36.2 F 3.4 37.6 F 5.9 36.7 F 5.3 35.4 F 3.9 34.3 F 6.7
PS 10.7 F 2.0 5.8 F 3.3* 6.8 F 2.0* 6.0 F 1.6* 6.3 F 1.7* 12.2 F 4.8 8.4 F 1.8 9.1 F 2.1 8.0 F 3.1
a
Values are percentage of time (mean F S.D.) spent in wakefulness (W), slow wave sleep (SWS) and paradoxical sleep (PS), during baseline (B),
deprivation days (D) and recovery days (R). SP = small platform; LP = Large platform. Ten animals were tested in each group. Differences from basal level are
indicated by *P < 0.05.

time (usually 23 h) for each day. Separate two-way repeated
measures ANOVAs were conducted using Group (Small
Platform, Large Platform or Control) as a between-subject
factor and Days (Baseline, day 1– 4) as a within subject
factor. Post-hoc pairwise comparisons were performed using
Tukey’s HSD tests.
3. Results
3.1. Experiment 1: modified multiple platform
Results are summarized in Table 1. Fig. 1 shows a
representative hypnogram illustrating changes in waking,
PS and SWS in a rat (small platform group) during baseline
(A), on the fourth day of sleep deprivation (B) and on the
first day of sleep recovery (C).
3.1.1. Changes during the sleep deprivation period
Wakefulness: Significant ANOVA effects were obtained
for Group ( F(2,27) = 13.9, P < 0. 001), Days ( F(4,108) =
59.4, P < 0. 001), and the Group  Day interaction ( F
(8,108) = 5.3, P < 0.001). Tukey HSD tests revealed that
both small and large platform groups exhibited increased
wakefulness as compared to the baseline throughout the 4
days of deprivation ( P < 0.001). The grid control group
showed small but significant increases in wakefulness on
days 3 and 4 compared to baseline ( P < 0.005). No
differences were detected between groups at baseline
( P >0.05).
Slow wave sleep: Significant main effects of Group ( F
(2,27) = 5.6, P < 0.01) and Days ( F(4,108) = 21.7, P < 0.001)
were detected, in addition to a significant Group  Days
interaction ( F(8,108) = 2.5, P < 0.05). Percent of time spent
in SWS significantly decreased in both small and large
platform groups on all 4 days of sleep deprivation compared
to baseline ( P < 0.05). Comparisons between baseline and grid
controls showed no difference in SWS percentage at any point
( P >0.05). At baseline, no differences were detected between
groups ( P >0.05).
Paradoxical sleep: As shown in Table 1, PS was com-
pletely absent in the small platform group on each of the 4
deprivation days. Significant Group ( F(1,18) = 28.6, P <
0.001), Days ( F(4,72) = 39.3, P < 0.001) and interaction
( F(4,72) = 3.5, P < 0.05) effects were detected. Post-hoc
tests indicated that both the large platform and grid groups
had reduced PS compared to baseline throughout the 4 days
of sleep deprivation ( P < 0. 001 and P < 0.005, respective-

Fig. 1. Hypnogram showing sleep – wake states in a representative rat (small platform group) in modified multiple platform condition during baseline (A), on
the fourth day of sleep deprivation (B) and on the first day of sleep recovery (C). W—awake; PS—paradoxical sleep; SWS—slow wave sleep.
48 R.B. Machado et al. / Brain Research 1004 (2004) 45–51

ly). No differences were detected between groups at base-
line ( P >0.05).
3.1.2. Sleep recovery period
Wakefulness: Significant effects were detected for Group
( F(2,27) = 5.0, P < 0.05) and Days ( F(4,108) = 15.1, P <
0.001), but not for the Group  Days interaction ( F(8,108)
= 1.7, P = 0.10). Compared to baseline, both small and large
platform groups showed decreased wakefulness on Day 1 of
recovery ( P < 0.05).
Slow wave sleep: There was no significant Group effect
( F(4,108) = 1.0, P = 0.39), a significant Days effect ( F
(4,108) = 4.5, P < 0.005) and no Group  Days interaction
( F(8,108) = 1.5, P = 0.17). Post-hoc tests indicated that SWS
percentage was decreased in the small platform group on
Day 1 ( P < 0.05) compared to baseline.
Paradoxical sleep: There was a significant Group effect,
( F(2,27) = 18.9, P < 0.001), a significant Days effect
( F(4,108) = 59.1, P < 0.001) and a significant Group  Days
Days interaction ( F(8,108) = 10.1, P < 0.001). For both
small and large platform groups the percentage of time
spent in PS was higher as compared to the baseline only in
the first 24 h of the sleep recovery period ( P < 0.001).
3.2. Experiment 2: single platform method
3.2.1. Sleep deprivation period
Wakefulness: There was a significant effect of Group,
( F(1,22) = 46.6, P < 0.001), a significant Days effect
( F(4,88) = 69.8, P < 0.001) and a significant Group Days
interaction ( F(4,88) = 30.0, P < 0.001). Post-hoc tests con-
firmed that wakefulness increased in both small and large
platform groups on all days ( P < 0.005). No differences
were detected between groups at baseline ( P >0.05).
Slow-wave sleep: Analyses of variance revealed a signif-
icant effects for Groups, ( F(1,22) = 29.7, P < 0.001), Days
( F(4,88) = 24.8, P < 0.001) and the Group  Days interac-
tion ( F(4,88) = 11.8, P < 0.001). Pairwise comparisons indi-
cated that SWS percentage was decreased in the small
platform group on all days ( P < 0.001) and in the large
platform group on Day 1 ( P < 0.001) of sleep deprivation
compared to the baseline. No differences were detected
between groups at baseline ( P >0.05).
Paradoxical sleep: Paradoxical sleep was completely
eliminated in the small platform group. There was a signif-
icant effect of days of sleep deprivation ( F(4,44) = 50.8,
P < 0.001). Compared to baseline, PS percentage was de-
creased in the large platform group on each of the 4 sleep
deprivation days ( P < 0.001). No differences were detected
between groups at baseline ( P >0.05).
3.2.2. Sleep recovery period
Wakefulness: There was no significant main effect for
Groups, ( F(1,22) = 2.1, P = 0.16), a significant Days effect
( F(4,88) = 8.7, P < 0.001) and no interaction ( F(4,88) = 0.4,
P = 0.79). Pairwise comparisons revealed that wakefulness
did not change in small platform at any point during the
recovery period; in the large platform group wakefulness
decreased on Day 1 ( P < 0.05). No differences were de-
tected between groups at baseline ( P >0.05).
Slow-wave sleep: The main effect of Groups was signif-
icant ( F(1,22) = 6.8, P < 0.05), but not the Days ( F(4,88)
= 0.1, P = 0.96) or interaction ( F(4,88) = 1.9, P = 0.12)
effects. Pairwise analyses confirmed that SWS percentage
did not change at any point during the 4 days.
Paradoxical sleep: The ANOVA revealed no significant
Group effect ( F(1,22) = 0.8, P = 0.38), a main effect of Days
( F(4,88) = 50.5, P < 0.001) and a significant Group  Days
interaction ( F(4,88) = 2.5, P < 0.05). Pairwise tests indicated
that in both small and large platform groups, PS percentage
was higher as compared to baseline only in the first 24 h of
sleep recovery period ( P < 0.001) (Table 2).
3.3. Comparisons between multiple and single platform
effects
As illustrated in Fig. 2A, the multiple and single
platform procedures resulted in similar levels of SWS loss
across the 4 deprivation days. During recovery, the MMP
procedure resulted in slightly higher rebound levels

Table 2
Sleep parameters in rats subjected to the single-platform procedurea
B D1 D2 D3 D4 R1 R2 R3 R4
SP
W 48.5 F 4.7 73.6 F 6.7* 71.1 F 7.6* 75.1 F 5.8* 78.7 F 7.3* 40.2 F 5.8 48.3 F 7.0 46.7 F 6.7 48.2 F 4.9
SWS 40.5 F 4.6 26.4 F 6.8* 28.9 F 7.6* 24.9 F 5.8* 21.3 F 7.3* 37.7 F 4.9 39.2 F 6.7 41.7 F 6.2 41.3 F 3.7
PS 11.0 F 1.9 0.0 F 0.0 0.0 F 0.0 0.0 F 0.0 0.0 F 0.0 22.1 F 5.3* 12.5 F 2.2 11.7 F 2.5 10.5 F 2.7

LP
W 45.5 F 7.1 68.5 F 9.7* 57.5 F 5.9* 55.4 F 5.5* 56.3 F 5.0* 35.9 F 9.7* 44.5 F 7.9 45.9 F 4.3 47.8 F 10.8
SWS 42.5 F 5.9 29.8 F 9.1* 39.0 F 4.9 40.5 F 4.9 39.0 F 4.1 45.8 F 5.8 43.3 F 6.3 43.0 F 2.8 42.6 F 7.9
PS 12.0 F 3.3 1.6 F 1.4* 3.5 F 1.4* 4.1 F 1.5* 4.7 F 1.6* 18.2 F 5.4* 12.1 F 3.2 11.1 F 3.0 9.6 F 4.1
a
Values are percentage of time (mean F S.D.) spent in wakefulness (W), slow wave sleep (SWS) and paradoxical sleep (PS), during baseline (B),
deprivation days (D) and recovery days (R). SP = small platform; LP = Large platform. Twelve animals were tested in each group. Differences from basal level
are indicated by *P < 0.05.
 R.B. Machado et al. / Brain Research 1004 (2004) 45–51
Fig. 2. Mean percentage changes in slow-wave (A) and paradoxical sleep
(B) in rats subjected to multiple or single platform procedures. Each bar is a
mean of 4 days of deprivation or first day of recovery. #P = 0.057;
*P < 0.05.
( 11% vs. 0.2%, P = 0.058). When only the first day of
recovery was considered, a very similar pattern was ob-
served ( 17% vs. 6%, P = 0.057).
Both procedures resulted in total PS abolition during
the deprivation period (Fig. 2B). During recovery, there
was trend toward higher rebound levels in the multiple
platform procedure (68% vs. 31%) although this differ-
ence did not reach statistical significance ( P = 0.13). On
the first day of recovery, the average level of PS was
significantly higher in the MMP group than in the SP
group, as illustrated in Fig. 2B ( + 187% vs. + 104%,
P = 0.047).
4. Discussion
The present study provides the first direct quantification
of changes in vigilance state in rats subjected to sleep
deprivation using the modified multiple platform (MMP)
technique. The continuous recording approach also made it
possible for the first time to evaluate the grid procedure as a
control for the MMP method.
4.1. Sleep deprivation period
Although the multiple platform procedure was effective
in producing a total suppression of paradoxical sleep,
significant SWS loss also occurred throughout the 4 days
of sleep deprivation ( 31%). In addition, animals placed
49
on larger diameter platforms in the MMP procedure also
showed significant reductions in both PS ( 77%) and SWS
( 26%) throughout the 4 days of sleep deprivation.
Previous analyses of sleep recording using platform
techniques with isolated animals have produced conflicting
results with respect to effects on SWS. Our findings are
consistent with some studies that have reported that platform
procedures suppress all phases of sleep [10]. On the other
hand, some previous electrophysiological examinations
found that platform techniques affect SWS only in the first
24 h [16,19]. In contrast to these studies, Porkka-Heiskanen
et al. [22] reported selective PS suppression in small
platform groups, with no significant variation in SWS
percentage for either small or large platform groups.
The reasons for significant SWS reductions in our study
are not clear, but it is possible that social interaction among
animals plays a role. Informal observation indicated that it
was not uncommon for two animals to share the same
platform at different points and animals constantly awak-
ened each other in the tank. It is noted that animals
subjected to the single platform technique, which were
tested in isolation, also showed SWS losses. However,
animals on large platforms in the single platform setup
showed SWS reductions only in the first 24 h of sleep
deprivation ( 29%), whereas their counterparts subjected
to the MMP method showed SWS reductions throughout the
4 days of sleep deprivation.
It is also conceivable that the surface diameter of the
large platforms used in this study were not ideal to fully
accommodate the body in a state of relaxation (SWS
involves a substantial degree of muscular relaxation). In-
deed, animals placed on the grid, which allowed complete
freedom to lie down, showed PS reduction ( 39%) but no
loss of SWS during the 4 test days.
Our results suggest that it is appropriate to refer to MMP-
exposed animals as being sleep deprived rather than being
exclusively deprived of paradoxical sleep. However, while
SWS was clearly affected during the deprivation period, it
did not rebound to levels higher than baseline, suggesting
the possibility that the loss during deprivation may not have
been of sufficient physiological significance to engage
compensatory mechanisms.
4.2. Sleep loss in control animals
The present results also suggest that large platforms are
not an adequate environmental control for the MMP tech-
nique, as animals in this group also show significant sleep
losses. Sleep loss was also seen in the grid control group.
These may be residual effects of stress that remains in these
procedures. In fact, it has been demonstrated that cortico-
sterone levels are elevated over basal levels in grid control
animals, although to a lesser degree than in small platform
animals [26].
A great variety of stressful situations are known to
decrease sleep duration and efficiency [4,12]. Many of these
50 R.B. Machado et al. / Brain Research 1004 (2004) 45–51
studies have reported decreases in active waking, deep
sleep, and paradoxical sleep. In contrast, we observed a
selective PS reduction in the grid control group. Further, the
PS loss observed in this group was not followed by PS sleep
rebound, suggesting that the PS reduction in these animals
was not severe enough to engage compensatory effects.
There are other instances in the literature where control rats
were reported to suffer a partial loss of paradoxical sleep
during the deprivation period, but did not show significantly
greater PS during recovery, as compared to baseline levels
[32].
4.3. Sleep rebound period
With respect to the recovery period, the present results
show that 24 h was enough for a complete recovery of PS in
MMP and well as in SP groups. Paradoxical sleep rebound
after sleep deprivation by the platform method has been
observed in several previous studies [10,16,19,21,28,34].
However, some of these studies also reported SWS rebound
[21,34].
The lack of SWS rebound in the present study for both
SP and MMP techniques may be due to several factors. One
possibility, as noted earlier, is simply that the SWS reduc-
tion during deprivation period was not severe enough to
engage compensatory controls and cause rebound. Alterna-
tively, SWS rebound could have been offset by the large
increase in PS during the recovery period, i.e. PS rebound
would predominate over SWS rebound [7,8,31]. The issue
of sleep stage priorities during recovery has been extensive-
ly discussed in a recent review [24]. Our results are in
accordance with several studies that used total or selective
and different sleep deprivation procedures. Endo et al. [7],
using several selective protocols for sleep deprivation,
showed that slow wave or delta activity, which is charac-
teristic from SWS, was inhibited by PS increase during the
recovery period.
It is also conceivable that finer analyses might unearth
differences within SWS with respect to rebound effects.
Two phases can be distinguished within SWS, one
characterized by low to medium voltage EEG and another
characterized by high-amplitude slow-wave (HASW) EEG
activity (for review, see Ref. [25]). Many reports have
described SWS rebound in terms of augmented HASW
only, both in humans [15] and rats [7] (for review, see
Ref. [24]). We did not score SWS phases separately in
the present study, and this may conceivably have masked
a selective rebound effect on one of these two SWS
subphases.
Another point that needs to be considered is the length of
sleep deprivation. It seems that rebound from 1 day of sleep
deprivation produces subsequent rebound of slow-wave or
high-amplitude EEG activity in SWS, usually accompanied
by substantial PS rebound (for review, see Ref. [24]). In
contrast, chronic sleep deprivation by the disk-over-water
method produces long-lasting rebound of PS with no
rebound of high-amplitude SWS [24]. Palma et al. [21]
demonstrated significant PS and SWS rebound ( + 16.9%
and + 9.4%, respectively) after 18 h of sleep deprivation
induced by the MMP method. On the other hand, Suchecki
et al. [28] found that 90 h of sleep deprivation produced PS
rebound ( + 184.7%) accompanied by a reduction in SWS
time ( 12.2%)—the same pattern as observed in the
present study using 96 h of sleep deprivation ( + 187%
and 17% for PS and SWS, respectively).
4.4. Rebound differences between SP and MMP methods
A surprising observation was that PS rebound was higher
in MMP- than in SP-deprived rats. The reasons for this are
not clear. Sleep rebound could in theory be affected by stress
factors associated with deprivation procedures, as several
studies have documented an increased need for sleep im-
mediately after a stressful experience. Stress per se can
produce alterations in sleep patterns and these alterations
seem to depend on stress modality, duration and intensity
[1 – 4,12,13,17,18,21,23]. By eliminating social isolation
and movement restriction, the MMP procedure was ex-
pected to be less stressful than the SP procedure and hence
cause less PS rebound. However, animals deprived by the
SP or by MMP technique show the similar corticosterone
levels [29], which suggests that differential stress effects
cannot account for the differences in sleep rebound ob-
served for the two procedures.
4.5. Use of rebound as an index of deprivation
Recent work from our laboratory has focused on sleep
rebound parameters as an index of the effectiveness PS-
specificity of the MMP procedure. In addition, comparisons
between baseline and grid group during 6 h of sleep rebound
after 90 h of sleep deprivation revealed no sleep parameters
differences [28]. Although sleep rebound patterns in these
studies were similar to the pattern observed in the present
study, our results suggest that analyses of sleep rebound in
sleep deprivation studies do not necessarily indicate that
only paradoxical sleep was suppressed. Moreover, the
absence of PS rebound in the grid control group does not
serve as evidence that PS deprivation did not occur. The
present result shows that the rats placed on a grid over the
water tank floor may not serve as adequate environmental
controls for the MMP method as far as sleep loss-related
variables are considered.
4.6. Conclusions
In summary, the present study shows that the MMP
method of sleep deprivation is efficient in producing total
but not exclusive PS suppression, since SWS is also
compromised. However, the absence of SWS rebound
suggests that caution should be used in assessing the
significance of SWS loss during deprivation. With respect
 R.B. Machado et al. / Brain Research 1004 (2004) 45–51
to controls, we observed that both control procedures
induce at least partial paradoxical sleep deprivation. The
present results also showed that 24 h was enough to
recover PS loss in all groups, except for grid controls
which PS deprivation observed was not sufficient to
induce sleep rebound. While these techniques remain
useful to model the loss or fragmentation of sleep that
occurs in a number of sleep disorders and in many human
conditions, their usefulness as tools to generate inferences
about the specific contribution of selective sleep phases in
various conditions may be limited by the lack of specific-
ity that is attendant to the procedures.
Acknowledgements
The authors would like to tank Dr. José N. Nobrega and
Beatriz D. Palma for their revision and comments on the
manuscript. R.B. Machado was fellowship from CAPES
and D.C. Hipólide is fellowship from CNPq, both from
Brazil government. This work was supported by Associacßão
Fundo de Incentivo a Psicofarmacologia-AFIP.
References
[1] J. Adrien, C. Dugovic, P. Martin, Sleep – wakefulness patterns in the
helpless-rat, Physiol. Behav. 49 (1991) 257 – 262.
[2] J.J. Bouyer, J.M. Deminiere, W. Mayo, M. Le Moal, Inter-individual
differences in the effects of acute stress on the sleep – wakefulness
cycle in rat, Neurosci. Lett. 225 (1997) 193 – 196.
[3] R. Cespuglio, S. Marinesco, B. Baubet, C. Bonnet, B. El Kafi, Evi-
dence for a sleep-promoting influence of stress, Adv. Neuroimmunol.
5 (1995) 145 – 154.
[4] S.G. Cheeta, J. Ruigt, P. Vanproosdij, P. Willner, Changes in sleep
architecture following chronic mild stress, Biol. Psychiatry 41 (1997)
419 – 427.
[5] A.M.L. Coenen, E.L.J.M. Van Luijtellar, Stress induced by three pro-
cedures of deprivation of PS, Physiol. Behav. 35 (1985) 501 – 504.
[6] H.B. Cohen, W.C. Dement, Sleep: changes in threshold to electrocon-
vulsive shock in rats after deprivation of paradoxical phase, Science
150 (1965) 1318 – 1319.
[7] T. Endo, B. Schwierin, A.A. Borbély, I. Tobler, Selective and total
sleep deprivation: effect on the sleep EEG in the rat, Psychiatry Res.
66 (1997) 97 – 110.
[8] I. Feinberg, I.G. Campbell, Total sleep deprivation in the rat transient-
ly abolishes the delta amplitude response to darkness: implications for
the mechanism of the ‘‘negative delta rebound’’, J. Neurophysiol. 70
(1993) 2695 – 2699.
[9] W. Fishbein, B.M. Gutwein, Paradoxical sleep and memory storage
processes, Behav. Biol. 19 (1977) 425 – 464.
[10] S. Grahnstedt, R. Ursin, Platform sleep deprivation affects deep SWS
in addition to REM sleep, Behav. Brain Res. 18 (1985) 233 – 239.
[11] D. Jouvet, E. Vimont, F. Delorme, M. Jouvet, Étude de la privation
sélective de la phase paradoxale de sommeil chez le chat, C. R. Soc.
Biol. 158 (1964) 756 – 759.
[12] G.J. Kant, R.H. Pastel, R.A. Bauman, G.R. Meininger, K.R.
Maughan, T.N. Robinson, W.L. Wright, P.S. Covington, Effects
51
of chronic stress on sleep in rats, Physiol. Behav. 57 (1995)
359 – 365.
[13] M. Koehl, J.J. Bouyer, M. Darnaudery, M. Le Moal, W. Mayo, The
effect of restraint stress on paradoxical sleep is influenced by the
circadian cycle, Brain Res. 937 (2002) 45 – 50.
[14] V.M. Kovalzon, V.L. Tsibulsky, REM-sleep deprivation, stress and
emotional behavior in rats, Behav. Brain Res. 14 (1984) 235 – 245.
[15] F. Lucidi, A. Devoto, C. Violani, P. Mastracci, M. Bertini, Effects of
different sleep duration on delta sleep in recovery nights, Psychophy-
siology 34 (1997) 227 – 233.
[16] K.J. Maloney, L. Mainville, B.E. Jones, Differential C-Fos expression
in cholinergic, monoaminergic, and GABAergic cell groups of the
pontomesencephalic tegmentum after PS deprivation and recovery,
J. Neurosci. 19 (1999) 3057 – 3072.
[17] S. Marinesco, C. Bonnet, R. Cespuglio, Influence of stress duration
on the sleep rebound induced by immobilization in the rat: a possible
role for corticosterone, Neuroscience 92 (1999) 921 – 933.
[18] P. Meerlo, B.J. Pragt, S. Daan, Social stress induces high intensity
sleep in rats, Neurosci. Lett. 225 (1997) 41 – 44.
[19] W.B. Mendelson, R.D. Guthrie, G. Frederick, R.J. Wyatt, The flower
pot technique of rapid eye movement (REM) sleep deprivation, Phar-
macol. Biochem. Behav. 2 (1974) 553 – 556.
[20] G.P. Nunes Jr., S. Tufik, Validation of the modified multiple platform
method (MMP) of paradoxical sleep deprivation in rats, Sleep Res. 23
(1994) 419.
[21] B.D. Palma, D. Suchecki, S. Tufik, Differential effects of acute cold
and footshock on the sleep of rats, Brain Res. 861 (2000) 97 – 104.
[22] T. Porkka-Heiskanen, S.E. Smith, T. Taira, J.H. Urban, J.E. Levine,
F.W. Turek, D. Stenberg, Noradrenergic activity in rat brain during
rapid eye movement sleep deprivation and rebound sleep, Am. J.
Physiol. 268 (1995) R1456 – R1463.
[23] C. Rampin, R. Cespuglio, N. Chastrette, M. Jouvet, Immobilization
stress induces a paradoxical rebound in rat, Neurosci. Lett. 126 (1991)
113 – 118.
[24] A. Rechtschaffen, B.M. Bergmann, M.A. Gilliland, K. Bauer, Effects
of method, duration, and sleep stage on rebounds from sleep depri-
vation in the rat, Sleep 22 (1999) 11 – 31.
[25] C. Robert, C. Guilpin, A. Limoge, Automated sleep staging systems
in rats, J. Neurosci. Methods 88 (1999) 111 – 122.
[26] D. Suchecki, S. Tufik, Social stability attenuates the stress in the
modified multiple platform method for PS deprivation in the rat,
Physiol. Behav. 68 (2000) 309 – 316.
[27] D. Suchecki, L.L. Lobo, D.C. Hipólide, S. Tufik, Increased ACTH
and corticosterone secretion induced by different methods of PS dep-
rivation, J. Sleep Res. 7 (1998) 276 – 281.
[28] D. Suchecki, B.D. Palma, S. Tufik, Sleep rebound in animals deprived
of PS by modified multiple platform method, Brain Res. 875 (2000)
14 – 22.
[29] D. Suchecki, P.A. Tiba, S. Tufik, Hormonal and behavioural res-
ponses of paradoxical sleep-deprived rats to the elevated plus maze,
J. Neuroendocrinol. 14 (2002) 549 – 554.
[30] C. Timo-Iaria, C.N. Negrão, R.W. Schimidek, K. Hoshino, C.E.L.
Menezes, T.L. Da Rocha, Phases and states of sleep in the rat, Phys-
iol. Behav. 5 (1970) 1057 – 1062.
[31] I. Tobler, A.A. Borbély, The effect of 3-h and 6-h sleep deprivation on
sleep and EEG spectra of the rat, Behav. Brain Res. 36 (1990) 36 – 73.
[32] L.L. Tsai, B.M. Bergmann, A. Rechtschaffen, Sleep deprivation in the
rat: XVI. Effects in a light – dark cycle, Sleep 15 (1992) 537 – 544.
[33] Z.J.M. Van Hulzen, A.M.L. Coenen, PS deprivation and locomotor
activity in rats, Physiol. Behav. 27 (1981) 741 – 744.
[34] E.L.J.M. Van Luijtelaar, A.M.L. Coenen, Electrophysiological evalu-
ation of three PS deprivation techniques in rats, Physiol. Behav. 36
(1986) 603 – 609.