Full text of DOI:10.1007/BF03019666

866
REPORTS OF INVESTIGATION
Midazolam attenuates
ketamine-induced
Manzo Suzuki MD,*
Kentaro Tsueda MD,*
Peter S. Lansing MD*
Merritt M. Tolan MD,*
Thomas M. Fuhrman MD,*
abnormal perception
and thought process but Rachel A. Sheppard BS,*
not mood changes
Harrell E. Hurst PhD,†
Steven B. Lippmann MD‡
Purpose: To determine the effects of midazolam, 30 ng·mL–1, on altered perception, mood, and cognition
induced by ketamine.
Methods: After ketamine was administered to achieve target concentrations of 50, 100, or 150 ng·mL–1 in11
volunteers, perception, mood, and thought process were assessed by a visual analog scale. Mini-Mental State
examination (MMSE) assessed cognition. Boluses of midazolam, 30, 14.5, and 12 µg·kg–1, were injected every 30
min to maintain the plasma concentration at 30 ng·mL–1, which was reached 30 min after each injection.
Results: Ketamine produced changes in perception about the body (P < 0.01, 0.001, and 0.0001 at 30, 60, and
90 min), surroundings (P < 0.01 and 0.0001 at 60 and 90 min), time (P < 0.002 and 0.0001 at 60 and 90 min),
reality (P < 0.001 and 0.0001 at 60 and 90 min), sounds (P < 0.002 at 90 min), and meaning (P < 0.05 at 90
min). Subjects felt less energetic and clearheaded (P < 0.02 and 0.05) during ketamine, midazolam, and their co-
administration. Ketamine impaired thought process (P < 0.003 and 0.0001 at 60 and 90 min). Ketamine and
midazolam decreased mean total MMSE and recall scores (P < 0.001 for both). Co-administration reduced the
number of subjects with perceptual (body, P < 0.01 and 0.001 at 30 and 60 min) and thought process abnor-
malities. Within the range of observation, co-administration did not affect the changes in mood or recall.
Conclusion: Midazolam attenuates ketamine-induced changes in perception and thought process.
Objectif : Déterminer les effets de 30 ng·mL–1 de midazolam sur les changements de perception, d’humeur et
de fonction cognitive induits par la kétamine.
Méthode : Après l’administration de kétamine visant à obtenir des concentrations cibles de 50, 100, ou 150
ng·mL–1 chez 11 volontaires, la perception, l’humeur et la fonction cognitive ont été évaluées à l’aide d’une
échelle visuelle analogique. L’examen MMS de Folstein et coll. (MMS) a servi à évaluer la fonction cognitive. Des
bolus de midazolam de 30, 14,5 et 12 µg·kg–1 ont été injectés à toutes les 30 min afin de maintenir la concen-
tration plasmatique à 30 ng·mL–1, concentration atteinte 30 min après chaque injection.
Résultats : La kétamine a modifié la perception du corps (P < 0,01; 0,001 et 0,0001 à 30, 60 et 90 min), de
l’environnement (P < 0,01 et 0,0001 à 60 et 90 min), du temps (P < 0,002 et 0,0001 à 60 et 90 min), de la
réalité (P < 0,001 et 0,0001 à 60 et 90 min), des sons (P < 0,002 à 90 min) et du sens (P < 0,05 à 90 min).
Les sujets se sentaient moins énergiques et moins lucides (P < 0,02 et 0,05) pendant l’administration de kéta-
mine, de midazolam et pendant leur co-administration. La kétamine a altéré le processus cognitif (P < 0,003 et
0,0001 à 60 et 90 min). La kétamine et le midazolam ont fait baisser les scores totaux moyens de MMS et de
mémoire (P < 0,001 pour les deux). La co-administration a réduit le nombre de sujets dont les perceptions
(corps, P < 0,01 et 0,001 à 30 et 60 min) et la fonction cognitive étaient modifiées. Pendant le temps d’obser-
vation, la co-administration n’a pas eu d’effet sur les changements d’humeur ou de mémoire.
Conclusion : Le midazolam diminue les changements de perception et de cognition induits par la kétamine.
From the Departments of Anesthesiology,* Pharmacology and Toxicology,† and Psychiatry,‡ University of Louisville School of Medicine,
Louisville, KY, USA.
Address correspondence to: Kentaro Tsueda MD, Department of Anesthesiology, University of Louisville, Louisville, KY 40292, USA.
Phone: 502-852-5851; Fax: 502-852-6056; E-mail: rashep01@gwise.louisville.edu
Accepted for Publication June 15, 2000.
CAN J ANESTH 2000 / 47: 9 / pp 866–874

Suzuki et al.: MIDAZOLAM A N D KETAMINE
867
concentration of approximately 30 ng·mL–1; and 4)
during co-administration of ketamine at target con-
centrations and the midazolam target concentration,
30 ng·mL–1. Each subject was randomly assigned,
according to a computer generated schedule, to a dif-
ferent order in which drug regimens were studied.
One of the investigators, who did not participate in
the assessment of patients, prepared 10 mL syringes
containing normal saline or normal saline with mida-
zolam, and a 30 mL syringe containing normal saline
with or without ketamine 0.25%.
To obtain target plasma concentrations, ketamine at a
concentration of 2,500 µg·mL– 1 was infused using a
Harvard pump 22™ (Harvard Apparatus, Holliston, MA)
controlled by the Stanpump program (Steven L. Shafer,
M.D., Department of Anesthesiology, Stanford
University) based on pharmacokinetic data.1 0 Three target
concentrations were maintained for 30 min at each level.
Four boluses of midazolam, 30, 14.5, and 12 µg·kg– 1,
were injected every 30 min, to attain a plasma concentra-
tion of 30 ng·mL– 1, 30 min after each injection.1 1
Perceptual change was assessed using a visual ana-
log scale (VAS) in eight categories (i.e., body, sur-
roundings, time, reality, colours, sound, voices [e.g., I
hear voices that are unreal], and meaning [e.g., events,
objects, and people have particular meaning for me]).9
The VAS was anchored by “not at all” at one end and
“extremely” at the other. Mood was assessed using a
similarly scaled VAS in 6 subsets of mood states (anx-
ious-composed, hostile-agreeable, depressed-elated,
unsure-confident, tired-energetic, and confused-clear-
headed).9 Cognition was evaluated using the Mini-
Mental State Examination (MMSE, 0 - 30),1 2 which
assessed five areas of cognitive function (i.e., orienta-
tion [0 - 10], registration [0 - 3], recall [0 - 3, atten-
tion [0 - 5], and language fluency [0 - 9]). Thought
process and the content of thought were assessed
using a similarly anchored VAS (i.e., I have difficulty
in concentrating on a thought and/or have flight of
ideas).9 Paranoia was assessed using the VAS to mea-
sure suspicion (i.e., I had suspicious ideas or beliefs
that others were against me).9
Sedation was assessed by the Observer’s Assessment
of Alertness/Sedation (OAA/S) score: 5 = responds
readily to name spoken in normal tone; 4 = lethargic
response to name spoken in normal tone; 3 = responds
after name is called loudly and/or repeatedly; 2 =
responds after mild prodding or shaking; and 1 = does
not respond to mild shaking.1 3 Subjective sense of
drowsiness was assessed by a VAS in which the worst
drowsiness was defined as when subjects could barely
keep their eyes open.2,9
- METHYL-D-ASPARTATE (NMDA)
receptors are located predominantly in the
cortex, basal ganglia, and the structures
associated with sensory systems and are
N
important in corticofugal and corticocortical interac-
tion.1 It has been suggested that NMDA-receptor
mediated synaptic potentiation may be an essential
process involved in memory and learning1 as well as the
development of sensory systems.2 Ketamine, an NMDA
receptor antagonist, activates the thalamic and limbic
systems with concomitant depression of thalamo-neo-
cortical pathways,3 and increases cerebral oxygen con-
sumption4 and hippocampal glucose utilization.5 At
anesthetic doses (1 - 3 mg·kg–1), ketamine may produce
unpleasant dreams or psychotomimetic symptoms in
more than one-third of patients on emergence from
anesthesia.6 At subanesthetic doses (100 - 500 µg·kg–1),
ketamine impairs some domains of cognition (vigilance
and memory),2,7,8 alters mood states,7–9 and produces a
dose-related impairment of sensory perception or sen-
sory integration in healthy human volunteers.7–9
Perceptual and mood changes at higher ranges of anal-
gesic doses (e.g., 500 µg·kg– 1) are shown to resemble
some aspects of psychosis, as in schizophrenia, and are
associated with dysphoria.7
Recently, low-dose ketamine has been used togeth-
er with midazolam, a competitive agonist for the sub-
receptor of the (-aminobutyric acid (GABA )
receptor, and an opioid for sedation and analgesia duAr-
ing monitored anesthesia care. Benzodiazepines are
reported to reduce psychotomimetic manifestations
during the emergence from ketamine anesthesia.6
However, it is not clear whether benzodiazepines
attenuate changes in perception, mood, and amnesia
or affect the level of sedation induced by low-dose
ketamine. We studied the effects of midazolam on the
changes in perception, mood, and cognition, as well as
on the sedation produced by low-dose ketamine.
Materials and Methods
Eleven male volunteers were recruited for this ran-
domized, double-blind, cross-over study. Age ranged
from 27 to 45 yr. Body weight was 91 ± 14 (SD) kg
and height was 180 ± 7 cm. None of the subjects had
psychiatric diseases, psychological problems, systemic
illness, drug dependence, or used medicines that affect
the central nervous system. Institutionally approved,
written informed consent was obtained from each
subject. All subjects had four sessions, each at least
one week apart, in which perception, mood, and cog-
nition were evaluated: 1) while receiving normal
saline; 2) at plasma ketamine target concentrations of
50, 100, and 150 ng·mL– 1; 3) at a plasma midazolam
Plasma ketamine and midazolam concentrations

868
CANADIAN JOURNAL OF ANESTHESIA
TABLE I Blood pressure (BP) and heart rate (mean ± SD) during placebo, ketamine, midazolam, and coadministration (n = 11)
Time (min)
0
30*
60†
90‡
Systolic BP (mmHg)
Diastolic BP (mmHg)
Heart rate (bpm)
Midazolam
Ketaminea
Co-admin
Placebo
Midazolam
Ketamineb
Co-admin
Placebo
Midazolam
Ketamine
Co-admin
Placebo
126 ± 16
131 ± 12
128 ± 16
125 ± 11
65 ± 13
72 ± 7
69 ± 11
64 ± 5
67 ± 11
72 ± 12
71 ± 14
69 ± 11
114 ± 15
129 ± 14
120 ± 13
119 ± 13
59 ± 10
68 ± 8
56 ± 8
62 ± 9
63 ± 12
68 ± 11
68 ± 10
68 ± 12
113 ± 15
130 ± 10
121 ± 19
119 ± 8
60 ± 10
74 ± 8
64 ± 13
62 ± 8
61 ± 9
116 ± 14
135 ± 16
124 ± 16
123 ± 8
62 ± 11
76 ± 11
66 ± 12
65 ± 6
58 ± 11
71 ± 11
73 ± 13
72 ± 10
67 ± 9
71 ± 11
69 ± 13
b
aP = 0.0001. P = 0.001.
*ketamine, 45 ± 15 (SD) ng·mL– 1; midazolam, 32 ± 6 ng·mL– 1; and ketamine, 59 ± 13 ng·mL– 1 and midazolam, 32 ± 7 ng·mL– 1 during
coadministration. †ketamine, 93 ± 29 ng·mL– 1; midazolam, 29 ± 11 ng·mL– 1; and ketamine, 103 ± 25 ng·mL– 1 and midazolam, 35 ± 9
ng·mL– 1 during coadministration. ‡ketamine, 146 ± 34 ng·mL– 1; midazolam, 29 ± 12 ng·mL– 1; and ketamine, 158 ± 35 ng·mL– 1 and
midazolam, 32 ± 11 ng·mL– 1 during coadministration.
were measured using capillary gas chromatography
with nitrogen-selective detection by adapting previ-
ously published methods for ketamine1 4 and midazo-
lam.1 5 Phencyclidine and flurazepam, respectively,
were used as internal standards. Limits of detection for
both drugs were 1 ng·mL–1. Limits for quantification
were set at 10 ng·mL–1; at this low concentration the
assay exhibited coefficients of variation of 4 and 12%
for the drugs, respectively.
electrode impedance kept below 5 kOhm (at 12 Hz).
Electrocardiogram and arterial oxygen saturation were
monitored. Non-invasive blood pressure (BP, mm
Hg), heart rate (HR, bpm), and respiratory rate (RR,
/min) were recorded every five minutes. Dysphoria
was to be treated by 1-2 mg midazolam, i.v.
Differences among groups in VAS scores for mood
states, subjective sense of drowsiness, MMSE scores,
OAA/S scores, vital signs, and EEG measures (mean val-
ues of 60 epochs) were tested by two-way within sub-
jects, multivariate, repeated measures analysis of variance
using baseline values as a covariate. Where applicable, the
data were further analyzed using the Sidak multiple com-
parison method.1 6 Mood and BIS scores were further
tested for trend over time using the SAS-PROC-MIXED
(repeated measures) test. VAS values measuring percep-
tion, thought process, and suspicion were dichotomized
into scores of zero (non-responder) and one (respon-
der). The dichotomized values were analyzed using
In a quiet isolation room, an antecubital vein was
cannulated for infusion in an arm in the first four
patients. Both antecubital veins were cannulated for
infusion and blood samples in the last seven subjects.
After a rest for 20 to 30 min, baseline data were
obtained on perception, mood states, cognition, para-
noia, and sedation. According to the assignment, sub-
jects received a 10 ml bolus injection over 20 sec of
normal saline with or without 30 µg·kg–1 midazolam,
and infusion of normal saline or ketamine 0.25% to
attain a plasma concentration of 50 ng·mL– 1. A bolus
of normal saline or midazolam, 14.5 and 12.0 µg·kg–1,
was repeated every 30 min. Ketamine infusion rate
was increased step-wise every 30 min to attain target
plasma concentrations of 100 and 150 ng·mL–1.
Assessments of perception, mood, cognition, para-
noia, and sedation were repeated 24 min after each
injection. A blood sample was obtained after each
assessment. A measure of electroencephalogram
(EEG) (i.e., bispectral index [BIS]) was monitored
using Aspect A-1000 EEG Brain Monitoring
System™ (Aspect Medical Systems, Natick, MA) with
the system bandpass set at 1 Hz and 30 Hz and the
Cochran’s Q statistics at 30, 60, and 90 min.
was considered to be statistically significant.
P < 0.05
Results
There was a small increase in systolic and diastolic BP
during ketamine infusion (P < 0.0001 and P < 0.001)
(Table I). There was no change in HR during keta-
mine infusion. Midazolam reduced HR (P < 0.05),
but produced no change in BP. There were no
changes in BP or HR during co-administration. There
was no group difference in RR.
Plasma concentrations of ketamine during keta-
mine infusion were 45 ± 15 (mean ± SD), 93.0 ± 29,

Suzuki et al.: MIDAZOLAM A N D KETAMINE
869
TABLE II Differences (P values by Cochran Q test) in the num-
ber of subjects having perceptual changes between 4 treatment
regimes (i.e., placebo, ketamine, midazolam, and coadministration
at 30, 60, and 90 min assessment) (n = 11)
Time (min)
Perception
30*
60†
90‡
Body
Surrounding
Time
Reality
Color
0.007
NS
NS
NS
NS
0.001
0.008
0.002
0.001
NS
0.0001
0.0001
0.0001
0.0001
NS
Sound
Voice
Meaning
NS
NS
NS
NS
NS
NS
0.002
NS
0.029
*ketamine, 45 ± 15 (SD) ng·mL– 1; midazolam, 32 ± 6 ng·mL– 1;
and ketamine, 59 ± 13 ng·mL– 1 and midazolam, 32 ± 7 ng·mL– 1
during coadministration. †ketamine, 93 ± 29 ng·mL– 1; midazo-
lam, 29 ± 11 ng·mL– 1; and ketamine, 103 ± 25 ng·mL– 1 and
midazolam, 35 ± 9 ng·mL– 1 during coadministration. ‡ketamine,
146 ± 34 ng·mL– 1; midazolam, 29 ± 12 ng·mL– 1; and ketamine,
158 ± 35 ng·mL– 1 and midazolam, 32 ± 11 ng·mL– 1 during coad-
ministration.
and 146 ± 34 ng·mL– 1 at 30, 60, and 90 min, respec-
tively. Plasma midazolam concentrations during mida-
zolam administration were 32 ± 6, 29 ± 11, and 29 ±
12 ng·mL–1 at 30, 60, and 90 min. During the coad-
ministration, ketamine concentrations were 59 ± 13,
103 ± 25, and 158 ± 35 ng·mL– 1, and midazolam con-
centrations were 32 ± 7, 35 ± 9, and 32 ± 11 ng·mL– 1
at 30, 60, and 90 min, respectively.
FIGURE 1 the number of subjects having perceptual abnormali-
ty. = midazolam administration, = ketamine infusion, and
= coadministration. aP < 0.01 and bP = 0.001 vs placebo, mida-
zolam, and coadministration. cP = 0.0001, dP <0.01, and eP =
0.01 vs placebo and midazolam. fP < 0.05 vs placebo, midazolam,
and coadministration.
*ketamine, 45 ± 15 (SD) ng·mL– 1; midazolam, 32 ± 6 ng·mL– 1;
ketamine, 59 ± 13 ng·mL– 1 and midazolam, 32 ± 7 ng·mL– 1 during
coadministration. †ketamine, 93 ± 29 ng·mL–1; midazolam, 29 ±
11 ng·mL– 1; and ketamine, 103 ± 25 ng·mL– 1 and midazolam, 35
± 9 ng·mL– 1 during coadministration. ‡ketamine, 146 ± 34
ng·mL– 1; midazolam, 29 ± 12 ng·mL– 1; and ketamine, 158 ± 35
ng·mL–1 and midazolam, 32 ± 11 ng·mL–1 during coadministration.
There were differences among the drug regimens in
the number of subjects having perceptual abnormality
in six out of eight categories (i.e., body, surrounding,
time, reality, sound, and meaning) (Table II). More
subjects experienced the abnormality at higher plasma
ketamine concentrations. They felt heavy, light, mov-
ing, floating and/or undulating, and stated that the
entire or a part of the body was tingling, or that a part
of the body was out of proportion in size, displaced or
absent. One subject thought his right hand was his left
hand and vice versa. Two subjects felt their body was so
heavy that they were unable to lift a hand. Perceptions
of surroundings were described as spinning, in slow
motion, smaller in appearance, or that the surroundings
appeared as if on a scroll. Time was felt to pass faster or
slower, or to have stopped. The subjects described
colour as brighter, less bright, or blending. Sound was
described as far away, louder, clearer, or amplified. Our
subjects felt “weird,” “strange,” and/or “detached,”
and some felt that the reality was a slow-motion dream.
Two subjects thought they were dreaming of riding a
comet and a spaceship. Neither placebo nor midazolam
produced perceptual change, except for one subject
receiving midazolam who felt sounds were unusually
loud. Coadministration of midazolam attenuated the
ketamine-induced change in body perception (P < 0.01
and 0.001 at higher ketamine target concentrations).
Although differences were not statistically significant in
the other five categories, the numbers of subjects with
perceptual changes were lower during the coadminis-
tration than they were during the infusion of ketamine
alone (Figure 1). None had dysphoria, hallucinations,
or psychotic symptoms.
The VAS score during ketamine infusion was lower
in two subsets of mood states (i.e., tired-energetic [P
< 0.02], and confused-clearheaded [P < 0.05]). The
SAS-PROC-MIXED (repeated measures) test showed

870
CANADIAN JOURNAL OF ANESTHESIA
TABLE III The total Mini-Mental State Examination (MMSE) score (mean ± SD) and the scores for five domains (i.e., orientation, reg-
istration, recall, attention, and language fluency) during placebo, ketamine, midazolam, and coadministration (n = 11)
Placebo
Ketamine
Midazolam
Co-administration
Orientation
Registration
Recall
Attention
Language
10.00 ± 0.00
3.00 ± 0.00
2.93 ± 0.03b
5.00 ± 0.00
9.00 ± 0.00
29.93 ± 0.03b
10.00 ± 0.00
3.00 ± 0.00
2.44 ± 0.14a
5.00 ± 0.00
9.00 ± 0.00
29.44 ± 0.14a
10.00 ± 0.00
3.00 ± 0.00
1.93 ± 0.22a
4.98 ± 0.02
9.00 ± 0.00
28.92 ± 0.23a
9.98 ± 0.02
3.00 ± 0.00
1.86 ± 0.25a
4.89 ± 0.05
8.96 ± 0.02
28.61 ± 0.30a
Total MMS score
aP < 0.01 vs placebo.
TABLE IV OAA/S scores, Drowsiness VAS scores, and BIS values (mean ± SD) during placebo, ketamine, midazolamm, and coadminis-
tration (n = 11)
Time (min)
0
30*
60†
90‡
OAA/S
Drowsiness
BIS
Midazolam
Ketamine
Coadmin
Placeboa
Midazolam
Ketamine
Coadmin
Placebob
Midazolam
Ketamine
Co-admin
Placeboc
5.0 ± 0.0
5.0 ± 0.0
5.0 ± 0.0
5.0 ± 0.0
0 ± 0
0 ± 0
0 ± 0
0 ± 0
96 ± 2
97 ± 1
97 ± 1
97 ± 1
4.4 ± 0.9
4.8 ± 0.4
4.4 ± 0.8
5.0 ± 0.0
58 ± 22
36 ± 23
55 ± 17
0 ± 0
81 ± 19
96 ± 3
79 ± 12
97 ± 1
3.8 ± 1.0
4.5 ± 0.5
4.0 ± 1.2
5.0 ± 0.0
64 ± 20
61 ± 27
62 ± 23
0 ± 0
75 ± 18
97 ± 1
88 ± 10
97 ± 1
4.3 ± 0.7
3.9 ± 0.3
4.3 ± 0.9
5.0 ± 0.0
62 ± 20
72 ± 26
74 ± 22
0 ± 0
78 ± 21
97 ± 1
90 ± 15
96 ± 1
aP = 0.001 vs the average scores for the midazolam, ketamine, and coadministration groups.
bP = 0.0001 vs the average scores for the midazolam, ketamine, and coadministration groups.
cP = 0.002 vs the average scores for the midazolam and coadministration groups.
*ketamine, 45 ± 15 (SD) ng·mL– 1; midazolam, 32 ± 6 ng·mL– 1; and ketamine, 59 ± 13 ng·mL– 1 and midazolam, 32 ± 7 ng·mL– 1 during
coadministration. †ketamine, 93 ± 29 ng·mL– 1; midazolam, 29 ± 11 ng·mL– 1; and ketamine, 103 ± 25 ng·mL– 1 and midazolam, 35 ± 9
ng·mL– 1 during coadministration. ‡ketamine, 146 ± 34 ng·mL– 1; midazolam, 29 ± 12 ng·mL– 1; and ketamine, 158 ± 35 ng·mL– 1 and
midazolam, 32 ± 11 ng·mL– 1 during coadministration.
significant trends over time (i.e., plasma concentra-
tion) (mood state = A × time [min] + B × time2).
There were positive changes over plasma concentra-
tions in three subsets (anxious-composed [B = 0.002,
P < 0.02], depressed-elated [B = 0.003, P < 0.001],
and hostile-agreeable [B = 0.002, P < 0.01]), and neg-
ative changes in two subsets of mood states (tired-
energetic and confused-clearheaded) during ketamine
infusion. The subjects described their mood as more
positive, high, uplifted, and euphoric, and stated that
they felt relaxed, although they were less energetic and
increasingly less clearheaded. Midazolam produced a
negative change over time in one subset (confused-
clearheaded [A = -0.406, P < 0.0002 and B = 0.0023,
P < 0.009]). The subjects expressed their mood as less
energetic but relaxed. One subject felt lethargic.
During co-administration, there was a positive change
over time in one subset (depressed-elated [B = 0.002,
P < 0.05]) and negative changes in two subsets (tired-
energetic and confused-clear-headed). Ten of 11 sub-
jects stated that they felt good and relaxed. One said
that he felt tired.
There was a difference in the mean total MMSE
score (P < 0.01). The mean recall score was lower dur-
ing administration of ketamine, midazolam, or co-
administration than during the placebo run (P < 0.01)
(Table III). There were no changes in the scores for
registration, orientation, attention, and language fluen-
cy. The number of subjects having abnormal thought
process increased, however, during ketamine infusion in
a dose-dependent manner (P < 0.003 and P < 0.0001
at 60 min and 90 min assessment, respectively). The

Suzuki et al.: MIDAZOLAM A N D KETAMINE
871
scores, P = 0.001). The average drowsiness score dur-
ing the placebo trial was lower and the average
OAA/S score was higher than those during the
administration of midazolam and ketamine, as well as
during co-administration.
Discussion
In this study, low-dose ketamine produced dose-relat-
ed alterations in somesthetic and proprioceptive per-
ceptions as well as auditory and visual perceptual
changes, and positive relationships for subsets of
mood states (i.e., elation, composure, and agreeable-
ness), described as relaxed, better, uplifted, “high,”
and/or euphoric. Midazolam did not produce percep-
tual alterations or similar mood changes. Co-adminis-
tration of midazolam attenuated the perceptual
alterations but did not affect the mood changes. Both
ketamine and midazolam impaired post-distraction
but not immediate recall. Midazolam did not increase
ketamine-induced memory impairment significantly.
Ketamine induced a thought disorder. Midazolam
attenuated the thought process abnormality.
Ketamine, midazolam, and their coadministration all
produced similar levels of sedation and drowsiness.
Dose-related perceptual alterations observed dur-
ing ketamine infusion were not quantitative but were
all qualitative. A sensation that a part of the body was
missing, displaced, or the inability to tell the left hand
from the right in our subjects is consistent with tactile
agnosia, which is associated with a lesion in the poste-
rior parietal lobe.1 7These alterations indicate that ket-
amine affected a large area of sensory association
cortices in our subjects, and that perceptual impair-
ment induced by low-dose ketamine may involve the
process of sensory integration.7 It has been shown that
ketamine produces electroencephalographic ictal
activity of the neocortex, thalamus, and hippocampus,
and markedly increases the evoked potentials of the
sensorimotor and visual cortex of cats.3 Ketamine
increases porcine cerebral oxygen consumption4 and
hippocampal glucose metabolism in rats,5 and has
been reported to induce seizure activity in patients
with temporal lobe epilepsy.1 8Midazolam antagonizes
the increase in oxygen consumption4 and the seizure
activity produced by ketamine. These findings as well
as the midazolam-induced attenuation of perceptual
alterations observed in our subjects appear to be con-
sistent with the concept that NMDA receptors may
modulate functions of cortical and subcortical struc-
tures in adults.2
FIGURE 2 The number of subjects having abnormal thought
processes. = midazolam administration, = ketamine infusion,
and = coadministration. aP < 0.01 and bP = 0.001 vs placebo
and midazolam.
*ketamine, 45 ± 15 (SD) ng·mL– 1; midazolam, 32 ± 6 ng·mL– 1;
ketamine, 59 ± 13 ng·mL– 1 and midazolam, 32 ± 7 ng·mL– 1 during
coadministration. †ketamine, 93 ± 29 ng·mL–1; midazolam, 29 ±
11 ng·mL– 1; and ketamine, 103 ± 25 ng·mL– 1 and midazolam, 35
± 9 ng·mL– 1 during coadministration. ‡ketamine, 146 ± 34
ng·mL– 1; midazolam, 29 ± 12 ng·mL– 1; and ketamine, 158 ± 35
ng·mL–1 and midazolam, 32 ± 11 ng·mL–1 during coadministration.
subjects described their thoughts as “flight of ideas,”
“my thoughts are faster, busy, racing, or jumping,” or
“scenes from movies or past events keep popping into
my head.” Two subjects stated their thoughts slowed
down transiently. Neither placebo nor midazolam pro-
duced a thought disorder. Co-administration reduced
the number of subjects having ketamine-induced
changes in thought process by 50% (Figure 2). Those
who continued to have abnormal thought processes
described their thoughts as floating, fleeting, blending,
or finding it difficult to concentrate. Neither midazo-
lam, ketamine, nor the co-administration produced
changes in the suspicion (i.e., paranoia) score.
There was a difference in the average score of BIS
over time (P < 0.02) and among the drug regimens (P
= 0.002). The BIS did not change during placebo or
ketamine infusion but was lower during midazolam
administration. The BIS appeared to increase during
coadministration as plasma ketamine concentration
increased (constant = 96.3; B = -0.439 [P = 0.0001];
and B = 0.0032 [P = 0.0001]) (Table IV). There were
differences over time and among the drug regimens in
the average score of drowsiness VAS (P = 0.0001 for
both) and OAA/S (time, P = 0.003, and the averaged
The striatal complexes (i.e., dorsal and ventral stria-
tum and pallidum) have been recognized as a powerful
inhibitory structure on the thalamus and the ascending

872
CANADIAN JOURNAL OF ANESTHESIA
reticular formation. Stimulation of the glutamatergic
corticostriatal projection inhibits the thalamus and
reticular nuclei, leading to reducted sensory transmis-
sion to cerebral cortices as well as a reduced state of
arousal, whereas dopamine exerts an inhibitory influ-
ence on the striatal neurons, thus counteracting the
inhibitory influence of the striatal complexes on the
thalamus and the reticular formation.19,20 When the
sensory transmission becomes excessive (e.g., the
NMDA-receptor antagonist), the integral capacity of
the cortex may break down, and confusion or psychosis
impairment was not related to concomitant attention-
al and behavioural changes.8 Thus, ketamine appears to
impair early phases of memory formation, relatively
sparing encoding and short-term memory. A benzodi-
azepine, on the other hand, has been considered to
impair consolidation of newly acquired information
through reduction in long-term potential, mediated by
its action on the GABAA receptor.23 In our subjects, the
MMSE showed impairment in post-distraction recall
during ketamine or midazolam administration.
Although midazolam did not increase ketamine-
induced amnesia, a recent study reported that
lorazepam increased the amnestic effect of ketamine.2 4
The small number of subjects in our study may be a
factor in the lack of a statistical significance.
Ketamine has been shown to induce frontal lobe dys-
function (i.e., executive cognitive function, verbal flu-
ency, and vigilance) in volunteers.7, 8A thought disorder
observed in our subjects during ketamine infusion was
consistent with flight of ideas, a form of disorganized
thought, which occurs in hyperadrenergic states (e.g.,
stimulant overdose and sedative withdrawal), anxiety
and agitation, as well as in psychotic and manic presen-
tations,2 5 suggesting that the thought disorder in our
subjects may have been a manifestation of excitatory
phenomenon. Benzodiazepines are usually effective for
the treatment of hyperadrenergic states, anxiety, and
agitation.2 6 The thought disorder in our subjects was
attenuated by midazolam. Of interest in this context
was a recent study that showed amelioration of keta-
mine-induced impairments in executive cognitive func-
may ensue. Reduction in dopaminergic (D ) function
2
(e.g., the D-receptor antagonist) results in inhibition
of motor an2d mental activity. Recent studies have also
shown antagonistic interactions between the NMDA-
receptor and the serotonergic, as well as noradrenergic
and muscarinic activities.2 0
Subanesthetic doses of ketamine and its enan-
tiomers have been shown to produce a feeling of
“high” and to be anxiolytic at low doses, but anxio-
genic at higher doses.2, 7The similarity in the “highs”
produced by ketamine and alcohol intoxication has
been attributed to the NMDA receptor antagonist
property of both drugs.7 Our subjects described mood
during ketamine infusion as “high,’ while midazolam
did not produce a state of “high,” and the co-admin-
istration did not affect ketamine-induced positive
mood, suggesting that ketamine and midazolam
involve different mechanisms for mood alteration. The
mechanism for ketamine-induced mood alteration is
not clear. It is possible that increased thalamic sensory
output to cerebral cortices secondary to ketamine-
induced blockade of the corticostriatal glutamatergic
pathway19,20 and a possible interaction with the sero-
tonergic system, may have contributed to the keta-
mine-induced mood alteration.2 0
tions by co-administration of haloperidol, a D -receptor
2
antagonist, suggesting the possibility that ketamine-
induced frontal lobe symptoms and disorganized,
hyperactive thought may, at least in part, involve the
striatal system.2 7
It has been suggested that NMDA receptor-induced
long-term synaptic potentiation in the hippocampus
and subsequent protein synthesis may be the mecha-
nism underlying the formation of short-term, interme-
diate-term, and long-term memory in vertebrates.2 1
Reports in humans have shown that keta-
mine impairs post-distraction and delayed word
recall,2,7,8,22 but may2 2 or may not7,8 affect immediate
recall. A 24-word task has shown impaired immediate
recall but preserved recognition memory, suggesting
that ketamine may impair retrieval of information.2 2
Although it was suggested that observed discrepancies
on immediate recall may be due to the level of the chal-
lenge the tasks impose and also ketamine-induced
attention deficit,7 a later study with a less challenging
task failed to detect impaired retrieval.2 Further, a
recent study showed that ketamine-induced memory
The absence of additive effect on the level of seda-
tion during the co-administration suggests that keta-
mine and midazolam induce sedation by different
mechanisms. In our subjects, ketamine produced active
thoughts and awake level BIS, while midazolam-
induced sedation was associated with quiet thought and
a decline in BIS, although the levels of sedation were
similar. During the coadministration, the thought status
and BIS were between those obtained during the
administration of ketamine and midazolam alone. The
mechanism by which ketamine induces sedation is not
known. However, association of hyperactive thought
process with sedation observed in our subjects and the
observations that subjects were rather awake than sleepy
while receiving subanesthetic doses of ketamine7 may
be consistent with arousal and increase in the thalamo-
cortical activity mediated by the striatal mechanism.17,18

Suzuki et al.: MIDAZOLAM A N D KETAMINE
873
Presynaptic inhibition of the GABA receptor2 8may also
be a possible contributor to the lack of increase in seda-
tion during the coadministration.
Our results suggest that midazolam attenuates per-
ceptual abnormalities and thought disorder induced
by low-dose ketamine without changes in sedation or
drowsiness. The results also suggest that midazolam
may not affect ketamine-induced amnesia or changes
in mood.
10 Domino EF, Domino EE, Smith RE, et al. Ketamine
kinetics in unmedicated and diazepam-premedicated
subjects. Clin Pharmacol Ther 1984; 36: 645–53.
11 Greenblatt DJ, Ehrenberg BL, Gunderman J, et al.
Pharmacokinetic and electroencephalographic study of
intravenous diazepam, midazolam, and placebo. Clin
Pharmacol Ther 1989; 45: 356–65.
12 Folstein MF, Folstein SE, McHugh PR. “Mini-mental
state.” A practical method for grading the cognitive
state of patients for the clinician. J Psychiatr Res 1975;
12: 189–98.
Acknowledgments
13 Chernik DA, Gillings D, Laine H, et al. Validity and
reliability of the observer’s assessment of
The authors gratefully acknowledge Hoffman-La
Roche for a gift of analytical grade midazolam, which
was used for standardization of the gas chromato-
graphic assay. We thank Robert L. Vogel PhD for his
help in statistical analysis and Hassan Shahab MD for
editorial advice.
alertness/sedation scale: study with intravenous mida-
zolam. J Clin Psychopharmacol 1990; 10: 244–51.
14 Stiller RL, Dayton PG, Perel JM, Hug CC Jr. Gas chro-
matographic analysis of ketamine and norketamine in
plasma and urine: nitrogen-sensitive detection. J
Chromatogr 1982; 232: 305–14.
References
15 Vasiliades J, Owens C. Determination of midazolam in
serum by gas chromatography with a nitrogen-sensitive
detector. J Chromatogr 1980; 182: 439–44.
16 Westfall PH, Young SS. Resampling-Based Multiple
Testing. Examples and Methods for p-Value
Adjustment. New York: John Wiley & Sons Inc., 1993:
64–6.
1 Cotman CW, Monaghan DT. Excitatory amino acid
neurotransmission: NMDA receptors and Hebb-type
synaptic plasticity. Ann Rev Neurosci 1988; 11: 61–80.
2 Øye I, Paulsen O, Maurset A. Effects of ketamine on
sensory perception: evidence for a role of N-methyl-
D-asparate receptors. J Pharmacol Exp Ther 1992;
260: 1209–13.
17 Gilroy J, Meyer JS. Medical Neurology, 3rd ed. New
York: MacMillan Publishing Co., Inc., 1979.
18 Ferrer-Allado T, Brechner VL, Dymond A, Cozen H,
Crandall P. Ketamine-induced electroconvulsive phe-
nomena in the human limbic and thalamic regions.
Anesthesiology 1973; 38: 333–44.
19 Carlsson M, Carlsson A. Schizophrenia: a subcortical
neurotransmitter imbalance syndrome? Schizophr Bull
1990; 16: 425–32.
20 Carlsson A, Hansson LO, Waters N, Carlsson ML.
Neurotransmitter aberrations in schizophrenia: new
perspectives and therapeutic implications. Life Sci
1997; 61: 75–94.
21 Rosenzweig MR, Leiman AL, Breedlove SM. Biological
Psychology. An Introduction to Behavioral, Cognitive
and Clinical Neuroscience, 2nd ed. Sunderland:
Sinauer Associates, Inc., 1999.
22 Ghoneim MM, Hinrichs JV, Mewaldt SP, Petersen RC.
Ketamine: behavioral effects of subanesthetic doses. J
Clin Psychopharmacol 1985; 5: 70–7.
23 Barbee JG. Memory, benzodiazepines, and anxiety:
integration of theoretical and clinical perspectives. J
Clin Psychiatr 1993; 54: 86–97.
24 Krystal JH, Karper LP, Bennett A, et al. Interactive
effects of subanesthetic ketamine and subhypnotic
lorazepam in humans. Psychopharmacology 1998; 135:
213–29.
3 Kayama Y, Iwama K. The EEG, evoked potentials,
and single unit activity during during ketamine anes-
thesia in cats. Anesthesiology 1972; 36: 316–28.
4 Âkeson J, Björkman S, Messeter K, Rosén I. Low-dose
midazolam antagonizes cerebral metabolic stimulation
by ketamine in the pig. Acta Anaesthesiol Scand 1993;
37: 525–31.
5 Oguchi K, Arakawa K, Nelson SR, Samson F. The influ-
ence of droperidol, diazepam, and physostigmine on
ketamine-induced behavior and brain regional glucose
utilization in rat. Anesthesiology 1982; 57: 353–8.
6 White PF, Way WL, Trevor AJ. Ketamine - its pharma-
cology and therapeutic uses. Anesthesiology 1982; 56:
119–36.
7 Krystal JH, Karper LP, Seibyl JP, et al. Subanesthetic
effects of noncompetitive NMDA antagonist, ketamine,
in humans. Psychotomimetic, perceptual, cognitive,
and neuroendocrine responses. Arch Gen Psychiatr
1994; 51: 199–214.
8 Malhotra AK, Pinals DA, Weingartner H, et al.
NMDA receptor function and human cognition: the
effects of ketamine in healthy volunteers.
Neuropsychopharmacology 1996; 14: 301–7.
9 Bowdle TA, Radant AD, Cowley DS, Kharasch ED,
Strassman RJ, Roy-Byrne PP. Psychedelic effects of keta-
mine in healthy volunteers. Relationship to steady-state
plasma concentrations. Anesthesiology 1998; 88: 82–8.

874
CANADIAN JOURNAL OF ANESTHESIA
25 Gitlin MJ. Clinical manifestations of psychiatric disor-
ders. In: Kaplan HI, Sadock BJ (Eds.). Comprehensive
Textbook of Psychiatry, 6th ed. Baltimore: Williams &
Wilkins., 1995: 637–69.
26 Yudofsky SC, Silver JM, Hales RE. Treatment of agitation
and aggressive disorders. In: Schatzberg AF, Nemeroff
CB (Eds.). Textbook of Psychopharmacology, 1st ed.
Washington, D.C.: American Psychiatric Press, Inc.,
1995: 735–51.
27 Krystal JH, D’Souza DC, Karper LP, et al. Interactive
effects of subanesthetic ketamine and haloperidol in
healthy humans. Psychopharmacology 1999; 145:
193–204.
28 Murugaiah KD, Hemmings HC Jr. Effects of intra-
venous general anesthetics on [3H]GABA release from
rat cortical synaptosomes. Anesthesiology 1998; 89:
919–28.