1. Consider the Lewis two-sector model. What are the 2 main assump

1. Consider the Lewis two-sector model. What are the 2 main
assumptions related to the traditional sectors? How these
assumptions help explaining the constant flow of workers from
the traditional to the modern sector?
2. Consider the Child Labor Multiple Equilibria Model. How is
it possible to have 2 equilibria? What the government can
possibly do to deal with it? Explain your answers.
Original Investigation | Psychiatry
Assessment of Relationship of Ketamine Dose With Magnetic
Resonance
Spectroscopy of Glx and GABA Responses in Adults With
Major Depression
A Randomized Clinical Trial
Matthew S. Milak, MD; Rain Rashid, BS; Zhengchao Dong,
PhD; Lawrence S. Kegeles, MD, PhD; Michael F. Grunebaum,
MD; R. Todd Ogden, PhD; Xuejing Lin, MA;
Stephanie T. Mulhern, BA; Raymond F. Suckow, PhD; Thomas
B. Cooper, MA; John G. Keilp, PhD; Xiangling Mao, MS;
Dikoma C. Shungu, PhD; J. John Mann, MD
Abstract
IMPORTANCE A single subanesthetic dose of ketamine
produces an antidepressant response in
patients with major depressive disorder (MDD) within hours,
but the mechanism of antidepressant
effect is uncertain.
OBJECTIVE To evaluate whether ketamine dose and brain
glutamate and glutamine (Glx) and

γ-aminobutyric acid (GABA) level responses to ketamine are
related to antidepressant benefit and
adverse effects.
DESIGN, SETTING, AND PARTICIPANTS This randomized,
parallel-group, triple-masked clinical trial
included 38 physically healthy, psychotropic medication–free
adult outpatients who were in a major
depressive episode of MDD but not actively suicidal. The trial
was conducted at Columbia University
Medical Center. Data were collected from February 2012 to
May 2015. Data analysis was conducted
from January to March 2020.
INTERVENTION Participants received 1 dose of placebo or
ketamine (0.1, 0.2, 0.3, 0.4, or 0.5
mg/kg) intravenously during 40 minutes of a proton magnetic
resonance spectroscopy scan that
measured ventro-medial prefrontal cortex Glx and GABA levels
in 13-minute data frames.
MAIN OUTCOMES AND MEASURES Clinical improvement
was measured using a 22-item version of
the Hamilton Depression Rating Scale (HDRS-22) 24 hours
after ketamine was administered.
Ketamine and metabolite blood levels were measured after the
scan.
RESULTS A total of 38 individuals participated in the study,
with a mean (SD) age of 38.6 (11.2) years,
23 (60.5%) women, and 25 (65.8%) White patients.
Improvement in HDRS-22 score at 24 hours
correlated positively with ketamine dose (t36 = 2.81; P = .008;
slope estimate, 19.80 [95% CI, 5.49
to 34.11]) and blood level (t36 = 2.25; P = .03; slope estimate,
0.070 [95% CI, 0.007 to 0.133]). The

lower the Glx response, the better the antidepressant response
(t33 = −2.400; P = .02; slope
estimate, −9.85 [95% CI, −18.2 to −1.50]). Although GABA
levels correlated with Glx (t33 = 8.117;
P < .001; slope estimate, 0.510 [95% CI, 0.382 to 0.638]),
GABA response did not correlate with
antidepressant effect. When both ketamine dose and Glx
response were included in a mediation
analysis model, ketamine dose was no longer associated with
antidepressant effect, indicating that
Glx response mediated the relationship. Adverse effects were
related to blood levels in men only
(t5 = 2.606; P = .048; estimated slope, 0.093 [95% CI, 0.001 to
0.186]), but Glx and GABA response
were not related to adverse effects.
(continued)
Key Points
Question What is the relationship
between the antidepressant effect of
ketamine and ketamine dose and blood
level, and is its antidepressant effect
mediated by an effect on ventro-medial
prefrontal cortical glutamate and
glutamine or γ-aminobutyric acid
response?
Findings This randomized clinical trial
of 38 patients with major depression

found a relationship of ketamine dose
and blood level with antidepressant
response at 24 hours. Ketamine
suppression of glutamate and glutamine
in the ventro-medial prefrontal cortex
mediated the relationship of ketamine
dose and level with antidepressant
effect but was unrelated to
psychotomimetic side effects.
Meaning The findings of this study
suggest that glutamate and glutamine
suppression by ketamine may be a
potential biomarker of rapid
antidepressant effect and that a fast-
acting antidepressant without
psychotomimetic adverse effects may
be possible.
+ Visual Abstract
+ Supplemental content

Author affiliations and article information are
listed at the end of this article.
Open Access. This is an open access article distributed under
the terms of the CC-BY License.
JAMA Network Open. 2020;3(8):e2013211.
doi:10.1001/jamanetworkopen.2020.13211 (Reprinted) August
12, 2020 1/14
Downloaded From: https://jamanetwork.com/ by Agata Boxe on
01/21/2021
https://jama.jamanetwork.com/article.aspx?doi=10.1001/jamanet
workopen.2020.13211&utm_campaign=articlePDF%26utm_med
ium=articlePDFlink%26utm_source=articlePDF%26utm_content
=jamanetworkopen.2020.13211
Abstract (continued)
CONCLUSIONS AND RELEVANCE In this study, intravenous
ketamine dose and blood levels
correlated positively with antidepressant response. The Glx
response correlated inversely with
ketamine dose and with antidepressant effect. Future studies are
needed to determine whether the
relationship between Glx level and antidepressant effect is due
to glutamate or glutamine.
TRIAL REGISTRATION ClinicalTrials.gov Identifier:
NCT01558063

doi:10.1001/jamanetworkopen.2020.13211
Introduction
Major depressive disorder (MDD), a leading cause of disability
worldwide,1 affects more than 16
million adults in the United States,2 with estimated costs of
$210.5 billion in 2010, a 21.5% increase
from 2005.3 Response to currently marketed antidepressants
generally requires treatment for 6 to 8
weeks,4 and they are ineffective in 30% to 50% of patients.5,6
Faster-acting and more effective
antidepressants are needed.
A single intravenous (IV) subanesthetic dose of ketamine can
produce an antidepressant
response in hours instead of weeks, even in medication-resistant
MDD.7-16 Ketamine has
antidepressant benefit in both patients with MDD17,18 and
those with bipolar depression.19 Adverse
effects of ketamine include transient depersonalization and
derealization, among other
psychotomimetic and dissociative symptoms.20,21
Understanding the mechanism of its rapid-onset
antidepressant action may aid identification of alternative
medications that can be used orally, have
fewer adverse effects, and have less abuse potential.
We previously reported on an open pilot MDD study that
showed, consistent with glutamate
increases in rodent studies,22,23 that ketamine induces an acute
increase in ventro-medial prefrontal
cortex (mPFC) levels of the combined resonance of glutamate
and glutamine (Glx), measured with

proton magnetic resonance spectroscopy (1H MRS).24
Unexpectedly, we also observed an increase in
γ-aminobutyric acid (GABA) levels that correlated positively
with the increase in Glx levels. Preclinical
studies suggest that ketamine’s antidepressant mechanism of
action may be mediated by glutamate
activation of the glutamatergic α-amino-3-hydroxy-5-methyl-4-
isoxazolepropionic acid (AMPA)
receptors25-28 and the downstream induction of the
neurotrophin29,30 and mammalian target of
rapamycin29,30 (mTOR; also known as mechanistic TOR)
signaling pathways. An alternative
model31,32 suggests that it is not glutamate that mediates
ketamine’s antidepressant action; rather, it
is a direct consequence of ketamine's inhibitory effect on the N-
methyl-D-aspartate (NMDA)
receptors. The role of GABA is unknown, but an increase in its
level reverses the reported GABA
deficit in major depression.33-37
In the current study, we sought to determine whether the dose of
ketamine or blood levels of
ketamine and its metabolites are correlated with antidepressant
effect and adverse effects. We also
explored whether ketamine or metabolite levels correlate with
mPFC Glx or GABA levels and, in turn,
whether Glx or GABA levels mediate the acute antidepressant
or adverse effects of ketamine among
individuals with MDD. To do this, we conducted a randomized,
placebo-controlled, dose-finding,
clinical trial in patients with MDD not currently receiving
medication, who underwent MRS
measurements of mPFC Glx and GABA during the IV
administration of ketamine. Blood levels of
ketamine and metabolites were assayed after the scan.

JAMA Network Open | Psychiatry Relationship of Ketamine
Dose With Depression Symptoms in Adults With Major
Depression
12, 2020 2/14
01/21/2021
https://www.clinicaltrials.gov/ct2/show/NCT01558063
Methods
Patients
Figure 1 shows the study flow. Of the 43 patients, 5 were
unable to complete the study because of
scanner-related discomfort or technical issues. For the final
analysis, we included 38 physically
healthy patients, aged 18 to 59 years, who met the Diagnostic
and Statistical Manual of Mental
Disorders (Fourth Edition) criteria for a major depressive
episode (MDE) in the context of MDD and
scored at least 22 on the Montgomery-Åsberg Depression Rating
Scale (Table), which was only used
to establish depression severity for determini ng eligibility to
prevent inflation of baseline scores38-40
on the primary outcome measure (a 22-item version of Hamilton
Depression Rating Scale

[HDRS-22]). Patients were not taking any psychotropic
medications and had not been taking
medications likely to interact with GABA or glutamate for at
least 14 days, neuroleptics for at least 1
month, or fluoxetine for at least 6 weeks before receiving
ketamine. The detailed protocol is available
in Supplement 1.
Exclusion criteria included lifetime history of bipolar disorder,
schizoaffective disorder,
schizophrenia, or any other psychotic disorder, including MDD
with psychotic features; a first-degree
relative with bipolar disorder, schizoaffective disorder, or
schizophrenia, with the potential
participant younger than 33 years (ie, still at age of risk for a
psychotic disorder); receipt of
electroconvulsive therapy within 3 months of enrolling in the
study; history of IV drug use;
nonresponse or intolerance to ketamine (defined as participation
in another ketamine study and
reporting to us less than a robust response); pregnancy,
planning to conceive, or sexually active but
not using adequate birth control; and contraindications to
magnetic resonance imaging (MRI).
Patients who were actively suicidal were also excluded. All
patients were outpatients. All patients
were enrolled after a psychiatric and medical screening,
conducted between February 2012 through
May 2015, determined that they met the entrance criteria and
after they provided written informed
consent. The study was approved by the New York State
Psychiatric Institute institutional review
board and was completed in October 2019, before recruitment
goals were met (see details in Study
Design section). This study followed the Consolidated
Standards of Reporting Trials (CONSORT)

reporting guideline.
Figure 1. Study Flow Diagram
422 Patients assessed for eligibility
5 Allocated to and received
placebo
38 Allocated to and received
ketamine
32 Analyzed5 Analyzed
43 Randomized
379 Excluded
117 Were treatment naive
37 Had psychiatric comorbidity
38 Had lack of interest
30 Had medical comorbidity
27 Were lost to follow-up after screening
80 Had other reasons
6 Excluded
1 Discontinued intervention because of
adverse effects
4 Experienced scanner technical issues
1 Had poor quality spectra
Depression

12, 2020 3/14
01/21/2021
http://www.equator-network.org/reporting-guidelines/consort/
Study Design
Randomization and assignment of ketamine dose (ie, 0.1, 0.2,
0.3, 0.4 or 0.5 mg/kg) or placebo were
performed by the statistician. An adaptive randomization
strategy was used to optimize group size
in terms of dose response curves. The randomization was
supposed to have 11 participants per group,
but the study was stopped before recruiting goals were met
because funding ran out and an interim
analysis found robust statistical effects. Randomized ketamine
doses were shared via sealed
nontranslucent envelopes with the research pharmacy and a
clinician who was not involved in
patient study ratings. Study patients and raters were masked
until study termination unless
unmasking was clinically warranted. No patient required
unmasking in the course of the ketamine
infusion. Patients received a single 40-minute infusion and were
observed for 24 hours after
ketamine administration. All scans and data were obtained at
Columbia University Medical Center.
Prior to treatment infusion, patients underwent structural MRI

and baseline 1H MRS scans. Scan
analysis was performed on coded data sets masked to the subject
dosing and treatment response.
Patients were administered placebo or ketamine intravenously
during approximately 40 minutes. Six
1H MRS data frames of approximately 13 minutes each were
acquired: 1 prior to ketamine infusion
and 5 during and immediately after ketamine infusion. We
determined clinical response 24 hours
after ketamine administration using the HDRS-22. The masking
procedures and randomization
ensured minimal possibility of biasing based on Cochrane
criteria.41
Safety and Tolerability
Vital signs, including blood pressure and heart rate, were
monitored 5 minutes prior to ketamine
infusion, every 5 minutes during the 40-minute ketamine
infusion, and every 5 minutes after for the
duration of the 1H MRS scan or until vital signs returned to
clinically acceptable levels. Patients were
evaluated by a physician for blood pressure, psychosis and other
psychotomimetic adverse effects
(using the Brief Psychiatric Rating Scale [BRPS]), and suicidal
ideation (using the Columbia–Suicide
Severity Rating Scale) 230 minutes after the initiation of the
ketamine infusion. Four serious adverse
events occurring during study participation were reported to the
institutional review board: 1 for
Table. Demographic and Clinical Characteristics of Participants
by Assigned Intravenous Ketamine Dose
Characteristic
Mean (SD), by dose group Statistical differences

0.0 mg/kg
(n = 5)
0.1 mg/kg
(n = 5)
0.2 mg/kg
(n = 6)
0.3 mg/kg
(n = 8)
0.4 mg/kg
(n = 5)
0.5 mg/kg
(n = 9) F5,32 P value
Age, y 46.8 (12.3) 37.4 (12.3) 37.8 (8.2) 38.1 (7.2) 30.6 (9.4)
40.2 (14.5) 1.119 .37a
Women, No. (%) 2 (40) 4 (80) 5 (83) 4 (50) 3 (60) 5 (56) NA
.69b
White race, No. (%) 4 (80) 1 (20) 4 (67) 7 (88) 2 (40) 7 (78) NA
.14b
Hispanic ethnic group, No. (%) 1 (20) 1 (20) 1 (17) 0 (0) 1 (20)
2 (22) NA .84b
Age of MDD first episode onset, y 28.8 (16.9) 20.6 (10.3) 13.3
(2.7) 20.6 (8.2) 19.4 (5.6) 16.6 (8.1) 1.778 .15a
Duration of current MDE, y 5.3 (3.1) 4.1 (4.4) 13.7 (15.9) 8.1
(12.5) 8.8 (11.2) 16.2 (19.7) 0.790 .57a

Duration of MDD, y 18.0 (14.7) 16.8 (14.4) 24.6 (9.6) 17.3
(11.6) 10.9 (8.8) 23.7 (16.6) 0.876 .51a
Previous MDEs, No. 0.4 (0.5) 1.0 (1.4) 1.7 (1.9) 1.5 (2.1) 1.0
(1.4) 0.8 (0.8) 0.609 .69a
Baseline MADRS total score 32.8 (4.9) 31.0 (6.3) 28.0 (3.6)
31.1 (3.0) 32.8 (1.3) 30.9 (4.1) 0.935c .47a
Baseline HDRS-22 total score 25.0 (3.4) 25.0 (5.5) 24.2 (4.3)
25.9 (7.7) 27.2 (5.0) 27.0 (5.3) 0.290 .93a
Change in HDRS-22 total score 24 h
after ketamine infusion, %
−2.8 (3.7) −6.2 (13.0) −6.7 (7.3) −8.3 (6.3) −10.8 (6.8) −13.3
(7.9) NA NA
AUC blood level, ng/mLd
Ketamine 0.0 (0.0) 27.2 (7.0) 44.5 (18.9) 59.1 (12.2) 71.6 (16.8)
112.8 (27.9) NA NA
Norketamine 0.0 (0.0) 30.2 (17.2) 31.8 (6.8) 56.6 (18.4) 71.0
(21.8) 95.2 (25.6) NA NA
Dehydronorketamine 0.0 (0.0) 17.6 (4.8) 31.8 (11.7) 39.9 (9.2)
42.8 (9.6) 52.1 (17.5) NA NA
Abbreviations: AUC, area under the curve; HDRS-22, 22-item
Hamilton Depression
Rating Scale; MADRS, Montgomery-Åsberg Depression Rating
Scale; MDD, major
depressive disorder; MDE, major depressive episode; NA, not
applicable.
a Results of between groups analysis of variance.

b Result of Fisher exact test.
c Result for F5,31 statistic.
d Area under the curve calculated for ketamine and metabolite
blood levels as the sum
of their respective blood levels at 90 and 120 minutes
postinitiation of ketamine
infusion.
Depression
12, 2020 4/14
01/21/2021
suicide, 1 for active suicidal ideation, 1 for antidepressant
misuse, 1 for unrelated medical illness (see
eTable in Supplement 2).
Pharmacokinetic Assessments
Blood samples were collected at 90 and 120 minutes after
initiation of ketamine infusion to assay
plasma concentrations of ketamine and metabolites.24 Blood
samples were not collected for
technical reasons during the approximately 80-minute 1H MRS
acquisition. We used the sum of the
2 ketamine and metabolite levels to get a more stable
measurement of their plasma concentrations

across participants because ketamine is metabolized rapidly and
variably.42
Clinical Assessments
The 24-item HDRS was administered within 60 minutes before,
230 minutes after, and 24 hours
after initiation of ketamine infusion. We modified the 24-item
HDRS, excluding items 16 (loss of
weight) and 18 (diurnal variation) because a change in these
measures could not be assessed within
24 hours. Response was defined as percentage change on this
22-item HDRS 24 hours after ketamine
administration.
Magnetic Resonance Neuroimaging Scans
All the neuroimaging data were acquired on a GE Signa
EXCITE 3.0T MR scanner equipped with an
8-channel surface coil, as previously described.24 Briefly, a 3-
plane localizer imaging series was
obtained, followed by a 2-dimensional fast spoiled gradient-
recalled echo MRI scan in sagittal planes
(echo time, 2.1 milliseconds; repetition time, 75 milliseconds;
flip angle, 75°; field of view, 256 × 256
mm2; slice thickness, 5 mm; 8 slices) and a volumetric T1-
weighted spoiled gradient-recalled MRI scan
prescribed in the oblique axial planes parallel to the anterior
commissure–posterior commissure line
(echo time, 2.86 milliseconds; repetition time, 7.12
milliseconds; flip angle, 9°; field of view,
256 × 256 mm2; image matrix size, 256 × 256; slice thickness,
1 mm; voxel size, 1 × 1 × 1 mm3). A
voxel of 3.0 × 2.5 × 2.5 cm3 was placed in the ventral mPFC
region based on the sagittal and oblique
axial images, with the center of the posterior side of the voxel
close to the front tip of the cingulate
gyrus (eFigures 1A and 1B in Supplement 2). In vivo brain

spectra of the combined resonance of Glx
and GABA were recorded from the voxel using the standard J-
edited spin echo difference
method.43,44 Data were acquired as six 13-minute frames. The
levels of Glx and GABA in the edited
spectra were fitted in the frequency domain as previously
described (eFigure 1C in
Supplement 2)43,44 and then expressed as peak area ratio
relative to the synchronously acquired and
similarly fitted unsuppressed voxel water signal—a commonly
used45-49 method with reasonable
test-retest reliability.44 Data from 1 participant were excluded
from further analysis because of poor-
quality spectra. Details of the 1H MRS data quality assessment
criteria used to retain or reject spectra
for inclusion in group analyses can be found in previously
published supplemental online material.44
Statistical Analysis
Ketamine, norketamine, and dehydronorketamine blood levels
(ng/mL) were measured as the mean
of the levels at 90 and 120 minutes. Levels of Glx and GABA in
the mPFC were measured as the area
under the curve from 0 to approximately 42 minutes following
initiation of ketamine infusion. Clinical
improvement was measured as the percentage change in HDRS-
22 score from baseline to 24 hours
after dose. Psychotomimetic effects were measured with the
BPRS total score.
To examine the effect of ketamine dose on clinical
improvement, we fit a simple linear
regression model. Next, we fit separate simple linear regression
models to test the effect of ketamine
dose on ketamine blood level, Glx level, and GABA level. With
similar models, we explored the effect

of ketamine blood level on Glx level and on GABA level. Next,
we fit separate simple linear regression
models with clinical improvement as outcome and ketamine
blood level, Glx level, and GABA level
as independent variables. Finally, to determine whether Glx
level mediates the effect of ketamine
dose on clinical improvement, we also fit a multiple regression
model with clinical improvement as
Depression
12, 2020 5/14
01/21/2021
outcome and both ketamine dose and Glx level as independent
variables.50-52 For each analysis we

reported the test statistic value with degrees of freedom, P
value, and parameter estimate with its
95% CI.
In the analysis of the adverse effects data, we fit separate
simple linear regression models with
psychotomimetic effects as outcome and as predictors: blood
ketamine level, Glx, GABA,
norketamine, dehydronorketamine, and clinical improvement.
These analyses were repeated
separately for men and women. Data analysis was conducted in
R version 3.6.3 (R Project for
Statistical Computing) from January to March 2020. Statistical
significance was set at P < .05, and all
test were 2-tailed but not corrected for multiple comparisons
over the entire analytic set.
Results
A total of 38 individuals participated in the study, with a mean
(SD) age of 38.6 (11.2) years, 23
(60.5%) women, and 25 (65.8%) White patients. The Table
describes the patient population
demographic characteristics, depression severity, and clinical
response to each ketamine dose. The
ketamine dose and placebo treatment groups did not differ
statistically in demographic
characteristics, baseline depression severity or duration of
current episode, number of lifetime
episodes, or years since onset of MDD. Notably, the entire
patient sample had a long duration of their
current major depressive episode.
Effects of Ketamine IV Dose and Blood Level on Clinical
Improvement
Ketamine dose effects on clinical improvement are summarized

in the Table and Figure 2A. Injected
dose of ketamine had a positive relationship with clinical
improvement (t36 = 2.81; P = .008; slope
estimate, 19.80 [95% CI, 5.49 to 34.11]), as did the ketamine
blood level (t36 = 2.25; P = .03; slope
estimate, 0.070 [95% CI, 0.007 to 0.133]). Metabolite levels did
not correlate statistically with
clinical improvement (norketamine: t35 = 1.88; P = .07; slope
estimate, 0.07 [95% CI, −0.006 to
0.146]; dehydronorketamine: t35 = 1.13; P = .27; slope
estimate, 0.075 [95% CI, −0.060 to 0.211]).
Relationship of Ketamine Dose With Blood Level
Ketamine dose correlated positively with ketamine blood level
(t36 = 12.08; P < .001; slope estimate,
10.5 [95% CI, 175.2 to 245.9]). The results of this test are
illustrated in eFigure 2 in Supplement 2.
Figure 2. Changes in Modified 22-Item Hamilton Depression
Rating Scale (HDRS) Score 24 Hours After Ketamine
Intravenous Dose
–55
5
Ch
an
ge
in
sc
or
e,

% –15
–35
Predose score Day 1 score
Percentage change in 22-item HDRS total scoreA
10
0
Ch
an
ge
in
H
DR
S
to
ta
l s
co
re
–10
–20
–30

–8.8 –8.5 –8.3 –8.0 –7.8 –7.5
Change in Glx level
Association of change in 22-item HDRS total score
and Glx level change
B
10
0
Ch
an
ge
in
H
DR
S
to
ta
l s
co
re
–10
–20
–30

–8.0 –7.8 –7.6 –7.4 –7.2
Change in GABA level
Association of change in 22-item HDRS total score
and GABA level change
C
0.0
Ketamine dose, mg/kg
0.1
0.2
0.3
0.4
0.5
GABA indicates γ-aminobutyric acid; Glx, glutamate and
glutamine.
Depression
12, 2020 6/14
01/21/2021
https://jama.jamanetwork.com/ar ticle.aspx?doi=10.1001/jamanet

Effects of Ketamine Blood Level on Glx and GABA Levels
The injected dose of ketamine showed a negative relationship
with Glx (t33 = −4.120; P < .001; slope
estimate, −1.088 [95% CI, −1.626 to −0.551]) and, in a separate
model, with GABA (t33 = −4.450;
P < .001; slope estimate, −0.714 [95% CI, −1.041 to −0.388]).
The ketamine blood level also had a
negative relationship with Glx (t33 = −4.087; P < .001; slope
estimate, −0.005 [95% CI, −0.007 to
−0.002]) (Figure 3B) and GABA (t33 = −4.334; P < .001; slope
estimate, −0.003 [95% CI, −0.004 to
−0.002]) (Figure 3C).
Effects of Glx and GABA on Clinical Improvement
We found a positive relationship between Glx and clinical
improvement (t33 = −2.400; P = .02; slope
estimate, −9.85 [95% CI, −18.2 to −1.50]) (Figure 2B). When
both ketamine blood level and Glx were
used as predictors in a multiple regression analysis, neither the
effect of Glx on clinical improvement
(t32 = −1.417; P = .17; slope estimate, −7.153 [95% CI,
−17.436 to 3.130]) nor the effect of ketamine
blood level on clinical improvement (t32 = 0.922; P = .36; slope
estimate, 0.037 [95% CI, −0.045 to
0.119]) remained significant. Similarly, when both injected
ketamine dose and Glx were used in a
multiple regression analysis, neither the effect of Glx on
clinical improvement (t32 = −1.103; P = .28;
slope estimate, −5.465 [95% CI, −15.554 to 4.624]) nor the
effect of injected ketamine dose on
clinical improvement (t32 = 1.519; P = .14; slope estimate,
14.053 [95% CI, −4.787 to 32.893])

remained significant. Although Glx had a significant positive
relationship with GABA (t33 = 8.117;
P < .001; slope estimate, 0.510 [95% CI, 0.382 to 0.638])
(Figure 3A), GABA was not associated with
clinical improvement (t33 = −1.552; P = .13; slope estimate,
−10.67 [95% CI, −24.66 to 3.32])
(Figure 2C).
Adverse Effects
Psychotomimetic effects were measured with the BPRS.
Ketamine blood level was not related to
BPRS score when analyzing the full sample (t23 = 1.084; P =
.29; estimated slope, 0.29 [95% CI,
−0.26 to 0.084]). In men only, ketamine blood level had a
positive relationship with BPRS score
(t5 = 2.606; P = .048; estimated slope, 0.093 [95% CI, 0.001 to
0.186]), while there was no effect in
women (t16 = −0.362; P = .72; estimated slope, −0.013 [95%
CI, −0.086 to 0.061]). In men, both
norketamine and dehydronorketamine blood levels had a
positive relationship with BPRS score
(norketamine: t5 = 3.944; P = .01; estimated slope, 0.139 [95%
CI, 0.049 to 0.230];
dehydronoketamine: t5 = 2.589; P = .049; estimated slope,
0.167 [95% CI, 0.001 to 0.334]).
Figure 3. Correlations Between Glutamate and Glutamine (Glx),
γ-Aminobutyric Acid (GABA), and Ketamine Blood Levels
–7.2
GA
BA
le
ve

l c
ha
ng
e
–7.4
–7.6
–7.8
–8.0
0 20 40 60 80 100 120 140 160
Correlation between GABA level change and
ketamine blood level
C
–7.2
GA
BA
le
ve
l c
ha
ng
e

–7.4
–7.6
–7.8
–8.0
–8.8 –8.5 –8.3 –8.0 –7.8 –7.5
Glx level change
Correlation between GABA and Glx level changes A
0 20 40 60 80 100 120 140 160
–7.6
–8.0
–7.8
Gl
x
le
ve
l c
ha
ng
e
–8.3

–8.5
–8.8
Ketamine blood level, ng/mL Ketamine blood level, ng/mL
Correlation between Glx level change and
ketamine blood level
B
0.0
Ketamine dose, mg/kg
0.1
0.2
0.3
0.4
0.5
Level change was measured as area under the curve 0 to
approximately 42 minutes following initiation of ketamine
infusion.
Depression
12, 2020 7/14
01/21/2021

We did not find a correlation between Glx response and BPRS
score in the full sample
(t22 = −0.015; P = .99; estimated slope, −0.054 [95% CI, −7.35
to 7.24]), in men only (t5 = −0.051;
P = .96; estimated slope, −0.401 [95% CI, −20.42 to 19.62]), or
in women only (t15 = 0.156; P = .88;
estimated slope, 0.630 [95% CI, −7.98 to 9.24]). GABA did not
correlate with BPRS score (data
not shown).
Clinical improvement was not related to BPRS score (t23 =
−1.01; P = .32; estimated slope, −0.32
[95% CI, −0.97 to 0.34]). Other significant adverse effects,
including vomiting, rise in blood pressure
requiring intervention, and so on, did not occur.
Discussion
In this study, we found a positive correlation of ketamine IV
dose and blood level with antidepressant
effect 24 hours after ketamine administration. We also observed
that ketamine produced a dose-
dependent decrease in mPFC Glx level and that a lower mean
Glx level was associated with better
antidepressant response. Although Glx and GABA levels were
correlated, the latter was unrelated to
antidepressant effect, and both were not related to
psychotomimetic adverse effects.
Psychotomimetic side effects correlated with blood levels of
ketamine in men only. Metabolites did
not appear to be statistically related to antidepressant effect but
were correlated with
psychotomimetic adverse effects in men.

Effect of Ketamine Dose on Clinical Improvement
Our finding that IV ketamine dose correlated positively with
improvement in the HDRS-22 score
(Figure 2A, Table) is supported by a meta-analysis53 of 9
ketamine trials—6 of which used the
standard IV dose of 0.5 mg/kg per 40 minutes and 3 of which
used lower doses—that found lower
doses to be less effective. Our findings are also suppor ted by a
multicenter ketamine dose-finding
randomized clinical trial in 99 patients with treatment-resistant
depression,54 which found that 1.0
mg/kg and 0.5 mg/kg doses were effective but 0.1 mg/kg and
0.2 mg/kg doses were not. Although an
American Psychological Association consensus statement
concluded that there are insufficient data
to draw firm conclusions about alternative doses,21 our findings
suggest otherwise. Doses less than
0.3 mg/kg do not appear to be as effective as higher doses.
There is debate regarding whether ketamine’s (2R,6R)-
hydroxynorketamine (HNK) metabolite
may be as or more important than ketamine itself in mediating
the antidepressant effect.28,55,56 We
found ketamine blood level correlated with antidepressant
response but no statistical relationship
for norketamine or dehydronorketamine blood levels and
antidepressant response. We did not assay
HNK metabolites.
Effect of Ketamine Dose on Brain Glx and GABA Levels
The effects of ketamine on brain Glx (Figure 3B) and GABA
(Figure 3C) levels were dose-dependent.
We did not measure glutamate or GABA release but overall
brain levels. A positron emission
tomography study57 measured metabotropic glutamate receptor
subtype 5 (mGluR5) receptor

binding and found that ketamine lowered tracer binding, an
indicator of increased glutamate release
or receptor internalization. Although preclinical rodent
studies23,58,59 have reported a ketamine-
induced increase in glutamate overflow using microdialysis,
that can be unrelated to overall brain
tissue glutamate levels and is not detectable by MRS glutamate
level measurement, which cannot
distinguish between intracellular and extracellular glutamate
levels. It is unclear how ketamine may
affect Glx or GABA levels or release, although the mGluR2
receptor may mediate a glutamate-related
effect.60 Perhaps ketamine blocks autoinhibitory presynaptic
NMDA receptors that regulate
glutamate release,23,58,59,61 and inhibition of this negative
feedback loop alters glutamate release and
production.23 By blocking NMDA receptors on GABAergic
neurons, ketamine may also alter the
inhibitory activity of GABAergic neurons that project to
glutamatergic neurons, leading to
less GABA.23,62
Depression
12, 2020 8/14
01/21/2021
In the present study, Glx was measured as the combined

resonances of glutamate and
glutamine at approximately 3.75 ppm in the edited MR
spectrum.63 Because most of the Glx peak is
owing to glutamate as opposed to glutamine, it is less likely that
any change in Glx is driven by
glutamine, but this remains to be confirmed by future studies.
We were unable to replicate the
increase in Glx (eFigure 3A in Supplement 2) or GABA
(eFigure 3B in Supplement 2) detected in our
open pilot study in MDD.22,23 The biggest increase in Glx
observed in the present study was with
placebo, and our original pilot study did not have a placebo
group. It is possible that the Glx increase
during the scan was a stress response and that increasing doses
of ketamine diminish that stress
response. This interpretation is supported by the finding that
cholecystokinin-induced panic in
healthy volunteers raises Glx bilaterally in the anterior
cingulate.64
Relationship of Glx and GABA Response to Clinical
Improvement
To our knowledge, no previous controlled study has examined
the relationship of ketamine-induced
Glx and GABA effects to clinical response. In this study, the
smaller the increase in Glx, the better the
clinical response. Ketamine blockade of NMDA receptors would
shift the balance of glutamatergic
signaling toward metabotropic glutamate (mGlu) and
AMPA/kainate receptors.62 This effect fits with
preclinical studies that suggest that the antidepressant effects of
ketamine are related to the effect
of glutamate on the AMPA/kainate and/or some of the mGlu
receptors, leading to activation of
downstream targets, such as the neurotrophin and mTOR
signaling pathways, and resulting in the

rapid production of mushroom spines, dendritic arborization and
synaptogenesis, perhaps to restore
and protect neuronal networks that are deficient or
dysfunctional in MDD.30,65,66 Preclinical studies
have also shown that the ketamine-induced activation of the
mTOR pathway and subsequent
increase in protein synthesis and synaptogenesis, are associated
with reduced adenosine
triphosphate to adenosine diphosphate ratio and increased
phosphorylated AMP–activated protein
kinase levels as well as fewer reactive oxygen species measured
as carbonylated or damaged
proteins, all of which suggest that there is indeed an increase in
energy utilization and mitochondrial
energy metabolism.67 A 2018 study68 failed to find that
rapamycin blocked the antidepressant effect
of ketamine, so we are unsure of its mechanism of action in
patients with depression. Mice studies
indicate that ketamine may produce a resilience effect that is
manifested over a longer time frame,69
and through the attenuation of Glx response, we may be
observing the onset of this effect. We did
not find GABA to be related to antidepressant effect.
Effect of Ketamine and Its Metabolites on Psychotomimetic
Adverse Effects
In this study, the severity of ketamine-induced psychotomimetic
adverse effects was related to
plasma concentration of ketamine, norketamine, and
dehydronorketamine in men but not in women.
There may be a sex-based difference in the way ketamine exerts
its psychotomimetic adverse effects.
Women may preferentially metabolize ketamine through the
hydroxynorketamine pathway, as
described in an animal study.28 Glx and GABA levels did not

correlate with psychotomimetic adverse
effects, consistent with the importance of NMDA receptor
antagonism in mediating this adverse
effect. Because psychomimetic side effect severity is also
unrelated to antidepressant efficacy, it may
be possible to develop newer, comparably fast-acting
antidepressants that lack the psychotomimetic
adverse effects of ketamine.
Limitations
This study has limitations. The study sample was small, but the
observed effects were robust. A larger
sample size is needed to evaluate sex differences in adverse
effects. We did not assay HNK
metabolite levels and so cannot comment on a possible
relationship to antidepressant response. Glx
combines glutamate and glutamine, and future studies should try
to measure the separate peaks.
Active suicidal ideation was an exclusion criterion, so we could
not examine the relationship of dose
or MRS indices to suicidal ideation response. Given that this
study was performed on medication-free
research participants, the results need to be reproduced in a
clinical treatment–seeking population.
Depression
12, 2020 9/14
01/21/2021

Conclusions
This study found a robust relationship between ketamine dose
and blood levels with antidepressant
response. We found that ketamine attenuation of Glx response
mediated the relationship between
ketamine dose and antidepressant response. Future research
should seek to identify the
pharmacological effects responsible for Glx response to
ketamine, determining if glutamate response
accounts for the Glx mediation of antidepressant response, and
clarifying the downstream effects
of ketamine. Finding new, comparably fast-acting
antidepressants without psychotomimetic effects
that can be orally administered may be a realistic goal for future
research, given that
psychotomimetic adverse effects did not correlate with
antidepressant or Glx effects of ketamine.
ARTICLE INFORMATION
Accepted for Publication: June 1, 2020.
Published: August 12, 2020.
doi:10.1001/jamanetworkopen.2020.13211
Open Access: This is an open access article distributed under

the terms of the CC-BY License. © 2020 Milak MS et al.
JAMA Network Open.
Corresponding Author: J. John Mann, MD, Department of
Psychiatry, College of Physicians and Surgeons,
Columbia University, NYSPI Unit 42, New York City, NY
10032 ([email protected]).
Author Affiliations: Department of Psychiatry, College of
Physicians and Surgeons, Columbia University, New
York, New York (Milak, Rashid, Dong, Kegeles, Grunebaum,
Mulhern, Suckow, Cooper, Keilp, Mann); Molecular
Imaging and Neuropathology Division, The New York State
Psychiatric Institute, New York (Milak, Rashid, Dong,
Kegeles, Grunebaum, Ogden, Lin, Mulhern, Suckow, Cooper,
Keilp, Mann); Department of Radiology, College of
Physicians and Surgeons, Columbia University, New York, New
York (Kegeles, Mann); Department of Biostatistics,
Mailman School of Public Health, Columbia University, New
York, New York (Ogden, Lin); Analytical
Psychopharmacology Laboratory, The Nathan S. Kline Institute
for Psychiatric Research, Orangeburg, New York
(Suckow, Cooper); Department of Radiology, Weill Cornell
Medicine, New York, New York (Mao, Shungu).
Author Contributions: Dr Milak and Mr Rashid had full access
to all of the data in the study and take responsibility
for the integrity of the data and the accuracy of the data
analysis.
Concept and design: Milak, Kegeles, Mann.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Milak, Rashid, Grunebaum, Ogden,
Lin, Mulhern, Suckow, Cooper, Shungu, Mann.

Critical revision of the manuscript for important intellectual
content: Milak, Rashid, Dong, Kegeles, Keilp, Mao,
Shungu, Mann.
Statistical analysis: Milak, Rashid, Ogden, Lin, Keilp.
Obtained funding: Milak, Mann.
Administrative, technical, or material support: Milak, Rashid,
Grunebaum, Mulhern, Suckow, Cooper, Mao,
Shungu, Mann.
Supervision: Milak, Mann.
Conflict of Interest Disclosures: Dr Kegeles reported receiving
grants from the National Institute of Mental Health
outside the submitted work. Dr Ogden reported receiving grants
from the National Institutes of Health during the
conduct of the study and outside the submitted work. Dr Cooper
reported receiving grants from the National
Institute for Mental Health during the conduct of the study. Dr
Mann reported receiving grants from the National
Institute of Mental Health during the conduct of the study and
having a patent to the Columbia Suicide Severity
Rating Scale, with royalties paid. No other disclosures were
reported.
Funding/Support: Support for this study was provided by grant
5R01MH093637 from the National Institute of
Mental Health (Dr Milak).
Role of the Funder/Sponsor: The funder had no role in the
design and conduct of the study; collection,
management, analysis, and interpretation of the data;
preparation, review, or approval of the manuscript; and

decision to submit the manuscript for publication.
Data Sharing Statement: See Supplement 3.
Depression
12, 2020 10/14
01/21/2021
https://jamanetwork.com/journals/jamanetworkopen/pages/instr
uctions-for-
authors#SecOpenAccess/?utm_campaign=articlePDF%26utm_m
edium=articlePDFlink%26utm_source=articlePDF%26utm_cont
ent=jamanetworkopen.2020.13211
mailto:[email protected]
REFERENCES
1. World Health Organization. Depression and other common
mental disorders: global health estimates. Accessed
July 13, 2020.
https://www.who.int/mental_health/management/depression/pre

valence_global_health_estimates/en/
2. National Institute of Mental Health. Major Depression
Among Adults. National Institute of Mental Health; 2015.
3. Greenberg PE, Fournier AA, Sisitsky T, Pike CT, Kessler RC.
The economic burden of adults with major depressive
disorder in the United States (2005 and 2010). J Clin
Psychiatry. 2015;76(2):155-162. doi:10.4088/JCP.14m09298
4. Wisniewski SR, Rush AJ, Nierenberg AA, et al. Can phase III
trial results of antidepressant medications be
generalized to clinical practice? a STAR*D report. Am J
Psychiatry. 2009;166(5):599-607. doi:10.1176/appi.ajp.
2008.08071027
5. Fava M, Davidson KG. Definition and epidemiology of
treatment-resistant depression. Psychiatr Clin North Am.
1996;19(2):179-200. doi:10.1016/S0193-953X(05)70283-5
6. Akil H, Gordon J, Hen R, et al. Treatment resistant
depression: a multi-scale, systems biology approach. Neurosci
Biobehav Rev. 2018;84:272-288.
doi:10.1016/j.neubiorev.2017.08.019
7. Berman RM, Cappiello A, Anand A, et al. Antidepressant
effects of ketamine in depressed patients. Biol
Psychiatry. 2000;47(4):351-354. doi:10.1016/S0006-
3223(99)00230-9
8. Phelps LE, Brutsche N, Moral JR, Luckenbaugh DA, Manji
HK, Zarate CA Jr. Family history of alcohol
dependence and initial antidepressant response to an N-methyl-
D-aspartate antagonist. Biol Psychiatry. 2009;65
(2):181-184. doi:10.1016/j.biopsych.2008.09.029

9. Mathew SJ, Murrough JW, aan het Rot M, Collins KA, Reich
DL, Charney DS. Riluzole for relapse prevention
following intravenous ketamine in treatment-resistant
depression: a pilot randomized, placebo-controlled
continuation trial. Int J Neuropsychopharmacol. 2010;13(1) :71-
82. doi:10.1017/S1461145709000169
10. Kudoh A, Takahira Y, Katagai H, Takazawa T. Small-dose
ketamine improves the postoperative state of
depressed patients. Anesth Analg. 2002;95(1):114-118.
doi:10.1097/00000539-200207000-00020
11. Zarate CA Jr, Singh JB, Carlson PJ, et al. A randomized
trial of an N-methyl-D-aspartate antagonist in treatment-
resistant major depression. Arch Gen Psychiatry.
2006;63(8):856-864. doi:10.1001/archpsyc.63.8.856
12. Price RB, Nock MK, Charney DS, Mathew SJ. Effects of
intravenous ketamine on explicit and implicit measures
of suicidality in treatment-resistant depression. Biol Psychiatry.
2009;66(5):522-526. doi:10.1016/j.biopsych.
2009.04.029
13. Hu YD, Xiang YT, Fang JX, et al. Single IV ketamine
augmentation of newly initiated escitalopram for major
depression: results from a randomized, placebo-controlled 4-
week study. Psychol Med. 2016;46(3):623-635. doi:
10.1017/S0033291715002159
14. Singh JB, Fedgchin M, Daly E, et al. Intravenous
esketamine in adult treatment-resistant depression: a double-
blind, double-randomization, placebo-controlled study. Biol
Psychiatry. 2016;80(6):424-431. doi:10.1016/j.
biopsych.2015.10.018
15. Singh JB, Fedgchin M, Daly EJ, et al. A double-blind,

randomized, placebo-controlled, dose-frequency study of
intravenous ketamine in patients with treatment-resistant
depression. Am J Psychiatry. 2016;173(8):816-826.
doi:10.1176/appi.ajp.2016.16010037
16. Murrough JW, Iosifescu DV, Chang LC, et al.
Antidepressant efficacy of ketamine in treatment-resistant major
depression: a two-site randomized controlled trial. Am J
Psychiatry. 2013;170(10):1134-1142. doi:10.1176/appi.ajp.
2013.13030392
17. DiazGranados N, Ibrahim LA, Brutsche NE, et al. Rapid
resolution of suicidal ideation after a single infusion of
an N-methyl-D-aspartate antagonist in patients with treatment-
resistant major depressive disorder. J Clin
Psychiatry. 2010;71(12):1605-1611.
doi:10.4088/JCP.09m05327blu
18. Grunebaum MF, Galfalvy HC, Choo TH, et al. Ketamine for
rapid reduction of suicidal thoughts in major
depression: a midazolam-controlled randomized clinical trial.
Am J Psychiatry. 2018;175(4):327-335. doi:10.1176/
appi.ajp.2017.17060647
19. Grunebaum MF, Ellis SP, Keilp JG, et al. Ketamine versus
midazolam in bipolar depression with suicidal
thoughts: a pilot midazolam-controlled randomized clinical
trial. Bipolar Disord. 2017;19(3):176-183. doi:10.1111/
bdi.12487
20. Newport DJ, Carpenter LL, McDonald WM, Potash JB,
Tohen M, Nemeroff CB; APA Council of Research Task
Force on Novel Biomarkers and Treatments. Ketamine and other
NMDA antagonists: early clinical trials and
possible mechanisms in depression. Am J Psychiatry.
2015;172(10):950-966. doi:10.1176/appi.ajp.2015.15040465

21. Sanacora G, Frye MA, McDonald W, et al; American
Psychiatric Association (APA) Council of Research Task
Force on Novel Biomarkers and Treatments. A consensus
statement on the use of ketamine in the treatment of
mood disorders. JAMA Psychiatry. 2017;74(4):399-405.
doi:10.1001/jamapsychiatry.2017.0080
Depression
12, 2020 11/14
01/21/2021
https://www.who.int/mental_health/management/depression/pre
valence_global_health_estimates/en/
https://dx.doi.org/10.4088/JCP.14m09298
https://dx.doi.org/10.1176/appi.ajp.2008.08071027
https://dx.doi.org/10.1016/S0193-953X(05)70283-5
https://dx.doi.org/10.1016/j.neubiorev.2017.08.019
https://dx.doi.org/10.1016/S0006-3223(99)00230-9
https://dx.doi.org/10.1016/j.biopsych.2008.09.029
https://dx.doi.org/10.1017/S1461145709000169
https://dx.doi.org/10.1097/00000539-200207000-00020
https://jama.jamanetwork.com/article.aspx?doi=10.1001/archpsy
c.63.8.856&utm_campaign=articlePDF%26utm_medium=article
PDFlink%26utm_source=articlePDF%26utm_content=jamanetw
orkopen.2020.13211

https://dx.doi.org/10.1017/S0033291715002159
https://dx.doi.org/10.4088/JCP.09m05327blu
https://dx.doi.org/10.1111/bdi.12487
https://dx.doi.org/10.1111/bdi.12487
https://jama.jamanetwork.com/article.aspx?doi=10.1001/jamaps
ychiatry.2017.0080&utm_campaign=articlePDF%26utm_mediu
m=articlePDFlink%26utm_source=articlePDF%26utm_content=j
amanetworkopen.2020.13211
22. Kim SY, Lee H, Kim HJ, et al. In vivo and ex vivo evidence
for ketamine-induced hyperglutamatergic activity in
the cerebral cortex of the rat: potential relevance to
schizophrenia. NMR Biomed. 2011;24(10):1235-1242. doi:10.
1002/nbm.1681
23. Moghaddam B, Adams B, Verma A, Daly D. Activation of
glutamatergic neurotransmission by ketamine:
a novel step in the pathway from NMDA receptor blockade to
dopaminergic and cognitive disruptions associated
with the prefrontal cortex. J Neurosci. 1997;17(8):2921-2927.
doi:10.1523/JNEUROSCI.17-08-02921.1997
24. Milak MS, Proper CJ, Mulhern ST, et al. A pilot in vivo
proton magnetic resonance spectroscopy study of amino
acid neurotransmitter response to ketamine treatment of major
depressive disorder. Mol Psychiatry. 2016;21(3):
320-327. doi:10.1038/mp.2015.83

25. Koike H, Chaki S. Requirement of AMPA receptor
stimulation for the sustained antidepressant activity of
ketamine and LY341495 during the forced swim test in rats.
Behav Brain Res. 2014;271:111-115. doi:10.1016/j.bbr.
2014.05.065
26. Koike H, Iijima M, Chaki S. Involvement of AMPA receptor
in both the rapid and sustained antidepressant-like
effects of ketamine in animal models of depression. Behav
Brain Res. 2011;224(1):107-111. doi:10.1016/j.bbr.2011.
05.035
27. Maeng S, Zarate CA Jr, Du J, et al. Cellular mechanisms
underlying the antidepressant effects of ketamine: role
of alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid
receptors. Biol Psychiatry. 2008;63(4):349-352.
doi:10.1016/j.biopsych.2007.05.028
28. Zanos P, Moaddel R, Morris PJ, et al. NMDAR inhibition-
independent antidepressant actions of ketamine
metabolites. Nature. 2016;533(7604):481-486.
doi:10.1038/nature17998
29. Yang C, Hu Y-M, Zhou Z-Q, Zhang G-F, Yang J-J. Acute
administration of ketamine in rats increases
hippocampal BDNF and mTOR levels during forced swimming
test. Ups J Med Sci. 2013;118(1):3-8. doi:10.3109/
03009734.2012.724118
30. Li N, Lee B, Liu RJ, et al. mTOR-dependent synapse
formation underlies the rapid antidepressant effects of
NMDA antagonists. Science. 2010;329(5994):959-964.
doi:10.1126/science.1190287
31. Kavalali ET, Monteggia LM. Synaptic mechanisms

underlying rapid antidepressant action of ketamine. Am J
Psychiatry. 2012;169(11):1150-1156.
doi:10.1176/appi.ajp.2012.12040531
32. Monteggia LM, Gideons E, Kavalali ET. The role of
eukaryotic elongation factor 2 kinase in rapid antidepressant
action of ketamine. Biol Psychiatry. 2013;73(12):1199-1203.
doi:10.1016/j.biopsych.2012.09.006
33. Bhagwagar Z, Wylezinska M, Jezzard P, et al. Low GABA
concentrations in occipital cortex and anterior
cingulate cortex in medication-free, recovered depressed
patients. Int J Neuropsychopharmacol. 2008;11(2):
255-260. doi:10.1017/S1461145707007924
34. Hasler G, van der Veen JW, Tumonis T, Meyers N, Shen J,
Drevets WC. Reduced prefrontal glutamate/
glutamine and gamma-aminobutyric acid levels in major
depression determined using proton magne tic resonance
spectroscopy. Arch Gen Psychiatry. 2007;64(2):193-200.
doi:10.1001/archpsyc.64.2.193
35. Sanacora G, Gueorguieva R, Epperson CN, et al. Subtype-
specific alterations of gamma-aminobutyric acid and
glutamate in patients with major depression. Arch Gen
Psychiatry. 2004;61(7):705-713. doi:10.1001/archpsyc.
61.7.705
36. Sanacora G, Mason GF, Rothman DL, et al. Reduced
cortical gamma-aminobutyric acid levels in depressed
patients determined by proton magnetic resonance spectroscopy.
Arch Gen Psychiatry. 1999;56(11):1043-1047.
doi:10.1001/archpsyc.56.11.1043
37. Song Z, Huang P, Qiu L, et al. Decreased occipital GABA
concentrations in patients with first-episode major

depressive disorder: a magnetic resonance spectroscopy study.
Article in Chinese. Sheng Wu Yi Xue Gong Cheng
Xue Za Zhi. 2012;29(2):233-236.
38. Kobak KA, Leuchter A, DeBrota D, et al. Site versus
centralized raters in a clinical depression trial: impact on
patient selection and placebo response. J Clin Psychopharmacol.
2010;30(2):193-197. doi:10.1097/JCP.
0b013e3181d20912
39. Kobak KA, Mundt JC, Greist JH, Katzelnick DJ, Jefferson
JW. Computer assessment of depression: automating
the Hamilton Depression Rating Scale. Drug Inf J.
2000;34(1):145-156. doi:10.1177/009286150003400120
40. Kobak KA. Inaccuracy in clinical trials: effects and methods
to control inaccuracy. Curr Alzheimer Res. 2010;7
(7):637-641. doi:10.2174/156720510793499057
41. Sterne JAC, Savović J, Page MJ, et al. RoB 2: a revised tool
for assessing risk of bias in randomised trials. BMJ.
2019;366:l4898. doi:10.1136/bmj.l4898
42. Zanos P, Moaddel R, Morris PJ, et al. Ketamine and
ketamine metabolite pharmacology: insights into
therapeutic mechanisms. Pharmacol Rev. 2018;70(3):621-660.
doi:10.1124/pr.117.015198
Depression
12, 2020 12/14

01/21/2021
https://dx.doi.org/10.1002/nbm.1681
https://dx.doi.org/10.1523/JNEUROSCI.17-08-02921.1997
https://dx.doi.org/10.1038/mp.2015.83
https://dx.doi.org/10.1016/j.bbr.2014.05.065
https://dx.doi.org/10.1038/nature17998
https://dx.doi.org/10.3109/03009734.2012 .724118
https://dx.doi.org/10.3109/03009734.2012.724118
https://dx.doi.org/10.1126/science.1190287
https://dx.doi.org/10.1017/S1461145707007924
orkopen.2020.13211
orkopen.2020.13211
orkopen.2020.13211
c.56.11.1043&utm_campaign=articlePDF%26utm_medium=artic
lePDFlink%26utm_source=articlePDF%26utm_content=jamanet
workopen.2020.13211
https://www.ncbi.nlm.nih.gov/pubmed/22616164

https://dx.doi.org/10.1097/JCP.0b013e3181d20912
https://dx.doi.org/10.1097/JCP.0b013e3181d20912
https://dx.doi.org/10.1177/009286150003400120
https://dx.doi.org/10.2174/156720510793499057
https://dx.doi.org/10.1136/bmj.l4898
https://dx.doi.org/10.1124/pr.117.015198
43. Rothman DL, Petroff OA, Behar KL, Mattson RH. Localized
1H NMR measurements of gamma-aminobutyric
acid in human brain in vivo. Proc Natl Acad Sci U S A.
1993;90(12):5662-5666. doi:10.1073/pnas.90.12.5662
44. Shungu DC, Mao X, Gonzales R, et al. Brain γ-aminobutyric
acid (GABA) detection in vivo with the J-editing (1)
H MRS technique: a comprehensive methodological evaluation
of sensitivity enhancement, macromolecule
contamination and test-retest reliability. NMR Biomed.
2016;29(7):932-942. doi:10.1002/nbm.3539
45. Murck H. Ketamine, magnesium and major depression—
from pharmacology to pathophysiology and back.
J Psychiatr Res. 2013;47(7):955-965.
doi:10.1016/j.jpsychires.2013.02.015
46. Douglas BG, Dagirmanjian R. The effects of magnesium
deficiency of ketamine sleeping times in the rat. Br J
Anaesth. 1975;47(3):336-340. doi:10.1093/bja/47.3.336
47. Wilson K, Brakoulias V. Magnesium intake and depression.
Aust N Z J Psychiatry. 2009;43(6):580.
48. Derom ML, Sayón-Orea C, Martínez-Ortega JM, Martínez-
González MA. Magnesium and depression:
a systematic review. Nutr Neurosci. 2013;16(5):191-206.

doi:10.1179/1476830512Y.0000000044
49. Derom ML, Martínez-González MA, Sayón-Orea MdelC,
Bes-Rastrollo M, Beunza JJ, Sánchez-Villegas A.
Magnesium intake is not related to depression risk in Spanish
university graduates. J Nutr. 2012;142(6):1053-1059.
doi:10.3945/jn.111.155572
50. James LR, Brett JM. Mediators, moderators, and tests for
mediation. J Appl Psychol. 1984;69(2):307-321. doi:10.
1037/0021-9010.69.2.307
51. Judd CM, Kenny DA. Process analysis—estimating
mediation in treatment evaluations. Eval Rev. 1981;5(5):
602-619. doi:10.1177/0193841X8100500502
52. Baron RM, Kenny DA. The moderator-mediator variable
distinction in social psychological research:
conceptual, strategic, and statistical considerations. J Pers Soc
Psychol. 1986;51(6):1173-1182. doi:10.1037/0022-
3514.51.6.1173
53. Xu Y, Hackett M, Carter G, et al. Effects of low-dose and
very low-dose ketamine among patients with major
depression: a systematic review and meta-analysis. Int J
Neuropsychopharmacol. 2016;19(4):pyv124. doi:10.1093/
ijnp/pyv124
54. Fava M, Freeman MP, Flynn M, et al. Double-blind,
placebo-controlled, dose-ranging trial of intravenous
ketamine as adjunctive therapy in treatment-resistant depression
(TRD). Mol Psychiatry. 2018. doi:10.1038/
s41380-018-0256-5
55. Suzuki K, Nosyreva E, Hunt KW, Kavalali ET, Monteggia
LM. Effects of a ketamine metabolite on synaptic

NMDAR function. Nature. 2017;546(7659):E1-E3.
doi:10.1038/nature22084
56. Pham TH, Defaix C, Xu X, et al. Common
neurotransmission recruited in (R,S)-ketamine and (2R,6R)-
hydroxynorketamine-induced sustained antidepressant-like
effects. Biol Psychiatry. 2018;84(1):e3-e6. doi:10.
1016/j.biopsych.2017.10.020
57. Esterlis I, DellaGioia N, Pietrzak RH, et al. Ketamine-
induced reduction in mGluR5 availability is associated with
an antidepressant response: an [11C]ABP688 and PET imaging
study in depression. Mol Psychiatry. 2018;23(4):
824-832. doi:10.1038/mp.2017.58
58. Lorrain DS, Baccei CS, Bristow LJ, Anderson JJ, Varney
MA. Effects of ketamine and N-methyl-D-aspartate on
glutamate and dopamine release in the rat prefrontal cortex:
modulation by a group II selective metabotropic
glutamate receptor agonist LY379268. Neuroscience.
2003;117(3):697-706. doi:10.1016/S0306-4522(02)
00652-8
59. Razoux F, Garcia R, Léna I. Ketamine, at a dose that
disrupts motor behavior and latent inhibition, enhances
prefrontal cortex synaptic efficacy and glutamate release in the
nucleus accumbens. Neuropsychopharmacology.
2007;32(3):719-727. doi:10.1038/sj.npp.1301057
60. Krystal JH, Sanacora G, Duman RS. Rapid-acting
glutamatergic antidepressants: the path to ketamine and
beyond. Biol Psychiatry. 2013;73(12):1133-1141.
doi:10.1016/j.biopsych.2013.03.026
61. Duguid I, Sjöström PJ. Novel presynaptic mechanisms for
coincidence detection in synaptic plasticity. Curr Opin

Neurobiol. 2006;16(3):312-322. doi:10.1016/j.conb.2006.05.008
62. Duman RS, Li N, Liu RJ, Duric V, Aghajanian G. Signaling
pathways underlying the rapid antidepressant actions
of ketamine. Neuropharmacology. 2012;62(1):35-41.
doi:10.1016/j.neuropharm.2011.08.044
63. Ramadan S, Lin A, Stanwell P. Glutamate and glutamine: a
review of in vivo MRS in the human brain. NMR
Biomed. 2013;26(12):1630-1646. doi:10.1002/nbm.3045
64. Zwanzger P, Zavorotnyy M, Gencheva E, et al. Acute shift
in glutamate concentrations following
experimentally induced panic with cholecystokinin
tetrapeptide—a 3T-MRS study in healthy subjects.
Neuropsychopharmacology. 2013;38(9):1648-1654.
doi:10.1038/npp.2013.61
Depression
12, 2020 13/14
01/21/2021
https://dx.doi.org/10.1073/pnas.90.12.5662
https://dx.doi.org/10.1016/j.jpsychires.2013.02.015
https://dx.doi.org/10.1093/bja/47.3.336
https://dx.doi.org/10.1179/1476830512Y.0000000044
https://dx.doi.org/10.3945/jn.111.155572

https://dx.doi.org/10.1037/0021-9010.69.2.307
https://dx.doi.org/10.1037/0021-9010.69.2.307
https://dx.doi.org/10.1177/0193841X8100500502
https://dx.doi.org/10.1037/0022-3514.51.6.1173
https://dx.doi.org/10.1037/0022-3514.51.6.1173
https://dx.doi.org/10.1093/ijnp/pyv124
https://dx.doi.org/10.1093/ijnp/pyv124
https://dx.doi.org/10.1038/s41380-018-0256-5
https://dx.doi.org/10.1038/s41380-018-0256-5
https://dx.doi.org/10.1038/nature22084
https://dx.doi.org/10.1038/mp.2017.58
https://dx.doi.org/10.1016/S0306-4522(02)00652-8
https://dx.doi.org/10.1016/S0306-4522(02)00652-8
https://dx.doi.org/10.1038/sj.npp.1301057
https://dx.doi.org/10.1016/j.conb.2006.05.008
https://dx.doi.org/10.1016/j.neuropharm.2011.08.044
https://dx.doi.org/10.1038/npp.2013.61
65. Moda-Sava RN, Murdock MH, Parekh PK, et al. Sustained
rescue of prefrontal circuit dysfunction by
antidepressant-induced spine formation. Science.
2019;364(6436):eaat8078. doi:10.1126/science.aat8078
66. Dong C, Zhang JC, Yao W, et al. Rapid and sustained
antidepressant action of the mGlu2/3 Receptor antagonist
MGS0039 in the social defeat stress model: comparison with
ketamine. Int J Neuropsychopharmacol. 2017;20(3):
228-236.
67. Weckmann K, Deery MJ, Howard JA, et al. Ketamine’s
antidepressant effect is mediated by energy metabolism

and antioxidant defense system. Sci Rep. 2017;7(1):15788.
doi:10.1038/s41598-017-16183-x
68. Yang C, Ren Q, Qu Y, et al. Mechanistic target of
rapamycin-independent antidepressant effects of
(R)-ketamine in a social defeat stress model. Biol Psychiatry.
2018;83(1):18-28. doi:10.1016/j.biopsych.2017.05.016
69. Brachman RA, McGowan JC, Perusini JN, et al. Ketamine as
a prophylactic against stress-induced depressive-
like behavior. Biol Psychiatry. 2016;79(9):776-786.
doi:10.1016/j.biopsych.2015.04.022
SUPPLEMENT 1.
Trial Protocol
SUPPLEMENT 2.
eFigure 1. Proton Magnetic Resonance Spectroscopy (1H-MRS)
Example Voxel Placement and J-Edited Spectra
eFigure 2. Plasma Concentration of Ketamine
eFigure 3. Change in Glx and GABA Levels Following
Initiation of Ketamine Infusion, Corrected for Internal Water
and Expressed as a Percentage of Preketamine Glx
eTable. Serious Adverse Events Occurring During Study
Participation
SUPPLEMENT 3.
Data Sharing Statement
Depression
12, 2020 14/14

01/21/2021
https://dx.doi.org/10.1126/science.aat8078
https://dx.doi.org/10.1038/s41598-017-16183-x
184 | wileyonlinelibrary.com/journal/jocn J Clin Nurs.
2020;29:184–194.© 2019 John Wiley & Sons Ltd
Received: 10 June 2019 | Revised: 30 August 2019 |
Accepted: 29 September 2019
DOI: 10.1111/jocn.15076
O R I G I N A L A R T I C L E
Sleep and fatigue in newly graduated nurses—Experiences and
strategies for handling shiftwork
Majken Epstein BSc1 | Marie Söderström PhD, Licensed
Psychologist1,2 |
Maria Jirwe PhD, Associate Professor, Registered Nurse3,4 |
Philip Tucker PhD,
Associate Professor5,6 | Anna Dahlgren PhD1
1Division of Psychology, Department of
Clinical Neuroscience, Karolinska Institutet,
Stockholm, Sweden
2Stressmottagningen/Stress clinic,

Stockholm, Sweden
3Department for health promoting
science, Sophiahemmet University,
Stockholm, Sweden
4Division of Nursing, Department
of Neurobiology, Care Sciences and
Society, Karolinska Institutet, Huddinge,
Sweden
5Stress Research Institute, Stockholm
University, Stockholm, Sweden
6Department of Psychology, Swansea
University, Swansea, UK
Correspondence
Anna Dahlgren, Department of Clinical
Neuroscience, Division of Psychology,
Karolinska Institutet, Nobels väg 9, 171 65
Solna, Stockholm, Sweden.
Email: [email protected]
Funding information
AFA Försäkring, Grant/Award Number:
150024
Abstract
Aims and objectives: To explore newly graduated nurses'
strategies for, and experi-
ences of, sleep problems and fatigue when starting shiftwork. A
more comprehensive
insight into nurses' strategies, sleep problems, fatigue
experiences and contributing
factors is needed to understand what support should be
provided.
Background: For graduate nurses, the first years of practice are
often stressful, with
many reporting high levels of burnout symptoms. Usually,
starting working as a nurse

also means an introduction to shiftwork, which is related to
sleep problems. Sleep
problems may impair stress management and, at the same time,
stress may cause
sleep problems. Previously, sleep problems and fatigue have
been associated with
burnout, poor health and increased accident risk.
Design and Methods: Semi‐ structured interviews were
conducted with nurses
(N = 11) from four different Swedish hospitals, and qualitative
inductive content
analysis was used. The study was approved by the Regional
Ethical Review Board in
Stockholm. The COREQ checklist was followed.
Results: Many nurses lacked effective strategies for managing
sleep and fatigue in re-
lation to shiftwork. Various strategies were used, of which some
might interfere with
factors regulating and promoting sleep such as the homeostatic
drive. Sleep prob-
lems were common during quick returns, often due to
difficulties unwinding before
sleep, and high workloads exacerbated the problems. The
described consequences
of fatigue in a clinical work context indicated impaired
executive and nonexecutive
cognitive function.
Conclusion: The findings indicate that supporting strategies and
behaviours for sleep
and fatigue in an intervention for newly graduated nurses
starting shiftwork may be
of importance to improve well‐ being among nurses and
increase patient safety.
Relevance to clinical practice: This study highlights the
importance of addressing
sleep and fatigue issues in nursing education and work

introduction programmes to
increase patient safety and improve well‐ being among nurses.
K E Y W O R D S
fatigue, newly graduated nurses, patient safety, shiftwork, sleep
https://orcid.org/0000-0001-8397-5986
https://orcid.org/0000-0001-7676-7919
https://orcid.org/0000-0003-4570-4047
https://orcid.org/0000-0002-8105-0901
mailto:
https://orcid.org/0000-0001-8252-3961
mailto:[email protected]
http://crossmark.crossref.org/dialog/?doi=10.1111%2Fjocn.1507
6&domain=pdf&date_stamp=2019-10-31
| 185EPSTEIN ET al.
1 | INTRODUC TION
Approximately 20% of all Swedish nurses experiences very high
levels of burnout symptoms at some point during the first years
of
practice (Rudman & Gustavsson, 2011). In addition to stressors
that
are evident for the whole nursing population, such as high
work-
load, staff shortage and emotional demands (McVicar, 2003),
newly
graduated nurses are facing specific challenges associated with
the
adaptation to a new professional role, a phenomenon referred to
as
a transition shock (Duchscher, 2009). When developing methods

to
facilitate the nurse's transition from education into working life,
a
variety of stressors at both the organisational and individual
level
have to be considered (Sun, Ji, Zhou, & Liu, 2019).
The stress response promotes adaptation through bodily and
behavioural changes, but may have damaging effects if
prolonged
or repeated without sufficient recovery in between (McEwen,
2004). Previously, incomplete recovery has been suggested to
mediate the relation between stressful working conditions and
health impairment (Geurts & Sonnentag, 2006). For many,
starting
working as a nurse also means an introduction to shiftwork,
which
inevitably affects opportunities for sleep and recuperation due
to interference with the circadian and homeostatic regulation of
sleep (Åkerstedt, 2003). Not surprisingly, sleep problems are
more
common among nurses working on shift schedules compared to
nurses working regular daytime (McDowall, Murphy, &
Anderson,
2017). In a cohort of Swedish nurses, subjective sleep quality
was
shown to decrease from the last semester of the education and
during the first 3 years of employment (Hasson & Gustavsson,
2010). Sleep problems and fatigue among nurses may threaten
both the nurses' own health and patient safety (Hughes &
Rogers,
2004; McEwen, 2006).
2 | BACKGROUND
Sleep deprivation is a stressor for the brain and body (McEwen,

2006), and insufficient sleep has been identified as a main risk
factor for clinical burnout (Söderström, Jeding, Ekstedt, Perski,
&
Åkerstedt, 2012). At the same time, stress is known to
negatively
influence sleep (Linton et al., 2015). High workload and high
per-
ceived stress at work have been associated with heightened
stress
at bedtime and shorter sleep (Dahlgren, Kecklund, & Åkerstedt,
2005). It is possible that sustained physiological activation
partly
explains the impact of stress and workload on sleep, but perse-
verative cognition is also likely to play a role (Brosschot,
Gerin, &
Thayer, 2006).
Shiftwork, in general, is associated with sleep problems and
fatigue (Kecklund & Axelsson, 2016). A key issue is the timing
of
the shifts. Night shifts and early morning shifts have been
associ-
ated with more sleep problems and fatigue, compared to evening
shifts (Åkerstedt, 2003). Recent studies have shown that shift
combinations involving so‐ called “quick returns,” that is less
than
11 hr between shifts, are associated with shorter sleep, impaired
sleep quality, increased sleepiness and fatigue (Vedaa et al.,
2015).
The frequency of quick returns is also associated with difficul-
ties unwinding (Dahlgren, Tucker, Gustavsson, & Rudman,
2016).
Two more or less ineluctable consequences of disturbed sleep
are
sleepiness and fatigue. Fatigue is a major safety hazard and has

been associated with impaired performance, higher error rates
and reduced safety (Dawson & McCulloch, 2005).
2.1 | Theoretical framework
Given the vital role of sleep and recovery in the relationship be -
tween stress and development of illness (Geurts & Sonnentag,
2006; McEwen, 2004; Söderström et al., 2012), effective strate-
gies for sleep and recovery are hypothesised to be crucial in
pre-
venting the development of stress‐ related illnesses (Colombo &
Cifre Gallego, 2012). They are likely to be particularly
important
for newly graduated nurses who are exposed to a large number
of
stressors in their daily work at the same time as they are
starting
to work shifts. Strategies for managing fatigue are also of great
importance, not least considering the potential impact on patient
safety.
2.2 | Aim
Little is known about the actual strategies for managing sleep
and
fatigue that nurses adopt when they start working shifts. Such
knowledge, together with more comprehensive understanding of
new nurses' sleep problems and the nature of their fatigue
problems,
is important to know which support (or interventions) nurses
need
when starting out in their new career. By focusing on newly
gradu-
ated nurses and their challenges, appropriate support can be put
in
place before more serious problems arise. Thus, the aim of the

cur-
rent study was to explore the strategies used by newly graduated
nurses for managing sleep and fatigue problems, identify which
fac-
tors they experienced as contributing to sleep problems and
ascer-
tain the nature of the fatigue problems that they experience. In
this
study, a newly graduated nurse is defined as a nurse in their
first year
of employment.
What does this paper contribute to the wider global
clinical community?
• Newly graduated nurses lack effective strategies for
sleep and fatigue, and the strategies used are often
counterproductive.
• Newly graduated nurses experience cognitive deficits
due to fatigue, which may have implications for patient
safety.
• Newly graduated nurses need to learn effective strate-
gies for managing sleep and fatigue, to promote patient
safety and nurses' health.
186 | EPSTEIN ET al.
3 | METHODS
3.1 | Participants
Eleven newly graduated nurses (ten women, one man) between

22–51 years old (M = 29.1, S = 8) were included in the study.
The
nurses were recruited from four, university and county
hospitals.
The inclusion criterion was a maximum of 1 year's employment
as
a registered nurse. In Sweden, the education for registered
nurses
consists of about 60% theoretical education and 40% clinical
prac-
tice (Ulfvarson, Oxelmark, & Jirwe, 2018). The sampling was
pur-
posive, and recruitment was made in two ways; either through a
lecture about working hours led by the researchers, included in
the
hospitals' introduction programmes for nurses; or through infor-
mation disseminated by the hospitals' HR departments. All
nurses
who signed up for the study (N = 17) were contacted by e‐ mail
or
telephone to make an appointment for the interview. Eleven
nurses
answered and agreed to participate in the study (see Table 1).
The study was approved by the Regional Ethical Review Board
in Stockholm (2016/1395‐ 31/2). Before starting each interview,
participants gave written informed consent. It was emphasised
that
participation in the study was voluntary and that the participant
could cancel the interview at anytime or choose not to answer
some
questions without giving further explanation. No compensation
was
given for participation. All data were stored on protected
servers
accessible only for researchers involved in the project.

3.2 | Materials
Semi‐ structured interviews were conducted by either MS
(female) or
AD (female) and lasted approximately 45 min. The interviews
were held
between October 2016–April 2017. The participants had not met
the
researcher before, apart from those who had attended the lecture
at
the introduction programme. Participants were told that the aim
of the
interview was to gain a deeper insight into the participant's
work situa-
tion with focus on work hours and recuperation, in order to gain
knowl-
edge about how to design an intervention for newly graduated
nurses.
Interviews were conducted either by telephone (eight nurses) or
face‐
to‐ face in the vicinity of the introduction programme (three
nurses),
focusing on sleep and fatigue in relation to shiftwork. No one
but the
participant and the researcher was present during the interview.
The
participants were told that the results would be presented at a
group
level only and that no individual answers would be
communicated to
managers. Specific questions addressed the nurses' experiences
of
sleep and fatigue in connection to different shifts (morning,
evening,

night and quick returns). Strategies used for recovery, such as
managing
sleep and fatigue when working the different shifts, were also
explored.
Also, more general questions like “Do you experience sleep
problems?”
“Are you sometimes feeling fatigued or exhausted at work?”
and “How
do you recover during spare time?” were used and
complemented by
appropriate follow‐ up questions. The questions were based on,
and
further developed from, a previous interview guide (Frögeli,
Rudman,
Ljótsson, & Gustavsson, 2018). During the last interviews, no
new in-
formation was identified and the research team considered data
satu-
ration to have been achieved in relation to the content of the
data, that
is when no new information was obtained from the interviews.
Sleepiness and fatigue can be described as two interrelated, but
distinct, phenomena. Sleepiness could be broadly defined as the
“drive” to fall asleep. Fatigue, on the other hand, could be
defined as
a subjective experience of overwhelming tiredness, lack of
energy
or exhaustion, that could be alleviated after rest, acti vity or
stress
management. However, the terms are commonly used
interchange-
ably or merged under the broader term “tiredness” (Shen,
Barbera, &
Shapiro, 2006). In this study, the term fatigue was used as a
broader

concept including also “sleepiness” and “tiredness”, since it is
likely
that the participating nurses did not make any distinction when
de-
scribing their experiences.
3.3 | Analysis
All interviews were audio‐ recorded, transcribed and analysed
using
inductive content analysis (Elo & Kyngäs , 2008). The analysis
was
not software‐ assisted. The analysis process was conducted in
three
phases: preparation, organising and reporting.
TA B L E 1 Background information
about participants
Participant Gender Age Workplace
Employment
as nurse
(months) Shift schedule
1 Female 29 Geriatrics 4 Morning and evening
2 Male 26 Psychiatry 4 Morning, evening and night
3 Female 35 Stroke 4 Morning and evening
4 Female 25 Infection 9 Morning, evening and night
5 Female 29 Palliative care 5 Morning and evening
6 Female 27 (Missing) 4 Morning and evening

7 Female 24 Infection 11 Morning, evening and night
8 Female 51 Rehabilitation 5 Morning and evening
9 Female 22 Neonatal care 5 Morning, evening and night
10 Female 26 Patient hotel 12 Morning and evening
11 Female 26 Hand surgery 12 Morning, evening and night
| 187EPSTEIN ET al.
In the preparation phase, the transcripts were read several times
to make sense of the whole and units of analysis were selected.
In
this study, the unit of analysis was, as suggested by Graneheim
and
Lundman (2004), the interview as a whole. The data were then
or-
ganised in three steps: open coding, creation of categories and
ab-
straction. During the open coding, notes and headings were
written
in the text to describe all aspects of the content. The headings
were
then put into groups with similar content, and categories were
gen-
erated. At the last step, abstraction, the subject under study was
described by generating subcategories and naming them using
con-
tent‐ characteristic words. The subcategories were grouped
together
into generic categories, and generic categories were grouped

into
wider main categories (see Table 2).
To increase the trustworthiness, the data collection, analysis
and
presentation of the results have involved a continuous process
of
discussions within the research group. The varied professional
back-
ground and expertise in the research group was such that the re-
searchers' pre‐ understandings were unlikely to affect data
collection
or analysis. Hence, the risk of preconceived ideas and
conclusions in-
truding was limited. Also, the questions asked during the
interviews
were open‐ ended and the respondents were asked to clarify
their re-
sponses, so as to minimise reinterpretation of the answer by the
in-
terviewer. MJ has extensive knowledge and experience of
qualitative
research/analysis and guided the others in the research team.
The anal-
ysis was initially conducted by ME and AD, with and through
regular
discussions with MS and MJ, until common agreement was
reached. To
increase credibility of the results, MJ read all the interview
transcripts
and confirmed the analysis after it was completed. Many
categories
are exemplified with literal quotes from the nurses, which
further in-
creases the credibility. The article adheres to the consolidated

criteria
for reporting qualitative research (COREQ) checklist, see
Appendix S1
(Tong, Sainsbury, & Craig, 2007).
4 | RESULTS
Analyses led to four main categories: (a) factors contributing to
sleep
problems, (b) strategies for sleep, (c) experiences of fatigue and
(d)
strategies for fatigue. Each main category consists of different
ge-
neric categories, of which some consist of subcategories (Figure
1).
4.1 | Factors contributing to sleep problems
The analyses identified two generic categories characterising
the
nurses' experiences of factors contributing to sleep problems:
(a)
shift factors and (b) cognitive arousal (with two subcategories).
4.1.1 | Shift factors
Sleep quality was commonly discussed in relation to type of
shift/
shift sequence. Most nurses experienced impaired sleep quality
in connection with shift combinations featuring an evening shift
TA B L E 2 Example of coding and categorisation
Data from interviews Codes from interviews Subcategory
Generic category Main category

It's harder when I'm on an evening‐ day shift, but
of course this is about winding down, but I find it
harder to switch off from work when I get home.
I just think, I have to do this tomorrow, and this,
and this, and this … That doesn't happen when I'm
on day‐ day shifts, because then I can let go in a
different way. (Nurse 4)
Hard to switch off, thinking about
what has to be done tomorrow
Anticipatory stress Cognitive arousal Factors contrib-
uting to sleep
problems
Mainly if you've been working an evening shift
and then you have to work in the day. Then I find,
well, I always sleep really badly. So it's a bit like,
you feel that you sleep and then you have to get
back to work, and if it was a busy night you sort
of know, “I can't forget that,” and things pop up in
my head, like there are certain things you have to
do. (Nurse 7)
Things that I have to do the next
day pop up in my head
Thinking about everything and then winding down
at home is incredibly difficult. The only thing buzz-
ing around in my head is did I do the right thing,
did I give the right medication, things like that, and
then I dream about these things. (Nurse 3)
Things like “did I do the right thing,”
“did I give the right medication”
are buzzing around in my head

Ruminative
thinking
If you compare with my previous workplace, that's
a bit why I resigned from my previous workplace,
because it was such an incredibly stressed work-
place. And there I sort of felt that it was almost
impossible to wind down, because there was so
much happening during the days, and you weren't
sure, did I do everything I was supposed to do.
(Nurse 11)
It was almost impossible to unwind,
I was not sure if I had done every-
thing I was supposed to
followed by a morning shift (hereafter referred to as “quick
returns”).
While some stated that sleep between the evening and the morn-
ing shift was always bad, a few noted that high workload during
the
evening shift worsened subsequent sleep.
Also, sleep duration was discussed in connection with shift se-
quence. The majority noted that quick returns were associated
with
shortened sleep duration. For many, the already short time
available
for sleep became even shorter due to difficulties unwinding, and
consequently a longer sleep latency. A few also experienced
shorter

sleep duration before regular morning shifts.
The majority of nurses working night shifts expressed dissatis-
faction with their sleep in relation to night work.
4.1.2 | Cognitive arousal
Cognitive arousal was identified as an important cause of sleep
problems, highlighted by most nurses. Worry or persistent
thoughts
in the evening, particularly related to work, often led to
problems
unwinding before sleep. This was especially common in relation
to
quick returns. Two main types of cognitive arousal were
described:
ruminative thinking and anticipatory stress. A majority noted
that high
workload during the previous or upcoming shift exacerbated
rumina-
tion and anticipation processes, and worsened problems
unwinding.
Ruminative thinking
Some nurses were often worrying or thinking about things that
had
happened at work, which impaired unwinding. This was, for
example,
expressed as persistent thoughts about a particular patient or
wor-
rying about not having done things correctly:
Thinking about everything and then winding down
at home is incredibly difficult. The only thing buzzing
around in my head is did I do the right thing, did I give

the right medication, things like that, and then I dream
about these things.
(Nurse 3)
Anticipatory stress
In relation to quick returns, some respondents also reported that
an-
ticipatory stress, such as thinking or worrying about the next
day's
work, impeded unwinding. This was typically expressed as
“think-
ing about what has to be done tomorrow” or as a feeling of
already
“being mentally at work the next day.” Some respondents
pointed
out that merely thinking about the shorter time available for
sleep
was stressful.
And if there's something in particular at work, you
know you've got a lot of patients or some patients
who sort of require a lot of attention, maybe that's
something that you lie awake thinking about, and then
sometimes you might not get to sleep until one or two
in the morning. (…) And because you then have to get
up so early the following morning, you start preparing
for it and sort of start thinking about what you'll be
doing during the day ahead.
(Nurse 2)
4.2 | Strategies for sleep
The analyses led to two generic categories of sleep strategies:

(a)
circadian rhythm and homeostasis and (b) unwinding (with two
subcategories).
4.2.1 | Circadian rhythm and homeostasis
A few nurses reported they were trying to keep regular sleeping
hours, and thus a stable circadian rhythm, despite irregular work
hours. Others reported trying to adjust the circadian rhythm to
their
schedule, for example through preparing for a morning shift by
get-
ting up earlier in the mornings 2 days in advance, despite not
work-
ing those mornings. Before night shifts, nurses reported
strategies
for rotating the rhythm forward by staying up late the preceding
night and then sleeping as long as possible in the morning
before the
F I G U R E 1 Newly graduated nurses' sleep and fatigue
experiences and strategies
Factors contributing to sleep
problems
Shift factors Cognitive arousal
Ruminative
thinking
Anticipatory
stress
Strategies for sleep

Circadian rhythm
and homeostasis Unwinding
General
unwinding
Detaching from
work
Experiences of fatigue
Manifestations of
fatigue
Cognitive
manifestations
Emotional
manifestations
Physical
manifestations
Temporal aspects
Strategies for fatigue
Proactive
strategies
Reactive
strategies
| 189EPSTEIN ET al.

night shift. When turning back to day or evening shifts after
night
work, two nurses described a strategy of setting the alarm
earlier
in the afternoon, to shorten the sleep and increase sleepines s in
the
evening.
4.2.2 | Unwinding
Nurses used different unwinding strategies to facilitate sleep:
(a)
general unwinding and (b) detaching from work.
General unwinding
Before sleep, many nurses were undertaking behaviours aimed
at
unwinding and relaxation, such as listening to music, watching
TV,
having a shower or using a breathing exercise. Some nurses de-
scribed that they usually stayed up for a while after an evening
shift,
to unwind before bedtime, but avoided this during quick returns.
However, two nurses used the strategy of staying up a while
dur-
ing quick returns; one had learned by experience that this made
her
sleep better:
In the beginning I think I tried to sort of go to bed as
soon as I got home, but I realised that that wouldn't
work. (…) I usually get home and, well, then perhaps I'll
get something to eat or I'll lie down and read or watch
TV. I wind down a bit and don't go to bed straight

away. (…) Perhaps I'd get to sleep sooner if I went
straight to bed, but I think maybe it would be slightly
worse quality of sleep.
(Nurse 10)
Detaching from work
Most nurses reported that they sometimes found it hard to
detach
from thoughts of work during leisure time and that this
sometimes
disturbed sleep. Different strategies were used to facilitate
detach-
ment. Some tried to distract themselves through focusing on
other
things, such as TV or mobile phone. Other strategies included
just
trying to think about something else and phoning a colleague.
Some
reported making a phone call from home to the ward, to ensure
that
everything was alright. One nurse had a strategy of telling her
col-
leagues (before leaving work) to call her if something seemed
un-
clear when she had left. Another had a strategy of writing
reminders
to herself while at home when she started thinking about things
to
do at work the next day.
Two nurses usually experienced problems unwinding from
thoughts of work during quick returns and lacked strategies for
han-
dling it. Another nurse reported that she always lacked

strategies
when thoughts of work occupied her mind.
4.3 | Experiences of fatigue
The experiences of fatigue consisted of two generic categories:
(a)
manifestations of fatigue (with three subcategories) and (b)
temporal
aspects.
4.3.1 | Manifestations of fatigue
Three subcategories of fatigue manifestations were identified:
cog-
nitive, emotional and physical manifestations. The majority of
the
nurses highlighted that fatigue was more noticeable during
certain
periods of high workload. These were, for example, complex or
highly emotional situations, bed crowding or unexpected
situations
with many things happening at the same time. In contrast, some
nurses experienced more fatigue during low workload or
inactivity.
Cognitive manifestations
A majority of the nurses reported various kinds of cognitive
manifes-
tations of fatigue. Some experienced problems concentrating,
for in-
stance during shift handovers; others described memory
difficulties,
such as problems remembering the care plan or appointed times
for

medication. Also problems prioritising, making decisions or
keeping
track of the overall picture were described.
When you're calculating medication, or performing
medical techniques, maybe taking samples, or when
you really do have to focus a bit more. I can feel
there's something missing. Then I can't focus as well,
and I have to prepare a medication. And it's important
not to make a mistake.
(Nurse 9)
Emotional manifestations
Nurses also experienced emotional manifestations of fatigue,
for ex-
ample feelings of being more emotionally sensitive, more easily
an-
noyed, unengaged or not as happy as usual. Also, feelings of
shame
were reported in connection to fatigue:
It is hard to keep up with everything. I lose… I can for-
get what was planned for a patient if I'm tired, so I can
forget the entire plan, even though we've done the
rounds and have discussed and planned. If someone
asks me something, I don't have a clue. (…) It's almost
embarrassing, because you yourself think that you
should know.
(Nurse 1)
Physical manifestations
For some nurses, fatigue was also manifested in physical

symptoms,
like headache, dizziness, feeling cold or physically exhausted.
4.3.2 | Temporal aspects
Fatigue varied with time of day and was dependent on type of
shift.
In relation to morning shifts, many nurses described fatigue at
the
beginning of the shift, while others experienced more fatigue in
the
afternoon. During evening shifts, only a few nurses experienced
fatigue. The majority described a period of severe fatigue
during
the early morning hours of the night shifts. Some also reported
a
sustained feeling of exhaustion following a night shift. During
quick
returns, while some nurses experienced fatigue at the beginning
of
the morning shift, the majority reported the most severe fatigue
afterwards:
It usually comes afterwards, that you're sort of struck
that… Well, you become really tired. When I get home,
I'm so tired that I could almost sleep sitting down.
(Nurse 2)
4.4 | Strategies for fatigue

Two generic categories of strategies for fatigue were identified
in the analyses: (a) proactive and (b) reactive strategies.
Proactive
strategies were behaviours aimed at preventing fatigue, and
reactive
strategies were behaviours for coping with fatigue.
4.4.1 | Proactive strategies
To prevent fatigue during morning shifts, some nurses described
strategies related to sleep the preceding night. While some
priori-
tised longer sleep in the morning instead of breakfast, others
got up
earlier to have time to “wake up” at home, or arrived early at
work to
have time to get started. One nurse, who could influence her
sched-
ule, usually planned an evening shift after a couple of
consecutive
morning shifts to enable a lie‐ in for recovery. In relation to
even-
ing shifts, some nurses reported having lie‐ ins in the morning
before
work to save energy and prevent fatigue:
Before an evening shift, I'll sleep almost the entire
morning because I'm thinking that I have to cope at
work. I really can't get anything done before work. It's
probably more psychological, that if I do anything in
the morning, like going and having lunch with someone
or whatever, I think I'll be so tired when I get to work.
(Nurse 1)
To prevent fatigue during the night shifts, most nurses reported

resting or trying to have a nap in the daytime before the shift.
4.4.2 | Reactive strategies
Various reactive strategies were reported aimed at coping with
fa-
tigue at work. One strategy was keeping active and being sure
of
having things to do at all times. One nurse had learned to ask
for
help when feeling tired. Another nurse said that she usually
tried to
pause for a minute and prioritise, make a plan and then follow
the
plan. One nurse described that, when experiencing fatigue at
work,
she accepted it. Another reported that she sometimes stayed
home
from work due to fatigue.
I function best when I'm a bit stressed, because then I
feel that I have to be doing something all the time – as
soon as I have nothing to do, I'm finished. So I always
try to find something.
(Nurse 3)
Regarding strategies to handle accumulated fatigue after quick
returns, some nurses went straight home after the morning shift
and
used the leisure time for resting. Other strategies included
taking a nap
when coming home and down‐ prioritising of physical activity
or meet-
ing up with friends.

5 | DISCUSSION
The aim of this study was to explore strategies used for
managing
sleep and fatigue problems, factors contributing to sleep
problems
and the nature of fatigue problems, among newly graduated
nurses.
Results showed that new nurses sometimes lack effective strate-
gies for addressing problems with sleep and fatigue.
Furthermore, as
discussed below, the results indicate that some strategies that
were
used may interfere with factors regulating sleep and
wakefulness.
High workloads were experienced as exacerbating cognitive
arousal
processes that disturb sleep, especially in relation to quick
returns.
Cognitive, emotional and physical manifestations of fatigue
were de-
scribed, and nurses often experienced severe fatigue after a
quick
return shift combination.
5.1 | Shift factors and cognitive arousal as
contributing factors to sleep problems
The results confirm previously observed associations between
quick
returns and poor sleep quality, longer sleep latency and
shortened
sleep duration (Dahlgren et al., 2016; Vedaa et al., 2015). They
are
also consistent with previous research that has identified

cognitive
arousal as a principal cause of disturbed sleep (Åkerstedt, 2006;
Åkerstedt, Kecklund, & Axelsson, 2007). Two main
sleep‐ disturbing
cognitive processes emerged, ruminative thinking about
previous
work and anticipatory stress regarding work next day, both of
which
have previously been linked to disturbed sleep (Åkerstedt, 2006;
Kecklund & Åkerstedt, 2004). Problems with both ruminative
think-
ing and anticipatory stress regarding work, during off‐ working
hours,
have also previously been described among newly graduated
nurses
(Duchscher, 2009).
A new finding was that these sleep‐ disturbing cognitive pro-
cesses were especially common in relation to quick returns and
were
highlighted as being even bigger problems during high
workload.
Rumination and anticipation are plausible explanations of
previous
findings that quick returns are associated with problems
unwinding
and longer sleep latencies (Dahlgren et al., 2016; Vedaa et al.,
2017).
It is also possible that problems with unwinding and rumination
ex-
acerbate anticipation. However, previous studies have not
examined
how these processes are linked to workload and it should be
explored
further.

1. Consider the Lewis two-sector model. What are the 2 main assump

1. Consider the Lewis two-sector model. What are the 2 main assump

Recommended

Recommended

More Related Content

Similar to 1. Consider the Lewis two-sector model. What are the 2 main assump

Similar to 1. Consider the Lewis two-sector model. What are the 2 main assump (20)

More from MartineMccracken314

More from MartineMccracken314 (20)

Recently uploaded

Recently uploaded (20)

1. Consider the Lewis two-sector model. What are the 2 main assump