increased neurosteroid sensitivity – an explanation to symptoms associated with chronic work...

Increased neurosteroid sensitivity — An explanationto symptoms associated with chronic work relatedstress in women?

Torbjorn Backstrom a,*, Marie Bixo b, Sigrid Nyberg a, Ivanka Savic c

aUmea Neurosteroid Research Centre, Department of Clinical Science, Umea University, SE-901 85 Umea, SwedenbKarolinska Institute, Department of Clinical Science and Education Sodersjukhuset, SE-118 83, Stockholm, SwedencKarolinska Institute, Stockholm Brain Institute, Department of Women’s and Children’s Health, Division of Pediatric Neurology,Stockholm, Sweden

Received 30 April 2012; received in revised form 18 October 2012; accepted 18 October 2012

Psychoneuroendocrinology (2013) 38, 1078—1089

KEYWORDSAllopregnanolonechallenge;Burnout syndrome;Saccadic velocity

Summary Work related psychosocial stress can be accompanied by so called burnout syndromewith symptoms of mental exhaustion, physical fatigue, and cognitive dysfunction. Underlyingmechanisms for acquiring burnout syndrome are not clear. Animal studies show that chronic stressis associated with altered release of GABA-A receptor modulating steroids (GAMS), alteredcomposition of the GABA-A receptor and altered sensitivity to GAMS. In the present study weinvestigated if such changes occur in women with burnout syndrome. We further asked whetherflumazenil (a benzodiazepine antagonist, but with positive modulating effects on GABA-Areceptors with altered subunit composition) can block the effect of the GAMS allopregnanolone.

Ten women with occupational psychosocial stress and burnout syndrome were compared withtwelve healthy controls in an experimental setting. Saccadic eye velocity (SEV) was measuredafter an injection of allopregnanolone, followed by an injection of flumazenil and a secondinjection of allopregnanolone.

The sensitivity to allopregnanolone was significantly higher in the patients compared tocontrols after the first injection ( p = 0.04) and the difference increased when the response perallopregnanolone concentration unit was compared ( p = 0.006). Following the flumazenil injec-tion the burnout patients ( p = 0.016), but not controls, showed a decrease in SEV and flumazenilacted like a positive modulator that is agonistic. There was no significant difference between thegroups after second allopregnanolone injection.

In conclusion, patients with work related psychosocial stress and burnout syndrome show adifferent response to GABA-A receptor modulators than controls suggesting a changed GABA-Areceptor function in these patients. More precisely we hypothesize that the a4 and delta subunits

* Corresponding author at: Umea Neurosteroid Research Center, Department of Clinical Sciences, Norrlands University Hospital, Building 5B5th floor, SE-901 85 Umea, Sweden. Tel.: +46 90 785 2144.

E-mail address: [email protected] (T. Backstrom).

Available online at www.sciencedirect.com

jou rn a l home pag e : ww w. el sev ie r. com/ loca te /psyn eu en

0306-4530/$ — see front matter # 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.psyneuen.2012.10.014

mailto:[email protected]

http://www.sciencedirect.com/science/journal/03064530


are up-regulated elevating the responsiveness to allopregnanolone and change the effect offlumazenil, which provides a potential explanation to the burnout syndrome. Flumazenil doesnot block the effect of allopregnanolone.# 2012 Elsevier Ltd. All rights reserved.

Neurosteroid sensitivity in burnout syndrome 1079

1. Introduction

The aim of the present study is to investigate the sensitivityin a GABA-A receptor system in patients having burnoutsyndrome due to chronic occupational and psychosocialstress. We use challenges with a GABA-A receptor modulat-ing steroid, allopregnanolone and the benzodiazepineantagonist flumazenil as our tools. In addition we wantedto investigate if the benzodiazepine antagonist flumazenilcould interact with the allopregnanolone effect. The read-out of the dose response effects is the maximal saccadiceye velocity tightly controlled by the GABA-A receptors.The sensitivity of a GABA-A system has, to the best of ourknowledge, never been studied in patients reporting expo-sure to chronic psychosocial stress and having burnoutsyndrome.

Western societies are facing increasing reports of stress-related illness among otherwise healthy and high-performingindividuals (Ahola et al., 2006; Fernandez Torres et al., 2006;Rydmark et al., 2006). The symptoms of these persons aregenerally stereotypical, characterized by memory and con-centration problems, sleeplessness, diffuse aches, fatigue,irritability, and anxiety. They are frequently attributed tooccupational stress, and when becoming chronic the condi-tion is often associated with emotional exhaustion, feeling ofdepersonalization, and reduced professional competence,and usually labeled as occupational burnout (Maslachet al., 2001). Occupational burnout is recognized in theICD-10 (World Health Organization, 2004) as ‘‘Problemsrelated to life-management difficulty’’. It is, however, nota recognized disorder in the DSM (American Psychiatric Asso-ciation, 1994) and, is, because of partly overlapping symp-toms (anxiety, fatigue, attention problems), sometimesconsidered as a sort of depression. However, the majorityof persons with occupational burnout have no dysthymia, andtheir response to SSRIs is much lower than in depression(Asberg et al., 2010). Furthermore in our recent PET studyof persons suffering from long standing occupational stresswe found localized limbic reductions of the 5-HT1A receptorbiding potential, despite absence of depression and history ofmajor life traumas (Jovanovic et al., 2011).

Acute stress is characterized by increased stimulation ofsteroid production from the adrenals. What is less recognizedis that in parallel to cortisol there is a production of neu-roactive steroids which in nM concentrations positively andallosterically modulates the action of the GABA-A receptor,so called GABA-A receptor modulating steroids (GAMS). Twovery potent GAMS are allopregnanolone (3a-hydroxy-5a-pregnan-20-one) and THDOC (3a-5a-tetrahydrodeoxycorti-costerone). Resent results suggest that the GABA-A receptorsystem is involved in the physiological stress response (Gunnet al., 2011; Sarkar et al., 2011). Studies in both humans androdents show that GAMS increase during acute stress andcorrelate to cortisol (Purdy et al., 1991; Droogleever-Fortuynet al., 2004). In addition there is an interaction between

cortisol and allopregnanolone actions, in that cortisol’s 3a-hydroxy-5a-reduced metabolites enhance the effect of allo-pregnanolone on the GABA-A receptor (Stromberg et al.,2005).

To study GABA-A receptor sensitivity in this category ofpatients is of particular interest for several reasons amongthem they below:

1. The GABA-A receptor system is reportedly engaged in theresponse to acute stress (Skilbeck et al., 2010; Gunnet al., 2011; Sarkar et al., 2011), as well as in chronicstress conditions (Geuze et al., 2008).

2. Long term exposure to GABA-A receptor active steroids asduring chronic stress changes the receptor subunit com-position and thus the function of the receptor, see below.

3. Studies in both humans and rodents show that concen-trations of two very potent GAMS are produced duringacute stress from the adrenal, allopregnanolone andTHDOC (Purdy et al., 1991; Droogleever-Fortuyn et al.,2004). The levels increase upon stimulation with cortisolreleasing hormone (CRH) and ACTH (Meczekalski et al.,2000). But in some chronic conditions, like post traumaticstress disorder reduced levels of allopregnanolone hasbeen detected (Rasmusson et al., 2006). So the allopreg-nanolone production/concentrations show varyingresults depending on condition, and stress situation.Therefore the sensitivity to allopregnanolon may insteadbe a crucial factor in burnout syndrome, se further below.

4. There is an interaction between stress related cortisoland allopregnanolone release, in that cortisol’s 3a-hy-droxy-5a-reduced metabolites enhance the effect ofallopregnanolone on the GABA-A receptor (Stromberget al., 2005).

5. At least some of the symptoms e.g. memory and concen-tration problems, sleeplessness, irritability, anxiety, ex-treme fatigue (Schaufeli and van Dierendonck, 1995;Rydmark et al., 2006) could be due to an increasedrelease or altered activity of GABA-A receptor activecompounds, e.g. the endogenous GAMS; allopregnano-lone and tetrahydrodeoxycorticosterone (THDOC).

We therefore set out to study a function that can reflectpotential differences in GABA-A receptor sensitivity betweenburnout and control subjects namely the maximal saccadiceye velocity.

The maximal saccadic eye velocity (SEV) is strictly con-trolled by the GABA-A system and as the saccade velocity isnot under voluntary control it is considered to be an indirectbut objective measure of GABA activity per se in the circuitscontrolling the saccade velocity (Baloh et al., 1975; Griffithset al., 1984; Marshall et al., 1985; Becker, 1989). SEV would,thus, be a suitable method to measure the sensitivity toallopregnanolone of the GABA-A receptor. To measure SEV isof particular interest in burnout syndrome as several symp-toms like memory and concentration problems, insomnia,

1080 T. Backstrom et al.

irritability, and anxiety, are typical symptoms in patientgroups showing changes in saccadic eye velocity like pre-menstrual dysphoric disorder and panic disorder (Roy-Byrneet al., 1990; Sundstrom et al., 1997). The method to measureeffects on SEV as a result of challenge with positive GABA-Areceptor modulating compounds has for several years beenused by our group to investigate GABA-A receptor sensitivity(Sundstrom et al., 1997; Nyberg et al., 2004; Timby et al.,2006).

We have not studied the cortisol production or sensitivityin our subjects with work-related psychosocial stress andburnout syndrome, but according to the literature the find-ings on cortisol concentrations are not consistent. Studiesshow awakening and morning cortisol concentrations to behigh, not different and low in certain individuals with sighs ofchronic stress and burnout syndrome (Melamed et al., 1999;Pruessner et al., 1999; Morgan et al., 2002; Moch et al., 2003;Grossi et al., 2003, 2005; Lupien et al., 2005; Langelaanet al., 2006; Mommersteeg et al., 2006; Sonnenschein et al.,2007; Chida and Steptoe, 2009; Danhof-Pont et al., 2011;Juster et al., 2011; Izawa et al., 2012). One reason for thediscrepancy in the literature may be when in the diseaseprogress the study was made. Early in the disease progressthe stress may give high adrenal activity while later in thedisease development a down regulation of the steroid pro-duction may occur, however, not in all individuals (Lupienet al., 2005; Herbert et al., 2006). In chronic stressedindividuals there are also indications of an increase in thesensitivity to cortisol, and a dysfunction in the hypothalamic-pituitary-adrenal axis (HPA) (Yehuda and Seckl, 2011). Othersystem, e.g. the 5-HT1A receptor binding potential wassignificantly reduced in hippocampus, insular and anteriorcingulated cortex in patients with chronic stress (Jovanovicet al., 2011). However, these areas were not studied in thepresent paper.

We know that changes in GABA-A receptor sensitivity andin the subunit composition may occur due to concentrationchanges of endogenous ligands e.g. changes in the secretionof GAMS (Mody and Maguire, 2012). It has been reported thathigh exposure to GAMS during long time will eventually leadto tolerance development and a decreased effect of thesubstance (Turkmen et al., 2011). Tolerance to anestheticdoses of allopregnanolone is known to relate to changes inthe subunit composition of the GABA-A receptor especiallyexpression of the a4 subunit in thalamus was shown to berelated to the degree of allopregnanolone induced tolerance(Birzniece et al., 2006). An up-regulation of especially thea4, b, and d receptor subunits is known to occur in hippo-campus as a result of increased neurosteroid productionduring e.g. stress and pregnancy (Weizman et al., 1989;Smith et al., 1998; Maguire et al., 2009). In male micesocially isolated for more than four weeks (chronic stress)studies have demonstrated that: (a) responsiveness of GABA-A receptors to GABA mimetic drugs is reduced; (b) biosynth-esis of allopregnanolone is down regulated; and (c) theexpression of GABA-A receptor subunits are altered (i.e. adecrease of a1/a2 and g2 subunits and an increase of a4 anda5 subunits) (Matsumoto et al., 2007).

Now if a tolerance has occurred on the receptor level,flumazenil, an antagonist to the benzodiazepine (BZ) bindingsite of the GABA-A receptor, has been suggested to reverse BZreceptor tolerance (Savic et al., 1991). The mechanism

behind this effect of flumazenil is not known and might bedifferent from its BZ antagonistic effect. If this is the case aflumazenil pretreatment might change the sensitivity toallopregnanolone after flumazenil injection and flumazenilhas been shown to block the anxiolytic action of allopregna-nolone. (Fernandez-Guasti and Picazo, 1995). In addition itwas shown that flumazenil can prevent a4 unit up-regulationand BZ tolerance (Biggio et al., 2007). Observations fromanimal experiments show that flumazenil has a differentaction depending on the GABA-A receptor subunit composi-tion. In receptors containing a4 subunit flumazenil has apositive modulating effect that is like a benzodiazepinebut in GABA-A receptors with other a subunits flumazenilalone is inert (Smith et al., 1998; Gangisetty and Reddy,2010). On the reverse flumazenil has been indicated toinduce panic attacks in patients with panic disorder (Nuttet al., 1990) but this has been questioned (Strohle et al.,1999). On the other hand flumazenil has been tested as anantagonist to allopregnanolone in cultured hippocampal cellsand has not shown any antagonism (Ahboucha et al., 2006).However, to our knowledge flumazenil has never been testedas an antagonist to allopregnanolone in humans.

The aim of the study was to investigate the sensitivity toallopregnanolone in patients with chronic psychosocial stressand controls, to test flumazenils own effect and if flumazenilmay inhibit the effect of allopregnanolone on saccadic eyevelocity.

2. Methods

2.1. Subjects

Twenty-two persons were recruited, 10 patients and 12 con-trols. Demographic data are given in Table 1. Inclusion: Thepatients were diagnosed as having had a ‘reaction to severestress, and adjustment disorder’ according to the Interna-tional Classification of Diseases (ICD-10, F43). They werereferred by their physician to the Stress Research Institute,Stockholm University, and recruited from there by one of theauthors (IS) after discussion with the physician at the StressResearch Institute (Dr Water Ossika, see acknowledgements)if they fulfilled the inclusion and exclusion criteria for thestudy, as follows.

2.1.1. Inclusion criteria for patientsWe selected only subjects who attributed their illness toprolonged work-related stress, after 60—70 working hours/week continuously over, on average, 2—3 years before symp-tom onset. All of them were previously high performingindividuals, but reported a reduced work capacity (confirmedby the employers). Furthermore, they described a character-istic symptom course with sleeplessness, diffuse aches, pal-pitations and fatigue, a subsequent onset of irritability,anxiety, memory and concentration problems. They alsoreported emotional exhaustion, feeling of depersonalization,and reduced professional competence. All of them attributedtheir symptoms to chronic stress, and had no other knownetiology for their distress. Based on this information, thepatients were deemed to have occupational burnout(Maslach et al., 2001). A further requirement to be includedin the ‘patient’ group was �50% sick leave for stress related

Table 1 Demographic data, baseline characteristics and subjective scorings of anxiety symptoms in women with work relatedpsychosocial stress and burn-out symptoms (BO, n = 10) and controls (n = 12). Data presented as median (range) or mean � SEM.

BO Controls p-Value between groups

Age, years 37(32—57) 37.5(23—53) nsWeight, kg 63.5(55—110) 66.5(57—81) nsParity, n 2(0—3) 0(0—3) nsTobacco use, n 0/7 4/11 nsS-allopregnanolone, nmol/L 1.50 � 0.75 0.87 � 0.31 nsSEV, degrees/sec 487 � 13 486 � 18 nsSedation, VAS, m.m. 49.5 � 8.0 22.7 � 7.0 0.04ASI 20(7—44) 9.5(4—22) 0.011STAI 1a 42(38—65) 28(24—43) 0.001PSS 1a 7.5(0—19) 1.0(0—5) 0.022STAI 2a 37(31—75) 27(22—30) 0.001PSS 2a 4.0(0—43) 0.0(0—2) 0.011SADS 1.0(0—3) 0.0(0—0) 0.004

SEV, saccadic eye velocity; VAS, visual analog scale of 100 mm; ASI, Anxiety Sensitivity Index; STAI, State-Trait Anxiety Inventory; PSS, PanicSymptoms Scale; SADS, State Anxiety and Discomfort Scale. Differences between groups were tested with the Mann—Whitney U-test and withingroups with the Wilcoxon paired-sign Rank test. No differences were detected within the groups between baseline (1) and after 24 h (2).a There was no significant difference between test occasions within the groups at baseline (1) and after 24 h (2).


symptoms during a minimum of 6 months before entering thestudy, and an average stress-burnout score of �3.0 on theMaslach Stress-Burnout Inventory — General Survey (MBI-GS)(Schaufeli and van Dierendonck, 1995). This 7-point ratingscale, ranging from 0 (never) to 6 (daily), consists of threesubscales: exhaustion (five items), cynicism (five items) andlack of professional efficacy (six items). When rating per-ceived stress, subjects were asked to take into considerationthe last six months, and not only the actual time-point. Theaverage scores found in Scandinavian populations are around2 for MBI-GS (Stenlund et al., 2007).

The possible presence of psychiatric disorders or person-ality disturbances were assessed according to the Diagnosticand Statistical Manual of the American Psychiatric Associa-tion, 4th Edition (DSM—IV), and included a Structured Ques-tionnaire for DSM-IV1 Axis I and II (Structured ClinicalInterview for DSM-IV1 (SCID-I, and II), along with a test fordepression using the Montgomery and Asberg (1979) Depres-sion scale (MADRS).

2.1.2. Inclusion criteria for controlsAs controls 12 healthy female volunteers, with no history ofchronic stress, or heredity for neuropsychiatric disordersmatched for age and weight were used (for demographicdata c.f. Table 1). All the controls were tested with MADRSscale, and for stress exposure with MBI-GS scale.

2.1.3. Exclusion criteria for all subjectsSubjects were excluded if they had heredity or previoushistory of psychosis, personality disorder, major or bipolardepression, alcohol or substance abuse, chronic fatigue,chronic pain, fibromyalgia, neurological, endocrine diseaseor had experienced prominent stress factors in private life ora major traumatic life event as assessed through a clinicalpsychiatric interview based on the non-work related items ofthe Holmes and Rahe (1967) Scale. Data about family historyand psychiatric disorders were based on self reports. Noongoing daily medication was allowed during two monthsprior to the study, except contraceptives. Patients also

received a medical screening, (physical examination, testof thyroid and liver function).

The women gave written informed consent prior to inclu-sion in the study. The study procedures were in accordancewith ethical standards for human experimentation, estab-lished by the Declaration of Helsinki. The Regional EthicalReview Board at Karolinska Institute, the Regional EthicalReview Board at University of Umea, and the Medical Pro-ducts Agency of Sweden approved the study. The controlswere compensated for donation of blood.

2.2. Study protocol

The experiments were performed at the neurological out-patient department, Karolinska University Hospital, Hud-dinge, Stockholm. The subjects arrived at 0730 h and theelectrodes for saccadic eye velocity recordings (SEV) wereplaced on the temples and forehead. An IV cannula wasinserted in each forearm, one for drug administration andthe other for drawing blood samples. Baseline blood samplesfor allopregnanolone were drawn and initial scorings ofanxiety symptoms (ASI, PSS, STAI, SADS, for clarificationsee below) were done. SEV were measured at baseline aswell as subjective ratings of sedation. After an IV injection of0.04 mg/kg allopregnanolone, SEV was measured after 5, 13,18, 25 and 30 min. At 35 min an IV injection of flumazenil1.5 mg was administered, SEV was measured at 45 and 50 minand thereafter the second IV injection of allopregnanolone0.04 mg/kg was given and SEV was measured at 60, 68, 73, 80and 85 min (Fig. 1). The dose of Flumazenil was chosen on thebasis of our previous data (Savic et al., 1991), showing that1.5 mg Flumazenil occupies about 55% of the benzodiazepinereceptors, but does not have any behavioral effects. Bloodsamples for allopregnanolone concentrations were drawn at5, 18, 60, 73 and 85 min after the first injection. Apart frombaseline, SADS was rated at 13, 25, 45, 68 and 80 min.Approximately 24 h after the experiment, the subjects ratedanxiety symptoms on the PSS, STAI and SADS at home. Theinternal consistency of the sample’s were measured with

Figure 1 Saccadic eye velocity and serum levels of allopreg-nanolone in women with work related psychosocial stress andburnout syndrome (n = 10) and healthy controls (n = 12) throughthe entire observation time. Injections of allopregnanolone andflumazenil are indicated by vertical lines. Fig. 1 is only added fordescriptive reasons and to show the actual SEV measurements.No statistical analysis is made on the actual SEV data.


Chronbach’s alpha and it varied between 0.796 and 0.943 forthe different tests.

2.3. Saccadic eye velocity (SEV)

Measurement of saccadic eye velocity is a validated andobjective way to estimate sensitivity to GABA-A receptoractive drugs with small intra-individual variations (Marshallet al., 1985; Marshall and Richens, 1989). SEV is measured usingelectro-oculography (EOG) with the CSGAAS5 system, fullydocumented elsewhere (Griffiths et al., 1984; Marshallet al., 1985; Marshall and Richens, 1989). The basic conceptis to measure the difference in electric potential between thecornea and the retina. The computer program that generatesand calculates the saccades was developed by Cardiff ClinicalTrials Ltd and is well documented (Marshall et al., 1985;Marshall and Richens, 1989). The subjects were instructedto watch the target being light-emitting diodes (LEDs) placedat eye-level, 67 cm from the glabella. They were asked to keeptheir head still and move the eyes to look at the illuminated LEDwhich move with fixed distance apart producing saccades of108, 208, 308 and 408, 308 measurements are used in this study.EEG cup electrodes (Synetics AB, Stockholm, Sweden) withelectrode gel (Elefix, Nihon Kohden Europe, Rosbach, Ger-many) were placed 1 cm lateral of the outer canthus of botheyes, with one common electrode in the center of the fore-head. Before application of the electrodes, the skin was

exfoliated with Skinpure cream (Nihon Kohden). Electrodeimpedance was confirmed to be less than 5 k?. The first foursteps of each session were not included in the subsequentanalyses, in order to allow the subject to adjust to the testprocedure. A computer controlled the target movements andrecorded the eye movements. The individual EOGs, resultingfrom the target steps, were stored and analyzed off-lineaccording to the method of Marshall and Richens (1989).Preemptive saccades and blinking artifacts were avoided bythe filtering of the program. Each saccade was analyzed todetermine the size of the saccade in degrees, the peak saccadicvelocity and latency from target movement to onset of thesaccade. Saccade accuracy was determined by comparing theactual eye position at the end of the saccade with theattempted target. SEV was further processed by plotting avelocity-saccade size curve, known as the main sequence(Baloh et al., 1975). The relationship between saccade sizeand peak velocity is important since it remains constant evenwhen voluntary control of saccades is attempted. The mainsequence is fitted by a quadratic equation to the peak velocitydata using the calculated saccade angle as the independentvariable. Carrying out the fitting procedure twice and weighingthe second fit with the inverse of the square of the residualsfrom the first fit minimize the influence of outliers in the data.The values of peak velocity for 108, 208, 308 and 408 saccadeswere calculated by interpolation. Saccades with amplitudes of308 were chosen for further analyses as SEV (Baloh et al., 1975).

2.4. Visual analog ratings of sedation

Subjective ratings of sedation were done by the test personsusing a visual analog scale (VAS) (McCormack et al., 1988).The scale measured from 0 to 10 cm, where 0 representedcomplete absence of sleepiness and 10 represented fallingasleep. The ratings were made at baseline and at the sametime points as the saccadic eye measurements.

2.5. Anxiety scales

Rating scales measuring different aspects of anxiety and panicwere used to detect differences between the groups at base-line and to find paradoxical reactions during and after the testsession. The Anxiety Sensitivity Index (ASI) measures fear forpanic or anxiety (Reiss et al., 1986). The overall score rangesfrom 0 to 64 and a mean ASI score of 19.01 has been reportedfor a normal sample (Peterson and Reiss, 1992) while the meanscore for subjects with panic disorder was 36.2 (Taylor et al.,1992). The Panic Symptoms Scale (PSS) is based on DSM-IVcriteria total scores range from 0 to 72. A mean PSS score of1.02—2.27 has been measured in healthy women (Bell et al.,2004). The State-Trait Anxiety Inventory (STAI) has a range ofscores from 20 to 80, and a total score below 40 indicate low,between 40 and 59 moderate, and 60 or more severe stateanxiety (Spielberger et al., 1970). The State Anxiety andDiscomfort Scale (SADS) ranges from 0 to 5 and measuresfluctuations in anxiety level when administered repeatedly.

2.6. Allopregnanolone analysis

The allopregnanolone analysis method has been described indetail earlier (Timby et al., 2006). Briefly, the samples

Figure 2 Saccadic eye velocity (delta values from baseline),serum levels of allopregnanolone and delta SEV divided by theallopregnanolone concentration in women with work relatedpsychosocial stress and burnout syndrome (n = 10) comparedto healthy controls (n = 12). The first allopregnanolone injectionis indicated by a vertical line.


(0.4 ml) were extracted with diethylether (Merck KGaA,Damstadt, Germany). Allopregnanolone was separated fromcross-reacting steroids with celite chromatography. Allopreg-nanolone was measured by RIA using a polyclonal rabbitantiserum raised against 3a-hydroxy-20-oxo-5a-pregnan-11-yl-carboxymethyl-ether coupled to bovine serum albu-min, kindly provided by RH. Purdy, (The Scripps ResearchInstitute, La Jolla, CA, USA). The sensitivity of the assay was25 pg, the intra-assay coefficient of variation 6.5% and inter-assay coefficient of variation 8.5%.

2.7. Statistics

SEV parameters and VAS self-rating scores were calculated asthe difference from baseline at each time-point, e.g. deltadegrees/second and delta sedation scores. The SEV para-meters were analyzed using two way analyses of variance(ANOVA) with repeated measures, and least significant dif-ference (LSD) as post hoc test. Changes in allopregnanoloneserum concentrations from baseline at each time point aredescribed as mean � SEM. In each group, changes in serumconcentrations were analyzed by ANOVA with repeated mea-sures and LSD. Differences in baseline serum allopregnano-lone concentration, anxiety scales and demographic data ofthe groups were explored by the Mann—Whitney U-test. TheSPSS statistical package (version 17.0) was used for allstatistical analyses. P values less than 0.05 were consideredto be statistically significant.

3. Results

3.1. Demographic data and subjective ratings

Demographic data of the 10 patients with work relatedpsychosocial stress and burnout and 12 controls are pre-sented in Table 1. The mean stress score in patients was4.1 � 0.8, and in controls 2.7 � 0.6. ( p = 0.006 Mann—Whit-ney U-test). MADRS (Montgomery Asberg Depression RatingScale (1979) scores were within the normal range in bothgroups (7.2 � 3.1 in controls and 10.1 � 6.3 in patients) andshowed no group difference. However, as expected baselinetiredness/sedation was higher in the patient group( p = 0.04, Mann—Whitney U-test) (Table 1). The patientswith work related psychosocial stress and burnout scoredsignificantly higher on all anxiety scales at base-line as wellas 24 h after the experiment compared to healthy controls(Table 1). There were no significant differences betweenbaseline scorings of STAI and PSS and scorings 24 hours afterthe test session in any of the groups (Table 1). Overall, smalland non-significant fluctuations in SADS scorings weredetected.

3.2. Allopregnanolone concentration

Baseline concentrations of allopregnanolone and baselinesaccadic eye velocity were not different between the groups(Table 1). The allopregnanolone concentration after theinjection was however higher in the controls compared tothe patients (F(1,20) = 5.06; p = 0.036, Two-way ANOVAgroup � time repeated measures 2 � 3 points) during thefirst 30 min after the injection even though the dosage

was adjusted for the weight (Fig. 2) but were not differentafter the second injection (Fig. 4).

3.3. Saccadic eye velocity data

Fig. 1 shows the actual saccadic eye velocity through theentire observation time and all three injections. Fig. 1 is onlyadded for descriptive reasons and to show the actual SEVmeasurements. No statistical analysis is made on the actualSEV data except for at the first baseline measurement (Table1). In the further analysis of effects of the different injectionsof allopregnanolone and flumazenil the data has been nor-malized around the nearest measured time point prior to theinjection.

The saccades at baseline were very similar betweenpatients and controls but differed as the study went on.After the first injection of allopregnanolone the maximalsaccadic eye velocity decreased significantly both in patients(F(5,45) = 9.24; p < 0.001, one-way ANOVA Repeatedmeasures, 6 time points) and controls (F(5,55) = 14.4;p < 0.001, one-way ANOVA Repeated measures, 6 time

Figure 3 Changes in saccadic eye velocity (delta values fromthe measurement just before the Flumazenil injection), in wom-en with work related psychosocial stress and burnout syndrome(n = 10) compared to healthy controls (n = 12) following theflumazenil injection (indicated by a vertical line).

Figure 4 Changes in saccadic eye velocity (delta values fromthe measurement just before the second allopregnanolone in-jection) in women with work related psychosocial stress andburnout syndrome (n = 10) compared to healthy controls (n = 12)following the second allopregnanolone injection (indicated by avertical line).


points). However, the patents had a significantly greaterresponse (F(1,20) = 4.83; p = 0.040, Two-way ANOVA group xtime repeated measures 2 � 5 points) than the controlsindicating that the patients were more sensitive to allopreg-nanolone compared to the controls (Fig. 2). Moreover, whendelta SEV (degrees/sec) was divided by the allopregnanoloneconcentration (nmol/L) at each time point, the ratio showedan even greater difference between the groups ( p = 0.006;Two-way ANOVA group � time repeated measures 2 � 3points, Fig. 2).

The patients decreased their SEV significantly followingthe flumazenil injection ( p = 0.016, one-way ANOVARepeated measures, 3 time points) but this was not seenin the controls (Fig. 3). There was a trend towards a differ-ence between the groups ( p = 0.072 Two-way ANOVA group xtime repeated measures 2 � 2 points,).

After the second allopregnanolone injection there was asignificant decrease in SEV in both patients (F(5,55) = 11.65;p < 0.001, one-way ANOVA Repeated measures, 6 timepoints) and controls (F(5,45) = 2.69; p = 0.033, one-wayANOVA Repeated measures, 6 time points), however, nodifference between the groups (Fig. 4). In the controls therewas a significant increase in SEV following the first measure-ment (at 5 min) after the injection ( p < 0.001, one-wayANOVA Repeated measures, 5 time points), but not in thepatients. The finding suggests that the controls developedacute tolerance while this was not the case in the patients.

3.4. Sedation ratings

Subjective feelings of sedation were recorded by the patientsand controls on a visual analog scale (VAS; Fig. 5). Both groupsincreased their sedation ( p < 0.001 and p = 0.003, resp. one-way ANOVA Repeated measures, 5 time points) but there wasno significant difference between the groups.

4. Discussion

The present study was aimed at discovering differences insensitivity to allopregnanolone between healthy controls andpatients with work related psychosocial stress and burnoutsyndrome. Further aim was to investigate the effect of

flumazenil on patients and controls and on following allo-pregnanolone challenge. The initial hypothesis was thatburnout patients might have signs of allopregnanolone tol-erance (deceased sensitivity) similar to other earlier inves-tigated conditions e.g. premenstrual dysphoric disorder(Sundstrom et al., 1997; Nyberg et al., 2004) and panicdisorder (Roy-Byrne et al., 1990). This hypothesis was notconfirmed in the present study. The second hypothesis was totest if flumazenil could inhibit the action of allopregnanolonethat is if there is an interaction between benzodiazepine andallopregnanolone binding sites at the GABA-A receptor. Thishypothesis was not confirmed. The reason for the firsthypothesis was that long-term exposure to allopregnanolonegive tolerance (Turkmen et al., 2011). The reason for thesecond hypothesis was that flumazenil has been suggested toreverse BZ receptor tolerance perhaps with another mechan-ism than pure BZ antagonism (Savic et al., 1991). As burnoutis a chronic stress disorder one could expect that in thebeginning of the syndrome high production of adrenal ster-oids, including allopregnanolone may occur and give a tol-erance after some time.

4.1. Surprising findings

The present study revealed surprising findings contrary to thehypothesis described above. The patients with work related

Figure 5 Subjectively rated sedation (delta values from thebaseline measurement) in women with work related psychosocialstress and burnout syndrome (n = 10) compared to healthy con-trols (n = 12) following the first allopregnanolone injection (in-dicated by a vertical line). Statistical analyses within groups aremade over the entire observation periods.


psychosocial stress and burnout were more sensitive to allo-pregnanolone than healthy controls. An explanation to thisunexpected finding is not directly obvious. According to someliterature patients with burnout syndrome may have a lowsteroid production from the adrenals (Danhof-Pont et al.,2011) and one speculative explanation might be that theirsensitivity increased as a compensation for decreased allo-pregnanolone exposure. However, in our study the baselineallopregnanolone concentration was not decreased in thepatients compared to controls. The allopregnanolone concen-trations in both groups were in the upper normal follicularphase range and higher than during mid follicular phase inwomen when using the same assay method for allopregnano-lone (Nyberg et al., 2007). However, we know that the situa-tion before start of an allopregnanolone challenge test and SEVmeasurement can be somewhat stressful, and therefore thebaseline values might be higher than allopregnanolone mea-sured during relaxed situations (Nyberg et al., 2007).

4.2. GABA-A receptor sensitivity measurements

Allopregnanolone is a very potent GABA-A receptor allostericmodulator and potentiates the effect of GABA at the GABA-Areceptor. The effect is similar to effects of benzodiazepinesand barbiturates (Majewska et al., 1986), but allopregnano-lone is more potent per mole than the most potent barbiturate

known today (Korkmaz and Wahlstrom, 1997). As technique tomeasure the sensitivity we have used the saccadic eye velocity(SEV). SEV is an objective measure not possible to influencevoluntarily and strictly regulated by the action of GABA via theGABA-A receptor. Therefore one can interpret the data in thepresent study as indirect measures of the GABA-A receptorsensitivity to allopregnanolone (Baloh et al., 1975; Marshallet al., 1985; Becker, 1989). Flumazenil is a well-known ben-zodiazepine antagonist considered to have no own effect, butbeing able to block the effect of benzodiazepines. The effectsof flumazenil will be further discussed below.

4.3. Allopregnanolone plasma concentrationsafter challenge

The patients had lower allopregnanolone concentrationsafter the first injection compared to the controls even thoughthe dosage was based on the bodyweight. This might be a signof a greater distribution volume in women with burnoutsyndrome compared to controls. Allopregnanolone has alarge distribution volume as it is accumulated in fat tissueand with larger unsaturated fat mass the plasma concentra-tion will decrease on a given IV dosage (Timby et al., 2006).As the weight was not increased in our patient group anincreased distribution volume might be the explanation.

4.4. Increased allopregnanolone sensitivitysuggests changed GABA-A receptor subunitcomposition

However, the lower concentration strengthens the conclusionthat women with work related psychosocial stress are moresensitive than controls as the ratio SEV/allopregnanoloneconcentration between the groups was even larger thanthe SEV change itself. The reason for the different reactionin this patient group compared to the previous studies is notknown. However, one interpretation is that patients withburnout syndrome have a different GABA-A receptor compo-sition, and thus sensitivity, in the region that regulates theSEV, see below. The velocity of the saccade is controlled bythe frontal eye field, substantia nigra, superior colliculus,pontine reticular formation and cerebellum (Becker, 1989).

4.5. Unexpected response to flumazenilindicates up-regulation of subunit a4

Concerning the response to flumazenil, the controls showedno reaction, which was expected. Flumazenil is known to bean antagonist to compounds binding at the benzodiazepinesite of the GABA-A receptor. In normal situations it has noeffect by itself or only weak partial anxiogenic/antagonisticeffects (Klotz and Kanto, 1988; Bell et al., 2004). Thepatients unexpectedly reacted with decreased SEV and simi-lar to what would be expected from a positive allostericmodulator to the GABA-A receptor like a benzodiazepine andnot an antagonist. The explanation for this finding could befound in animal studies showing that flumazenil changes itseffect from being a benzodiazepine antagonist or inert com-pound alone to becoming an agonist due to an up-regulationof the a4 subunit of the GABA-A receptor (Smith et al., 1998;Gangisetty and Reddy, 2010). This suggests that burnout


patients have an up-regulation of the a4 subunit in thecircuits controlling SEV. It is known that the a4 subunit isoften combined with the d subunit in an a4, b3, and d GABA-Areceptor, especially frequently expressed in thalamus andhippocampus (Sur et al., 1999; Backstrom et al., 2008). Thissubunit combination is of interest in relation to the increasedsensitivity to allopregnanolone seen in patients with workrelated psychosocial stress as receptors containing the d

subunit are known to be much more sensitive to GAMS thanreceptors containing the g subunit (Brown et al., 2002).

4.6. A new hypothesis on GABA-A receptorchanges in burnout syndrome emerges

� Increased allopregnanolone sensitivity can be due toGABA-A delta subunit up regulation.� Changed flumazenil action from a BZ antagonist to GABA-Areceptor positive modulator agonist can be due to subunita4 up-regulation.� a4 prefer co-localization with delta subunit in the GABA-Areceptor.

Thus, one possibility is that burnout patients have an up-regulation of the a4, b, and d containing receptors in thebrain areas controlling SEV. Such receptor subtype up-regu-lation would explain the increased sensitivity to allopregna-nolone as well as the agonistic effect of flumazenil,demonstrated for the first time in the present study. Anup-regulation of the a4, b, and d receptor subunits is aplausible explanation since this receptor subtype is knownto up-regulate as a result of increased neurosteroid produc-tion during e.g. stress and pregnancy (Weizman et al., 1989;Smith et al., 1998; Maguire et al., 2009).

4.7. No difference in sedation between thegroups

The subjective scorings of sedation did not show any differentpattern between the groups, in the response to the allopreg-nanolone and flumazenil challenges but there was anincreased sedation in both groups. A possible explanationto the discrepancy could be that SEV is processed by differentneural networks compared to sedation. It is known fromearlier studies that SEV and sedation can show differentresponses to challenges (Glue et al., 1992). Sedation is alsoa more subjective measure and patients with burnout syn-drome were more tired at baseline and that could haveprecipitated a sealing effect among the patients thus givingless space for change.

4.8. Response in controls after secondallopregnanolone injection suggests acutetolerance

After the second injection there was an equal responsemagnitude in patients and controls. This can be interpretedas if flumazenil has changed the sensitivity to allopregnano-lone in the patients as was presented in the hypothesis.However, the interpretation is blurred as in the patients,the SEV response remained on the same level during theobservation time but in the controls there was a significant

increase in SEV following the first observation at 5 min afterthe second allopregnanolone injection. This pattern is dif-ferent from after the first injection were the peak in SEVresponse occurred at 5—20 min but the increase in SEVstarted 25 min after the injection. The short and weakerresponse in controls after the second injection could be a signof acute tolerance, which is seen as compensation to a strongeffect of an anesthetic drug. We have seen acute tolerance,detected in the SEV response, earlier after the second injec-tion of midazolam, when given to healthy female controls(Sundstrom et al., 1997). This phenomenon could also be asign of an a4, b, d up-regulation in the patients as the a4, d

subunit composition is known to be less prone to desensitiza-tion (a possible mechanism for acute tolerance) compared tointrasynaptic receptors such as the a1 (Brown et al., 2002).

4.9. No effect on anxiety scores

As expected, the patients scored higher on all the anxietyscales at baseline compared to the healthy controls. Other-wise only small, and insignificant, fluctuations in anxietylevels were detected in both group during the course ofthe experiment and afterwards. No anxiogenic effect wasnoted after the flumazenil injection.

4.10. Limitations

The major limitation is the small group size and that the studyis limited to women. However, the results are quite clear andconsistent within the subjects participating.

5. Conclusion

In conclusion, patients with work related psychosocial stressand burnout show a different response to GABA-A receptormodulators than controls suggesting a changed GABA-Areceptor function in these patients. More precisely wehypothesize that the a4 and d subunits are up-regulatedelevating the responsiveness to allopregnanolone and chan-ging the response to flumazenil, which is the first time apossible explanation to the symptoms is provided. Further-more, flumazenil could not block the effect of allopregna-nolone but after the flumazenil administration there is was nodifference in the allopregnanolone SEV response in burn outpatients compared to controls likely because of an acutetolerance in controls. One plausible interpretation is thatpatients with work related psychosocial stress and burnouthave an up-regulated a4, b, d subunit combination in thebrain areas regulating SEV.

Role of the funding source

Swedish research council demands open access publication.The other funding bodies have no demands on the manu-script.

Authors disclosures

This study was funded by grants from Swedish researchcouncil, project number 4X-11198, and the EU structural


fund, objective 1, Svenska lakaresallskapet, Visare Norr,Umea University Foundations and ALF medel from Vaster-bottens Lans Landsting.

Contributors

All authors designed the study and wrote the protocol man-aged the literature searches and analyses. Authors T.B. andM.B. undertook the statistical analysis, and author T.B. wrotethe first draft of the manuscript. All authors contributed toand have approved the final manuscript.

Conflict of interest

M.B., S.N. and I.S. have nothing to declare. T.B. has equityinterests in Umecrine A.B. and has received grant supportsfrom Swedish research council and Vasterbottens lans land-sting.

Acknowledgements

This study was funded by grants from Swedish researchcouncil, project number 4X-11198, and the EU structuralfund, objective 1, Svenska lakaresallskapet, Visare Norr,Umea University Foundations and ALF medel from Vaster-bottens Lans Landsting. The authors thank Elisabeth Zing-mark for the allopregnanolone assays.

References

Ahboucha, S., Coyne, L., Hirakawa, R., Butterworth, R.F., Halliwell,R.F., 2006. An interaction between benzodiazepines and neuro-active steroids at GABA-A receptors in cultured hippocampalneurons. Neurochem. Int. 48, 703—707.

Ahola, K., Honkonen, T., Isometsa, E., Kalimo, R., Nykyri, E., Koski-nen, S., Aromaa, A., Lonnqvist, J., 2006. Burnout in the generalpopulation. Results from the Finnish Health 2000 Study. Soc.Psychiatr. Psychiatr. Epidemiol. 41, 11—17.

American Psychiatric Association (APA), 1994. Diagnostic and Statis-tical Manual of Mental Disorders, 4th edition (DSM IV). USDepartment of Health and Human Services, Washington, DC,pp. 714—718.

Asberg, M., Grape, T., Krakau, I., Nygren, A., Rohde, M., Wahlberg,A., Wahrborg, P., 2010. Stress as the cause of mental illness.Lakartidningen 107, 1307—1310.

Backstrom, T., Andreen, L., Bixo, M., Bjorn, I., Fernandez, G.,Johansson, I.M., Lundgren, P., Lofgren, L., Nyberg, S., Ragagnin,G., Poromaa-Sundstrom, I., Stromberg, J., van Broekhoven, F.,van Wingen, G., Wang, M.D., 2008. Neuroactive steroids in brainand mood. In: Ritsner, Weizman, (Eds.), Neuroactive Steroids inBrain Functions, and Mental Health: New Perspectives for Re-search and Treatment. Springer, New York Inc, Chapter 20, pp.423—433.

Baloh, R.W., Sills, A.W., Kumley, W.E., Honrubia, V., 1975. Quantita-tive measurement of saccade amplitude, duration, and velocity.Neurology 25, 1065—1070.

Becker, W., 1989. The neurobiology of saccadic eye movements.Metrics. Rev. Oculomot. Res. 3, 13—67.

Bell, E.C., Baker, G.B., Poag, C., Bellavance, F., Khudabux, J., LeMelledo, J.M., 2004. Response to flumazenil in the lateluteal phase and follicular phase of the menstrual cycle inhealthy control females. Psychopharmacology (Berl) 172,248—254.

Biggio, F., Gorini, G., Caria, S., Murru, L., Sanna, E., Follesa, P., 2007.Flumazenil selectively prevents the increase in alpha(4)-subunitgene expression and an associated change in GABA(A) receptorfunction induced by ethanol withdrawal. J. Neurochem. 102,657—666.

Birzniece, V., Turkmen, S., Lindblad, C., Zhu, D., Johansson, I.M.,Backstrom, T., Wahlstrom, G., 2006. GABA-A receptor mRNAchanges in acute allopregnanolone tolerance. Eur. J. Pharmacol.535, 125—134.

Brown, N., Kerby, J., Bonnert, T.P., Whiting, P.J., Wafford, K.A.,2002. Pharmacological characterization of a novel cell lineexpressing human alpha(4)beta(3)delta GABA(A) receptors. Br.J. Pharmacol. 136, 965—974.

Chida, Y., Steptoe, A., 2009. Cortisol awakening response and psy-chosocial factors: a systematic review and meta-analysis. Biol.Psychol. 80, 265—278.

Danhof-Pont, M.B., van Veen, T., Zitman, F.G., 2011. Biomarkersin burnout: a systematic review. J. Psychosom. Res. 70,505—524.

Droogleever-Fortuyn, H.A., van Broekhoven, F., Span, P.N., Back-strom, T., Zitman, F.G., Verkes, R.J., 2004. Effects of PhD exami-nation stress on allopregnanolone and cortisol plasma levels andperipheral benzodiazepine receptor density. Psychoneuroendo-crinology 29, 1341—1344.

Fernandez-Guasti, A., Picazo, O., 1995. Flumazenil blocks theanxiolytic action of allopregnanolone. Eur. J. Pharmacol. 281,113—115.

Fernandez Torres, B., Roldan Perez, L.M., Guerra Velez, A., RoldanRodrıguez, T., Gutierrez Guillen, A., De las Mulas Bejar, M., 2006.Prevalence of burnout among anesthesiologists at Hospital Uni-versitario Virgen Macarena de Sevilla. Rev. Esp. Anestesiol. Rea-nim. 53, 359—362.

Gangisetty, O., Reddy, D.S., 2010. Neurosteroid withdrawal regu-lates GABA-A receptor alpha4-subunit expression and seizuresusceptibility by activation of progesterone receptor-indepen-dent early growth response factor-3 pathway. Neuroscience170, 865—880.

Geuze, E., van Berckel, B.N., Lammertsma, A.A., Boellaard, R., deKloet, C.S., Vermetten, E., Westenberg, H.G., 2008. ReducedGABAA benzodiazepine receptor binding in veterans with post-traumatic stress disorder. Mol. Psychiatr. 13, 74—83.

Glue, P., Bailey, J., Wilson, S., Hudson, A., Nutt, D.J., 1992. Thyro-tropin-releasing hormone selectively reverses lorazepam-in-duced sedation but not slowing of saccadic eye movements.Life Sci. 50, 25—30.

Griffiths, A.N., Marshall, R.W., Richens, A., 1984. Saccadic eyemovement analysis as a measure of drug effects on humanpsychomotor performance. Br. J. Clin. Pharmacol. 18 (Suppl.1), 73S—82S.

Grossi, G., Perski, A., Evengard, B., Blomkvist, V., Orth-Gomer, K.,2003. Physiological correlates of burnout among women. J. Psy-chosom. Res. 55 (4), 309—316.

Grossi, G., Perski, A., Ekstedt, M., Johansson, T., Lindstrom, M.,Holm, K., 2005. The morning salivary cortisol response in burnout.J. Psychosom. Res. 59, 103—111.

Gunn, B.G., Brown, A.R., Lambert, J.J., Belelli, D., 2011. Neuro-steroids and GABA(A) receptor interactions: a focus on stress.Front. Neurosci. 131, 1—20.

Herbert, J., Goodyer, I.M., Grossman, A.B., Hastings, M.H., de Kloet,E.R., Lightman, S.L., Lupien, S.J., Roozendaal, B., Seckl, J.R.,2006. Do corticosteroids damage the brain? J. Neuroendocrinol.18, 393—411.

Holmes, T.H., Rahe, R.H., 1967. The social readjustment ratingscale. J. Psychosom. Res. 11, 213—218.

Izawa, S., Saito, K., Shirotsuki, K., Sugaya, N., Nomura, S., 2012.Effects of prolonged stress on salivary cortisol and dehydroepi-androsterone: a study of a two-week teaching practice. Psycho-neuroendocrinology 37, 852—858.


Jovanovic, H., Perski, A., Berglund, H., Savic, I., 2011. Chronic stressis linked to 5-HT(1A) receptor changes and functional disintegra-tion of the limbic networks. Neuroimage 55, 1178—1188.

Juster, R.-P., Sindi, S., Marin, M.-F., Perna, A., Hashemi, A., Pruess-ner, J.C., Lupien, S.J., 2011. A clinical allostatic load index isassociated with burnout symptoms and hypocortisolemic profilesin healthy workers. Psychoneuroendocrinology 36 (6), 797—805.

Klotz, U., Kanto, J., 1988. Pharmacokinetics and clinical use offlumazenil (Ro 15-1788). Clin. Pharmacokinet. 14, 1—12.

Korkmaz, S., Wahlstrom, G., 1997. The EEG burst suppression thresh-old test for the determination of CNS sensitivity to intravenousanesthetics in rats. Brain Res. Brain Res. Protoc. 1, 378—384.

Langelaan, S., Bakker, A.B., Schaufeli, W.B., van Rhenen, W., vanDoornen, L.J., 2006. Do burned-out and work-engaged employeesdiffer in the functioning of the hypothalamic-pituitary-adrenalaxis? Scand. J. Work Environ. Health 32, 339—348.

Lupien, S.J., Fiocco, A., Wan, N., Maheu, F., Lord, C., Schramek, T.,2005. Stress hormones and human memory function across thelifespan. Psychoneuroendocrinology 30, 225—242.

Maguire, J., Ferando, I., Simonsen, C., Mody, I., 2009. Excitabilitychanges related to GABAA receptor plasticity during pregnancy. J.Neurosci. 29, 9592—9601.

Majewska, M.D., Harrison, N.L., Schwartz, R.D., Barker, J.L., Paul,S.M., 1986. Steroid hormone metabolites are barbiturate-likemodulators of the GABA receptor. Science 232, 1004—1007.

Marshall, R.W., Griffiths, A.N., Richens, A., 1985. A microcomputersystem to assess CNS depression from analysis of the dynamics ofsaccadic eye movements. Br. J. Pharmacol. 20, 304—305.

Marshall, R.W., Richens, A., 1989. An IBM-based system for thegeneration, collection and analysis of saccadic and smooth pur-suit eye movements. Br. J. Clin. Pharmacol. 20, 304—305.

Maslach, C., Schaufeli, W., Leiter, B.M.P., 2001. Job burnout. Annu.Rev. Psychol. 52, 397—422.

Matsumoto, K., Puia, G., Dong, E., Pinna, G., 2007. GABA(A) receptorneurotransmission dysfunction in a mouse model of social isola-tion-induced stress: possible insights into a non-serotonergicmechanism of action of SSRIs in mood and anxiety disorders.Stress 10, 3—12.

McCormack, H.M., Horne, D.J., Sheather, S., 1988. Clinical applica-tions of visual analogue scales: a critical review. Psychol. Med. 18,1007—1019.

Meczekalski, B., Tonetti, A., Monteleone, P., Bernardi, F., Luisi, S.,Stomati, M., Luisi, M., Petraglia, F., Genazzani, A.R., 2000.Hypothalamic amenorrhea with normal body weight: ACTH, allo-pregnanolone and cortisol responses to corticotropinreleasinghormone test. Eur. J. Endocrinol. 142, 280—285.

Melamed, S., Ugarten, U., Shirom, A., Kahana, L., Lerman, Y.,Froom, P., 1999. Chronic burnout, somatic arousal and elevatedsalivary cortisol levels. J. Psychosom. Res. 46, 591—598.

Moch, S.L., Panz, V.R., Joffe, B.I., Havlik, I., Moch, J.D., 2003.Longitudinal changes in pituitary-adrenal hormones in SouthAfrican women with burnout. Endocrine 21, 267—272.

Mody, I., Maguire, J., 2012. The reciprocal regulation of stresshormones and GABA(A) receptors. Front. Cell. Neurosci. 6, 4,http://dx.doi.org/10.3389/fncel.2012.00004 (Epub Jan 30).

Mommersteeg, P.M., Heijnen, C.J., Verbraak, M.J., van Doornen,L.J., 2006. Clinical burnout is not reflected in the cortisol awak-ening response, the day-curve or the response to a low-dosedexamethasone suppression test. Psychoneuroendocrinology31, 216—225.

Montgomery, S.A., Asberg, M., 1979. A new depression scale designedto be sensitive to change. Br. J. Psychiatr. 134, 382—389.

Morgan 3rd., C.A., Cho, T., Hazlett, G., Coric, V., Morgan, J., 2002.The impact of burnout on human physiology and on operationalperformance: a prospective study of soldiers enrolled in thecombat diver qualification course. Yale J. Biol. Med. 75, 199—205.

Nutt, D.J., Glue, P., Lawson, C., Wilson, S., 1990. Flumazenil provo-cation of panic attacks. Arch. Gen. Psychiatr. 47, 917—925.

Nyberg, S., Backstrom, T., Zingmark, E., Purdy, R.H., Sundstrom-Poromaa, I., 2007. Allopregnanolone decrease with symptomimprovement during placebo and gonadotropin-releasing hor-mone agonist treatment in women with severe premenstrualsyndrome. Gynecol. Endocrinol. 23, 257—266.

Nyberg, S., Wahlstrom, G., Backstrom, T., Sundstrom-Poromaa, I.,2004. Altered sensitivity to alcohol in the late luteal phase amongpatients with premenstrual dysphoric disorder. Psychoneuroen-docrinology 29, 767—777.

Peterson, R.A., Reiss, S., 1992. Anxiety Sensitivity Index Manual, 2ndedition. International Diagnostic Systems, Worthington, OH.

Pruessner, J.C., Hellhammer, D.H., Kirschbaum, C., 1999. Burnout,perceived stress, and cortisol responses to awakening. Psycho-som. Med. 61, 197—204.

Purdy, R.H., Morrow, A.L., Moore Jr., P.H., Paul, S.M., 1991. Stress-induced elevations of gamma-aminobutyric acid type A receptor-active steroids in the rat brain. Proc. Natl. Acad. Sci. U.S.A. 88,4553—4557.

Rasmusson, A.M., Pinna, G., Paliwal, P., Weisman, D., Gottschalk, C.,Charney, D., Krystal, J., Guidotti, A., 2006. Decreased cerebro-spinal fluid allopregnanolone levels in women with posttraumaticstress disorder. Biol. Psychiatr. 60, 704—713.

Reiss, S., Peterson, R.A., Gursky, D.M., McNally, R.J., 1986. Anxietysensitivity, anxiety frequency and the prediction of fearfulness.Behav. Res. Ther. 24, 1—8.

Roy-Byrne, P.P., Cowley, D.S., Greenblatt, D.J., Shader, R.I., Hom-mer, D., 1990. Reduced benzodiazepine sensitivity in panic dis-order. Arch. Gen. Psychiatr. 47, 534—538.

Rydmark, I., Wahlberg, K., Ghatan, P.H., Modell, S., Nygren, A.,Ingvar, M., Asberg, M., Helling, M., 2006. Neuroendocrine, cogni-tive and structural imaging characteristics of women on longtermsickleave with job stress-induced depression. Biol. Psychiatr. 60,867—873.

Sarkar, J., Wakefield, S., MacKenzie, G., Moss, S.J., Maguire, J.,2011. Neurosteroidogenesis is required for the physiological re-sponse to stress: role of neurosteroid-sensitive GABAA receptors.J. Neurosci. 31, 18198—18210.

Savic, I., Widen, L., Stone-Elander, S., 1991. Feasibility of reversingbenzodiazepine tolerance with flumazenil. Lancet 337, 133—137.

Schaufeli, W.B., van Dierendonck, D., 1995. A cautionary note aboutthe cross-national and clinical validity of cut-off points for theMaslach Burnout Inventory. Psychol. Rep. 76, 1083—1090.

Skilbeck, K.J., Johnston, G.A., Hinton, T., 2010. Stress and GABAreceptors. J. Neurochem. 112, 1115—1130.

Smith, S.S., Gong, Q.H., Hsu, F.C., Markowitz, R.S., ffrench-Mullen,J.M., Li, X., 1998. GABA(A) receptor alpha4 subunit suppressionprevents withdrawal properties of an endogenous steroid. Nature392, 926—930.

Sonnenschein, M., Mommersteeg, P.M., Houtveen, J.H., Sorbi, M.J.,Schaufeli, W.B., van Doornen, L.J., 2007. Exhaustion and endo-crine functioning in clinical burnout: an in-depth study using theexperience sampling method. Biol. Psychol. 75, 176—184.

Spielberger, C.D., Gorusch, R.L., Lushene, R.E., 1970. Manual for theState-Trait Anxiety Inventory. Consulting Psychology Press, PaloAlto, CA.

Stenlund, T., Ahlgren, C., Lindahl, B., Burell, G., Knutsson, A.,Stegmayr, B., 2007. Patients with burnout in relation to genderand a general population. Scand. J. Public Health 35, 516—523.

Strohle, A., Kellner, M., Holsboer, F., Wiedemann, K., 1999. Behav-ioral, neuroendocrine, and cardiovascular response to flumaze-nil: no evidence for an altered benzodiazepine receptorsensitivity in panic disorder. Biol. Psychiatr. 45, 321—326.

Stromberg, J., Backstrom, T., Lundgren, P., 2005. Rapid non-geno-mic effect of glucocorticoid metabolites and neurosteroids onthe gamma-aminobutyric acid-A receptor. Eur. J. Neurosci. 21,2083—2088.

Sundstrom, I., Nyberg, S., Backstrom, T., 1997. Patients with pre-menstrual syndrome have reduced sensitivity to midazolam

http://dx.doi.org/10.3389/fncel.2012.00004

http://dx.doi.org/10.3389/fncel.2012.00004


compared to control subjects. Neuropsychopharmacology 17,370—381.

Sur, C., Farrar, S.J., Kerby, J., Whiting, P.J., Atack, J.R., McKernan,R.M., 1999. Preferential coassembly of alpha4 and delta subunitsof the gamma-aminobutyric acidA receptor in rat thalamus. Mol.Pharmacol. 56, 110—115.

Taylor, S., Koch, W.J., McNally, R.J., 1992. How does anxietysensitivity vary across the anxiety disorders? J. Anxiety Disord.6, 249—259.

Timby, E., Balgard, M., Nyberg, S., Spigset, O., Andersson, A.,Porankiewicz-Asplund, J., Purdy, R.H., Zhu, D., Backstrom, T.,Sundstrom-Poromaa, I., 2006. Pharmacokinetic and behavioraleffects of allopregnanolone in healthy women. Psychopharma-cology (Berl). 186, 414—424.

Turkmen, S., Backstrom, T., Wahlstrom, G., Andreen, L., Johansson,I.M., 2011. Tolerance to allopregnanolone with focus on theGABA-A receptor. Br. J. Pharmacol. 162, 311—327.

Weizman, R., Weizman, A., Kook, K.A., Vocci, F., Deutsch, S.I.,Paul, S.M., 1989. Repeated swim stress alters brain benzodiaz-epine receptors measured in vivo. J. Pharmacol. Exp. Ther. 249,701—707.

World Health Organization, WHO, 2004. ICD-10: International Statis-tical Classification of Diseases and Related Health Problems, 2nded., 10th rev. WHO Press, World Health Organization, Geneva,Switzerland.

Yehuda, R., Seckl, J., 2011. Minireview: stress-related psychiatricdisorders with low cortisol levels: a metabolic hypothesis. Endo-crinology 152, 4496—4503.

increased neurosteroid sensitivity – an explanation to symptoms associated with chronic work...

Documents