Mm

DVa

b

a

ARRA

KEPSSV

1

atsmbbrar

((

h0

Psychoneuroendocrinology 70 (2016) 108–117

Contents lists available at ScienceDirect

Psychoneuroendocrinology

jo ur nal ho me p ag e: www.elsev ier .com/ locate /psyneuen

odulation of spatial and response strategies by phase of theenstrual cycle in women tested in a virtual navigation task

ema Hussain a, Sarah Hanafi b, Kyoko Konishi b, Wayne G. Brake a,∗,éronique D. Bohbot b

Centre for Studies in Behavioral Neurobiology, Department of Psychology, Concordia University, Montréal, CanadaDouglas Mental Health University Institute, McGill University, Montréal, Canada

r t i c l e i n f o

rticle history:eceived 21 August 2015eceived in revised form 5 May 2016ccepted 6 May 2016

eywords:strogenrogesteronepatial memorytimulus-response memoryirtual navigation

a b s t r a c t

Different memory systems are employed to navigate an environment. It has been consistently shown inrodents that estrogen impacts multiple memory system bias such that low estradiol (E2) is associated withincreased use of a striatal-mediated response strategy whereas high E2 increases use of a hippocampal-dependent spatial memory. Low E2 also enhances performance on a response-based task whereas highE2 levels improve learning on a spatial task. The purpose of the present cross-sectional study was toinvestigate navigational strategies in young, healthy, naturally cycling women. Participants were splitinto either an early follicular (i.e., when E2 levels are low), ovulatory (i.e., when E2 levels are high) ormid/late luteal (i.e., end of the cycle, when E2 levels decrease and progesterone levels rise) phase group,using self-reported date of the menstrual cycle. Serum hormone level measurements (E2, progesterone,testosterone) were used to confirm cycle phase assignment. Participants were administered a verbalmemory task as well as a virtual navigation task that can be solved by using either a response or spatialstrategy. Women tested in the ovulatory phase, under high E2 conditions, performed better on a verbal

memory task than women tested during the other phases of the cycle. Interestingly, women tested inthe mid/late luteal phase, when progesterone is high, predominantly used a spatial strategy, whereasthe opposite pattern was observed in the early follicular and ovulatory groups. Our data suggest thatthe specific memory system engaged differs depending on the phase of the menstrual cycle and may bemediated by both E2 and progesterone, rather than E2 alone.

© 2016 Elsevier Ltd. All rights reserved.

. Introduction

Multiple memory systems can be engaged to solve a task andid in navigating a complex environment. Different learning sys-ems were first documented by Tolman and colleagues (1946) whohowed that rats utilize different strategies to find their way in aaze (Tolman et al., 1946). Namely, several learning strategies can

e used: one is response strategy, which is a strategy that relies onody turns at specific points in the environment forming stimulus-

esponse associations, and the second is spatial strategy, which isllocentric, i.e. independent of the position of the observer andelies on forming stimulus–stimulus associations between land-

∗ Corresponding author.E-mail addresses: dema.hu@gmail.com (D. Hussain), shanafi9@gmail.com

S. Hanafi), kyoko.konishi@mail.mcgill.ca (K. Konishi), wayne.brake@concordia.caW.G. Brake), veronique.bohbot@mcgill.ca (V.D. Bohbot).

ttp://dx.doi.org/10.1016/j.psyneuen.2016.05.008306-4530/© 2016 Elsevier Ltd. All rights reserved.

marks in order to create a cognitive map of the environment. Thesesystems are dissociable, they can be competitive, and rely on dif-ferent brain regions to function optimally. The hippocampus isimplicated in spatial memory (O’Keefe and Nadel, 1978) whereasthe dorsal striatum (which includes the caudate nucleus) is crucialfor response memory (Packard et al., 1989). In a series of seminalstudies, spatial memory was significantly impaired when the hip-pocampal formation (fornix) was damaged and response memorywas impaired when the dorsal striatum was damaged (McDonaldand White, 1994, 1993). It has also been shown that rats ini-tially use hippocampus-dependent spatial memory early on in adual-solution maze task but this changes to striatum-dependentresponse memory with additional training, suggesting that the hip-pocampus and striatum have different temporal dynamics (Packard

and McGaugh, 1996). There is some evidence that multiple mem-ory system bias is influenced by dopamine (DA) levels in the dorsalstriatum, such that DA enhances response learning and DA blockadewithin the dorsal striatum impairs response learning (Daniel et al.,

oendo

2cta

aaw(v2a(ptrirsemSaubKmQ

thaleiHmcer2as

e(ihmomess(imFtcim

big

D. Hussain et al. / Psychoneur

006; Quinlan et al., 2013). Thus, these memory systems are disso-iable, such that as one is inactivated, other systems are engagedo navigate an environment (Packard and McGaugh, 1992; Packardnd White, 1991; Packard et al., 1989).

Evidence for the existence of multiple memory systems haslso been observed in humans; as in rodents, spatial memory isssociated with fMRI activity and grey matter in the hippocampushereas response memory is associated with the caudate nucleus

Bohbot et al., 2007, 2004; Iaria et al., 2003). Increased hippocampalolume is associated with navigational expertise (Maguire et al.,000; Woollett and Maguire, 2011), and individuals with a dam-ged hippocampus were shown to have impaired spatial memoryBohbot et al., 1998; Holdstock et al., 2000). Yet a damaged hip-ocampus does not impair the use of a response strategy to solvehe task (Bohbot et al., 2004). Though multiple memory systemsely on specific brain structures, there exist individual differencesn the type of strategies that are employed to navigate an envi-onment. However, with time and practice, spatial learners tend towitch to the less cognitively demanding and faster response strat-gy. This has been observed in both humans (Iaria et al., 2003) andale rats (Chang and Gold, 2003; Packard and McGaugh, 1996).

witching from a spatial strategy to a response strategy with timend practice is observed in male rodents; female rats persist insing a spatial strategy when estradiol (E2) levels are high. This haseen shown in naturally cycling rats (Korol et al., 2004; McElroy andorol, 2005) as well as ovariectomized rats receiving E2 replace-ent (Davis et al., 2005; Hussain et al., 2013; Korol and Kolo, 2002;uinlan et al., 2008).

Based on the rat literature, it is clear that E2 levels modulatehe use of memory systems; however, little is known about if andow this occurs in humans. It has been observed that E2 is associ-ted with changes in cognition in women; for example, E2 has beeninked with improved verbal memory (Maki et al., 2002; Mordecait al., 2008; Rosenberg and Park, 2002) whereas it is associated withmpaired performance on mental rotation tasks (Hampson, 1990;ausmann et al., 2000). Hippocampal volume changes across theenstrual cycle in women, i.e., high endogenous E2 levels are asso-

iated with an increase in hippocampal grey matter (Protopescut al., 2008). In addition, it has previously been found that estrogeneceptors are present in the human hippocampus (Osterlund et al.,000a,b). Thus, E2 could be structurally altering the hippocampusnd binding to estrogen receptors within this brain area to promotepatial memory.

Progesterone (P) has been shown to be associated with bothnhanced (Maki et al., 2001; Natale et al., 2001) and disruptingFreeman et al., 1992) effects on verbal memory in women. It ismportant to note that the majority of studies that are focused onormones and cognition in women, are carried out with a post-enopausal sample, taking hormone replacements. These samples

f women typically receive progestin with their hormone treat-ents that include E2, thus, very few studies have focused on the

ffects of P in isolation. P has been shown to increase hippocampalpine density when administered with E2, but these spine den-ities decrease more rapidly than when E2 is administered aloneWoolley and McEwen, 1993). P receptor function is dependent onnduction of E2 receptors (Lydon et al., 1995), which suggests that

any of the effects linked to P are also underscored by E2 action.urthermore, E2 and P are often studied separately so the interac-ion between the two hormones, and how this can potentially affectognitive function, is not well understood. E2 and P seem to workn concert to affect hippocampal function and, possibly, multiple

emory system bias.

Low levels of testosterone (T) produced by the ovaries have also

een found in some areas of the female brain, and can be convertednto E2 (Davis and Tran, 2001; Vierhapper et al., 1997). Andro-en receptors are also located in the hippocampus (Beyenburg

crinology 70 (2016) 108–117 109

et al., 2000), which indicates that T could be exerting an effect onhippocampus-dependent tasks, such as allocentric spatial memory.T has been shown to influence cognitive abilities in cycling women(Hausmann et al., 2000) and spatial cognition is affected by diurnalchanges in testosterone levels in women, such that T is correlatedwith enhanced performance on a visuo-spatial task (Moffat andHampson, 1996) and shorter latencies in finding a hidden platformin a virtual water maze task (Burkitt et al., 2007). Another studyalso demonstrated that performance on a virtual water maze taskin women is affected by the interaction between T function andandrogen receptor sensitivity (Nowak et al., 2014). These studiessuggest that testosterone could also be playing a role in multiplememory system bias.

In the current study, young, naturally cycling women weretested on the 4 on 8 virtual maze (4/8 VM), a dual-solution nav-igation task in which they could utilize either a spatial or responsestrategy to complete the experiment. Estradiol peaks towards theend of the follicular phase (ovulation) and then rises and plateausacross the luteal phase. The menstrual cycle is also marked bychanges in P levels such that they are low throughout the follic-ular phase while they peak and plateau in the luteal phase, beforedropping at the onset of menstruation (for review, see Ref.: Hussainet al., 2014; Mihm et al., 2011). E2 levels are higher in the ovulatoryphase compared to the early follicular phase, which is marked bylow E2 levels throughout menstruation. Based on animal studies, itwas hypothesized that women tested during the ovulatory phase,when E2 levels peak, would be more likely to use a spatial strategyand utilize more landmarks to navigate whereas those tested inthe early follicular phase, when E2 and P levels are low, would usea response strategy more often. Furthermore, it was expected thatthe ovulatory phase would be associated with a higher number oferrors and trials required to reach criterion, since spatial strategiesare more cognitively demanding and are associated with increasederrors on the 4/8 VM (Iaria et al., 2003).

In addition, a battery of standard neuropsychological tests thatmeasure verbal and visuo-spatial memory was administered inorder to test whether these functions change across the menstrualcycle; according to the existing literature, we would expect high E2levels to be associated with enhanced verbal memory (Maki et al.,2002; Mordecai et al., 2008; Rosenberg and Park, 2002) whereaslow E2 would be related to enhanced performance on a visuo-spatial task (Hampson, 1990; Hausmann et al., 2000). Finally, anIQ test as well as questionnaires measuring perceived stress andquality of sleep were carried out in order to ensure that participantstested in the early follicular, ovulatory, and mid/late luteal phaseof the menstrual cycle are similar in terms of cognitive function,stress levels, and sleep quality.

2. Methods

2.1. Participant characteristics

A total of 45 healthy, right-handed, regularly cycling (i.e., a men-strual cycle lasting between 25 and 34 days) women were tested(age: M = 30.31; SD = 3.38; range = 23-36; see Fig. 1 for the study out-line). All participants underwent a screening questionnaire over thephone to determine whether they were eligible to participate in thisstudy. Participants who reported a history of psychological or neu-rological illness, drug or alcohol abuse, had been pregnant withinthe past two years, currently breastfeeding, or had taken contra-ceptive medication within three months of testing were excluded.

Participants had, on average, 16.67 years of education (SD = 2.76)and a mean sleep score of 0.90 (SD = 0.15; sleep score = averagenumber of hours of sleep over the past week/ideal number of hoursof sleep for the individual). See Table 1 for all participant demo-

110 D. Hussain et al. / Psychoneuroendocrinology 70 (2016) 108–117

Fig. 1. Study outline. Dotted line denotes which steps were carried out on testing day. Asterisks denote tasks that were counterbalanced across participants (i.e., whetherneuropsychological tests were administered before the 4/8 VM task or vice versa).

Table 1Participant demographics, LSEQ, PSS, and neuropsychological measures.

Early Follicular Ovulatory Mid/Late Luteal

n = 11 n = 13 n = 21

M SD M SD M SD

Age (in years) 29.45 3.86 30.62 0.96 30.57 3.14Number of years of education 16.18 2.52 16.85 2.48 16.81 3.11Sleep ratiob 0.95 0.13 0.85 0.12 0.91 0.17Hours of exercise/week 4.64 3.50 3.14 1.82 3.10 3.46Number of alcoholic drinks/week 2.09 2.39 1.64 2.16 0.95 1.65Number of cigarettes/week 8.18 20.77 0.00 0.00 4.52 15.96

Menstrual cycle dataAge first period 12.73 1.62 12.81 1.55 12.52 1.33Mean cycle length 28.11 1.80 29.77 2.31 29.99 3.18Mean period length 4.67 1.02 5.13 0.87 5.45 0.99

Leeds Sleep Evaluation QuestionnaireGetting to Sleep 59.30 13.62 51.58 14.06 52.48 15.82Quality of Sleep 46.50 21.49 45.42 13.80 47.29 22.96Awakening from Sleep 43.91 22.21 47.12 14.41 51.40 22.97Behavior Following Wakefulness 50.21 21.30 54.40 14.75 56.29 22.03

Rey Auditory Verbal Learning Testc

Pre-Interference (total score over 5 trials)a 56.27 7.39 62.77 4.55 60.67 5.40Post-Interferencea 11.55 2.02 13.54 1.27 13.10 1.79Post-Delaya 11.82 1.89 13.92 1.26 13.38 1.43

Rey-Osterrieth Complex Figure Taskd

Copy 35.18 1.25 35.31 0.95 34.69 1.86Immediate Recall 22.82 9.12 26.69 5.85 24.31 4.88Delay 23.09 6.11 26.39 6.14 23.81 5.07IQ score 102.91 10.63 111.77 12.90 107.48 13.70Perceived Stress Scoree 22.82 9.09 24.04 5.94 23.55 6.37

Note. N = 45.a indicates statistically significant cycle phase difference.b Sleep Ratio = number of hours of sleep over last week/ideal number of hours of sleep.

).

gdwgeolDco

2q

rm

c RAVLT: maximum score per trial = 15.d RO: maximum score = 36.e PSS: maximum score = 56 (higher score indicates higher perceived stress levels

raphics split by cycle phase group. Comparison groups did notiffer in age, education, and sleep scores (p > 0.05). Participantsere administered a brief questionnaire pertaining to their video

ame experience; most participants reported no or very low lev-ls of prior experience with video games. Informed consent wasbtained from participants in accordance with local ethics guide-

ines. This study was approved by the Research Ethics Board at theouglas Mental Health University Institute and was carried out inollaboration with the Center for the Study of Behavioral Neurobi-logy (CSBN) at Concordia University in Montreal, Canada.

.2. Menstrual cycle phase measurement and hormonal profileuestionnaire

As part of the screening questionnaire, participants wereequired to have kept track of the start dates and durations of their

enses over at least the last six months and to provide the date of

their last period. Based on the information given, it was determinedwhether the participant had a regular menstrual cycle, which ischaracterized as a cycle lasting between 25 and 34 days from thestart of one period to just before the next one begins (Mihm et al.,2011).

On the day of testing, a hormonal profile questionnaire wasadministered to participants in order to gather more detailedinformation regarding regularity of menstrual cycles, past pregnan-cies, contraceptive and synthetic hormone history, and general lifehabits (e.g., exercise, smoking, drinking). The dates of the last sixperiods were used to determine any given woman’s average cyclelength, current day in cycle (testing date − date of last period) and,consequently, the phase of the cycle in which the participant was in

when tested. If the participant was tested when in days 1–7 of hercycle, she was in the early follicular phase of her menstrual cycle,which corresponds with low E2 and P levels (menses). If the partic-ipant was tested from days 13–17 of her cycle, she was considered

oendo

tatppmtpe2

2

pbtaq

2

tIwt12awp

d1ttaartdtwiOctfwTttsmt(noia

2

a

D. Hussain et al. / Psychoneur

o be in the ovulatory phase, when E2 levels are high and P levelsre low. If the participant was tested sometime between day 20 andhe end of the cycle, she was considered to be in the mid/late lutealhase of her cycle, when E2 levels are intermediate and P levelseak. T levels should be constant and low across all phases of theenstrual cycle. All women tested during days that fell outside of

hese ranges were excluded from all analyses that included cyclehase. For the final analysis, there were 11 women included in thearly follicular phase group, 13 in the ovulatory phase group, and1 included in the mid/late luteal phase group.

.3. Blood extraction and analysis

Blood samples (18 mL; 3 tubes of 6 mL) were collected fromarticipants following the completion of the testing session. Thelood samples were immediately placed in ice and later centrifugedo extract plasma. Samples were stored at −20 ◦C until furthernalysis. Finally, serum levels of E2, P, and T were analyzed anduantified.

.4. Neuropsychological tests, PSS, and sleep questionnaire

Participants were administered a battery of neuropsychologicalests in order to control for differences in cognitive function andQ. Furthermore, verbal and visuo-spatial memory was assessed

ith standard neuropsychological tests in order to compare levelso other studies in the literature (Bohbot et al., 1998; Hampson,990; Hausmann et al., 2000; Maki et al., 2002; Mordecai et al.,008; Rosenberg and Park, 2002). Questionnaires measuring stressnd quality of sleep were also given to ensure that these variablesould not differ across cycle phase group and have an effect on

erformance on the task.The Rey Auditory Verbal Learning Test (RAVLT) is a stan-

ard neuropsychological verbal memory test (Rey, 1941; Schmidt,996). A list of 15 words (list A) is read for five trials and after eachrial the participant is asked to verbally recall as many words ashey can remember. Next, an interference trial is provided whereby

different list of 15 words (list B) is read to the participant andgain the participant is asked to recall as many words as they canemember. Following this interference trial, the participant is askedo recall as many words as they can from list A. Finally, after a 30 minelay, the participant is again asked to recall as many words ashey can from list A. Performance was assessed with the number ofords recalled after interference and after the 30 min delay. Dur-

ng the 30 min delay, the Rey-Osterrieth Complex Figure task (RO;sterrieth, 1944) was administered. Participants were asked toopy a complex figure as accurately as possible. The figure was thenaken away and the participant was asked to immediately draw itrom memory (immediate recall). After a 30 min delay, participants

ere asked again to redraw the complex figure from memory. Theest for Non-Verbal Intelligence-3 (TONI; form A), a language-freeest of intelligence that avoids problems associated with adminis-ering verbal tests of intelligence to non-native English or Frenchpeakers (Brown et al., 1997), was also administered in order toeasure IQ scores. The perceived stress scale (PSS) was adminis-

ered in order to measure participants’ self-reported stress levelsCohen et al., 1983). Finally, the Leed’s Sleep Evaluation Question-aire (LSEQ) was administered to evaluate sleep quality by focusingn four domains: ease with which an individual falls asleep, qual-

ty of sleep, ease with which an individual awakens following sleep,nd behavior following wakening (Parrott and Hindmarch, 1978).

.5. Behavioral task (4/8 VM)

A virtual navigational task was adapted from a commerciallyvailable computer game (Unreal; Epic Games, Raleigh, NC). The

crinology 70 (2016) 108–117 111

virtual environment was composed of an eight-arm radial mazewith a central starting location (Bohbot et al., 2007; Iaria et al.,2003). The maze was surrounded by a landscape (mountains, sky,clouds), a tree, and a boulder. At the end of each arm, there wasa staircase leading to the location where, in some of the arms, anobject could be picked up. Objects were not visible from the cen-tral platform and only became visible when participants reachedthe end of the pathway. The experiment was carried out using astandard 17 in. computer screen while participants were seatedat a desk in a testing room. Participants used a keypad with for-ward, backward, left turn, and right turn buttons to move within theenvironment. Before testing, participants were given a habituationtrial in which they spent a few minutes moving in a virtual roomthat was different from the experimental environment to practicethe motor aspects of the task. Once the participants were comfort-able using the keypad and navigating the virtual environment, theexperimenter gave the instructions, and the experiment started.

The experiment consisted of successive trials that were com-posed of two parts. In Part 1, four of the eight arms were accessiblewith objects at the end of each arm; in Part 2, all arms were accessi-ble and objects were present in the four arms that had been blockedin Part 1 (see Fig. 2). The participants were told to retrieve all fourobjects from the accessible arms in Part 1 and remember whicharms they visited because they will need to avoid these arms inorder to find the objects in Part 2. An error consisted of an entryinto an arm that did not contain an object. As such, part 2 errorsconsisted of the dependent variable because it involved memoryfor the four previously visited pathways out of eight choices. Twodifferent sequences were used. In Part 1 of trial type A, arms 1, 3, 5,and 8 were accessible and contained an object; in Part 2, the fourobjects were located at the end of the four previously blocked arms(i.e., arms 2, 4, 6, and 7). In Part 1 of trial type B, arms 3, 4, 6, and8 were accessible; in Part 2, the objects were located at the end ofarms 1, 2, 5, and 7. A minimum of three trials were administered,with sequence order ABA. If the participant reached criterion (i.e.,made no errors during Part 2 of any of the first three trials), a probetrial was administered. If criterion was not reached during the firstthree trials, up to five additional trials were given (sequence A) untilthe participant made no errors during Part 2. For every trial, par-ticipants completed the trial once the four objects had been pickedup.

Part 1 of the probe trial was identical to trial type A (sequence A).In Part 2, however, the walls around the radial maze were raised toconceal the landscape, so that no landmarks were visible. Further-more, unbeknownst to the participant, eight objects were present(one at the end of each arm). If a spatial strategy was employed, thelandmarks present in the environment were necessary to performthe task; therefore, removing the landmarks from the environmentin Part 2 of the probe trial should result in an increase in errors. Con-versely, if participants were using a response strategy, no increasein errors should occur with this change (Iaria et al., 2003). The pat-tern of visited pathways was used to assess the number of probeerrors. Once the probe trial was completed, an additional type Atrial was given in order to measure whether participants shiftedtheir strategy following the probe trial.

At the end of the experiment, participants were debriefed andwere given a structured questionnaire to gather information onhow they solved the task from beginning to end. Participants werecategorized as using a response strategy if they used a sequenceor counting strategy (i.e., numbered the pathways) from a singlestarting point. If they used at least two landmarks and relied onthe spatial relationship among these landmarks and the objects,

without using a pattern, they were categorized as using a spatialstrategy. The initial strategy that was used by the participant wasassessed. Two raters independently evaluated the verbal reportsand assigned participants to a particular strategy group corre-

112 D. Hussain et al. / Psychoneuroendocrinology 70 (2016) 108–117

Fig. 2. Illustration of learning and probe trials of the 4/8 VM. Sample trial of the 4-on-8 virtual maze (4/8 VM) showing location of landmarks around 8-arm radial maze (left).In Part 1, 4 of the 8 pathways are blocked and participant must collect objects in open pathways. In Part 2, all pathways are now accessible and participant must rememberwhere the objects were located in part 1 and visit the previously blocked pathways to find objects. The arrow represents the starting position the participant begins eacht o to td rt 2, al

sTukieommdsmt

2

rt(fpa

ccntbsrtt

rial in. The stars represent the pathways that contain an object; participants must guring a probe trial (right). In Part 1, all landmarks are visible as in other trials. In Pa

andmarks to orient themselves.

ponding to the method used when navigating the environment.he independent judgments of the raters were correlated to eval-ate their consistency; inter-rater agreement was good (Cohen’sappa = 0.777). Raters discussed the cases that had conflicting rat-

ngs and adjusted them following agreement. Aside from proberrors, errors made during all Part 2 trials (which assess mem-ry for the location of objects in the maze), number of landmarksentioned in the verbal report, and trials to criterion were alsoeasured. Finally, the current study was a between-subjects design

ue to the strong practice effect associated with the standard ver-ion of the 4/8 VM task, which would decrease sensitivity of oureasures (e.g. probe trial) in participants tested multiple times on

his task.

.6. Statistical analysis

A series of one-way analysis of variance (ANOVA) tests were car-ied out in order to compare the early follicular, ovulatory, and mido late luteal groups on the neuropsychological measures: RAVLTtotal recall for first 5 trials, recall following interference, recallollowing delay) and RO (copy, immediate recall, delay). Similarly,articipants were compared across groups on IQ score (TONI), PSS,nd all LSEQ measures.

For the 4/8 VM task, one-way ANOVAs were carried out toompare the cycle phase groups on the average number of errorsommitted during all part 2 trials, number of trials to criterion,umber of probe errors, and number of landmarks mentioned inhe verbal report. These 4/8 VM task measures were also comparedetween response and spatial learners in order to observe whether

patial learners make more errors, require more trials to reach crite-ion, and mention more landmarks than response learners. Levene’sest for equality of variances was carried out for every ANOVA. Fur-hermore, Bonferroni post-hoc tests were carried out in order to

he end of the pathway in order to see the object. Sample screen shots of the 4/8 VM wall obsurs the landscape surrounding the maze so participants can no longer use

compare cycle phase groups on each of the dependent variables.Furthermore, a Chi square test was carried out to compare propor-tions of participants in the early follicular, ovulatory, and mid tolate luteal groups using an initial spatial or response strategy in thetask.

A series of Pearson’s correlation coefficients were computed inorder to test whether serum hormone levels (E2, P, T) correlatedwith any of the neuropsychological test measures, IQ, sleep mea-sures, or PSS. Additionally, Pearson’s correlation coefficients werecomputed in order to observe the relationships between hormonelevels and 4/8 VM task measures: average number of errors com-mitted during all part 2 trials, number of trials to criterion, numberof probe errors made during part 2 of the probe trial, and number oflandmarks mentioned in the verbal report. A one-way ANOVA wascomputed to compare hormone levels (E2, P, T) across the differ-ent cycle phases in order to assess accuracy of self-reported cyclephase information. Finally, three independent samples t-tests werecarried out to compare serum E2, P, and T levels between spatialand response learners.

3. Results

3.1. Neuropsychological tests, PSS, sleep questionnaires, andmenstrual cycle data

The average cycle day that women were tested on in theearly follicular group is 4.64 (SD = 1.80). For the ovulatory group,the average cycle day was 15.00 (SD = 1.41). For the mid/lateluteal group, the average cycle day was 25.86 (SD = 4.70). Results

reveal a statistically significant cycle group difference for all threeRAVLT measures: total recall during acquisition of the first 5 trials(F(2.42) = 3.97, p = 0.026), recall following interference (F(2.42) = 4.44,p = 0.018), and recall following a 30-min delay (F(2.42) = 6.24,

D. Hussain et al. / Psychoneuroendocrinology 70 (2016) 108–117 113

Fig. 3. Line graphs representing verbal memory across cycle phase. * Women tested in the ovulatory phase of the cycle recalled significantly more words than the earlyf s (left panel; p = 0.025), following interference (center panel; p = 0.021), and following ad recalled more words than the early follicular group follwing a delay (p = 0.024).

ptm(warwRdwt

3

mtVlpiswoarlicaotpIs

3

fr(l(dsv

Fig. 4. Bar graph representing proportion of participants using an initial spatial orresponse strategy in the 4-on-8 virtual maze in the early follicular, ovulatory, andmid/late luteal phases of the cycle. Women tested in the early follicular and ovultoryphases predominantly used a response strategy whereas the opposite was observed

ollicular group on the Rey Auditory Verbal Learning Test during the first five trialelay (right panel; p = 0.004). + Women tested during the mid/late luteal phase also

= 0.004). Post-hoc comparisons indicate that women tested inhe ovulatory phase, when E2 levels are high, recalled significantly

ore words in the first 5 trials (p = 0.025), following interferencep = 0.021), and following a time delay (p = 0.004), compared toomen tested during the early follicular phase, when E2 levels

re low. Furthermore, women tested in the early follicular phaseecalled significantly fewer words following a time delay thanomen tested in the mid/late luteal phase (p = 0.024; see Fig. 3 forAVLT). The early follicular, ovulatory, and mid/late luteal groupsid not differ in RO, LSEQ, IQ, and PSS scores (p > 0.05). In sum,omen learned and remembered significantly more words during

heir ovulatory phase when E2 levels are high.

.2. Behavioral task (4/8 VM)

Chi-square results revealed that participants tested in theid/late luteal phase predominantly used a spatial strategy (Spa-

ial = 76.2%) as opposed to using a response strategy in the 4/8M task, whereas the opposite was observed in the early follicu-

ar (Spatial = 36.4%) and ovulatory group (Spatial = 30.8%); X2 = 8.34, = 0.015 (see Fig. 4). This suggests that when women are tested dur-

ng the mid/late luteal phase, they were more likely to use a spatialtrategy than a response strategy to complete the task. Conversely,hen women are tested in the early follicular and ovulatory phases

f the cycle, they are more likely to use a response strategy than spatial strategy. These results also suggest that this pattern iseversed in the phases of the menstrual cycle that are marked byow or high E2 levels; thus, it would appear that spatial strategy uses not associated with high E2 levels. Furthermore, there were noycle phase group differences in terms of the general learning vari-bles, such as number of errors made during Part 2 trials, numberf landmarks used, probe errors, or trials to criterion. This indicateshat there were no differences in learning the 4/8 VM task acrosshase of the cycle, despite the differences observed in strategy use.

n sum, a significantly higher proportion of women used the spatialtrategy during the mid/late luteal phase, when P levels are high.

.3. Hormone levels

Serum hormone level data revealed that the self-report dataor determining participants’ cycle phase was accurate. Indeed, theesults show that E2 levels significantly differ across cycle phaseF(2.20) = 9.34, p = 0.001), with E2 levels being higher in the ovu-atory phase (M = 260.03 pg/ml) than in the early follicular phase

M = 82.53 pg/ml; p = 0.005); E2 levels in the mid/late luteal phaseid not differ from the other two phases (M = 161.31 pg/ml). Ithould be noted that Levene’s test for homogeneity of variance wasiolated (F(2.38) = 5.79, p = 0.006); thus, Welch’s corrected F-ratio

in women tested during the mid/late luteal phase; X2 = 8.34, p = 0.015. * signifies astatistically significant difference in proportion of women using a spatial or responsestrategy across the three cycle phases.

was used. The results also revealed that P levels differed signif-icantly across cycle phase group (F(2.18) = 8.17, p = 0.003), with Plevels being higher in the mid/late luteal phase (M = 6.35 ng/ml)than in the early follicular phase of the cycle (M = 0.22 pg/ml,p = 0.028). Again, Levene’s test for homogeneity of variance wasviolated for this ANOVA (F(2.38) = 10.28, p < 0.001); thus, Welch’scorrected F-ratio was used. As predicted, there was no difference inT values across cycle phase groups (see Fig. 5 for all hormone data).

Results revealed that hormone levels did not significantly cor-relate with any of the neuropsychological tests or the 4/8 VMmeasures. Finally, hormone levels did not statistically significantlydiffer between spatial and response learners, though there was atrend showing that spatial learners had lower E2 (p = 0.129) andT levels (p = 0.077) and higher P levels (p = 0.085) than responselearners (see Fig. 6).

4. Discussion

Consistent with other studies (Maki et al., 2002; Mordecaiet al., 2008; Rosenberg and Park, 2002), here, women learned andremembered more words during the ovulatory phase, when E2 lev-

114 D. Hussain et al. / Psychoneuroendocrinology 70 (2016) 108–117

Fig. 5. Serum estradiol and progesterone levels across menstrual cycle. Estradiol levels were higher in the ovulatory phase than in the early follicular phase (left panel;p = 0.005). * signifies statistically significant difference in estradiol levels between the early follicular and ovulatory phases. Progesterone levels were higher in the mid/lateluteal phase than in the early follicular phase (right panel; p = 0.028). * signifies statistically significant difference in progesterone levels between the early follicular andmid/late luteal phases.

F nse le( s. Non

eisoqhtrtriP

itotcdiitil

uatrs

ig. 6. Serum estradiol, progesterone, and testosterone levels in spatial and respocenter panel), and testosterone (right panel) levels in spatial and response learner

ls are high. Women tested in the mid/late luteal phase, when Ps high, use a spatial strategy significantly more than a responsetrategy. In contrast, during the early follicular (low E2 and P) andvulatory (high E2) phases, response strategy was used more fre-uently (see Figs. 4 and 5). Thus, these results do not support ourypothesis that a high E2 state would be associated with spa-ial strategy use and a low E2 state with higher proportion ofesponse strategy use, but they suggest that multiple memory sys-em bias in cycling women is mediated by changes in P such thatesponse memory is promoted when P is low, and spatial memorys enhanced in the phase of the cycle that is characterized by high

levels.No group differences were observed on any of the 4/8VM learn-

ng measures. Despite the difference observed in strategy acrosshe cycle, the menstrual cycle groups did not differ in the numberf trials needed to reach criterion or the number of errors made inhe task. This may be because the sample tested in this study wasomprised of young, healthy participants; perhaps the task was notifficult enough to observe a high number of errors or differences

n trials required to reach criterion. Indeed, out of the 45 partic-pants tested, only four of them required any extra trials (abovehe minimum 3 trials given before the probe) to reach criterion. Its therefore possible that group differences were obscured by theow variability in this measure.

The present findings indicate that ovarian hormones may mod-late memory systems such that the phase of the cycle favoring

high P state, promotes the use of hippocampus-dependent spa-ial strategies and decreases the use of caudate nucleus-dependentesponse strategies. To the best of our knowledge, this is the firsttudy to demonstrate a menstrual cycle phase-mediated effect

arners. This bar graph displays differences in estradiol (left panel), progesteronee of the differences were statistically significant.

on multiple memory systems in humans. The current results donot support previous findings observed in rodents showing thatE2 directly modulates multiple memory system bias in naturallycycling and ovariectomized females given E2 replacement, suchthat low E2 levels bias a female to use response memory and high E2levels are associated with the use of spatial memory (Hussain et al.,2013; Korol and Kolo, 2002; Korol et al., 2004; Quinlan et al., 2008;Zurkovsky et al., 2006, 2007, 2011). The current results suggest thathumans may be different from rodents in this regard. Here, boththe lowest and highest E2 levels observed across the cycle wereassociated with response strategy use. The difference observed inhow E2 impacts multiple memory system bias in rodents and inhumans could be explained by the significant difference betweenthe rat estrous cycle and the human menstrual cycle. In rats, E2and P peak concurrently, such that a high E2 phase is also markedby high P. Conversely, in women, the two hormones fluctuate dif-ferently and peak at different points in the cycle. It is possible thatovarian hormones interact differently in the human brain and, thus,affect cognitive functions, such as multiple memory system bias, ina unique way.

It has previously been demonstrated that a response strat-egy used in the 4/8 VM task corresponds with enhanced caudatenucleus activity and spatial strategy use correlates with enhancedhippocampal activity (Iaria et al., 2003). Similarly, response learnershave increased grey matter in the caudate nucleus and decreasedhippocampal grey matter; this pattern is reversed in spatial learn-

ers (Bohbot et al., 2007). It is possible that E2 affects spatial memoryin a curvilinear manner, such that low or high levels impair,whereas intermediate levels enhance, spatial memory. The way inwhich E2 and P fluctuate across the cycle could also be underscoring

oendo

ttsoma

fslip1Mt2ptats

sEciEbeEildDttrslEFDswiatst

dt2Ea((ctHhtdd

m

D. Hussain et al. / Psychoneur

his effect. As P levels peak in the mid/late luteal phase of the cycle,his could be interacting with E2 such that hippocampal spine den-ity increases, then decreases as hormone levels drop at the startf the next cycle. Thus, the hormonal milieu that characterizes theid/late luteal phase could be conducive to enhanced hippocampal

ctivity and promote spatial learning.Though the current results did not reveal any significant dif-

erences in T levels across the menstrual cycle, there was a trendhowing that spatial learners had lower T levels than responseearners (Fig. 6). These data are not consistent with previous find-ngs, where it was observed that T correlated with enhancederformance on spatial tasks in women (Moffat and Hampson,996; Burkitt et al., 2007). However, the spatial task used in theoffat and Hampson (1996) study relies on mental rotation abili-

ies, known to be dependent on the parietal cortex (Foroughi et al.,015; Harris and Miniussi, 2003). It is also possible that women’serformance on an allocentric memory spatial task is affected byhe interaction of circulating T and androgen receptor distributionnd sensitivity in the brain (Nowak et al., 2014). Thus, it is possiblehat T can have opposing effects on allocentric spatial memory andpatial cognition in women.

Recent findings hint at the possibility that this phase-mediatedhift in strategy could be occurring due to an interaction between2, P, and neurotransmitter systems. A recent study revealed thatycle-dependent changes in cognitive function could be due to E2nteracting with prefrontal DA levels, such that the direction of2′s effect on performance on a working memory task depends onaseline DA levels. Specifically, individuals with low basal DA lev-ls showed impaired performance on the task when endogenous2 levels were low whereas a low E2 state was associated with

mproved working memory in individuals with higher basal DAevels (Jacobs and D’Esposito, 2011). This finding is consistent withata from rodent studies that show an interaction between E2 andA transmission in the dorsal striatum and how this impacts mul-

iple memory system bias. Specifically, it has been demonstratedhat directly blocking DA binding with a locally administered DAeceptor antagonist in the dorsal striatum leads to a switch in thetrategy employed by rats in a dual solution maze task, such thatow E2 rats switched from response to spatial memory and high2 rats switched from spatial to response (Quinlan et al., 2013).urthermore, research has shown that P opposes the effect of E2 onA transmission; E2 is associated with increased DA levels, synthe-

is, and DA receptor density (Becker, 1999; Lévesque et al., 1989),hereas P has been shown to decrease E2 receptor density and

nhibit DA synthesis and activity (Dluzen and Ramirez, 1984; Luinend Rhodes, 1983; Mauvais-Jarvis et al., 1986). Thus, it is possiblehat differences in strategy use across cycle phase could be under-cored by cyclic changes in the combined effect of E2 and P on DAransmission.

Generally, when women are tested during periods of low E2 (e.g.,uring menstruation), they tend to perform better on tasks sensi-ive to a male advantage, such as mental rotation (Courvoisier et al.,013; Hausmann et al., 2000; Maki et al., 2002). Conversely, when2 is higher (e.g., luteal phase), performance on tasks sensitive to

female advantage, such as verbal fluency and memory, improvesMaki et al., 2002). However, many studies fail to find such effectsfor review, see: Sundström Poromaa and Gingnell, 2014). In theurrent study, the differences in verbal memory (as measured byhe RAVLT) across the menstrual cycle agree with previous studies.ere, women tested during ovulation, when E2 levels are at theirighest, recall significantly more words on the RAVLT than womenested in the early follicular phase (immediate, interference, and

elay; see Fig. 3). This suggests that verbal memory is enhanceduring the ovulatory phase, when E2 levels peak, in cycling women.

Functional neuroimaging studies have also revealed that theenstrual cycle-dependent flux in ovarian hormones is associated

crinology 70 (2016) 108–117 115

with changes in brain function and related behaviors. For instance,it has been shown that high blood E2 levels correlate with increasedcortical activity during completion of a word-stem and mental rota-tion tasks (Dietrich et al., 2001) and that brain activity in the parietaland prefrontal areas is correlated with E2 during a mental rota-tion task (Schöning et al., 2007). It has been shown that hormonalchanges across the cycle are also related with changes in lateraliza-tion during completion of a verbal task, such that lateralization ispronounced when E2 levels are low (Weis et al., 2008). E2 replace-ment in menopausal women has also been shown to alter brainactivity (for review, see Ref.: Comasco et al., 2014); the additionof a progestin to hormone therapy appears to oppose the effectsof E2 on functional brain activity. Though the current study didnot involve brain imaging, it could be hypothesized based on thesestudies that functional activity in the hippocampus and caudatenucleus could change across the menstrual cycle, and as a result ofthe combined effects of E2 and P, as women shift in strategy use.

5. Conclusion

The present study shows that menstrual cycle phase influencesthe type of memory system that is likely to be engaged in by womenwhen solving a task or effectively navigating a virtual environment.The mid/late luteal phase, when P is high, is associated with a signif-icant increase of spatial strategy. Conversely, a response strategy isused in the early follicular and ovulatory phase, when P is low. Thus,it would appear that multiple memory system bias is mediated bychanges in P and, possibly, how P and E2 interact. These findingssupport the growing body of research showing that cognitive func-tion is modulated by and change with fluctuating hormones acrossthe menstrual cycle.

Conflict of interest

The authors declare no conflicts of interest.

Contributors

All the authors have contributed equally in this work.

Role of funding source

The grant agencies that funded this research (NSERC, CIHR andFRQ-S) are at both at the federal and provincial levels in Canada.They have absolutely no influence on the study design, or the col-lection, analysis, or interpretation of the data. They had no influenceon the writing of the report or decision to submit the article.

Acknowledgements

Supported by a grant from the Natural Sciences and Engineer-ing Research Council of Canada (#06653; WGB) and the CanadianInstitutes of Health Research (#301763; VDB). The Douglas HospitalResearch Institute and the Centre for Studies in Behavioral Neuro-biology are both funded by the Fonds de Recherche du Québec −Santé.

References

Becker, J.B., 1999. Gender differenced in dopaminergic function in striatum andnucleus accumbens. Pharmacol. Biochem. 64, 803–812.

Beyenburg, S., Watzka, M., Clusmann, H., Blümcke, I., Bidlingmaier, F., Elger, C.E.,

Stoffel-Wagner, B., 2000. Androgen receptor mRNA expression in the humanhippocampus. Neurosci. Lett. 294, 25–28.

Bohbot, V.D., Kalina, M., Stepankova, K., Spackova, N., Petrides, M., Nadel, L., 1998.Spatial memory deficits in patients with lesions to the right hippocampus andto the right parahippocampal cortex. Neuropsychologia 36, 1217–1238.

1 oendo

B

B

B

B

C

C

C

C

D

D

D

D

D

F

F

H

H

H

H

H

H

I

J

K

K

L

L

L

M

M

16 D. Hussain et al. / Psychoneur

ohbot, V.D., Iaria, G., Petrides, M., 2004. Hippocampal function and spatialmemory: evidence from functional neuroimaging in healthy participants andperformance of patients with medial temporal lobe resections. Neuropsychol18, 418–425.

ohbot, V.D., Lerch, J., Thorndycraft, B., Iaria, G., Zijdenbos, A.P., 2007. Gray matterdifferences correlate with spontaneous strategies in a human navigation task.J. Neurosci. 27, 10078–10083.

rown, L., Johnsen, S.K., Sherbenou, R.J., 1997. Examiner’s Manual: Test ofNonverbal Intelligence, a language-free measure of cognitive ability, Thirdedition. Austin, Texas.

urkitt, J., Widman, D., Saucier, D.M., 2007. Evidence for the influence oftestosterone in the performance of spatial navigation in a virtual water mazein women but not in men. Horm. Behav. 51, 649–654.

hang, Q., Gold, P.E., 2003. Switching memory systems during learning: changes inpatterns of brain acetylcholine release in the hippocampus and striatum inrats. J. Neurosci. 23, 3001–3005.

ohen, S., Kamarck, T., Mermelstein, R., 1983. A global measure of perceived stress.J. Health Soc. Behav. 24, 385–396.

omasco, E., Frokjaer, V.G., Sundström Poromaa, I., 2014. Functional and molecularneuroimaging of menopause and hormone replacement therapy. Front.Neurosci. 8, 388.

ourvoisier, D.S., Renaud, O., Geiser, C., Paschke, K., Gaudy, K., Jordan, K., 2013. Sexhormones and mental rotation: an intensive longitudinal investigation. Horm.Behav. 63, 345–351.

aniel, J.M., Sulzer, J.K., Hulst, J.L., 2006. Estrogen increases the sensitivity ofovariectomized rats to the disruptive effects produced by antagonism of D2but not D1 dopamine receptors during performance of a response learningtask. Horm. Behav. 49, 38–44.

avis, S.R., Tran, J., 2001. Testosterone influences libido and well being in women.Trends Endocrin. Met. 12, 33–37.

avis, D.M., Jacobson, T.K., Aliakbari, S., Mizumori, S.J.Y., 2005. Differential effectsof estrogen on hippocampal- and striatal-dependent learning. Neurobiol. LearnMem. 84, 132–137.

ietrich, T., Krings, T., Neulen, J., Wilmes, K., Erberich, S., Thron, A., Sturm, W.,2001. Effects of blood estrogen level on cortical activation patterns duringcognitive activation as measured by functional MRI. Neuroimage 13, 425–432.

luzen, D.E., Ramirez, V.D., 1984. Bimodel effect of progesterone on in vitrodopamine function of the rat corpus striatum. Neuroendocrinology 39,149–155.

oroughi, C.K., Blumberg, E.J., Parasuraman, R., 2015. Activation and inhibition ofposterior parietal cortex have bi-directional effects on spatial errors followinginterruptions. Front. Syst. Neurosci. 8, 245.

reeman, E.W., Weinstock, L., Rickels, K., Sondheimer, S.J., Coutifaris, C., 1992. Aplacebo-controlled study of effects of oral progesterone on performance andmood. Br. J. Pharmacol. 33, 293–298.

ampson, E., 1990. Estrogen-related variations in human spatial andarticulatory-motor skills. Psychoneuroendocrinology 15, 97–111.

arris, I.M., Miniussi, C., 2003. Parietal lobe contribution to mental rotationdemonstrated with rTMS. J. Cogn. Neurosci. 15, 315–323.

ausmann, M., Slabbekoorn, D., Goozen, S.H.M., Cohen-Kettenis, P.T., Gunturkun,O., 2000. Sex hormones affect spatial abilities during the mentrual cycle.Behav. Neurosci. 114, 1245–1250.

oldstock, J.S., Mayes, A.R., Cezayirli, E., Isaac, C.L., Aggleton, J.P., Roberts, N., 2000.A comparison of egocentric and allocentric spatial memory in a patient withselective hippocampal damage. Neuropsychologia 38, 410–425.

ussain, D., Hoehne, A., Woodside, B., Brake, W.G., 2013. Reproductive experiencemodifies the effects of estradiol on learning and memory bias in female rats.Horm. Behav. 63, 418–423.

ussain, D., Shams, W.M., Brake, W.G., 2014. Estrogen and memory system bias infemales across the lifespan. Transl. Neurosci. 5, 35–50.

aria, G., Petrides, M., Dagher, A., Pike, B., Bohbot, V.D., 2003. Cognitive strategiesdependent on the hippocampus and caudate nucleus in human navigation:variability and change with practice. J. Neurosci. 23, 5945–5952.

acobs, E., D’Esposito, M., 2011. Estrogen shapes dopamine-dependent cognitiveprocesses: implications for women’s health. J. Neurosci. 31, 5286–5293.

orol, D.L., Kolo, L.L., 2002. Estrogen-induced changes in place and responselearning in young adult female rats. Behav. Neurosci. 116, 411–420.

orol, D.L., Malin, E.L., Borden, K.A., Busby, R.A., Couper-Leo, J., 2004. Shifts inpreferred learning strategy across the estrous cycle in female rats. Horm.Behav. 45, 330–338.

évesque, D., Gagnon, S., Di Paolo, T., 1989. Striatal D1 dopamine receptor densityfluctuates during the rat estrous cycle. Neurosci. Lett. 98, 345–350.

uine, V.N., Rhodes, J.C., 1983. Gonadal hormone regulation of MAO and otherenzymes in hypothalamic areas. Neuroendocrinology 36, 235–241.

ydon, J.P., DeMayo, F.J., Funk, C.R., Mani, S.K., Hughes, A.R., Montgomery Jr., C.A.,Shyamala, G., Coneely, O.M., O’Malley, B.W., 1995. Mice lacking progesteronereceptor exhibit pleiotropic reproductive abnormalities. Genes Dev. 9,2268–2278.

aguire, E.A., Gadian, D.G., Johnsrude, I.S., Good, C.D., Ashburner, J., Frackowiak,R.S., Frith, C.D., 2000. Navigation-related structural change in the hippocampiof taxi drivers. Proc. Natl. Acad. Sci. U.S.A. 97, 4398–4403.

aki, P.M., Zonderman, A.B., Resnick, S.M., 2001. Enhanced verbal memory innondemented elderly women receiving hormone-replacement therapy. Am. J.Psychiatry 158, 227–233.

crinology 70 (2016) 108–117

Maki, P.M., Rich, J.B., Rosenbaum, R.S., 2002. Implicit memory varies across themenstrual cycle: estrogen effects in young women. Neuropsychologia 40,518–529.

Mauvais-Jarvis, P., Kuttenn, F., Gompel, A., 1986. Antiestrogen action ofprogesterone in breast tissue. Breast Cancer Res. Treat. 8, 179–188.

McDonald, R.J., White, N.M., 1993. A triple dissociation of memory systems:hippocampus, amygdala, and dorsal striatum. Behav. Neurosci. 107, 3–22.

McDonald, R.J., White, N.M., 1994. Parallel information processing in the watermaze: evidence for independent memory systems involving dorsal striatumand hippocampus. Behav. Neural Biol. 61, 260–270.

McElroy, M.W., Korol, D.L., 2005. Intrahippocampal muscimol shifts learningstrategy in gonadally intact young female rats. Learn. Mem. 12, 150–158.

Mihm, M., Gangooly, S., Muttukrishna, S., 2011. The normal menstrual cycle inwomen. Anim. Reprod. Sci. 124, 229–236.

Moffat, S.D., Hampson, E., 1996. A curvilinear relationship between testosteroneand spatial cognition in humans: possible influence of hand preference.Psychoneuroendocrinology 21, 323–337.

Mordecai, K.L., Rubin, L.H., Maki, P.M., 2008. Effects of menstrual cycle phase andoral contraceptive use on verbal memory. Horm. Behav. 54, 286–293.

Natale, V., Albertazzi, P., Zini, M., Di Micco, R., 2001. Exploration of cyclical changesin memory and mood in postmenopausal women taking sequential combinedoestrogen and progestogen preparations. Brit. J. Obstet. Gynaec. 108, 286–290.

Nowak, N.T., Diamond, M.P., Land, S.M., Moffat, S.D., 2014. Contributions of sex,testosterone: and androgen Receptor CAG repeat number to virtual Morriswater maze performance. Psychoneuroendocrinology 41, 13–22.

O’Keefe, J., Nadel, L., 1978. The Hippocampus as a Cognitive Map. Clarendon PressOxford.

Osterlund, M.K., Gustafsson, J.A., Keller, E., Hurd, Y.L., 2000a. Estrogen receptorbeta (ERbeta) messenger ribonucleic acid (mRNA) expression within thehuman forebrain: distinct distribution pattern to ERalpha mRNA. J. Clin.Endocrinol. Metab. 85, 3840–3846.

Osterlund, M.K., Keller, E., Hurd, Y.L., 2000b. The human forebrain has discreteestrogen receptor alpha messenger RNA expression: high levels in theamygdaloid complex. Neuroscience 95, 333–342.

Osterrieth, P.A., 1944. Filetest de copie d’une figure complex: contribution a l’etudede la perception et de la memoire [The test of copying a complex figure: acontribution to the study of perception and memory]. Archi. Psychol. 30,286–356.

Packard, M.G., McGaugh, J.L., 1992. Double dissociation of fornix and caudatenucleus lesions on acquisition of two water maze tasks: further evidence formultiple memory systems. Behav. Neurosci. 106, 436–446.

Packard, M.G., McGaugh, J.L., 1996. Inactivation of hippocampus or caudatenucleus with lidocaine differentially affects expression of place and responselearning. Neurobiol. Learn Mem. 65, 65–72.

Packard, M.G., White, N.M., 1991. Dissociation of hippocampus and caudatenucleus memory systems by posttraining intracerebral injection of dopamineagonists. Behav. Neurosci. 105, 195–306.

Packard, M.G., Hirsh, R., White, N.M., 1989. Differential effects of fornix andcaudate nucleus lesions on two radial maze tasks: evidence for multiplememory systems. J. Neurosci. 9, 1465–1472.

Parrott, A.C., Hindmarch, I., 1978. Factor analysis of a sleep evaluationquestionnaire. Psychol. Med. 8, 325–329.

Protopescu, X., Butler, T., Pan, H., Root, J., Altemus, M., Polanecsky, M., McEwen, B.,Silbersweig, D., Stern, E., 2008. Hippocampal structural changes across themenstrual cycle. Hippocampus 18, 985–988.

Quinlan, M.G., Hussain, D., Brake, W.G., 2008. Use of cognitive strategies in rats: therole of estradiol and its interaction with dopamine. Horm. Behav. 53, 185–191.

Quinlan, M.G., Almey, A., Caissie, M., LaChappelle, I., Radiotis, G., Brake, W.G., 2013.Estradiol and striatal dopamine receptor antagonism influence memorysystem bias in the female rat. Neurobiol. Learn Mem. 106, 221–229.

Rey, A., 1941. L’examen psychologique dans les cas d’encéphalopathietraumatique. (Les problèmes). [The psychological examination in cases oftraumatic encephalopathy. Problems.]. Arch. Psychol. 28, 215–285.

Rosenberg, L., Park, S., 2002. Verbal and spatial functions across the menstrualcycle in healthy young women. Psychoneuroendocrinology 27, 835–841.

Schöning, S., Engelien, A., Kugel, H., Schäfer, S., Schiffbauer, H., Zwitserlood, P.,Pletziger, E., Beizai, P., Kersting, A., Ohrmann, P., Greb, R.R., Lehmann, W.,Heindel, W., Arolt, V., Konrad, C., 2007. Functional anatomy of visuo-spatialworking memory during mental rotation is influenced by sex, menstrual cycle,and sex steroid hormones. Neuropsychologia 5, 3203–3214.

Schmidt, M. 1996. Rey auditory verbal learning test: A handbook. Los Angeles,California.

Sundström Poromaa, I., Gingnell, M., 2014. Menstrual cycle influence on cognitivefunction and emotion processing-from a reproductive perspective. Front.Neurosci. 8, 380.

Tolman, E.C., Ritchie, B.F., Kalish, D., 1946. Studies in spatial learning; placelearning versus response learning. J. Exp. Psychol. 36, 221–229.

Vierhapper, H., Nowotny, P., Waldhäusl, W., 1997. Determination of testosteroneproduction rates in men and women using stable Isotope/Dilution and massspectrometry 1. J. Clin. Endocrinol. Metab. 82, 1492–1496.

Weis, S., Hausmann, M., Stoffers, B., Vohn, R., Kellerman, T., Sturm, W., 2008.

Estradiol modulates functional brain organization during the menstrual cycle:an analysis of interhemispheric inhibition. J. Neurosci. 10, 13401–13410.

Woollett, K., Maguire, E.A., 2011. Acquiring the knowledge of London’s layoutdrives structural brain changes. Curr. Biol. 21, 2109–2114.

oendo

W

Z

D. Hussain et al. / Psychoneur

oolley, C.S., McEwen, B.S., 1993. Roles of estradiol and progesterone in regulation

of hippocampal dendritic spine density during the estrous cycle in the rat. J.Comp. Neurol. 336, 293–306.

urkovsky, L., Brown, S.L., Boyd, S.E., Fell, J.A., Korol, D.L., 2007. Estrogen modulateslearning in female rats by acting directly at distinct memory systems.Neuroscience 144, 26–37.

crinology 70 (2016) 108–117 117

Zurkovsky, L., Serio, S.J., Korol, D.L., 2011. Intra-striatal estradiol in female rats

impairs response learning within two hours of treatment. Horm. Behav. 60,470–477.

Zurkovsky, L., Brown, S.L., Korol, D.L., 2006. Estrogen modulates place learningthrough estrogen receptors in the hippocampus. Neurobiol. Learn Mem. 86,336–343.