Physiological arousal in females with fragile X or Turner syndrome

14 Pages·2002·0.13 MB·English

Cynthia S. Keysor
Departmentof Psychiatry
Physiological Arousal in
KennedyKriegerInstitute
Baltimore,MD 21211and
Departmentof Psychiatryand
Females With Fragile X or
BehavioralSciences
JohnsHopkins UniversitySchool of
Medicine
Turner Syndrome
Baltimore, MD21205
Miche´le M. M. Mazzocco
KennedyKriegerInstitute
Baltimore, MD21205
Departmentof Psychiatryand
BehavioralSciences
JohnsHopkins UniversitySchool of
Medicine
Baltimore,MD 21205and
Departmentof Populationand
Family HealthSciences
JohnsHopkins UniversitySchool of
Public Health
Baltimore, MD21205
Daniel R. McLeod
Rudolf Hoehn-Saric
Departmentof Psychiatryand
BehavioralSciences
JohnsHopkins UniversitySchool of
Medicine
Baltimore, MD21205
Received12February2001;Accepted14August2001
ABSTRACT: Physiological arousal was measured in 12- to 22-year-old females with either
fragileX,Turnersyndrome,orneitherdisordertoexplorepotentialdifferencesinthemanifestation
of arousal and anxiety in adolescents and young women. Physiological arousal was measured at
baseline and during performance on mental arithmetic, divided attention, and risk-taking tasks.
Contrarytoprediction,femaleswithfragileXrarelyexhibitedhigherarousalthanfemalesineither
the Turner syndrome or comparison groups. On the Divided Attention Task, both the fragile X
and Turner syndrome groups exhibited higher arousal relative to one another based on different
physiological indices. Relative to the comparison group, the fragile X group presented with a
heightenedstateofarousalatbaseline,basedonmeanskinconductancerange,whichmayaccount
forlittleincreaseinarousalonthecognitivetasksforfemaleswithfragileX.FemaleswithTurner
syndromeexhibitedhigherarousalrelativetothecomparisongrouponallcognitivetasks,butnotfor
Correspondenceto:Miche´leM.M.Mazzocco
E-mail:[email protected]
Contractgrantsponsor:NIH
Contractgrantnumbers:NSP5035359andR01HD34061-04
(cid:1)2002WileyPeriodicals,Inc.
134 Keysoretal.
allphysiologicalmeasures.FactorspotentiallyassociatedwithheightenedarousalinfragileXand
Turnersyndromearediscussed.(cid:1)2002WileyPeriodicals,Inc.DevPsychobiol41:133–146,2002.
PublishedonlineinWileyInterScience(www.interscience.wiley.com). DOI10.1002/dev.10060
Keywords: psychophysiology; arousal;anxiety; fragileX; TurnerSyndrome
FragileXandTurnersyndromesaredistinctdisorders anxietydisordersinthispopulationhasvariedpartlyas
with different genotypes, phenotypes, and clinical afactorofmethodologicaldifferencesinstudydesign
manifestations. Both involve X chromosome gene (Keysor & Mazzocco, 2002. The reliability of self-
function, and there is some similarity between their reportmeasures,usedprimarilywithadultwomenwith
respective global psychological phenotype descrip- fragileX,hasbeenquestionedbecausehighLie-Scale
tions.Inparticular,socialskillsdifficulties,including scores on the Minnesota Multiphasic Personality
heightened anxiety or difficulty dealing with anxiety, Inventory–II (Hathaway & McKinley, 1989) suggest
havebeenreportedforgirlswithfragileXandTurner atendencyforthesewomentopresentthemselvesina
syndrome.Inthepresentstudy,thenatureofanxietyin positivethoughunrealisticlight(Sobesky,Pennington,
eachdisorderisindirectlyexploredthroughstudiesof Porter,Hull,&Hagerman,1994;Sobeskyetal.,1995).
physiologicalarousalin12-to22-year-oldfemales.Of Kovar (1995) cautions that underreporting may not
interestiswhetherarousallevelsdifferacrossthesetwo representdenialorunawarenessofsymptoms;instead,
populationsandwhetherthesedifferenceshaveimpli- women with fragile X may be aware of their deficits
cationsforhowanxietyineachdisorderismanifested. without recognizing the magnitude of their impair-
The rationale for this research approach is discussed ment.Interestingly,Kovar’shypothesisparallelsfind-
followingabriefreviewofeachdisorder. ings that patients with generalized anxiety disorder
accurately report the direction, butnot the degree, of
changes in skin conductance and heart rate experi-
FRAGILE X SYNDROME enced under stress (McLeod, Hoehn-Saric, & Stefan,
1986). Viewed as a potential indicator of anxiety,
FragileXsyndrometypicallyresultsfromamutation physiological arousal may be a useful indicator of
of a single X chromosome gene, referred to as the distress levels in this population, which could aid in
FMR1 (fragile X mental retardation 1) gene. FMR identificationofsituationsandfactorsassociatedwith
proteinproductionisdisruptedasaconsequenceofthis increasedanxiety.
gene mutation, and it is believed that the diminished Thereisevidencethathyperarousalandhypersens-
FMRproteinlevelsleadtotheclinicalmanifestations itivity characterize the phenotype of fragile X syn-
offragileXsyndrome(Pierettietal.,1991).Thepre- drome,particularlyformales(reviewedbyHagerman,
valence of this disorder is approximately 1 in 4,000 1996), which enhances their intolerance of physiolo-
malesand1in8,000females(Turner,Webb,Wake,& gical responses to environmental stimuli. In a pre-
Robinson,1996),withmalesbeingsignificantlymore liminarystudy,5-to37-year-oldmaleswithfragileX
impairedthanmostaffectedfemales.Themajorityof (N¼10) exhibited higher tonic levels of skin con-
maleswithfragileX(Bailey,Hatton,&Skinner,1998) ductance in response to eye contact during conversa-
and approximately half of females (Rousseau et al., tion than did males with either Down Syndrome or
1994) have mental retardation. Shyness and social attention deficit hyperactivity disorder (Belser &
anxiety reported for both males and females may Sudhalter, 1995). This finding suggests that arousal
contributetosocialskillsdeficits,suchasavoidanceof may underlie the social avoidant behavior of males
eye contact and difficulty initiating and maintaining with fragile X. Electrodermal response (EDR), an
conversation (Hagerman, 1996; Mazzocco, Kates, alternative measure of sympathetic nervous system
Baumgardner,Freund,&Reiss,1997;Sobesky,Porter, (SNS)activitythatmeasuresquick,phasicresponsesto
Pennington,&Hagerman,1995). stimuli,wasenhancedin4-to49-year-oldmaleswith
Symptomsofshynessandsocialanxietyhavebeen fragileX(N¼15)relativetoage-andgender-matched
retrospectively reported in females with fragile X as controls (Miller et al., 1999). The fragile X group
earlyasthepreschoolyears(Freund,Reiss,&Abrams, exhibitedheightenedarousalinresponsetoavarietyof
1993)andalsoinadulthood(Mazzocco,Baumgardner, stimuli and had lower rates of habituation. An
Freund,&Reiss,1998;Sobeskyetal.,1995).Areview association between higher sympathetic arousal and
oftherecentliteraturesuggeststhattheprevalenceof deficientFMRproteinalsowasreportedamongmales
Arousalin Fragile Xor Turner Syndrome 135
(n¼19) and females (n¼6) with fragile X. This have suggested that girls with Turner syndrome may
association suggests that FMR protein may affect havedifficultydealingwithanxiety.Inarecentstudy
the balance of sympathetic/parasympathetic systems, using self-report measures (McCauley et al., 2001),
leadingtoanimbalancethatfavorsgreatersympathetic 13- to 18-year-old girls with Turner syndrome rated
activity(Milleretal.,1999).However,inmorerecent themselves as slightly less anxious on the Revised
studies involvingyoung boys with fragile X(N¼29, Children’s Manifest Anxiety Scale (Reynolds &
range¼1–11 years), no such association was found Richmond, 1978) than girls in a comparison group,
between FMR protein levels and heart rate measures yettheyalsohadhigherLie-Scalescoressuggestiveof
(Roberts, Boccia, Bailey, Hatton, & Skinner, 2001). inaccurateself-reporting.Thisresponsepatternraises
Hyperarousalinthispopulationmaybeexplainedless questions about the reliability of self-report based
by enhanced sympathetic activity and more by de- assessments of anxiety in females with Turner syn-
creased parasympathetic activity (Boccia & Roberts, drome.Withnoknownstudiesofphysiologicalarousal
2000).Whenheartperiodmeasuresinyoungboyswith infemaleswithTurnersyndrome,thepresentresearch
fragile X (N¼20, mean¼4.03 years) were recorded measuresarousalasapotentialindicatorofanxietyin
during alternating passive or cognitive tasks, boys thispopulation.
with fragile X recorded different patterns of arousal
regulation across experimental phases, with a shorter
heartperiod(i.e.,fasterheartrate)butlowervagaltone ANXIETY, AROUSAL, AND
relative to chronologically age-matched peers. The PSYCHOPHYSIOLOGY
latter is indicative of a less active parasympathetic
systemandthusdecreasedabilitytoregulatebehavior Anxietyhasbeendefinedas‘‘anunpleasantsubjective
(Boccia&Roberts,2000).Consideredtogether,these experience of tension, apprehension, or anticipation,
findingssuggestthatmeasuresofphysiologicalarousal imposedbytheexpectationofdangerordistressorthe
may be useful for understanding autonomic nervous need for a special effort’’ (Kelly, Brown, & Shaffer,
systemfunctionanditsassociationwithanxietybeha- 1970,p.429).Italsohasbeendescribedas‘‘acomplex
viors in both males and females with fragile X syn- phenomenon that includes heightened arousal, incre-
drome,possiblyincludingshynessandsocialanxiety. asedmuscularandautonomictonus,aswellascognitive
Todate,theseassociationsareunclear. efforts to conceptualize its apparent cause’’ (Hoehn-
Saric,Hazlett,Pourmotabbed,&McLeod,1997,p.49).
Physiologically, anxiety is associated with increased
TURNER SYNDROME bloodflowinthemusclesanddecreasedbloodflowin
the skin (Blair, Glover, Greenfield, & Roddie, 1959),
Turnersyndromeresultsfromthepartialorcomplete withvariousstimulihavingdifferentialeffectsonthe
absenceofthesecondXchromosomethatispresentin sympathetic impulses that are discharged in the skin
normal female development.Itoccurspredominantly and muscle nerves (Gellhorn, 1965). For example,
in females, with a prevalence of approximately 1 in mental stress such as experienced on a mental arith-
2,000 to 1 in 5,000 live female births (Hook & metictaskresultsingreatersympatheticactivityinthe
Warburton,1983).Itsphysicalphenotypeisassociated, skinthanwithinthemuscles(Steptoe,1987).Heartrate
at least in part, with deficiencies in growth hormone is expected to become higher and more stabile under
and estrogen (Hall & Gilchrist, 1990), and these stress compared to normal resting conditions (Porges
deficiencies appear associated with aspects of the &Raskin,1969).
psychological phenotype (Ross, Roeltgen, Feuillan, Nonanxious individuals may experience a heigh-
Kushner, & Cutler, 1998, 2000). Turner syndrome tened physiological response to novel situations, but
leads to a lowering of global intellectual scores, but theygenerallyreturntoalowerautonomicstateearlier,
mentalretardationinfemaleswithTurnersyndromeis and habituate more quickly, than do individuals with
nomorefrequentthaninthegeneralpopulation.Social anxietydisorders;variationinthispatterndependson
skills deficits have been implicated as a component the nature of the stimuli and type ofanxiety disorder
ofthepsychologicalphenotype,particularlywithres- (Hoehn-Saric & McLeod, 1988). Among indivi-
pect to poor peer relationships and low self-esteem dualswithanxietydisorders,heightenedphysiological
(McCauley, Feuillan, Kushner, & Ross, 2001; arousal is notnecessarilyassociated with self-reports
McCauley,Ross,Kushner,&Cutler,1995;McCauley, ofincreasedanxiety.Thisresponsepatternwasfound
Sybert,&Ehrhardt,1986;Rovet&Ireland,1994). foradultpatientswithvariousanxietydisorderswhen
Althoughsignificantpsychopathologyhasnotbeen measures were taken at rest (Hofmann, Newman,
implicatedinTurnersyndrome,Mambellietal.(1996) Ehlers, & Roth, 1995; Kelly et al., 1970; McLeod,
136 Keysoretal.
Hoehn-Saric, Zimmerli, de Souza, & Oliver, 1990; at risk for later social behavior problems (Doussard-
Tyrer, Lee, & Alexander, 1980) and during perfor- Roosevelt,McClenny,&Porges,2001).
manceonamentalarithmetictask(Kellyetal.,1970). Thepurposeofthepresentstudywastodetermine
In children, cardiac measures have often been if adolescent and young adult females with fragile X
used as indicators of arousal in response to stressful or Turner syndrome differ in level of physiological
situations.Inonelongitudinalstudy,extremelylowor arousal, and whether these differences have potential
highscoresonassessmentclassified21-month-oldsas implications for how anxiety is manifested in each
eitherinhibitedoruninhibited(Garcia-Coll,Kagan,& disorder. Females in either the fragile X or Turner
Reznick, 1984). Inhibited children were consistently syndromegroup,andthosewithneitherdisorderinthe
shy, quiet, and timid, and compared to uninhibited comparison group, were assessed on several physio-
children,theyhadhigherbutmorevariableheartrates logical measures of arousal and on the self-report
on an information-processing task. Reassessment Subjective Units of Distress Scale (SUDS). These
1monthlaterfoundthat68%oftheinhibitedchildren assessments occurred following initial and task-
retained this classification; their heart rate remained interim baseline periods and during performance on
higher, but became more stabile than that of unin- eachofthreecognitivetasksbelievedtoinducestress.
hibited children on the cognitive task. The inhibited Itwashypothesizedthatphysiologicalmeasureswould
childrenwithhighandstabileheartrates(vs.highand reflect higher levels of arousal in the fragile X group
variable) were those most likely to maintain their relativetotheTurnerSyndromeorcomparisongroups
inhibitedstatuswhenreassessedatage4years(Kagan, becausetheevidenceforheightenedanxietyisstronger
Reznick,Clarke,Snidman,&Garcia-Coll,1984)and in the fragile X literature than in studies of Turner
at the final assessment at age 7.5 years (Kagan, syndrome. Similarly, higher SUDS ratings also were
Reznick,&Snidman,1988).Theauthorshypothesized predictedforthefragileXgrouprelativetotheTurner
that the ‘‘stress circuits’’ of inhibited children may orcomparisongroups.ForboththeTurnerandfragile
be more easily excitable, although it was unclear Xgroups,SUDSratingswerenotexpectedtocorrelate
whethertheirheightenedarousalwasachroniccondi- withmeasuresofphysiologicalarousal.
tion or a specific acute response to unfamiliarity,
potential harm, or a challenging cognitive task
(Reznicketal.,1986). METHOD
Othershavemeasuredcardiacvagaltone,asindex-
ed by respiratory sinus arrhythmia (RSA), in infants
Participants
and children, emphasizing the potential impact of
parasympathetic nervous system (PNS) function on Thetwoprimaryparticipantgroupsincludedfemales
arousal states. In some individuals, the PNS may be with fragile X (n¼13) or Turner syndrome (n¼11).
compromised,resultingindisruptionoftheregulation Femaleswithneitherdisordercomprisedthecompar-
ofhomeostaticfunction.Porges(1992)suggestedthat ison group (n¼14). The 38 participants were drawn
cardiacvagaltonemaybeareliableindicatorofPNS from a larger study of cognitive and behavioral phe-
tone,withlowervagaltonereflectinglessvariabilityin notypes of fragile X and Turner syndrome. Those
RSA and thus, less ability for physiologic self- includedinthepresentstudywereover12yearsofage
regulationintheeventofstress;thisinturnnegatively because human subjects approval for the procedures
impactsone’sabilitytoorganizebehavioralresponses. described later were approved only for this older age
Consistentwiththismodel,Porges(1992)foundthat, group.
asagroup,high-riskneonateshadsignificantlylower Thethreegroupsoffemaleswerewellmatchedon
vagaltonethanfull-termbabies.Thisalsowastrueina age.ThoseinthefragileXgroupwere13to22years
longitudinal study of very low birth weight (VLBW) old(mean¼16.50(cid:1)3.1years)whowereidentifiedby
neonates for whom lower cardiac vagal tone also DNA analysis to have the full mutation. Participants
predicted poorer social competence at age 3 years intheTurnersyndromegroupwere 12to20yearsof
(Doussard-Roosevelt, Porges, Scanlon, Alemi, & age (mean¼16.70(cid:1)3.1 years), and the comparison
Scanlon, 1997). By age 8 years, greater social com- group participants ranged from 12 to 17 years of age
petence in the VLBW group was best predicted by (mean¼14.96(cid:1)1.7 years). Scores on a measure of
greatermaturationofRSA,measuredasthedifference global intelligence differed across the groups, with
in RSA from 33 to 35 weeks’ gestational age. The femalesinthefragileXgroupscoringlowerthanthe
authorssuggestthatthedegreeofRSAmaturationmay remaining two groups. Girls with fragile X had IQ
beausefulphysiologicalmarkerofneuralmaturation scores between 59 and 125 (mean FSIQ¼88.5(cid:1)
andmaybeusefulintheearlyidentificationofchildren 17.4), participants in the Turner group had IQ scores
Arousalin Fragile Xor Turner Syndrome 137
between 65 and 126 (mean FSIQ¼96.5(cid:1)18.0), and messagethenwasdisplayedinformingtheparticipant
the comparison group participants had IQ scores ofthislossandurgingtheindividualtotryharder.After
between79to133(meanFSIQ¼108.9(cid:1)15.9).These eachanswer,theremainingtimewasdisplayedalong
IQ score means and ranges indicate that the partici- withthenumbercorrectthusfarandthepointvalueof
pants in the present study were representative of the the next correct answer. This information remained
populationsfromwhichtheyweredrawn. on the screen until the next answer was chosen. The
Participants were recruited through fragile X or number of problems attempted, the number correct,
Turnersyndromenewslettersandthroughpediatricand and percent accuracy were measured. Individuals
endocrinologyclinicsattheJohnsHopkinsUniversity whose percent accuracy was the same may have dif-
School of Medicine, Kennedy Krieger Institute, and feredinthe number of correct responses.This would
Thomas Jefferson University. The comparison group suggestthatparticipantswhoachievedahighernum-
included unaffected siblings of children who partici- ber of correct answers worked more quickly than
pated in other concurrent studies of genetic or beha- individualswhoansweredfewerproblemscorrectly.
vioralphenotypesthatwereconductedattheKennedy The Divided Attention Task (McLeod, Hoehn-
KriegerInstituteLearningDisabilitiesResearchCenter. Saric,Labib,&Greenblatt,1988)includedbothdivid-
Allparticipantsundertheageof18yearswereaccom- edattentionandreactiontimecomponents.Asequence
paniedbyparentswhogaveinformedconsent.Partici- of numbers, centrally located, was presented on the
pantsover18yearsofagegaveinformedconsent,and monitor screen. When one number left the screen,
for lower functioning participants over 18, parental anothernumberwaspresented.Followingtheappear-
consent also was obtained. Participants who were ance ofthenumber‘‘5,’’theparticipanthadtodeter-
unable to refrain from taking medications were ex- mine whether the number presented immediately
cluded from the study due to the potential confound afterwardwasanumbergreaterthanfive.Additionally,
withassessingphysiologicalarousal. participants wereinstructed topress a keyas soonas
possibleeachtimea‘‘0’’appearedinthesequence.The
numberofcorrecthits(detecting‘‘0’’),misses(failing
Measures
todetect‘‘0’’),andpercentaccuracyweremeasured.
On the Risk-Taking Task (McLeod et al., 1988),
PsychomotorPerformanceTasks. Psychomotorper- participants could lose or gain points depending on
formancetaskswerechosentoexaminephysiological theirresponse.Aleft-keyresponse(‘‘1’’)accumulated
changes as a result of engaging in stressful cognitive points and increased the width of a green bar on the
tasks. All tasks were presented by an Apple II Plus computermonitorscreen.Arightkeyresponse(‘‘3’’)
microcomputer. The computer monitor was within a savedtrialpointstoacumulativetasktotal,thusending
comfortable distance of the subject, and a numeric a given trial. With a probability of 0.05, any left-key
keypad was placed in the subject’s lap. The psycho- responsecouldproducearedbarsignalingthattwice
motor performance battery included the Mental thenumberofpointsaccumulatedfromthatparticular
Arithmetic Task, the Divided Attention Task, and trial would be subtracted from the cumulative task
theRisk-TakingTask,eachofwhichrequiredapproxi- total. To assess risk-taking behavior, the number of
mately 5 minutes tocomplete. left-keypressespriortoaright-keypressthatsavedthe
The Mental Arithmetic Task presented the parti- trialpointswasmeasured.
cipant with three one-digit numbers: one on the left,
oneinthecenter,andoneontherightofthecomputer Psychophysiological Measures. The laboratory ex-
monitorscreen.Theparticipant’sgoalwastoaddthe aminations included measures of skin conductance,
threenumbersmentally,withouttheuseofacalculator gastrocnemius electromyographic (GEMG) activity,
orpenandpaper.Toindicatecompletionofthemental heart rate, and cardiac vagal tone. Skin conductance
arithmetic task, participants pressed the center key measures of interest were mean skin conductance;
(‘‘2’’)onanumerickeypad.Theparticipantwasthen skin conductance range, reflecting the difference
asked to select the correct answer from among three between the largest and smallest within-subject skin
givenchoicesbypressingtheleft,center,orrightkey, conductance values; and skin conductance fluctua-
corresponding to ‘‘1,’’ ‘‘2,’’ or ‘‘3,’’respectively. For tions,which reflect the frequencyofchange indirec-
eachofthefirstthreecorrectanswers,theparticipant tionofskinconductanceovertime.Skinconductance
earned 20 points per answer. The number of points fluctuationsprovideagoodmeasureoflability.Lader,
earnedafterthesefirstthreecorrecttrialsincreasedby Gelder, and Marks (1967) found that the number of
10foreveryadditionaltwocorrectanswers.Allpoints spontaneousfluctuationsinskinconductancediffered
were lost if four mistakes were made in a row. A among patients with different anxiety disorders. All
138 Keysoretal.
measureswererecordedduringinitialandtask-interim Electrodeswerethenattachedinthemannerdescribed
periodsofrest,andduringperformanceontheMental earlier.
Arithmetic,DividedAttention,andRisk-TakingTasks Thebaselinerestingperiodlasted15minutes,with
described previously. With the exception of cardiac physiological measures recorded during the last
vagal tone,allphysiologicalmeasureswere expected 5minutes.Aftertheinitialbaselinerestperiod,parti-
toincrease inresponse toheightened arousal. cipants performed the Mental Arithmetic Task, the
All measures were recorded, amplified, and con- Divided Attention Task, and then the Risk-Taking
vertedfromanalogtodigitalformforallsubjectsusing Task. Each task lasted approximately 5 minutes and
a Coulborn Instruments Lablinc Interface System was preceded by its own 15-minute baseline period.
(Lehigh Valley, PA), a Modular Instruments Proces- The small sample size of the study prevented sys-
sing Center (Malvern, PA), and a Zytek 386 Tower tematic variation of task order. For this reason, the
Computer(ZytekEngineering, Inc., Baltimore,MD). orderoftaskswasheldconstantforallparticipantsso
Skinconductancewasrecordedbyplacingelectrodes that order effects would not confound the results of
(silver/silverchloride,1cmindiameter)onthevolar groupdifferences,iffound.Wechoseanorderconsi-
surfacesoftheindexandmiddlefingers(middleseg- stent with that used in previous studies of anxiety
ments)ofthenondominanthandandwasmonitoredby correlates conducted in our physiological laboratory
means of a constant voltage electrodermograph. The (D.R.M., R.H.-S.). Physiological measures were
electrodepastewaspreparedfromaneutralointment recorded throughout the duration of each of the
cream (Unibase, Park-Davis Co., Morris Plains, NJ) cognitivetasks, but only during the last 5 minutes of
and saline according to the procedure suggested by each baseline period. Participants were asked to rate
Fowles et al. (1981). GEMG activity (in mV) was themselvesontheSUDS,ananalogratingscale,atthe
assessedfollowingstandardpreparationoftheskinand end of each baseline period and after completion of
electrode placement (silver/silver chloride, 15mm in each of the three cognitive tasks. This scale is a
diameter) over the right gastrocnemius muscle. To thermometerlike illustration of states ranging from 0
obtain a measure (in ms) of heart interbeat interval (fullyrelaxed)to100(highlyanxious).Onthisscale,
(IBI), disposable electrodes were placed on the right theparticipantsmarktheirlevelofanxietyexperienced
andleftchestandontherightandleftabdomen.The atacurrenttime.
EKGleadthatbestdisplayedtheTwaveandtheonset
oftheQRScomplexwasused.Allphysiologicaldata
weresampledatarateof350timespersecond.Data RESULTS
werecollectedthroughtheserial interface andstored
bytheModularInstrumentssoftwareforlateranalysis. The primary variables of interest included both per-
Allsignalswereamplified,decoded,andaveragedby formanceandphysiologicaloutcomevariables.These
themicrocomputer. variables were examined for possible main effects of
The amplitude of RSA was quantified to assess groupstatus(Turner,fragileX,orcomparisongroup).
cardiacvagaltone.RSAreflectsthefunctionaleffects Possiblestatisticalinteractionbetweentaskorderand
ofthecardiacvagalefferentactivityatthelevelofthe diagnostic group could not be determined from the
heart.Toobtainameasureofcardiacvagaltone,each presentpreliminarystudy.However,maineffectsthat
seriesofsequentialIBIswasestimatedforsuccessive didemergedifferentiatedphysiologicalresponsepat-
500-mswindowsandanalyzedbycomputerprograms ternsbetweenthethreediagnosticgroups,asindicated
created in accordance with methods developed by in Table 1. When main effects were indicated for a
Porges(1985).Moredetailsonthisprocedurecanbe variable,allthreepossiblepairwiseposthoccompar-
foundinMcLeod,Hoehn-Saric,Porges,andZimmerli isons were carried out. In view of the small sample
(1992). sizes,normalityofdatacouldbeneitherassumednor
statisticallyexamined.Therefore,nonparametriccom-
parisons were used. All analyses were unpaired,
Procedure
corresponding to the research design. Thus, Kruskal-
Testing was conducted in the morning, beginning Wallis statistics were used to test for possible main
atapproximately9:00a.m.,tocontrolfordiurnalvaria- effectsofgroup.AllKruskal-Wallisresultscarriedout
tionsthatmayhaveconfoundedthedata.Participants werebasedonmorethanfiveobservationspersample;
had been advised to have a light breakfast without therefore,posthocpairedcomparisonswerebasedon
caffeine-containing beverages. Participants were di- Mann–Whitney U tests. Means and standard devia-
rectedtositinacomfortable,recliningchairinanair- tions for measures associated with performance on
conditioned room kept in constant dim illumination. thecognitivetasksappearinTable1;Table2presents
Arousalin Fragile Xor Turner Syndrome 139
Table1. Mean(þSD)ValuesforPerformanceMeasuresonCognitiveTasksbyGroup
Group
Task FragileXSyndrome TurnerSyndrome Comparison
Mentalarithmetic
#Attempted 30.62(10.65) 33.82(8.48) 39.57(10.11)
#Correct 25.08(12.71) 28.91(12.17) 37.29(10.52)
%Accuracy 0.82(0.28)a 0.83(0.29) 0.93(0.07)
Dividedattention
#0Hits 6.43(0.43) 6.42(0.67) 6.59(0.39)
#0Omissions 0.85(0.56) 0.84(0.63) 0.61(0.31)
%Accuracy 0.89(0.07) 0.88(0.09) 0.92(0.04)
Risk-taking
Riskescaperesponse 8.50(7.17) 14.54(7.70)a,b 8.21(2.88)
aSignificantlydifferentfromcomparisongroup,p<.02.bSignificantlydifferentfromfragileXgroup,p<.02.
means and standard deviations for physiological interest,duringtheinitialbaselineandduringadmin-
measures. istrationofallthreetasks.Forbaselineandallexperi-
Kruskal Wallis analyses showed main effects for mentaltasks,cardiacvagaltone,asreflectedbyRSA,
several physiological and performance variables of and the ratings on the SUDS failed to show a main
Table2. Mean(þSD)ValuesofPhysiologicalArousalatInitialBaselineandforEachTask
Group
ArousalIndex FragileXSyndrome TurnerSyndrome Comparison
Meanskinconductance
Baseline 6.12(4.57) 6.28(4.66) 3.95(4.35)
Mentalarithmetic 7.17(4.17) 10.33(5.00)a 6.08(4.30)
Dividedattention 6.60(4.14) 11.76(5.34)a,b 6.45(4.34)
Risk-taking 6.14(4.54) 9.94(6.11) 7.29(4.76)
Skinconductancerange
Baseline 3.59(3.02)a 3.13(3.97) 0.88(0.99)
Mentalarithmetic 4.08(1.8–) 4.89(2.73) 3.60(2.32)
Dividedattention 4.16(2.09) 4.30(1.85) 3.48(2.57)
Risk-taking 3.74(2.45) 2.96(1.84) 3.14(2.00)
Skinconductancefluctuation
Baseline 21.92(20.39) 14.00(12.99) 9.21(14.25)
Mentalarithmetic 31.15(16.72) 41.36(20.47)a 19.14(13.76)
Dividedattention 26.62(13.70) 60.46(22.95)a,b 19.93(11.91)
Risk-taking 22.23(17.93) 34.09(23.78) 28.21(16.78)
HeartIBI
Baseline 770.95(110.5) 741.90(120.3) 855.90(144.5)
Mentalarithmetic 756.39(114.8) 701.83(114.4)a 805.08(117.3)
Dividedattention 784.30(112.0) 709.12(120.1)a 840.83(108.7)
Risk-taking 806.02(115.2) 727.56(122.6)a 839.94(100.8)
GastrocnemiusEMG
Baseline 4.27(1.83) 3.29(1.37) 4.64(2.17)
Mentalarithmetic 4.21(1.31) 3.65(1.69) 4.43(2.45)
Dividedattention 4.23(1.28) 3.21(1.29)a,b 4.26(1.57)
Risk-taking 5.57(2.39) 3.79(1.48) 4.95(2.09)
aSignificantlydifferentfromtheComparisongroup,p<.05.bSignificantlydifferentfromthefragileXgroup,
p<.05.
140 Keysoretal.
effectofgrouporanysignificantpairwisecomparison. During the Divided Attention Task, there were
Therefore,cardiacvagaltoneandSUDSratingsarenot significant differences among four of the psychophy-
reported. siological measures. Mean GEMG measures, heart
IBI, and skin conductance fluctuations differed as a
functionofgroup(p<.03),asdidmeanskinconduc-
Initial Baseline Measures tance (p<.001). No significant differences on these
four measures were observed between girls in the
During initial baseline, skin conductance range dif-
fragile X and comparison groups (p>.30). When
fered as a function of group (H¼6.69, p<.04). Post
comparedtogirlsinthecomparisongroup,girlswith
hoc Mann–Whitney comparisons revealed that girls
TurnersyndromehadsignificantlylowermeanGEMG
with fragile X had a significantly higher mean skin
ratings (U¼36.5, p<.03), significantly shorter heart
conductancerangethangirlsinthecomparisongroup
IBI (U¼25.0, p<.01), and markedly higher skin
(U¼42.0, p<.02). Although girls with Turner syn-
conductance mean and fluctuations (U¼30, p<.02
drome also had higher mean skin conductance range
andU¼11.0,p<.001,respectively).Relativetogirls
scoresthangirlsinthecomparisongroup,thisdiffer-
withTurnersyndrome,girlswithfragileXhadhigher
ence was not statistically significant (U¼42.0,
meanGEMG(U¼33.0,p<.03),markedlyfewerskin
p¼.055). Girls with fragile X or Turner syndrome
conductance fluctuations (U¼14.0, p<.001), and
did not differ from each other on this measure
lower mean skin conductance values (U¼33.0,
(p¼.58). During baseline, the groups did not differ
p<.03).
significantlyonanyotherphysiologicalmeasures.
Risk-Taking Task
Mental Arithmetic Task
The Risk Taking Task yielded several significant
On the Mental Arithmetic Task, therewas no signifi-
main effects, including a significant group difference
cantdifferenceacrossthethreegroupsinthenumberof
intherateofriskescaperesponses(H¼9.85,p<.01).
problems attempted (p>.064) whereas accuracy of
The fragile X and comparison groups did not differ
problemsattempteddiddifferacrossgroups(p<.03).
on escape response rate (p¼.52). Girls with Turner
GirlswithfragileXhadloweraccuracythanthecontrol
syndromehadasignificantlyhigherriskescapemean
group(U¼37.0,p<.01);noothersignificantperfor-
score than did either the fragile X or comparison
mancedifferenceswerefound.
groups(U¼25.0and27.0,respectively,ps<.01).
Withrespect topsychophysiologicalmeasuresob-
When psychophysiological measures during the
tainedduringtheMentalArithmeticTask,therewasa
risk task were examined, group differences emerged
significantgroupdifferenceonlyforskinconductance
for mean GEMG value (H¼6.18, p<.05) and heart
fluctuations(H¼8.23,p<.02).Noothermaineffects
IBI (H¼7. 49, p<.03). No post hoc comparisons of
were reported for the remaining psychophysiological
thefragileXandcomparisongroupswerestatistically
measures on this task. However, in view of apparent
significant.GirlswithTurnersyndromehadashorter
trendsnoted(seeTable2),additionalposthocanalyses
mean IBI than those in the comparison group
werecarriedout.Therewerenosignificantdifferences
(U¼27.0,p<.01).
betweenthefragile Xandcomparison groups onany
measures (ps>.23). However, relativeto girls in the
comparison group, girls with Turner syndrome had a Correlational Analyses
shorterheartIBI(U¼40.0,p<.05),highermeanskin
Nonparametric correlations were conducted, within
conductance (U¼36.0, p<.03), and greater skin
eachgroup,toexaminethemagnitudeandsignificance
conductance fluctuations (U¼28.0, p<.01). There
of associations between degree of psychophysiolog-
were no significant differences between girls in the
ical arousal, the SUDS, and performance accuracy
fragile X and Turner syndrome groups on these
(Tables3&4).Onlysignificantcorrelationsatp<.05
measures.
are reported in the tables due to the large number of
correlations performed. For this reason, the correla-
tionsreportedalsoshouldbeinterpretedwithcaution,
Divided Attention Task
asnonewerestatisticallysignificantwhenadjustedbya
No significant group differences were found on the Bonferroni correction.InTable3,significantcorrela-
DividedAttentionTaskforthenumberofhits(H¼.79, tions at p<.05 occurred most frequently for the
p¼.67),omissions(H¼1.45,p¼.48),orthepercent comparison group. Longer heart IBI was associated
accuracy(H¼.92,p¼.63). with a greater number of correct responses (r¼61,
Arousalin Fragile Xor Turner Syndrome 141
Table 3. Spearman Rank Correlations of Cognitive (r¼(cid:2).67, p¼.035). No significant correlations be-
Performance Measures With Physiological Indices of tween the SUDS and performance measures were
ArousalandtheSUDS,byGroupandTask foundforthefragileXgroup.
InTable4,correlationsbetweentheSUDSratings
Task ComparisonGroup
and measures of psychophysiological arousal are re-
Mentalarithmetic
ported, with no evidence of a consistent pattern for
HeartIBI
group or task. Again, the reported findings may be
#Correct r¼.61
spurious and should be interpreted with caution. At
p¼.028
initial baseline, only girls with Turner syndrome
Accuracy r¼.55
reportedheightenedself-perceiveddistressinassocia-
p¼.046
Risk-taking tion with shorter heart IBI (r¼(cid:2).66, p¼.037). No
HeartIBI otherassociationsbetweentheSUDSandphysiologi-
Riskescape r¼(cid:2).66 calarousalwerefoundforgirlswithTurnersyndrome
Response p¼.018 at baseline or for any of the cognitive tasks. On the
MentalArithmeticTask,girlsinthecomparisongroup
Mentalarithmetic TurnerSyndrome reportedhigherlevelsofperceiveddistressinassocia-
MeanSkinConductance
tion with lower mean skin conductance (r¼(cid:2).69,
#Attempted r¼(cid:2).67
p¼.013) and fewer fluctuations in skin conductance
p¼.035
(r¼(cid:2).66, p¼.018). Thus, the SUDS ratings were
inconsistent in direction with the level of recorded
arousalontheMentalArithmeticTask.OntheDivided
Attention Task, girls in the comparison group did
p¼.028) and greater accuracy (r¼.55, p¼.046) on reportheightenedlevelsofdistressinassociationwith
theMentalArithmeticTask,andwithfewerriskescape increasedGEMGactivity(r¼.56,p¼.045).Noother
responses on the Risk-Taking Task (r¼(cid:2).66, significantassociationsforthecomparisongroupwere
p¼.018).ForgirlswithTurnersyndrome,heightened found. Finally, girls with fragile X reported greater
mean skin conductance was associated with fewer distress in relation to decreased GEMG activity only
attempted problems on the Mental Arithmetic Task ontheRisk-TakingTask(r¼(cid:2).58,p¼.045);noother
significantassociationsbetweentheSUDSandmeas-
uresofarousalwerefoundforthisgroup.
In summary, these findings reveal indicators of
Table 4. Spearman Rank Correlations of SUDS With
Measures of Physiological Arousal at Initial Baseline higher arousal in girls with fragile X or Turner
andFollowingEachCognitiveTask syndrome, although to varying degrees and different
inpattern.GirlswithfragileXshowedhigherarousalat
Baseline—significantforgirlswithTurnersyndromeonly initial baseline, relativeto the comparison group, but
SUDS onlywithrespecttomeanrangeinskinconductance.
HeartIBI r¼(cid:2).66 DuringtheMentalArithmeticTask,girlswithfragileX
p¼.037 hadlowerperformanceaccuracy,butdidnotmanifest
Mentalarithmetic—significantforcomparisongrouponly
morearousalthandidgirlsinthecomparisongroup.In
SUDS
contrast,girlswithTurnersyndromemanifestedsigns
Meanskinconductance r¼(cid:2).69
ofhigherarousalthandidthegirlsinthecomparison
p¼.013
group; although not significantly less accurate, their
Skinconductancefluctuation r¼(cid:2).66
overall accuracy was lower than that seen in the
p¼.018
Dividedattention—significantforcomparisongrouponly comparisongroup.Similarly,ontheDividedAttention
SUDS Task,girlswithfragileXhadpoorerperformance,but
GEMG r¼.56 girlswithTurnersyndromedemonstratedmoreindices
p¼.045 ofarousal.OntheRisk-TakingTask,girlswithTurner
Risk-taking—significantforgirlswithfragileXsyndrome syndromeshowedthehighestdegreeofrisktakingand
only higherindicesofarousalbasedonheartIBI.Overall,
SUDS
relative to girls in the comparison group, girls with
GEMG r¼(cid:2).58
Turner syndrome recorded significantly higher levels
p¼.045
ofarousalonseveraldifferentindiceswhilegirlswith
fragileXrecordedhigherphysiologicalactivityonlyat
Note. SUDS¼Subjective Units of Distress Scale; GEMG¼
gastrocnemiuselectromyographic. initialbaselineandonameasurethatdidnotindicate
142 Keysoretal.
heightened arousal for girls with Turner syndrome at because of their state of heightened arousal at rest
anyphaseoftheprocedure. (Kellyetal.,1970).
The heightened arousal at baseline in females
with fragile X is consistent with several other sets of
DISCUSSION findings. Preliminary evidence suggests abnormal
hypothalamic–pituitary–adrenal (HPA) function in
The aim of this study was to assess whether there children with fragile X. Normal cortisol diurnal pat-
are group differences in levels of psychophysiolo- terns reflect gradual and steady decreases in cortisol
gical arousal recorded for females with fragile X or levelsthroughout the day, excepting increasesduring
Turnersyndrome,orfemaleswithneitherdisorder,and stressful situations (Stansbury & Gunnar, 1994). In a
whetherthesedifferences havepotentialimplications smallsampleof8malesand7femaleswithfragileX,
for the manifestation of anxiety in each of these X corticallevelsdiddeclinethroughouttheday,although
chromosomegenedisorders. meancortisollevelswerehigherrelativetonormative
data during the middle and late part of the day
(Wisbecketal.,2000).Thesedatawerereportedwith
Fragile X Versus Turner Syndrome
parametricprocedures,andthusmayhavebeensubject
The hypothesis that physiological arousal would be to outlier effects; regardless, they are consistent with
significantly higher for females with fragile X than ourhypothesisofahigherbaselinelevelofarousalin
those with Turner syndrome received little support. females with fragile X syndrome. Similarly, height-
Significantly higher mean GEMG activity on the ened arousal, based on mean skin conductance, was
DividedAttentionTaskwastheonlyincidenceofgirls found in males with fragile X relative to males with
withfragileXrecordinghigherarousalthangirlswith Down Syndrome during conditions that involved
Turnersyndrome.Onthissametask,measuresofmean participants either having or not having eye contact
skin conductance and fluctuations indicated higher during a conversation. Although males with fragile
arousallevelsforgirlswithTurnersyndrome,relative X were less aroused during the condition with no
tothosewithfragileX.Thus,bothgroupsexperienced eye contact, their heightened state of arousal under
heightened arousal on the Divided Attention Task, both conditions suggests that hyperarousal may be
althoughhigherarousalwasdemonstratedbydifferent related to fragile X independent of associated devel-
physiological indices. Given that different regions of opmental delays (Belser & Sudhalter, 1995). Further
the SNS can be stimulated, resulting in potentially supportofthisisfoundinCohen(1995),whoreported
different patterns of physiological activity (Gellhorn, higher arousal and less tolerance to environmental
1965), it is possible that either the nature of the stimuli in males with both fragile X and autism
stimulus associated with arousal in each group dif- compared to males with autism who do not have
fered, or the nature of the stimulus was the same for fragile X, and in Roberts et al. (2001), who reported
both groups but different regions of the SNS were a high baseline heart rate among young boys with
stimulated. fragileX.
Fragile X Versus Comparison Group Turner Syndrome Versus
Comparison Group
InadditiontoalackofdifferencebetweenthefragileX
andTurnersyndromegroups,therewaslittledifference Contrarytoourprediction,girlswithTurnersyndrome
in physiological arousal between girls with fragile X recorded significantly higher levels of arousal than
and those in the comparison group. Only at initial girls in the comparison group during all cognitive
baseline did girls with fragile X present with signi- tasks, but not consistently on the same physiological
ficantly higher arousal than girls in the comparison indices nor for the initial resting assessment. Differ-
group, based on the mean skin conductance range. ences in arousal levels were found on the Mental
Theheightenedstateofarousalforfemaleswithfragile Arithmetic and Divided Attention Tasks, on which
Xatrestmayexplainwhytheyrecordedlittleincrease girls with Turner syndrome recorded significantly
inarousalonthecognitivetasks,relativetotheincrease higher mean skin conductance and greater skin con-
recordedbyfemalesineithertheTurnersyndromeor ductance fluctuations than girls in the comparison
comparisongroups.Thisresultparallelsthehypothesis group.ThesedifferencesarenotsustainedontheRisk-
that adult patients with anxiety showed less of an Taking Task; in fact, scores indicate that girls with
increase in physiological activity on a Mental Arith- Turnersyndromeexperienceadecreaseinarousallevel
meticTask,comparedtothoseinacomparisongroup, based on these measures. Although the scores do not