DTIC ADA538359: Prospective Risk Factors for New-Onset Post-Traumatic Stress Disorder in National Guard Soldiers Deployed to Iraq

0.12 MB·English

by  Defense Technical Information Center

PsychologicalMedicine(2011),41,687–698. fCambridgeUniversityPress2010 ORIGINAL ARTICLE
doi:10.1017/S0033291710002047
Prospective risk factors for new-onset post-traumatic
stress disorder in National Guard soldiers deployed
to Iraq
M.A.Polusny1,2,3*,C.R.Erbes1,3,M.Murdoch1,2,4,P.A.Arbisi1,3,P.Thuras1,3andM.B.Rath5
1MinneapolisVAHealthCareSystem,Minneapolis,MN,USA
2CenterforChronicDiseaseOutcomesResearch,Minneapolis,MN,USA
3DepartmentofPsychiatry,UniversityofMinnesotaMedicalSchool,Minneapolis,MN,USA
4DepartmentofMedicine,UniversityofMinnesotaMedicalSchool,Minneapolis,MN,USA
5MinnesotaArmyNationalGuard,StPaul,MN,USA
Background. NationalGuardtroopsareatincreasedriskforpost-traumaticstressdisorder(PTSD);however,littleis
knownaboutriskandresilienceinthispopulation.
Method. TheReadinessandResilienceinNationalGuardSoldiersStudyisaprospective,longitudinalinvestigation
of522ArmyNationalGuardtroopsdeployedtoIraqfromMarch2006toJuly2007.Participantscompletedmeasures
of PTSD symptoms and potential risk/protective factors 1 month before deployment. Of these, 81% (n=424)
completed measures of PTSD, deployment stressor exposure and post-deployment outcomes 2–3 months after
returning from Iraq. New onset of probable PTSD ‘diagnosis’ was measured by the PTSD Checklist–Military
(PCL-M).Independentpredictorsofnew-onsetprobablePTSDwereidentifiedusinghierarchicallogisticregression
analyses.
Results. At baseline prior to deployment, 3.7% had probable PTSD. Among soldiers without PTSD symptoms at
baseline, 13.8% reported post-deployment new-onset probable PTSD. Hierarchical logistic regression adjusted for
gender, age, race/ethnicity and military rank showed that reporting more stressors prior to deployment predicted
new-onset probable PTSD [odds ratio (OR) 2.20] as did feeling less prepared for deployment (OR 0.58). After
accounting for pre-deployment factors, new-onset probable PTSD was predicted by exposure to combat (OR 2.19)
andtocombat’saftermath(OR1.62).Reportingmorestressfullifeeventsafterdeployment(OR1.96)wasassociated
withincreasedoddsofnew-onsetprobablePTSD,whilepost-deploymentsocialsupport(OR0.31)wasasignificant
protectivefactorintheetiologyofPTSD.
Conclusions. Combat exposure may be unavoidable in military service members, but other vulnerability and
protectivefactorsalsopredictPTSDandcouldbetargetsforpreventionstrategies.
Received3February2010;Revised14September2010;Accepted17September2010;Firstpublishedonline10December2010
Keywords:Combat,militarypersonnel,prospectivestudies,PTSD,riskfactors.
Introduction reminders about those events, and hyperarousal
symptoms such as impaired sleep, irritability and
Over1.8millionUStroopshavebeendeployedtothe
decreased concentration (APA, 1994). While most
wars in Afghanistan (Operation Enduring Freedom;
combat-exposed troops will fortunately not develop
OEF) and Iraq (Operation Iraqi Freedom; OIF).
PTSD(Hogeetal.2004),thesubstantialminoritywho
Combatisassociatedwithconsiderablementalhealth
do will face considerable difficulties in interpersonal
risk, including elevated rates of post-traumatic stress
relationships,occupationalfunctioning,andqualityof
disorder (PTSD), a psychiatric disorder characterized
life as well as high rates of co-morbidity with other
by intrusive and distressing reliving of traumatic
psychiatric disorders (Kessler, 2000). PTSD occurs in
events (through memories or dreams), avoidance of
as many as 1 in 5 Vietnam veterans, in contrast to
rates of less than 1 in 10 for the general population
(Kessler et al. 2005; Dohrenwend et al. 2006). In
*Addressforcorrespondence:M.A.Polusny,Ph.D.,Minneapolis
military personnel deployed to OEF/OIF, about 1 in
VAHealthCareSystem(116A9),OneVeteransDrive,Minneapolis,
8servicemembersreturnwithPTSD(Hogeetal.2004;
MN55417,USA.
(Email:[email protected]) Schell&Marshall,2008).
Report Documentation Page Form Approved
OMB No. 0704-0188
Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington
VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it

does not display a currently valid OMB control number.
1. REPORT DATE 3. DATES COVERED
2010 2. REPORT TYPE 00-00-2010 to 00-00-2010
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
Prospective Risk Factors For New-Onset Post-Traumatic Stress Disorder
5b. GRANT NUMBER
In National Guard Soldiers Deployed To Iraq
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION
Minneapolis VA Health Care System,Minneapolis,MN,55401 REPORT NUMBER
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)
11. SPONSOR/MONITOR’S REPORT
NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT
Approved for public release; distribution unlimited
13. SUPPLEMENTARY NOTES
Psychological Medicine (2011), 41, 687-698
14. ABSTRACT

15. SUBJECT TERMS
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF
ABSTRACT OF PAGES RESPONSIBLE PERSON
a. REPORT b. ABSTRACT c. THIS PAGE Same as 13
unclassified unclassified unclassified Report (SAR)
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std Z39-18
688 M.A.Polusnyetal.
The US military has increasingly relied on large Furthermore, what is known about combat-related
numbersofNationalGuardandReserve(NGR)troop PTSD has largely been derived from cross-sectional
deployments to support OEF/OIF. Several reports studies of veterans from earlier wars, where retro-
indicate that NGR troops are at heightened risk for spective data were often collected a decade or more
post-deployment psychiatric distress compared with afterhostilitiesceased(Ozeretal.2003).Consequently,
regularactivedutytroops(USArmySurgeonGeneral, these studies have been limited by the potential
2005; Hotopf et al. 2006; Browne et al. 2007; Milliken for recall errors and ambiguity about the temporal
et al. 2007; Smith et al. 2008; Iversen et al. 2009), and sequence of events (i.e. direction of cause and effect)
thisheightenedriskappearstoincreaseevenmorein (King et al. 2000). More recent, prospective studies of
the months and years following combat deployment troops deployed to OEF/OIF suggest that new-onset
(Wolfe et al. 1999; Milliken et al. 2007). For example, PTSDfollowingdeploymentisassociatedwithfemale
Milliken et al. found that rates of positive screening gender, younger age, enlisted (non-officer) rank,
forPTSDsymptomsmorethandoubledamongNGR NGR status, being a smoker prior to deployment
soldiers from their immediate post-deployment (Smith et al. 2008), and reporting PTSD symptoms or
screening (12.7%) compared with when they were poorerphysicalhealthpriortodeployment(Ronaetal.
re-evaluated 6 months later (24.5%). In contrast, the 2009).Whilethesestudiesareanadvanceoverearlier
PTSDscreeningrateincreasedbyonly4.9%inregular research,theyevaluatedveryfewmodifiableriskand
active duty troops during the same time-frame protective factors for PTSD. Identifying risk and pro-
(Millikenetal.2007).Combatdeploymentmaybees- tective factors for PTSD is critical to understanding
peciallystressfulforNGRtroopsor‘civiliansoldiers’ the disorder’s etiology, identifying those most vul-
who may be unaccustomed to prolonged separations nerable, and informing prevention and treatment-
from family and who mayexperience harmful career developmentefforts.
disruptionsintheircivilianoccupations.NGRtroops’ Ourgoalsweretogobeyondgenerallyfixeddemo-
military training, perceptions of preparedness and graphicvariables(e.g.age,race,gender)andidentify
unit cohesion might also differ from regular compo- pre-trauma (pre-deployment), trauma (deployment-
nent troops in ways that increase NGR troops’ risks related),andpost-trauma(post-deployment)riskand
for harmful post-combat sequelae. In a recent cross- protectivefactorsfordevelopingnew-onsetPTSDina
sectionalstudyofUKtroopsdeployedtoIraq,reserve cohort of National Guard troops deployed to OIF.
component UK troops reported lower levels of unit Even after accounting for the influence of combat
cohesion and felt less informed than regular active exposure on PTSD, we hypothesized that new-onset
duty component UK troops (Browne et al. 2007). probable PTSD would be uniquely predicted by
Despite indications of elevated PTSD risk among soldiers’ pre-deployment reports of their childhood
NGR troops, however, most PTSD research focuses family environments, past exposure to potentially
on active duty component soldiers, and little is traumatic events, military preparedness, unit co-
known about risk and resilience in this important hesion, and worries about the impact of deployment
population. on civilian life and family. We also hypothesized,
Three decades of research has demonstrated that afteraccountingforpre-deploymentvariablesandthe
severity of trauma exposure robustly predicts PTSD impact of combat deployment stressors, that post-
(Ozer et al. 2003). However, other factors have also deploymentsocialsupportwouldbeuniquelyassoci-
beenimplicatedinPTSD’sdevelopment.Forexample, ated with lower risk of new-onset probable PTSD,
childhood trauma and adversity (King et al. 1999; whilesoldiers’experiencesofstressfullifeeventssince
Iversen et al. 2007), neuroticism (Miller et al. 2004; deploymentwouldbeuniquelyassociatedwithriskof
Rubinetal.2008;Lahey,2009),worriesaboutfamilies new-onsetprobablePTSD.
and civilian life while deployed (Vogt et al. 2005),
subsequentlifestressors(Kingetal.1998;Dirkzwager Method
etal.2003;Ozeretal.2003)andlackofsocialsupport
Studydesignandparticipants
(Ozeretal.2003)havebeenassociatedwithincreased
odds of developing PTSD. However, positive child- Thisprospectivepanelstudyfollowedtheclassicepi-
hood family environments (Foy et al. 1987; Schnurr demiological strategy of defining a panel of National
et al. 2004), unit cohesion (Brailey et al. 2007; Iversen Guard soldiers prior to deployment (i.e. prior to
et al. 2008; Rona et al. 2009), military preparedness ‘exposure’) and then excluding from analysis those
(Kingetal.2006)andgreatersocialsupportfollowing who were already symptomatic for the disorder of
deployment (Benotsch et al. 2000; Dirkzwager et al. interest–in this instance, PTSD. After deployment,
2003) have been associated with lower odds of de- all National Guard soldiers who had initially been
velopingPTSD. asymptomaticforPTSDwereexaminedfornew-onset
Riskfactorsfornew-onsetPTSD 689
probable PTSD. Hypothesized risk and protective Department of Veterans Affairs, University of
factorswerethencomparedacrossthosewhodidand Minnesota,andrelevantArmyNationalGuard(ARNG)
didnotdevelopnew-onsetprobablePTSD. command.Aftercompletedescriptionofthestudyto
Data were collected as part of the Readiness and subjects,writteninformedconsentwasobtained.
ResilienceinNationalGuardSoldiers(RINGS)study,
a prospective study of 522 Army National Guard
Assessmentofnew-onsetPTSD
soldiers (462 men and 60 women) from a Brigade
CombatTeam(BCT)deployedtoIraq.InMarch2006, New-onset probable PTSD was assessed using the
questionnaires assessing psychosocial risk/protective 17-item PTSD Checklist (PCL; Weathers et al. 1993),
factors and baseline psychiatric symptoms were col- which measures PTSD symptoms corresponding
lected 1 month prior to troops’ deployment to Iraq. to the Diagnostic and Statistical Manual of Mental
Troopswereinformedaboutthestudythroughflyers Disorders, 4th edition (DSM-IV) diagnostic criteria
as well as announcements by mid-level leadership. (APA, 1994). Respondents rated the severity of each
Although no specific time for participation was symptomduringthepastmonthonaLikertscalefrom
allotted in troops’ intense pre-deployment training 1 (not at all) to 5 (extremely) (range 17–85). Baseline
schedule, about 20% of the total BCT force met with PTSD symptoms were assessed prior to deployment
investigators for a group briefing and received infor- usingthecivilianversionofthePCL(PCL-C);military
mation about the study. Precise participation rates deployment-related PTSD symptoms were assessed
could not be obtained, but participation appeared to following deployment using the military version of
behighamongthoseunitsthatattendedgroupbrief- the PCL (PCL-M; Weathers et al. 1993). The PCL has
ings. Participants completed questionnaires in group excellent test–retest reliability and high overall con-
classroomsunderstandardizedconditions.Consistent vergentvalidity(Weathersetal.1993;Blanchardetal.
withmilitaryregulations,noincentiveswereprovided 1996).
pre-deployment when soldiers were on active duty. A range of criteria has been recommended for
Troops had just completed 5 months of intensive identifying probable PTSD among military personnel
mobilizationtrainingatCampShelby,Mississippiand usingthePCL(Weathersetal.1993;Blieseetal.2008;
werepoisedfora1-yeardeployment,whichwaslater Terhakopian et al. 2008). Based on our analytical
extendedby4months.TheBCTwasdeployedtoIraq strategy, we wanted to be as certain as possible that
fromMarch2006toJuly2007. our analysis was restricted only to panel members
Post-deployment data were collected via mailed who did not have PTSD prior to deployment. There-
survey. Approximately 2 months after the panel’s fore,weuseda‘liberal’screeningcut-off(PCLtotalof
return (September 2007), we mailed a follow-up o34) to identify soldiers with PTSD symptoms prior
questionnaire, cover letter containing the elements of to deployment. This cut-off score has 71% sensitivity
informed consent,and$50cashincentive toallpanel and 91% specificity for PTSD diagnosis based on
members. A postcard reminder and two additional the Mini International Neuropsychiatric Interview
mailings were sent to non-respondents at 2-week (MINI;Blieseetal.2008).Ittendstoerronthesideof
intervals. Survey tests were counterbalanced to being overly inclusive, which was acceptable for our
control for the potential influence of ordering effects purposes. Conversely, after deployment, we wanted
(Reddyetal.2009). tobeascertainaspossiblethatpanelmembersdefined
Of the original panel, 424(81%) returned post- ashavingPTSDtrulydidhaveanew-onsetdiagnosis.
deploymentquestionnaires.Panelmemberswhocom- Therefore, we used a ‘stringent’ definition for prob-
pleted the post-deployment questionnaire did not ablePTSDthatrequiredthefollowing:(1)participants
differfromthosewhodidnotcompleteitintermsof had to report DSM-IV diagnostic criteria (endorsing
gender, rank, pre-deployment measures of vulner- at least one intrusion symptom, three avoidance
abilityandprotectivefactors,orbaselinePTSDsymp- symptoms and two hyperarousal symptoms at the
toms. Panel members who did not complete the moderatelevel)and(2)haveatotalPCLscoreofo50
post-deployment questionnaire were younger [25.3 (Hoge et al. 2004). While this stringent method has
(S.D.=6.9)v.29.9(S.D.=8.8)years,p<0.0001]andmore beenwidelyusedinmilitarystudies(Ramchandetal.
likely to be non-white (11% v. 6%, p<0.05) or un- 2008), sensitivity and specificity has not been eval-
married(49%v.31%,p<0.001)comparedwiththose uated (Terhakopian et al. 2008). However, using
who completed the post-deployment questionnaire. pooled data from validation studies comparing the
Non-responders also reported fewer years of edu- PCL cut-off of 50 against ‘gold standard’ structured
cation[13.5(S.D.=1.7)v.14.4(S.D.=2.0)years,p<0.0001]. diagnostic interviews yields a weighted average sen-
The RINGS study was approved by the human sitivity of 54% and a weighted average specificity of
subject research review boards of the Army, 93%(Terhakopianetal.2008);theuseofcriteriausing
690 M.A.Polusnyetal.
DSM-IV symptom endorsement to identify PTSD accident, unemployment, legal problems) was as-
casenessbasedontheSCIDyields40%sensitivityand sessedusingthePost-deploymentStressorsscale(King
97%specificity(Widowsetal.2000). et al. 2006; Vogt et al. 2008). The Post-Deployment
SocialSupportscale(Kingetal.2006;Vogtetal.2008)
Riskandprotectivefactorsassessedat measured soldiers’ perceived emotional and instru-
pre-deployment mental support from family, friends, employers, co-
workers and community. Higher scores indicate
Five valid and reliable scales from the Deployment
greaterlevelsofeachconstruct.
RiskandResilienceInventory(DRRI)(Kingetal.2006;
Vogt et al. 2008) were used to prospectively measure
Statisticalanalyses
risk and protective factors prior to soldiers’ deploy-
ment.The17-itemPriorStressorsscalemeasuredsol- We describe the panel’s deployment experiences
diers’ exposure to stressful and potentially traumatic overalland,amongthosewithoutPTSDsymptomsat
events before deployment, i.e. sexual abuse, physical baseline,bypost-deploymentPTSDstatus(new-onset
assault and natural disaster. Responses (0=no, 1= probable PTSD versus not). Among those without
yes)weresummedtocreateapriorstressorsseverity PTSDsymptomsatbaseline,wetestedforunadjusted
score. A modified Concerns about Family/Life differencesinriskandprotectivefactorsbynew-onset
Disruptions scale (sum of 14 items rated on a Likert probable PTSD status using independent t tests. We
scale from 4=a great deal to 1=not at all) assessed used hierarchical logistic regression to determine the
soldiers’ pre-deployment worries about how the up- adjusted odds of association between these same
coming deployment might lead to losses and dis- factors and new-onset probable PTSD. All continu-
ruptions in their family and civilian career. The ouslydistributedvariableswereconvertedtoz-scores
ChildhoodFamilyEnvironmentscale(sumof15items prior to entry into regression models. Variables were
ratedonaLikertscalefrom1=almostnoneofthetime entered in three blocks. In the first block, we entered
to 5=almost all of the time) measured the quality of participants’sociodemographiccharacteristics(gender,
soldiers’ childhood family environments in terms of age, race/ethnicity and military rank), their pre-
cohesion, accord and closeness among family mem- deploymentPCLscore,andtheirpre-deploymentrisk
bers. The 14-item Preparedness scale measured the and protective factors. By adding participants’ base-
extent to which prior to deployment soldiers per- line PCL scores to this block, we control for any pre-
ceived they had mastered technical military skills existing, low-level, PTSD symptoms that might have
needed for combat operations and had adequate facilitated later development of full-syndrome PTSD.
knowledge of what to expect during deployment. The two pre-deployment risk factors entered in this
Finally,soldiers’perceptionsofmilitarycohesionand block were prior stressful life events and concerns
unit support were measured pre-deployment using abouttheimpactthatdeploymentmighthaveonpar-
the12-itemUnitSocialSupportscale. ticipants’ life and family. The three pre-deployment
protective factors were childhood family environ-
Deployment-relatedfactorsassessedatfollow-up
ment, military preparedness and unit social support.
When soldiers returned from deployment, we used Deployment-related risk factors (combat exposure,
three DRRI subscales to measure deployment-related witnessing the aftermath of battle, perceived life
stressor exposure (King et al. 2006; Vogt et al. 2008). threat) were entered in the second block. In the third
The Combat Experiences scale asked about specific andfinalblock,weenteredpost-deploymentvariables
combatexposuresduringdeployment;responseswere (post-deployment stressful life events and post-de-
summed to create a combat exposure severity score. ployment social support). Adjusted odds ratios for
The Aftermath of Battle scale measured exposure predictor variables were examined, with a two-tailed
to the consequences of combat including handling a of <0.05 used to determine statistical significance.
humanremains.ThePerceivedThreatscalemeasured All analyses were conducted using SPSS version 17
soldiers’subjectiveexperienceoffearorthreattowell- (SPSSInc.,USA).
being during deployment. Higher scores indicate
greaterlevelsofexposuretoeachdeployment-related
Results
riskfactor.
Samplecharacteristics
Post-deploymentriskandprotectivefactorsassessed
As shown in Table 1 and consistent with character-
atfollow-up
istics of the entire BCT, 88% of the cohort was male.
Soldiers’exposuretostressfullifeeventssincereturn Most were white, enlisted rank and aged <30 years.
from deployment (e.g. death of a loved one, serious There were no significant differences between cohort
Riskfactorsfornew-onsetPTSD 691
Table1.BaselinedemographiccharacteristicsoftheRINGScohortparticipantscomparedwithfollow-uppanelandBrigadeCombat
Teampopulation
Overallcohort Panelparticipants Brigadecombatteam
Demographiccharacteristic atbaseline(n=522) atfollow-up(n=424) population(n=2573)
Gender
Male 462(88.5) 372(87.7) 2339(90.9)
Female 60(11.5) 52(12.3) 234(9.1)
Meanage,years(S.D.)a 29.1(8.6) 29.9(8.8) –
18–29years 313(60.0) 235(55.4) 1672(65.0)
30+years 209(40.0) 189(44.6) 901(35.0)
Race/ethnicity
White,non-Hispanic 484(92.9) 398(93.9) 2407(94.7)
Non-white 37(7.1) 26(6.1) 134(5.3)
Rank
Enlisted 471(90.2) 377(88.9) 2301(89.4)
Officer 51(9.8) 47(11.1) 270(10.5)
Maritalstatus
Married 237(45.4) 207(48.8) 1006(39.9)
Notmarried 285(54.6) 217(51.2) 1515(60.1)
Educationa
Highschool 143(27.4) 107(24.1) –
Somecollege 215(41.2) 174(41.0) –
Collegeorgraduatedegree 164(31.4) 148(34.9) –
Militaryoccupationalspecialtya
Combatarms 249(48.1) 196(46.2) –
Combatsupport 81(15.7) 61(14.4) –
Combatservicesupport 187(36.2) 163(38.4) –
RINGS,ReadinessandResilienceinNationalGuardSoldiers.
Valuesaregivenasnumber(%).Numbersmightnotadduptototalsbecauseofmissingdata.Percentagesreportedarethe
proportionofindividualsendorsingeachdemographiccharacteristicadjustedtotakeaccountofsampleandmissingdata.
aComparabledemographicdataformeanage,educationandmilitaryoccupationalspecialtyatthebrigadelevelwerenot
available.
membersandtheBCTpopulationintermsofgender, participant had new-onset probable PTSD after de-
race/ethnicity,orrank;however,agreaterpercentage ployment.Overall,morethan90%ofpanelmembers
ofcohortmemberswereagedo30yearsandmarried. went on combat missions and patrols and received
hostile, incoming fire. More than half served in units
RatesofreportedPTSDsymptomsatbaselineand that suffered casualties, and about one-fifth believed
new-onsetprobablePTSDatfollow-up theymayhavekilledenemiesincombat.Despitethis
overallhighprevalenceofintensecombatexperiences,
Themajorityofsoldiers(430/516)wereasymptomatic
Table 2 also shows that, among those without PTSD
forPTSDsymptomspriortodeployment.Atbaseline,
at baseline, these and other experiences were signifi-
16.7%(86/516mettheliberalscreeningcut-offofo34
cantly more common for those who developed new-
on the PCL-C) had pre-existing PTSD symptoms,
onset probable PTSD compared with those who did
while3.7%(19/516soldiersmetstringentcriteria)had
notdevelopPTSD.Effectsizesfornew-onsetprobable
probable PTSD. Among those without PTSD symp-
PTSD were particularly large for activities related to
toms at baseline who were assessed at follow-up,
killingorkilling’saftermath,beinginavehicleunder
13.8%(48of349metstringentcriteriaonthePCL-M)
fire,andencounteringlandorwatermines.
developedpost-deployment,new-onsetprobablePTSD.
Combatdeploymentexperiencesoverallandamong Riskandprotectivefactorsoverallandamongthose
thosewithoutPTSDatbaseline withoutPTSDatbaseline
Table 2 reports panel members’ combat experiences Table 3 presents the mean scalescores for DRRI sub-
overall and, among those without PTSD at baseline, scales overall and, among those without PTSD at
comparestheseexperiencesaccordingtowhetherthe baseline, by post-deployment new-onset probable
692 M.A.Polusnyetal.
Table2.FrequencyofdeploymentstressorexposuresreportedbyNationalGuardtroopsdeployedtoIraqoverallandamongthose
withoutPTSDatbaseline(n=349)bynew-onsetprobablePTSDstatus
AmongthosewithoutPTSD
symptomsatbaseline
Overall New-onset
panel probable Effect
participants PTSDa NoPTSDb size,
Deploymentexperiences (n=424) (n=48) (n=301) p g
Combatexposure
Wentoncombatpatrolsormissions 380(90.5) 46(95.8) 263(88.6) 0.13 0.08
Encounteredlandorwaterminesand/orboobytraps 229(54.1) 37(78.7) 147(48.8) <0.001 0.21
Receivedhostileincomingfirefromsmallarms,artillery, 392(92.9) 47(100.0) 276(92.0) 0.04 0.11
rockets,mortarsorbombs
Received‘friendly’incomingfirefromsmallarms,artillery, 71(16.8) 16(34.0) 39(13.0) <0.001 0.20
rockets,mortarsorbombs
Vehicle(forexample,atruck,tank,armoredpersonnelcarrier, 239(56.5) 41(85.4) 152(50.5) <0.001 0.24
helicopter,plane,orboat)wasunderfire
Attackedbyterroristsorcivilians 237(56.4) 38(79.2) 154(51.7) <0.001 0.19
Tookpartofalandornavalartilleryunitthatfiredontheenemy 78(18.5) 9(18.8) 57(19.0) 0.97 0.00
Tookpartofanassaultonentrenchedorfortifiedpositions 45(10.6) 12(25.0) 21(7.0) <0.001 0.21
Tookpartonaninvasionthatinvolvednavaland/orlandforces 31(7.4) 4(8.5) 20(6.7) 0.65 0.03
Unitengagedinbattleinwhichitsufferedcasualties 240(57.3) 32(66.7) 170(57.0) 0.21 0.07
Personallywitnessedsomeonefromunitorallyunitbeing 179(42.5) 29(60.4) 111(37.1) 0.002 0.16
seriouslywoundedorkilled
Personallywitnessedsoldiersfromenemytroopsbeingseriously 161(38.3) 30(62.5) 94(31.5) <0.001 0.22
woundedorkilled
Woundedorinjuredincombat 49(11.7) 12(25.5) 26(8.7) <0.001 0.19
Firedweaponattheenemy 123(29.4) 24(51.1) 63(21.1) <0.001 0.24
Killedorbelievedtohavekilledenemyincombat 91(21.8) 23(48.9) 44(14.8) <0.001 0.30
Exposuretotheaftermathofbattle
Observedhomesorvillagesthathadbeendestroyed 283(67.2) 39(81.3) 187(62.3) 0.01 0.14
Sawrefugeeswholosttheirhomesandbelongingsasaresultof 177(42.1) 27(57.4) 115(38.3) 0.01 0.13
battle
Sawpeoplebeggingforfood 364(86.3) 48(100.0) 252(84.0) 0.003 0.16
Tookprisonersofwar 133(31.9) 19(39.6) 82(27.7) 0.09 0.09
Interactedwithenemysoldierswhoweretakenasprisonersofwar 145(34.4) 22(45.8) 94(31.4) 0.05 0.11
Exposedtosight,sound,smellofanimalsthathadbeenwounded 211(50.0) 39(81.3) 127(42.3) <0.001 0.27
orkilledfromwar-relatedcauses
Tookcareofinjuredordyingpeople 191(45.4) 27(56.3) 124(41.5) 0.06 0.10
Involvedinremovingdeadbodiesafterbattle 94(22.3) 15(31.3) 58(19.3) 0.06 0.10
Exposedtothesight,sound,smellofdyingmenandwomen 210(50.0) 32(66.7) 133(44.6) 0.005 0.15
Sawenemysoldiersaftertheyhadbeenseverelywoundedor 174(41.3) 33(68.8) 99(33.1) <0.001 0.25
disfiguredincombat
Sawciviliansaftertheyhadbeenseverelywoundedordisfigured 241(57.4) 42(89.4) 155(51.7) <0.001 0.26
Sawbodiesofdeadcivilians 219(52.4) 37(77.1) 140(47.0) <0.001 0.21
SawAmericansoralliesaftertheyhadbeenseverelywounded 242(57.5) 39(81.3) 159(53.2) <0.001 0.20
ordisfigured
SawthebodiesofdeadAmericansorallies 166(39.3) 26(54.2) 104(34.7) 0.01 0.14
Sawthebodiesofdeadenemysoldiers 153(36.3) 31(66.0) 87(29.0) <0.001 0.27
PTSD,Post-traumaticstressdisorder;PCL-C,civilianversionofthePTSDChecklist;PCL-M,militaryversionofthePTSD
Checklist.
Valuesaregivenasnumber(%).Percentagesreportedaretheproportionofindividualsendorsingeachresponseadjustedto
takeaccountofsampleandmissingdata.
aAmongthosewithoutPTSDsymptomsatbaseline(PCL-CtotalscorefellbelowliberalPTSDsymptomscreeningcut-offof
o34),participantshadnew-onsetprobablePTSDifstringentcriteriaweremet(totalPCL-Mscoreo50andendorsementofat
leastoneintrusionsymptom,threeavoidancesymptoms,andtwohyperarousalsymptomseachatthemoderateorhigherlevel)
atfollow-up/post-deployment.
bAmongthosewithoutPTSDsymptomsatbaseline,participantshadnoPTSDifstringentcriteriaforprobablePTSDwerenot
metatfollow-up/post-deployment.
Riskfactorsfornew-onsetPTSD 693
Table3.RiskandprotectivefactorsreportedbytheRINGScohortparticipantsoverallandamongthosewithoutPTSDatbaseline
(n=349)bynew-onsetprobablePTSDstatus
AmongthosewithoutPTSD
symptomsatbaseline
Overall Meandifference
cohort New-onsetprobable NoPTSDb (95%CIofthe
Riskorprotectivefactor (n=522) PTSDa(n=48) (n=301) p difference)
Pre-deploymentfactorsc
Childhoodfamilyenvironment 53.4(10.2) 52.5(10.91) 54.8(9.6) 0.13 x2.3(x5.3to0.7)
Priorlifestressors 5.6(3.2) 7.4(3.6) 5.1(3.0) <0.001 2.3(1.4to3.2)
Militarypreparedness 34.5(7.4) 31.7(7.1) 35.2(7.2) 0.002 x3.5(x5.7tox1.3)
Unitsocialsupport 40.6(9.9) 40.1(10.8) 41.4(9.6) 0.42 x1.2(x4.2to1.8)
Concernforlife/familydisruption 28.8(7.5) 30.6(7.2) 27.9(7.0) 0.01 2.7(0.5to4.8)
Deploymentfactorsd
Combatexposure 28.8(8.4) 34.8(8.7) 27.2(6.8) <0.001 7.6(5.5to9.8)
Aftermathofbattle 7.1(4.3) 10.0(3.8) 6.4(4.2) <0.001 3.6(2.3to4.8)
Perceivedlifethreat 44.5(9.7) 49.0(8.7) 43.1(9.5) <0.001 5.9(3.0to8.8)
Post-deploymentfactorsd
Post-deploymentsocialsupport 58.8(8.4) 51.7(8.3) 60.5(7.5) <0.001 x8.8(x11.1tox6.5)
Post-deploymentlifestressors 1.1(1.5) 2.0(1.9) 0.7(1.0) <0.001 1.3(1.0to1.7)
RINGS,ReadinessandResilienceinNationalGuardSoldiers;PTSD,post-traumaticstressdisorder;PCL-C,civilianversionof
thePTSDChecklist;PCL-M,militaryversionofthePTSDChecklist.
Valuesaregivenasmean(standarddeviation).
aAmongthosewithoutPTSDsymptomsatbaseline(PCL-CtotalscorefellbelowliberalPTSDsymptomscreeningcut-offof
o34),participantshadnew-onsetprobablePTSDifstringentcriteriaweremet(totalPCL-Mscoreo50andendorsementof
atleastoneintrusionsymptom,threeavoidancesymptoms,andtwohyperarousalsymptomseachatthemoderateorhigher
level)atfollow-up/post-deployment.
bAmongthosewithoutPTSDsymptomsatbaseline,participantshadnoPTSDifstringentcriteriaforprobablePTSDwerenot
metatfollow-up/post-deployment.
cAssessedatbaseline/pre-deployment.
dAssessedatfollow-up.
PTSD status. With the exception of supportive child- sociodemographics, baseline PTSD symptoms, and
hood family environments and unit social support, pre-deploymentriskandprotectivevariables,combat
participantswithnew-onsetprobablePTSDaveraged experiencesandtheaftermathofbattleeachindepen-
significantly higher scores on all hypothesized risk dentlypredictednew-onsetprobablePTSD.However,
factors and significantly lower scores on all hypo- even after controlling for deployment-related stres-
thesizedprotectivefactors thanpanelmemberswith- sors, prior exposure to potentially traumatic events
outnew-onsetprobablePTSD. andperceivedlackofmilitarypreparednessremained
significant independent predictors of new-onset
probable PTSD. Table 4 also shows adjusted post-
Predictorsofnew-onsetprobablePTSD
deployment correlates of new-onset probable PTSD
Among those without PTSD symptoms at baseline, (seeBlock3).Aftercontrollingforparticipants’socio-
we used hierarchical logistic regression analysis to demographic characteristics, baseline PTSD symp-
identify independent pre-deployment, deployment- toms and all other predictor variables, exposure to
related,andpost-deploymentpredictorsofnew-onset recent stressful events and post-deployment social
probable PTSD. As shown in Table 4 (see Block 1), support were significant independent correlates of
after adjusting for sociodemographic characteristics new-onsetprobablePTSD.
andcontrollingforpre-existing,low-levelPTSDsymp-
toms at baseline, unique pre-deployment determi-
Discussion
nants of new-onset probable PTSD assessed prior
to deployment included: prior stressful life events In this longitudinal panel of US National Guard
andperceivedmilitarypreparedness.Asexpectedand soldiers, we demonstrated a nearly 4-fold increase in
shown in Table 4 (see Block 2), after controlling for new-onset probable PTSD 3 months after soldiers
694 M.A.Polusnyetal.
Table4.Hierarchicallogisticregressionanalysispredictingnew-onsetprobablePTSDinNationalGuardsoldiersdeployedtoIraq
amongthosewithoutPTSDatbaseline(n=349)a,b
Block1 Block2 Block3
Pre-deploymentfactors
BaselinePTSDsymptoms 0.73(0.34–1.58) 0.79(0.34–1.85) 0.69(0.27–1.79)
Childhoodfamilyenvironment 1.03(0.71–1.49) 0.95(0.64–1.42) 1.01(0.64–1.59)
Priorlifestressors 2.20(1.47–3.28)*** 1.75(1.13–2.70)* 1.39(0.85–2.27)
Militarypreparedness 0.58(0.39–0.87)** 0.62(0.40–0.95)* 0.77(0.48–1.25)
Concernsaboutlife/familydisruptions 1.38(0.97–1.97) 1.31(0.88–1.95) 1.12(0.71–1.77)
Unitsupport 1.43(0.95–2.15) 1.15(0.73–1.79) 1.15(0.70–1.89)
Deploymentexposurefactors
Combatexperiences – 2.19(1.40–3.41)*** 2.35(1.41–3.92)**
Exposuretoaftermathofbattle – 1.62(1.04–2.53)* 1.81(1.08–3.06)*
Perceivedlifethreat – 1.21(0.81–1.81) 1.01(0.63–1.64)
Post-deploymentfactors
Post-deploymentsocialsupport 0.31(0.19–0.50)***
Post-deploymentlifestressors 1.96(1.17–3.28)*
PTSD,Post-traumaticstressdisorder;PCL-C,civilianversionofthePTSDChecklist;PCL-M,militaryversionofthePTSD
Checklist.
Valuesaregivenasoddsratio(95%confidenceinterval)aftercontrollingforage,gender,race,andmilitaryrank.
aNew-onsetprobablePTSDdefinedasnopre-existingPTSDsymptoms(baselinePCL-CtotalscorefellbelowliberalPTSD
symptomscreeningcut-offofo34)atbaselineandmetstringentcriteriaforprobablePTSD(totalPCL-Mscoreo50and
endorsementofatleastoneintrusionsymptom,threeavoidancesymptoms,andtwohyperarousalsymptomseachatthe
moderateorhigherlevel)atfollow-up/post-deployment.
bDatafromtheReadinessandResilienceinNationalGuardSoldiers(RINGS)study(baselinedataassessingpre-deployment
factorswerecollected1monthpriortotroops’deploymenttoIraqinMarch2006;follow-updataassessingdeployment
exposureandpost-deploymentfactorswerecollected2–3monthsfollowingtroops’returnfromdeployment).
*p<0.05,**p<0.01,***p<0.001.
returned from OIF deployment compared with their warriormayfeelespeciallyunprepared(e.g.killinga
pre-deployment base rates. This longitudinal cohort non-combatant) and may significantly contribute to
of US National Guard soldiers reported high combat PTSD risk (Litz et al. 2009). When examining specific
exposure, comparable with that reported by US aspects of combat stressors, we found that killing
active duty soldiers and Marines deployed to Iraq (e.g.‘killedorbelievedtohavekilledenem[ies]during
(Hoge et al. 2004). Not surprisingly, frequency and combat’) was an important predictor of new-onset
intensity of combat were potent predictors of new- probable PTSD. This association is consistent with
onsetprobablePTSD.However,exposuretothesight, others’ recent reports (Rona et al. 2009). One expla-
sound and smell of combat’s aftermath also inde- nationfortheassociationbetweensoldiers’reportsof
pendently predicted probable PTSD. Besides combat, killing in combat and new-onset PTSD is that killing
new-onsetprobablePTSDwasuniquelypredictedby reflectsintensecombatexposureandlifethreatdueto
soldiers’pre-deploymentstressorexposuresandtheir being in close contact with the enemy. However,
perceptions ofmilitary preparedness. Both thesepre- Maguen et al. found that killing was a significant
deployment factors remained significant even after predictor of PTSD symptoms even after controlling
controllingforparticipants’baselinePTSDsymptoms for combat exposure (Maguen et al. 2010). Litz et al.
and their deployment stressor exposures. However, havearguedthatwarriors’vulnerabilitytoPTSDafter
after controlling for post-deployment factors,noneof killing results not simply from exposure totraumatic
thepre-deploymentfactorssignificantlypredictedthe events,butfrommoralinjuriesthatmaybesignsofthe
developmentofprobablePTSD. warrior’s humanity (Litz et al. 2009). Consistent with
The wars in Iraq and Afghanistan have been char- this notion, these authors have suggested that self-
acterized by unconventional features (e.g. use of im- forgiveness may play an important role in recovery
provisedexplosivedevicesbyanindistinctiveenemy, aftermoralinjury.Whileitmaynotbepossibletofully
counterinsurgency, and urban warfare) that may prepareforthechallengesofcombat,preventionstrat-
produce ambiguous combat situations for which the egies aimed at enhancing soldiers’ sense of mastery
Riskfactorsfornew-onsetPTSD 695
and self-efficacy could buffer soldiers against the soldiers’ resilience and increase vulnerability to
stressful effects of combat exposure (Hobfoll, 1989) PTSD. These findings are consistent with other
andbyextensionreduceriskforlaterPTSD. studies (Benotsch et al. 2000;Browne et al. 2007),and
In this prospective study, we found that soldiers’ suggest that post-deployment interventions aimed
risk of developing new-onset probable PTSD follow- at enhancing soldiers’ interpersonal resources at
ingdeploymentwaspredictedbysoldiers’reportsof home, work, and in the community and alleviating
priorstressorexposuresevenaftercontrollingfortheir subsequent stressors (e.g. unemployment, family
baseline PTSD symptoms and deployment stressor distress) might enhance recovery and resiliency.
exposures. This finding is contrary to other recent Further research is needed to understand the role of
findingsthatpriortraumaexposureintheabsenceof military families in harnessing social support for
PTSDwasnotassociatedwithincreasedvulnerability soldiers and how stressors associated with impaired
of developing PTSD following a subsequent trauma familyfunctioningmayfurther increasevulnerability
(Breslau et al. 2008; Breslau & Peterson, 2010). Prior forPTSD.
stressful life events may sensitize soldiers to the Therateofnew-onsetprobablePTSD(13.8%)inour
deleterious effects of combat exposure and amplify panelwassimilartothePTSDprevalencereportedfor
previously traumatized soldiers’ vulnerability to de- ArmyinfantrysoldiersandMarinesabout3–4months
velopingPTSD(Schummetal.2005).Ourfindingthat after their return from deployment to Iraq (12.9%)
previous exposure to traumatic stressors increased (Hogeetal.2004),buthigherthanthatreportedforthe
soldiers’ risk for new-onset probable PTSD is con- Millennium Cohort Study (7.6%) (Smith et al. 2008)
cerninginlightofthehighratesofpre-militarytrauma and UK reservists deployed to Iraq (6.5%) in the
exposurereportedbymilitarypersonnel(Boltonetal. King’s Centre for Military Health Research Study
2001) and the large numbers of personnel serving (Hotopf et al. 2006). Given that these studies used
multiple deployments. With the US military’s sus- the same instrument and criteria to define PTSD, the
tained operations in Afghanistan and Iraq, it will be higher PTSD rate documented in this panel may be
important to understand how multiple combat de- duetogreatercombatexposure(Iversenetal.2009).
ploymentshaveanimpactonriskfornew-onsetPTSD Thisstudyhadseveralstrengths,includingitspro-
amongredeployingmilitarypersonnel. spectiveassessmentofriskandprotectivefactorsprior
Thefactthatpre-deploymentfactorswerenolonger to participants’ deployment to OIF, its focus on US
significant predictors of new-onset probable PTSD National Guard soldiers, and its focus on potentially
aftercontrolling fortheinfluenceofpost-deployment modifiable risk factors for PTSD. In terms of limi-
factorscannotbeduetopre-deploymentfactorsbeing tations, participants were self-selected, although the
causedbypost-deploymentfactors.Thisisbecauseof panel was representative of the overall brigade in
thetemporalnatureofvariablesinthisstudy(e.g.pre- terms ofgender,race/ethnicity andrank,and results
deployment factors were assessed prospectively be- may not generalize to active duty military personnel
foretheoccurrenceofpost-deploymentfactors).While ortoothermilitarybranches.Moreresearchisneeded
it is possible that the pre-deployment factors may totestwhetherpredictorsofnew-onsetPTSDdifferfor
have remained significant with a larger sample, it is active duty members. Although we obtained follow-
alsopossiblethatsomethirdvariable(e.g.personality) up data from more than 80% of the original panel
was predictive of both pre-deployment and post- and our analyses of responders and non-responders
deployment factors. For example, soldiers’ reporting showed few differences in pre-deployment measures
ofpriorstressfullifeeventsandotherriskfactorsmay of risk and protective factors, non-responders to
reflect, at least in part, individual differences in per- follow-upwereimportantlydifferentfromresponders
sonality(e.g.thetendencytobemore‘stressed’bylife (younger,morelikelytobenon-white,unmarriedand
circumstances). Future studies need to examine the less educated) and post-deployment findings could
roleofpersonalityfactors,suchasneuroticismwhich havebeeninfluencedbyresponsebiases.
is a robust risk factor for PTSD (Rubin et al. 2008), in WhileweusedavalidandreliablemeasureofPTSD
understandingtheserelationships. symptomatology, with highly sensitive and specific
Our findings also suggest that the development of definitions for probable PTSD diagnosis, self-report
PTSD following deployment is associated with lower data are susceptible to information biases, and mis-
perceived social support and experiencing a greater classificationerrorcouldhavedampenedassociations
number of recent stressful life events. As National betweensomevariables.Thissuggests,however,that
Guard soldiers transition from the combat zone to the associations between new-onset probable PTSD
their civilian lives, lack of post-deployment social and military preparedness, post-deployment social
support and additional life stressors represent two supportandothertraumaexposuresareevenstronger
important resource losses that appear to erode than our data suggest. An important limitation of