Lifetime Prevalence of DSM-IV Mental Disorders Among New Soldiers in the U.S. Army

Research article from Depression and Anxiety, published October 2014
of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  D  EPRESSION AND  A  NXIETY   00:1–12 (2014)  Research Article LIFETIME PREVALENCE OF DSM-IV MENTALDISORDERS AMONG NEW SOLDIERS IN THE U.S. ARMY:RESULTS FROM THE ARMY STUDY TO ASSESS RISK  AND RESILIENCE IN SERVICEMEMBERS (ARMY STARRS)  Anthony J. Rosellini, Ph.D., 1 Steven G. Heeringa, Ph.D., 2  Murray B. Stein, M.D., M.P.H., 3,4 Robert J. Ursano, M.D., 5  Wai Tat Chiu, A.M., 1 Lisa J. Colpe, Ph.D., M.P.H., 6 Carol S. Fullerton, Ph.D., 5 Stephen E. Gilman, Sc.D., 7 Irving Hwang, M.A., 1  James A. Naifeh, Ph.D., 5  Matthew K. Nock, Ph.D., 8  Maria Petukhova, Ph.D., 1 Nancy A. Sampson, B.A., 1  Michael Schoenbaum, Ph.D., 6  Alan M. Zaslavsky, Ph.D., 1 and Ronald C. Kessler, Ph.D. 1 ∗ Background:  The prevalence of 30-day mental disorders with retrospectivelyreported early onsets is significantly higher in the U.S. Army than among socio-demographically matched civilians. This difference could reflect high prevalenceofpreenlistmentdisordersand/orhighpersistenceofthesedisordersinthecontext of the stresses associated with military service. These alternatives can to some ex-tentbedistinguishedbyestimatinglifetimedisorderprevalenceamongnewArmyrecruits.  Methods:  The New Soldier Study (NSS) in the Army Study to Assess  Risk and Resilience in Servicemembers (Army STARRS) used fully structured measures to estimate lifetime prevalence of 10 DSM-IV disorders in new soldiers reporting for Basic Combat Training in 2011–2012 (n  =  38,507). Prevalencewas compared to estimates from a matched civilian sample. Multivariate re- gression models examined socio-demographic correlates of disorder prevalence and persistence among new soldiers.  Results:  Lifetime prevalence of having at least one internalizing, externalizing, or either type of disorder did not differ  significantly between new soldiers and civilians, although three specific disorders (generalized anxiety, posttraumatic stress, and conduct disorders) and multi-morbidity were significantly more common among new soldiers than civilians. Although several socio-demographic characteristics were significantly associated with disorderprevalence and persistence,these associationswere uniformly weak. Conclusions:  New soldiers differ somewhat, but not consistently, from civilians  1 Department of Health Care Policy, Harvard Medical School,Boston, Massachusetts 2 University of Michigan, Institute for Social Research, Ann Ar-bor, Michigan 3 Departments of Psychiatry and Family and PreventiveMedicine, University of California San Diego, La Jolla, Cali-fornia 4 VA San Diego Healthcare System, San Diego, California 5 Center for the Study of Traumatic Stress, Department of Psy-chiatry, Uniformed Services University of the Health Sciences,Bethesda, Maryland 6 National Institute of Mental Health, Bethesda, Maryland 7 Departments of Social and Behavioral Sciences, and Epi-demiology, Harvard School of Public Health, Boston, Mas-sachusetts 8 Department of Psychology, Harvard College, Cambridge,Massachusetts Contract grant sponsor: Department of the Army, U.S. DepartmentofHealthandHumanServices,andNIH/NIMH;contractgrantnum-ber: U01MH087981. ∗ Correspondenceto:RonaldC.Kessler,DepartmentofHealthCarePolicy, Harvard Medical School, 180 Longwood Avenue, Boston,MA 02115. E-mail: Kessler@hcp.med.harvard.eduReceived for publication 27 June 2014; Revised 15 August 2014;Accepted 24 August 2014DOI 10.1002/da.22316Published online in Wiley Online Library(wileyonlinelibrary.com). C  2014 Wiley Periodicals, Inc. UNDER EMBARGO UNTIL THURSDAY, OCTOBER 23, 4:00 PM ET  2   Rosellini et al. in lifetime preenlistment mental disorders. This suggests that prior findings of  higher prevalence of current disorders with preenlistment onsets among soldiers than civilians are likely due primarily to a more persistent course of early-onset disorders in the context of the special stresses experienced by Army personnel.Depression and Anxiety 00:1–12, 2014.  C  2014 Wiley Periodicals, Inc. Key words:  military personnel; mental disorders; prevalence; epidemiology;demographics  INTRODUCTION   M ental disorders are leading causes of U.S. military morbidity. [1]  This high relative burden of mental dis-orders could reflect the fact that soldiers are physically healthy at the time of enlistment due to serious physi-cal disorders being exclusions from military service, but might also be due partly to a high absolute burden of mental disorders in the military compared to civilians. The scant data on this issue suggest that military per-sonnel on active duty have higher rates of some mentaldisorders than civilians. [2]  The most rigorous study of thisissuetodatecomesfromaself-reportsurvey,theAll- Army Study (AAS), in the  Army Study to Assess Risk and  Resilience in Servicemembers   (Army STARRS). [3,4]  The AAS assessed a representative sample of nondeployedU.S. Army soldiers exclusive of those in Basic Com-bat Training (BCT) and found that the prevalence of having at least one common psychiatric disorder in the30daysbeforeinterviewwasconsiderablyhigheramongthese soldiers (25.1%) than a civilian sample calibratedto have similar socio-demographics as soldiers andnot to have exclusions for enlistment (11.6%). [5]  Al-though this higher prevalence in the Army might bedue to the unique stressors associated with military service, [6–9] another possibility is that differential selec-tion exists into military service on the basis of preenlist-ment mental disorders or risk factors for such disorders.Evaluatingtherelativeimportanceofthesetwopossibil-ities is important given recent discussions about optimalrecruitment, retention, and health care delivery strate-gies for an all-volunteer Army during times of war. [10,11]  The AAS provided some limited information on thisissue by asking respondents retrospectively to report theage-of-onset (AOO) of their 30-day mental disorders. Three-quarters (76.6%) of respondents reported onsetsprior to enlistment. This high proportion should not besurprising, as general population epidemiological stud-ies find most lifetime mental disorders have childhood-adolescence onsets. [12–14] But a more striking result wasthat a significantly higher proportion of respondents with 30-day disorders in the civilian comparison samplereported early onsets (91.2% vs. 76.6%,  χ 21  = 10.7,  P  = .001).However,absolute prevalence of 30-day disorders with preenlistment onsets was nonetheless significantly higher among soldiers than civilians (19.2% vs. 10.6%, χ 21  = 10.4,  P  = .001). The higher absolute prevalence of 30-day disorders with early onsets among soldiers than civilians could beduetoanyofthreeprocesses:(1)recallerrorinretrospec-tive AOO reports; (2) early-onset disorders and/or theirrisk factors being positively associated with Army enlist-ment; and (3) higher chronicity of early-onset disordersamong soldiers than civilians (possibly due to the specialstressors associated with Army service). The first pos-sibility is implausible because methodological researchsuggests that the tendency in such dating errors is to re-call first onset as more recent than actually occurred, [15] and there is no reason to think that recall error wouldbe greater among soldiers than civilians. However, theremaining possibilities are both plausible. Adjudication between these two possibilities couldbe important in helping to design Army preventive in-terventions, including early interventions for high-risk groups or early treatment if soldiers were found to havesignificantly higher rates of child-adolescent disordersthan civilians. The data in the AAS did not allow forthis analysis, as lifetime prevalence was assessed only among soldiers with 30-day disorders. However, usefulinformation on this issue could be obtained by compar-ing lifetime prevalence of mental disorders among new  Army recruits and civilians. We present the results of such a study in the current report, examining preenlist-ment prevalence and socio-demographic correlates of anumber of common mental disorders.  MATERIALS AND METHODS SAMPLE Data come from the Army STARRS New Soldier Study (NSS).Unlike the AAS, which did not include soldiers in BCT, the NSS wascarried out exclusively among new soldiers who had already been suc-cessful in passing the screening hurdles for Army enlistment (i.e., forhistories of criminal behaviors, severe physical disorders-handicaps,and severe mental illness) [16] and were about to begin BCT at one of three Army Installations (Fort Benning, GA; Fort Jackson, SC; andFort Leonard Wood, MO) between April 2011 and November 2012.Data collection occurred during the days immediately prior to start-ing BCT when new soldiers were processed (e.g., completing physicalexams; issuance of uniforms). Samples sizes were proportional to therelative size of the cohorts across installations. Recruitment began by selecting a weekly sample of 200–300 new soldiers in each installationto attend a study overview and informed consent presentation for thestudy. Army STARRS staff worked closely with Army coordinators to Depression and Anxiety   Research Article: Mental Disorders Among New U.S. Army Soldiers   3  guarantee that these samples were representative of all new soldiers ineach weekly cohort. The overview and informed consent presentationexplained study purposes, confidentiality, emphasized that participa-tion was voluntary, and answered all questions before seeking writteninformed consent to (i) complete a self-administered questionnaire(SAQ), (ii) link administrative records to SAQ responses, and (iii) par-ticipate in future data collections. Identifying information (e.g., name,SSN)wascollectedfromconsentingrespondentsandkeptinaseparatesecurefile.Theserecruitment,consent,anddataprotectionprocedures were approved by the Human Subjects Committees of the UniformedServices University of the Health Sciences for the Henry M. JacksonFoundation (the primary grantee), the Institute for Social Research at the University of Michigan (the organization collecting the data), andall other collaborating organizations. The 38,507 NSS respondents considered here represent all con-senting soldiers who completed the SAQ April 2011–November2012. All new soldiers selected to attend the informed consent session did so, virtually all (99.9%) provided consent, and most (93.7%) completed the full SAQ (see Appendix Table 5, available at  www.armystarrs.org/publications). Incomplete surveys were primarily due to time constraints (e.g., cohorts arriving late or having to leaveearly; certain respondents being unable to fully complete the surveysduringtheallottedtime).Mostsoldierswho completedthe surveyalsoprovided consent for and were successfully linked to their adminis-trative records (77.0%). All analyses reported here utilize a combinedanalysis weight that both adjusts for differential administrative recordlinkageconsentamongsoldierswhocompletedthesurveyandincludesapoststratificationoftheseconsentweightstoknowndemographicandservicecharacteristicsofthepopulationofnewsoldiersattendingBCTduring the study period. A detailed description of NSS clustering and weighting is available elsewhere. [17]  THE COMPARISON CIVILIAN SAMPLE Lifetime prevalence of DSM-IV disorders was compared to esti-mates from a subsample of the National Comorbidity Survey Repli-cation (NCS-R) [18] limited to respondents who lacked self-reportedexclusions for Army service (histories of criminal behaviors, se- vere physical disorders-handicaps, and severe mental illness) and was weighted to have the same multivariate distribution as the NSS on arange of socio-demographics separately among soldiers in the Regu-lar Army and in the Army National Guard or Army Reserve. A de-tailed discussion of the civilian sample and calibration is presentedelsewhere. [19]  MEASURES Diagnostic Assessment.  NSS respondents self-administered acomputerized version of the Composite International Diagnostic In-terview screening scales (CIDI-SC) [20] and a screening version of thePTSDChecklist(PCL) [21] toassess10lifetimeDSM-IVmentaldisor-ders.Wefocusedonlifetimeprevalenceratherthan30-dayprevalencebecause we were interested in studying differences in the rates of any preenlistment mental disorders rather than current disorders at thetime of accession. The NCS-R assessed the same lifetime disorders with the full CIDI, [20]  which means that between-survey comparisonsofprevalenceareinexact.RespondentsintheNSSbutnotNCS-Rwithlifetime disorders were also asked how many years each disorder hadbeen present at least some of the time. We examined these responsesto assess persistence of preenlistment disorders both by studying theabsolute number of years in which each disorder occurred beyond the year of onset and also the ratio of number of years in which each dis-order occurred beyond the year of onset divided by the total numberof years since onset. The latter ratios were calculated at the aggregatelevelforeachdisordertoadjustforsomedisordershavingmuchearlierages-of-onset than others. [13]  Although this is only a rough measure of persistence,itisnonethelessusefulinprovidingageneralsenseofhow often preenlistment disorders are persistent rather than short lived. Wedistinguishedbetweeninternalizingandexternalizingdisordersbased on empirical evidence for this distinction. [22] Five internaliz-ing disorders were assessed: major depressive episode (MDE), bipolarI-II or subthreshold bipolar disorder (BPD), generalized anxiety dis-order (GAD), panic disorder (PD), and posttraumatic stress disorder(PTSD),alongwithfiveexternalizingdisorders:intermittentexplosivedisorder (IED), conduct disorder (CD), oppositional defiant disorder(ODD), substance use disorder (SUD; alcohol or drug abuse or de-pendence),andattention-deficit/hyperactivity disorder(ADHD).TheSUD assessment included not only illicit drugs but also misused pre-scriptiondrugsbasedonevidencethatprescriptiondrugmisuseiscon-siderablymorecommonthanillicitdruguseintheArmy. [23] Diagnosesin both surveys were made without DSM-IV diagnostic hierarchy ororganic exclusion rules. As reported in detail elsewhere, [24] an Army STARRS clinical reappraisal study found good concordance betweenCIDI-SC and modified PCL diagnoses and independent clinical di-agnoses based on blinded Structured Clinical Interviews for DSM-IV (SCID). [25]  The clinical reappraisal study also found CIDI-SC andPCL prevalence estimates were unbiased relative to SCID estimates( χ 21  =  0.0–0.6,  P   =  .89–.43). The earlier report, [24]  which includeddetailed concordance results for each of the 10 disorders studied here,is available elsewhere (www.armystarrs.org/publications). Socio-Demographics.  Socio-demographics included respon-dentage,sex,race-ethnicity,soldiereducation,maritalstatus,religion,soldier and parent nativity, and parent education relative to respon-dent education. Separate questions were asked about Hispanic ethnic-ity(yes–no)andrace( White,BlackorAfrican-American,AmericanIndianor Native American, Asian [e.g., Chinese, Filipino, Indian], Native Hawai-ian or other Pacific Islander  , and  Other  ), with responses collapsed intosummary categories of Non-Hispanic Black, Non-Hispanic White,Hispanic, and Other.  ANALYSIS METHODS Cross-tabulationswere usedtoestimatedisorderprevalence. Com-parisonofprevalenceestimatesintheNSSandthecomparisonciviliansample was used to determine if preenlistment prevalence was higheramong new soldiers than civilians. Socio-demographic predictors of disorder onset and persistence were examined to determine if highpreenlistment disorder risk was isolated in a small subset of new sol-diers or widely distributed. Logistic regression was used to predict lifetimedisordersandnegative binomial regressiontopredictdisorderpersistence controlling for AOO and number of years since onset. Co-efficients and standard errors were exponentiated in logistic models tocreate odds ratios (ORs) with 95% confidence intervals and in nega-tive binomial models to create incident rate ratios (IRRs; the expecteddifference in mean number of years of persistence associated with a 1unitincreaseinthepredictor)with95%confidenceintervals.Strengthof associations was evaluated with Cramer’s V ( ϕ c ). All analyses were carried out using weighted data. Design effectsdue to weighting and implicit stratification by location and cluster-ing were handled by using the design-based Taylor series linearizationmethod [26] to estimate standard errors. Pseudo-strata were defined toimplement this method based on location and bi-weekly time win-dows treating each weekly time-space cluster as a separate samplingerror calculation unit. Significance of predictor sets was evaluated us-ingdesign-basedWald χ 2 tests.AllanalyseswerecarriedoutwithSAS Version 9.3, [27]  with  proc surveyfreq  to estimate prevalence,  proc survey-logistic  toestimatelogisticmodels,and  proc genmod  toestimatenegativebinomial models. Depression and Anxiety  4   Rosellini et al. RESULTS SOCIO-DEMOGRAPHIC DISTRIBUTIONS Distributions of socio-demographic variables in the weightedNSSRegularArmyandNationalGuard/Army Reserve (Guard/Reserve) were comparable to those inthe target population of all new soldiers (Table 1). LIFETIME DISORDER PREVALENCE  The estimated lifetime prevalence in the total NSSsample was 38.7% for any DSM-IV/CIDI-PCL disor-der, 19.8% for internalizing disorder, and 31.8% forany externalizing disorders. PTSD was the most com-mon internalizing disorder (12.6%) and IED the most common externalizing disorder (14.6%). These generalpatterns were very similar in the Regular Army andGuard/Reserve, although prevalence was consistently somewhat higher in the latter than former, with theGuard/Reserve having higher prevalence of any dis-order (40.0% vs. 37.6%;  χ 21  =  14.0,  P   <  .001), any internalizing disorder (21.0% vs. 18.8%;  χ 21  =  19.4,  P   <  .001), each internalizing disorder other than BPD(3.3–13.3% vs. 2.7–12.1%;  χ 21  = 7.2–19.7,  P   <  .001 to  P  = .007),andtwoexternalizingdisorders(IED,ADHD;7.0–15.1% vs. 5.9–14.2%;  χ 21  = 4.5–9.3,  P  = .002–.034; Table 2).Lifetime prevalence differences between all new sol-diers and the civilian sample were not significant for theaggregate variables representing any DSM-IV/CIDI-PCL disorder (38.7% vs. 36.5%;  χ 21  =  0.1,  P   =  .76),any internalizing disorder (19.8% vs. 20.3%;  χ 21  = 0.0,  P  = .93),oranyexternalizingdisorder(31.8%vs.28.8%; χ 21  = 0.2,  P  = .62). However, prevalence of three indi- vidual disorders (GAD, PTSD, CD) were significantly higher among soldiers than civilians. The differencesin GAD (8.2% vs. 1.2%;  χ 21  =  245.0,  P   <  .001) andPTSD(12.6%vs.2.5%; χ 21  = 44.5,  P  < .001)weremuchmore striking than the difference in CD (5.9% vs. 3.3%; χ 21 = 3.9,  P  = .048).Thesedifferencesresultedinasignif-icantly higher proportion of new soldiers than civilianshaving multi-morbidity (3 +  lifetime disorders; 11.3% vs. 6.5%;  χ 21  =  4.0,  P   =  .046). These soldier-versus-civilian differences were broadly similar when examinedseparately in the Regular Army and Guard/Reserve ex-cluding that the prevalence of CD was not significantly higher among new soldiers in the Guard/Reserve thancivilians (5.5% vs. 3.6%;  χ 21  = 1.6,  P  = .21). PERSISTENCE OF LIFETIME DSM-IV/CIDI-PCLDISORDERS  Mean years of disorder persistence (exclusive of  ADHD, for which persistence was not assessed) acrossdisorders was comparable among new soldiers inthe Regular Army (1.3–4.5) and Guard/Reserve (1.2–4.4) (Table 3). IED was the only disorder withmean persistence significantly different in the Regular Army than Guard/Reserve, although the difference wassubstantivelysmall(3.6vs.3.4; χ 21  = 4.9,  P  = .027).Meanpersistence ratios were in the range 33.2–60.7% for theRegular Army and 31.6–62.0% for the Guard/Reserveand BPD was the only disorder with a persistenceratio that significantly differed in the Regular Army than Guard/Reserve (41.9% vs. 48.3%,  χ 21  =  4.9,  P   = .027). Mean persistence was generally higher for ex-ternalizing (3.4–4.5) than internalizing (1.4–3.0) dis-orders with the exception of SUD. It is notewor-thy that the two highest persistence ratios were forPD (60.7–62.0%) and IED (57.0–58.4%), both of  which are characterized by repeated and uncontrol-lable attacks (of fear in the case of PD and anger inthe case of IED) out of proportion to precipitatingevents. SOCIO-DEMOGRAPHIC PREDICTORS OF PREVALENCE AND PERSISTENCE  The vast majority of associations between socio-demographics and lifetime disorders were statistically significant in multivariate models, including 22 of 24in pooled models (Table 4) and 51 of 80 in models forindividual disorders. (The tables for individual disor-ders are available at (www.armystarrs.org/publications). These associations were for the most part in the direc-tion predicted by previous research: higher rates of in-ternalizing disorders among women and soldiers withNon-Western religions; higher rates of externalizingdisordersamongmenandtheunmarried;andinverseas-sociations of age, minority status (Non-Hispanic Black and Hispanic), soldier and parent education, and immi-grant status with both internalizing and externalizingdisorders. However, these statistically significant asso-ciations were all small in substantive terms ( ϕ c  in therange .00–.07). The significant associations of socio-demographics with disorder persistence were less consistent: 16 of 27associations in pooled models and 29 of 81 in modelsfor individual disorders. Persistence was higher among women than men and Non-Hispanic Whites than mi-norities (only for externalizing disorders), lower amongimmigrantsthan1standlatergenerationAmericans,andinversely related both to AOO and to time-since-onset (see Table 5). Parent education was related inversely to persistence of internalizing disorders and positively to persistence of externalizing disorders. Religion, sol-dier education, and marital status were unrelated to per-sistence. As with prevalence, the statistically significant associations with persistence were small in substantiveterms ( ϕ c  in the range .02–.09) other than those involv-ing age-of-onset and time-since-onset ( ϕ c  in the range.06–.27). DISCUSSION   Theaboveresultsareimportantindemonstratingthat new soldiers in the U.S. Army during 2011–2012, al-though having higher rates of GAD, PTSD, CD, and Depression and Anxiety
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks