UBC Faculty Research and Publications

Suicide Attempts and Childhood Maltreatment Among Street Youth : A Prospective Cohort Study Hadland, Scott E.; Wood, Evan; Dong, Huiru; Marshall, Brandon David Lewis; Kerr, Thomas; Montaner, Julio; DeBeck, Kora Sep 30, 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-Hadland_S_et_al_Suicide_attempts_childhood.pdf [ 716.85kB ]
JSON: 52383-1.0339929.json
JSON-LD: 52383-1.0339929-ld.json
RDF/XML (Pretty): 52383-1.0339929-rdf.xml
RDF/JSON: 52383-1.0339929-rdf.json
Turtle: 52383-1.0339929-turtle.txt
N-Triples: 52383-1.0339929-rdf-ntriples.txt
Original Record: 52383-1.0339929-source.json
Full Text

Full Text

Suicide Attempts and ChildhoodMaltreatment Among Street Youth:A Prospective Cohort StudyScott E. Hadland, MD, MPHa,b, Evan Wood, MD, PhDc,d, Huiru Dong, MScc, Brandon D.L. Marshall, PhDe, Thomas Kerr, PhDc,d,Julio S. Montaner, MDc,d, Kora DeBeck, PhDc,fabstract BACKGROUND: Although suicide is a known leading cause of death among street youth, fewprospective studies have explored childhood experiences as risk factors for future suicideattempt in this population. We examined the risk of attempted suicide in relation to childhoodmaltreatment among street youth.METHODS: From September 2005 to November 2013, data were collected from the At Risk Youth Study(ARYS), a prospective cohort of street youth in Vancouver, Canada. Inclusion criteria were age 14 to26 years, past-month illicit drug use, and street involvement. Participants completed the ChildhoodTrauma Questionnaire, an instrument measuring self-reported sexual, physical, and emotional abuseand physical and emotional neglect. Suicide attempts were assessed semiannually. Using Coxregression, we examined the association between the 5 types of maltreatment and suicide attempts.RESULTS: Of 660 participants, 68.2% were male and 24.6% were Aboriginal. Median age was21.5 years. The prevalence of moderate to extreme childhood maltreatment ranged from16.8% (sexual abuse) to 45.2% (emotional abuse). Participants contributed 1841 person-years, with suicide attempts reported by 35 (5.3%) individuals (crude incidence density: 1.9per 100 person-years; 95% confidence interval [CI]: 1.4–2.6 per 100 person-years). In adjustedanalyses, types of maltreatment associated with suicide attempts included physical abuse(adjusted hazard ratio [HR]: 4.47; 95% CI: 2.12–9.42), emotional abuse (adjusted HR: 4.92; 95%CI: 2.11–11.5), and emotional neglect (adjusted HR: 3.08; 95% CI: 1.05–9.03).CONCLUSIONS: Childhood maltreatment is associated with subsequent risk of suicidal behavioramong street youth. Suicide prevention efforts should be targeted toward this marginalizedpopulation and delivered from a trauma-informed perspective.WHAT’S KNOWN ON THIS SUBJECT: Street youthdemonstrate elevated mortality compared withthe general adolescent and young adultpopulation. Suicide is a leading cause of deathamong street youth. Many street youth haveexperienced childhood maltreatment, includingabuse and neglect.WHAT THIS STUDY ADDS: In this prospectivecohort of street youth, self-reported attemptedsuicide and history of childhood maltreatmentwere common. Individuals who experiencedchildhood physical abuse, emotional abuse, oremotional neglect were at highest risk ofattempting suicide.aDivision of Adolescent/Young Adult Medicine, Department of Medicine, Boston Children’s Hospital, Boston, Massachusetts;bDepartment of Pediatrics, Harvard Medical School, Boston, Massachusetts; cBritish Columbia Centre for Excellence inHIV/AIDS, St Paul’s Hospital, Vancouver, British Columbia, Canada; dFaculty of Medicine, University of British Columbia,Vancouver, British Columbia, Canada; eDepartment of Epidemiology, Brown University School of Public Health, Providence,Rhode Island; and fSchool of Public Policy, Simon Fraser University, Vancouver, British Columbia, CanadaDr Hadland designed the study, wrote the protocol, conducted the literature review, provided inputfor statistical analyses, and wrote the first draft of the manuscript; Dr Wood designed the study,wrote the protocol, provided input for statistical analyses, contributed to the first draft of themanuscript, and contributed to the final manuscript; Ms Dong undertook data management andstatistical analyses, contributed to the first draft of the manuscript, and contributed to the finalmanuscript; Dr Marshall designed the study, wrote the protocol, provided input for statisticalanalyses, and contributed to the final manuscript; Drs Kerr and Montaner designed the study, wrotethe protocol, and contributed to the final manuscript; Dr DeBeck designed the study, wrote theprotocol, aided with the literature review, provided input for statistical analyses, and contributed tothe first draft of the manuscript; and all authors approved the final manuscript as submitted.www.pediatrics.org/cgi/doi/10.1542/peds.2015-1108DOI: 10.1542/peds.2015-1108Accepted for publication Jun 23, 2015ARTICLE PEDIATRICS Volume 136, number 3, September 2015Each year, .10 000 adolescents andyoung adults die of suicide in theUnited States, where it is the thirdleading cause of death among youthaged 15 to 24 years and accounts for20% of all deaths in this age group.1In Canada, suicide also accounts for20% of all deaths among those ,25years old and is the second leadingcause of death among adolescentsand young adults.2 Generalpopulation–based estimates ofsuicidal behavior may not accuratelyreflect rates among hidden youthpopulations, particularly those whoare homeless or street-involved andmay be at even higher risk of suicide.Street youth (young people living orworking all or part of their time onthe street) are a marginalizedpopulation with greatly elevatedmortality compared with the generalyouth population.3 In one of the onlylongitudinal studies of mortalityamong street youth conducted todate, the death rate was 11 times thatof the general youth population, andsuicide was the single leading causeof death.4Understanding why street youth areat high risk of suicide is critical todeveloping prevention efforts. Illicitdrug use, which is highly prevalentamong street youth,5,6 is stronglyassociated with suicidal ideation andbehavior.4,7–10 Illicit drug use is, inturn, associated with childhoodmaltreatment, which can includesexual, physical, and emotional abuseand physical and emotionalneglect.11,12 Among adult drug users,childhood physical and sexual abuseare determinants of suicidal ideationand attempts.13,14 Emotional abuseand physical and emotional neglecthave been the subject of less studybut are emerging as important riskfactors for suicide.15,16 A history ofchildhood maltreatment is commonamong street youth,17 but itsassociation with suicide in thisvulnerable population has receivedlittle study. In addition, thecontributions of different types ofabuse (sexual, physical, andemotional) and neglect (physical andemotional) to suicide risk have notbeen systematically examined.We conducted this prospective cohortstudy in street youth to examine therisk of attempted suicide in relationto childhood maltreatment. Wehypothesized that more severe levelsof maltreatment would be associatedwith an elevated risk of attemptingsuicide. Consistent with recent datafrom adults,16 we also hypothesizedthat emotional abuse and neglectwould be correlates of risk ofattempted suicide.METHODSStudy DesignThe At Risk Youth Study (ARYS) isa prospective cohort of street youthin Vancouver, Canada, which hasbeen described previously.18Participants were recruited betweenSeptember 2005 and November2013. Inclusion criteria were age 14to 26 years and past-month use ofillicit drugs other than marijuana.(Because marijuana use is highlyprevalent among Canadian youth,19inclusion criteria required use ofdrugs other than marijuana to enrollyouth at risk of initiating injection,a primary study outcome.) Eligibleyouth were street-involved, definedas having been without housing inthe preceding 6 months, or, if nothomeless, having accessed servicesfor street-involved youth during thattime.6,20,21Recruitment occurred duringdaytime and nighttime inneighborhoods where street youthwere known to congregate, withsnowball sampling to recruitadditional participants. Informedconsent was obtained, and baselineand semiannual follow-up interviewswere completed at a storefrontlocation in downtown Vancouver.Participants were listed in studyrecords by name, but identities weresafeguarded by institutional reviewboard–approved study protocols.18As part of consent procedures,participants were advised verballyand in writing of the legal duty toreport abuse of persons aged ,19years in British Columbia and wereassured of their right to refuse toanswer questions on this or anyother topic. Participants wereremunerated 20 Canadian dollars atbaseline and follow-up visits. TheARYS was approved by theUniversity of British Columbia andProvidence Health Care ResearchEthics Board.MeasuresParticipants underwent aninterviewer-administeredquestionnaire capturingsociodemographic characteristics,drug use, and sexual behaviors. Thequestionnaire included the Center forEpidemiologic Studies Depression(CES-D) scale, a validated instrumentwith a standardized cutoff of .21indicating high levels of depressivesymptoms.22,23 At baseline andfollow-up visits occurringsemiannually, participants alsocompleted a nurse-administeredinterview with questions on past-6-month suicide attempts. Emergencyreferral services were available forparticipants with active suicidalideation or suspected ongoing abuseor neglect.The primary outcome of interest wasthe response to the question, “In thelast 6 months, have you attemptedsuicide?” We also ascertained lifetimesuicidal ideation with the question,“Have you ever seriously thoughtabout taking your own life?” Nurseinterviewers were trained todifferentiate intentional suicideattempts from accidental drugoverdoses. The primary exposure wasself-reported childhood maltreatmentbased on responses to the ChildhoodTrauma Questionnaire (CTQ), whichparticipants completed at baseline.The CTQ is a validated 25-itemmeasure that has been successfullyused among street-involved youth,PEDIATRICS Volume 136, number 3, September 2015 441with subscales detecting 5 types ofmaltreatment: (1) sexual abuse, (2)physical abuse, (3) emotional abuse,(4) physical neglect, and (5)emotional neglect.24–26The CTQ provides a separate scorefor each type of maltreatment basedon 5-point Likert responses rangingfrom 1 (“never true”) to 5 (“veryoften true”). All questions beganwith the prompt, “When I wasgrowing up…,” to specificallyaddress experiences from childhood.Sample questions included, “…I waspunished with a belt, a board, a cord,or some other hard object” (physicalabuse) and “…People in my familysaid hurtful or insulting things tome” (emotional abuse).26 We usedvalidated standard cutoffs for each ofthe 5 CTQ subscales: “no or minimal”maltreatment (score of 5–8), “low tomoderate” (score of 9–12),“moderate to severe” (score of13–15), and “severe to extreme”(score .15).26,27Other measured characteristicsincluded the following: age, gender,Aboriginal ancestry, sexual identity(lesbian, gay, bisexual, ortransgender), current enrollment inor previous completion of highschool, past-6-month homelessness,use of street-based outreach services(outreach worker, street nurse, healthvan, home care worker/nurse, safeinjection facility, or youth drop-incenter), daily alcohol use, injectiondrug use, overdose, sex work (tradingsex for money/drugs/shelter/clothing), and CES-D score .21.SampleAlthough the ARYS continued openenrollment beyond the study periodfor the present analysis, the sampleincluded here was restricted to the660 (66.0%) participants of 1002individuals recruited who hadreturned for $1 follow-up visit byNovember 2013. Those who returnedfor follow-up were similar to thosewho did not with regard to key studyvariables, including past-6-monthsuicide attempts, CTQ subscalescores, and all other characteristics(P . .05 for all).Statistical AnalysesWe calculated the proportion of youthreporting a suicide attempt duringthe study period and, on the basis ofthe person-time method, calculatedthe crude incidence density ofattempted suicide. We next comparedthose who reported a suicide attemptduring follow-up with those who didnot with regard to baselinecharacteristics. Analyses used thePearson x2 test and Fisher’s exact test(for cell counts ,5) for categoricalvariables and the Wilcoxon test forcontinuous variables. We thencompared those with and withouta suicide attempt according to eachtype of childhood maltreatment usingthe Cochran-Mantel-Haenszel test forordinal data.Next, using the Kaplan-Meier method,we compared the cumulativeincidence of attempted suicideaccording to CTQ scores of 5 to 12(“none to moderate”) versus scores.12 (“moderate to extreme”) foreach type of maltreatment. (Althoughthese terminologies both included“moderate” severity of abuse, scoregroupings were mutually exclusiveand score ranges were based onstandardized cutoffs of the CTQ.26,27Scores were dichotomized because ofthe high prevalence of low-moderatechildhood maltreatment.16)Participants who attempted suicidewere right-censored at the time oftheir first attempt; the remainingparticipants were right-censored atthe time of their last follow-up visit.We compared survival distributionsusing the log-rank test.Then, using Cox proportional hazardsregression, we examined associationsbetween the 5 types of maltreatmentand suicide attempts in 5 separatemodels. We also generateda combined analysis in which weentered all types of maltreatment intothe same model to identify theindependent contributions of each.All multivariable analyses adjustedfor age and gender on the basis oftheir known association with risk ofsuicide among youth.28,29 In addition,because previous studies of mortalityamong street youth have shown thatdeath from drug overdose may bedifficult to differentiate from suicide,4we also adjusted for past-6-monthinjection drug use and drug overdose.Variables were time-updated.Because the outcome of interest,suicide attempt, might have beenassociated with loss to follow-up(ie, informative censoring), weconducted separate analyses in whichwe generated models using inverseprobability of censoring weights(IPCW) for comparison.30 IPCWanalyses reweighted the sample suchthat the contribution of participantswho did not return for follow-up wasincreased, thus generating effectestimates that would have beencalculated if every participant hadprovided full follow-up. As describedelsewhere,31,32 these analysesmodeled the probability of remaininguncensored according to CTQ scoreand other study variables with theuse of pooled logistic regression. Wealso linked to provincial registries toobtain information on deaths amongparticipants lost to follow-up.Finally, to examine the role ofdepressive symptoms as anintermediate in the pathway betweenchildhood maltreatment and suicideattempts, we used the Sobel methodto determine whether CES-D scoremediated the relationship betweeneach type of maltreatment andsuicide attempts.33Analyses were performed by usingSAS version 9.3 (SAS Institute, Cary,NC). All P values were 2-sided andtests were considered significant atP , .05.RESULTSOf 660 eligible participants, 68.2%were male, 24.6% were Aboriginal,442 HADLAND et alFIGURE 1Cumulative incidence of attempting suicide according to severity of sexual abuse (A), physical abuse (B), emotional abuse (C), physical neglect (D), emotionalneglect (E), and any type of abuse/neglect (F): ARYS (Vancouver, British Columbia; 2005–2013). N = 660. Although both groupings listed (“moderate to severeabuse,” “none to moderate abuse”) include “moderate” levels of abuse, categories used standardized cutoffs of the CTQ and score ranges were mutuallyexclusive.26,27 Categories were collapsed with “none to moderate” (score of 5–12) incorporating “none to minimal” (score of 5–8) combined with “low tomoderate” (score of 9–12), and “moderate to severe” (score.12) incorporating “moderate to severe” (score of 13–15) combined with “severe to extreme” (score.15). Where totals do not add to 660 at time zero, values are missing because the participant elected not to complete the CTQ at the baseline visit.PEDIATRICS Volume 136, number 3, September 2015 443and 15.9% identified as lesbian,gay, bisexual, or transgender, with amedian age of 22 years (interquartilerange [IQR]: 20–24 years) at baseline.A total of 320 (48.5%) participantsreported having ever lived in anorphanage, a foster home, or a grouphome. Lifetime suicidal ideationwas reported by 259 (39.2%)individuals.Participants contributed 1841person-years of total follow-up(median follow-up per participant:26 months; IQR: 16–48 months;median visits: 4; IQR: 2–7).Provincial records revealed that 7deaths were observed among thoseafter their last follow-up visit; causesof death included 1 suicide and 1overdose, and the cause was unclearin 5 cases. Suicide attempts werereported by 35 (5.3%) individualsduring follow-up, resulting in a crudeincidence density of 1.9 per 100person-years (95% confidenceinterval: 1.4–2.6 per 100 person-years). Six participants reported .1suicide attempt (median: 2;maximum: 4), resulting in a total of44 attempts for the 35 participantsreporting an attempt during follow-up. As shown in Table 1, individualswho reported a suicide attempt weremore likely to have a CES-D score.21 at baseline but otherwise didnot differ significantly on othercharacteristics.Although every participant in thesample completed all or part of theCTQ, scores were missing for thesexual abuse subscale from 33 (5.0%)participants, for the physical abusesubscale from 23 (3.5%), for theemotional abuse subscale from 26(3.9%), for the physical neglectsubscale from 21 (3.2%), and for theemotional neglect subscale from 29(4.4%). In addition, the CES-D scalewas not fully completed by 45 (6.8%)participants. For both instruments,respondents did not differ fromnonrespondents in terms of suicideattempts and all other variables(P . .05 for all). CTQ respondents didnot differ from nonrespondents withregard to CES-D score (P . .05). Eachof the CTQ subscales was correlatedwith one another (P , .001 for allpairs of subscales), with correlationcoefficients ranging from 0.21 (sexualabuse and emotional neglect) to 0.49(physical abuse and emotionalabuse).As shown in Table 2, increasingseverity of childhood maltreatmentwas associated with attemptingsuicide for all types of maltreatmentexcept for sexual abuse. Thecumulative incidence of suicideattempts was significantlyassociated with all types ofmaltreatment examined separately(Fig 1 A, B, C, D, and E) and together(Fig 1F). As shown in Table 3, whichlists crude and adjusted hazardratios according to type ofmaltreatment, we observeda significantly elevated risk ofsuicide attempts among thosereporting “moderate to extreme”physical abuse, emotional abuse, andemotional neglect but not amongTABLE 1 Baseline Characteristics of 660 Street Youth Who Did and Did Not Attempt Suicide DuringStudy Follow-up: ARYS (Vancouver, British Columbia; 2005–2013)Characteristic Suicide AttemptaYes (n = 35) No (n = 625) PMedian age, y (IQR) 21.5 (19.8–23.1) 21.8 (19.8–23.6) .518Gender .747Male 23 (65.7) 427 (68.3)Female 12 (34.3) 198 (31.7)Aboriginal ancestry .812Yes 8 (22.9) 154 (24.6)No 27 (77.1) 471 (75.4)Lesbian/gay/bisexual/transgender .508Yes 7 (20.0) 98 (15.7)No 28 (80.0) 523 (83.7)High school educationb .277Yes 9 (25.7) 215 (34.4)No 26 (74.3) 405 (64.8)Homelessc .140Yes 29 (82.9) 443 (70.9)No 6 (17.1) 178 (28.5)Use of outreach servicesc,d .252Yes 31 (88.6) 505 (80.8)No 4 (11.4) 120 (19.2)Daily alcohol usec .639Yes 7 (20.0) 105 (16.8)No 28 (80.0) 515 (82.4)Injection drug usec .445Yes 8 (22.9) 180 (28.8)No 27 (77.1) 444 (71.0)Drug overdosec .105Yes 7 (20.0) 68 (10.9)No 28 (80.0) 552 (88.3)Sex workc .554Yes 4 (11.4) 57 (9.1)No 31 (88.6) 568 (90.9)CES-De score .21 .037Yes 19 (54.3) 213 (34.1)No 16 (45.7) 367 (58.7)Data are presented as n (%) unless otherwise indicated; N = 660. Characteristics were reported at the time of studyenrollment.a Cells do not uniformly add up to column totals due to missing values.b Previous completion of or current enrollment in high school.c During the preceding 6 months.d Street-based outreach services include outreach worker, street nurse, health van, home care worker/nurse, safeinjection facility, or youth drop-in center.e CES-D scale using a standardized cutoff of .21.444 HADLAND et althose reporting “moderate toextreme” sexual abuse or physicalneglect. Effect sizes werecomparable in IPCW models, whichreweighted the sample to accountfor informative censoring (ie, toaccount for the possibility that thoselost to follow-up were at greatestrisk of suicide). In the combinedmodel, which included all types ofmaltreatment in the same model,only physical abuse retaineda statistically significant hazardratio.CES-D score did not appear tomediate the relationship betweenstatistically significant forms ofmaltreatment and suicide attempts.Sobel test statistics for mediationby CES-D scores .21 were asfollows: physical abuse, 1.75(P = .080); emotional abuse, 1.77(P = .077); and emotional neglect,1.55 (P = .122).DISCUSSIONIn this prospective cohort study instreet youth, we observed a highincidence of self-reported attemptedsuicide, with .1 in 20 youthreporting an attempt over the 7-yearstudy period. History of childhoodmaltreatment was strongly associatedwith risk of attempted suicide, withyouth who reported previous physicalabuse, emotional abuse, or emotionalneglect 3 to nearly 5 times as likely toreport attempting suicide. Among the5 types of self-reported childhoodmaltreatment examined ina combined model, only physicalabuse retained an independent effecton risk of attempted suicide.To date, most studies on suicidalityand childhood maltreatment havefocused on adult drug users in clinic-based samples and have useda cross-sectional design.13,14,34 Ourstudy is novel in examininga longitudinal cohort of community-recruited street youth. Although weobserved a high frequency ofattempted suicide in our sample,studies reveal that attempted suicideis alarmingly common throughoutthe general adolescent population.The 2013 Youth Risk BehaviorSurveillance Study of US high schoolstudents estimated the past-yearprevalence of attempted suicide tobe 8% (girls: 11%; boys: 5%).35 InBritish Columbia, the estimated past-year prevalence of attempted suicidein the general adolescent populationin 2013 was 6% (girls: 9%; boys:3%). Despite the similar prevalenceof attempted suicide among streetyouth and the general adolescentpopulation, the most commonmethods used among street youthare especially deadly and includewrist slitting, intentional drugoverdose, and hanging.36Lifetime suicidal ideation wasreported by 39% of participants inour study, a prevalence .3 times thatreported by the National ComorbiditySurvey of adolescents in the UnitedStates.37 This elevated prevalence inpart reflects the higher median age inour study sample, because suicidalitygenerally increases as adolescenceprogresses to early adulthood.1,2,28 Inany case, estimates of lifetime suicidalideation among adults in the generalpopulation are one-third that found inthe street youth in our study, andfactors such as childhoodmaltreatment are likely criticalcontributors.38,39Our results extend those from cross-sectional studies in clinic-based adultdrug users.13,14,34 Our results alsobuild on findings from a recentlongitudinal study of community-recruited adult drug users inVancouver showing elevated risk ofattempted suicide after childhoodmaltreatment.16 What is less clearamong these studies is which types ofchildhood maltreatment confer thegreatest risk. In studies examiningmultiple types of maltreatment,TABLE 2 Associations of Childhood Maltreatment With Suicide Attempts During Follow-up: ARYS(Vancouver, British Columbia; 2005–2013)Type of Maltreatment (CTQ Subscale Score)a Totalb (N = 660), n (%) Suicide Attempt, n (%)Yes (n = 35) No (n = 625) PcSexual abuse .062None to minimal (5–8) 476 (72.1) 22 (62.9) 454 (72.6)Low to moderate (9–12) 40 (6.1) 1 (2.9) 39 (6.2)Moderate to severe (13–15) 31 (4.7) 2 (5.7) 29 (4.6)Severe to extreme (.15) 80 (12.1) 8 (22.9) 72 (11.5)Physical abuse ,.001None to minimal (5–8) 345 (52.3) 9 (25.7) 336 (53.8)Low to moderate (9–12) 115 (17.4) 4 (11.4) 111 (17.8)Moderate to severe (13–15) 56 (8.5) 8 (22.9) 48 (7.7)Severe to extreme (.15) 121 (18.3) 13 (37.1) 108 (17.3)Emotional abuse ,.001None to minimal (5–8) 208 (31.5) 3 (8.6) 205 (32.8)Low to moderate (9–12) 128 (19.4) 4 (11.4) 124 (19.8)Moderate to severe (13–15) 98 (14.9) 8 (22.9) 90 (14.4)Severe to extreme (.15) 200 (30.3) 17 (48.6) 183 (29.3)Physical neglect .013None to minimal (5–8) 300 (45.5) 10 (28.6) 290 (46.4)Low to moderate (9–12) 171 (25.9) 9 (25.7) 162 (25.9)Moderate to severe (13–15) 87 (13.2) 6 (17.1) 81 (13.0)Severe to extreme (.15) 81 (12.3) 8 (22.9) 73 (11.7)Emotional neglect ,.001None to minimal (5–8) 179 (27.1) 1 (2.9) 178 (28.5)Low to moderate (9–12) 171 (25.9) 6 (17.1) 165 (26.4)Moderate to severe (13–15) 104 (15.8) 11 (31.4) 93 (14.9)Severe to extreme (.15) 177 (26.8) 14 (40.0) 163 (26.1)N = 660.a Categories used standardized cutoffs of the CTQ.26,27b Cells do not uniformly add up to column total due to missing values.c P values were obtained from Cochran-Mantel-Haenszel test for ordinal data.PEDIATRICS Volume 136, number 3, September 2015 445results often suggest that sexualabuse shows the strongestassociation with suicidality.15,34 Inour sample, sexual abuse wasassociated with increased cumulativeincidence of attempted suicide butwas not significant in Cox regressionmodels. Sexual abuse was the leastcommon form of childhoodmaltreatment reported in our study,although we cannot exclude thatparticipants may have feltuncomfortable disclosing suchexperiences.Nonetheless, our study builds onprevious studies by including lessstudied forms such as emotionalabuse and physical and emotionalneglect. Although less researched,emotional abuse has shown a strongassociation with suicidality instudies in which it has beenexamined.16,34 In our study,emotional abuse was reported tohave co-occurred in nearly half of allparticipants also reporting physicalabuse, and emotional neglect wassimilarly comorbid. Because physicalabuse was the only type ofmaltreatment to maintain anindependent effect in a combinedmodel, some of the harm ofemotional abuse and neglect may bedue to physical abuse that youthexperience simultaneously.The underlying mechanisms linkingchildhood maltreatment to suiciderequire further study. Although ourdata did not support a mediating rolefor depressive symptoms, childhoodmaltreatment and suicidality are bothlinked to depression, bipolar disorder,anxiety disorders, and posttraumaticstress disorder.37,40 Such psychiatricconditions, many of which areprevalent among street youth, mayresult from childhood maltreatmentand subsequently predispose tosuicidality.3,6,41 Prolonged streetinvolvement is also likely detrimentalto the mental health of youth.42Homelessness is an independentpredictor of substance use.43 Bothhomelessness and substance use areharmful yet common aspects of streetlife,5,44,45 and both are associatedwith suicidality.46,47 Qualitativeresearch shows that homelessness,substance use, and adverse childhoodexperiences converge on feelings ofhopelessness and of being “trapped,”which contribute to suicide amongstreet youth.48Regardless of mechanism, previouswork suggests that to delivereffective, trauma-informed suicidetreatment and prevention to streetyouth, providers should understandthat the impact of childhoodmaltreatment is far-reaching andinfluences child and adolescentdevelopment and copingstrategies.49,50 Street youthproviders should ensure a physicallyand emotionally safe treatmentenvironment to avoidretraumatization and, in deliveringservices, decrease hierarchicalapproaches and empower youth tocollaborate in their own care.49,51Limitations of our study include thereliance on self-report, which mayhave been subject to recall bias.Participants may have beenreluctant to disclose traumaticexperiences, suicide attempts, orother risk behaviors. Moreover,distinguishing suicide from drugoverdose may be difficult,particularly because overdose isa common method of suicide.4,8,13To account for this possibility, wetrained nurse interviewers todistinguish between overdose andsuicide and adjusted for co-occurring injection drug use andrecent overdose in our analyses. Inaddition, we studied childhoodmaltreatment and did not ascertainongoing, street-based abuse, whichmay have contributed to the risk ofattempted suicide.52 Conversely,we cannot rule out that ongoingcontact with research staff andreferrals to mental health servicesfor high-risk participants reducedsuicidality.CONCLUSIONSOur findings reveal an elevated risk ofsuicide attempts among street youthwith a history of childhoodmaltreatment compared with thoseTABLE 3 Crude and Adjusted Hazard Ratios for Attempting Suicide: ARYS (Vancouver, British Columbia; 2005–2013)Type of Maltreatmenta Crude Hazard Ratio (95% CI) Adjusted Hazard Ratiob (95% CI)Standard Model IPCW Model Combined ModelSexual abuse 1.71 (0.80–3.64) 1.63 (0.70–3.78) 1.43 (0.58–3.55) 1.00 (0.43–2.34)Physical abuse 4.62 (2.20–9.71) 4.47 (2.12–9.42) 5.09 (2.27–11.4) 2.74 (1.03–7.28)Emotional abuse 4.66 (1.93–11.2) 4.92 (2.11–11.5) 4.45 (1.82–10.9) 2.52 (0.88–7.25)Physical neglect 1.84 (0.89–3.82) 1.70 (0.80–3.59) 1.88 (0.84–4.19) 0.76 (0.36–1.60)Emotional neglect 3.36 (1.28–8.81) 3.08 (1.05–9.03) 2.95 (1.02–8.54) 1.42 (0.40–5.06)N = 660.a Categories used standardized cutoffs of the CTQ and were collapsed with “none to moderate” (score of 5–12; reference group) incorporating “none to minimal” (score of 5–8) combinedwith “low to moderate” (score of 9–12), and “moderate to severe” (score .12; comparison) incorporating “moderate to severe” (score of 13–15) combined with “severe to extreme”(score .15).26,27b Standard and IPCW models include the type of abuse/neglect listed (and no other), adjusted for age, gender, past-6-month injection drug use, and past-6-month drug overdose. IPCWmodels reweighted the sample to inflate the contribution of participants lost to follow-up, thus accounting for the possibility that those who did not return for full follow-up may havebeen at elevated risk of suicide. The combined model included all 5 types of abuse/neglect in same model without IPCW (and additionally adjusted for age, gender, past-6-month injectiondrug use, and past-6-month drug overdose).446 HADLAND et alwithout this history. These resultsbuild on preexisting research byspecifically examining the harms ofother, less studied forms ofmaltreatment, including emotionalabuse and physical and emotionalneglect. Additional studies are neededto show the mechanistic pathways andto identify protective factors.Nonetheless, the high prevalence ofself-reported childhood maltreatmentamong street youth in this settingdemonstrates the importance ofdelivering suicide prevention servicesfrom a trauma-informed perspective.ACKNOWLEDGMENTSWe thank the study participants fortheir contribution to the research, aswell as current and past researchersand staff. We specifically thankMs Sabina Dobrer, Mr Cody Callon,Ms Jennifer Matthews, Ms DeborahGraham, Mr Peter Vann, Mr SteveKain, Ms Kristie Starr, Ms TriciaCollingham, and Ms Carmen Rockfor their research andadministrative assistance. We alsoappreciate the support of Drs S. JeanEmans, Elizabeth R. Woods, andSarah A.B. Pitts and the Division ofAdolescent/Young Adult Medicineat Boston Children’s Hospital.ABBREVIATIONSARYS: At Risk Youth StudyCES-D: Center for EpidemiologicStudies DepressionCTQ: Childhood TraumaQuestionnaireIPCW: inverse probability ofcensoring weightsIQR: interquartile rangeAddress correspondence to Kora DeBeck, PhD, British Columbia Centre for Excellence in HIV/AIDS, St Paul’s Hospital, 608-1081 Burrard St, Vancouver, BC, Canada V6Z1Y6. E-mail: uhri-kd@cfenet.ubc.caPEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).Copyright © 2015 by the American Academy of PediatricsFINANCIAL DISCLOSURE: As an HIV physician, Dr Montaner has received limited unrestricted funding, paid to his institution, from Abbvie, Bristol-Myers Squibb,Gilead Sciences, Janssen, Merck, and ViiV Healthcare; the other authors have indicated they have no financial relationships relevant to this article to disclose.FUNDING: Supported by National Institutes of Health (NIH)/National Institute on Drug Abuse (NIDA) grants R01DA028532 and U01DA038886 as well as the CanadianInstitutes of Health Research (CIHR) operating grant MOP-102742. In addition, Dr Hadland is supported by the Maternal and Child Health/Health Resources andServices Administration Leadership Education in Adolescent Health Training Program (T71 MC00009) and by a National Research Service Award from NIH/NationalInstitute of Child Health and Human Development (1T32 HD075727). Dr Wood is supported by the Canada Research Chairs program through a Tier 1 CanadaResearch Chair in Inner City Medicine. Dr Marshall is supported by NIH/NIDA grant R03 DA037770. Dr Montaner is supported by the British Columbia Ministry ofHealth and by NIH/NIDA grant R01DA036307. Dr DeBeck is supported by a Michael Smith Foundation for Health Research/St Paul’s Hospital/Providence Health CareCareer Scholar Award and a CIHR New Investigator Award. Funded by the National Institutes of Health (NIH).POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.REFERENCES1. Johnson NB, Hayes LD, Brown K, Hoo EC,Ethier KA; Centers for Disease Controland Prevention. CDC National HealthReport: leading causes of morbidity andmortality and associated behavioral riskand protective factors—United States,2005-2013. MMWR Surveill Summ. 2014;63(suppl 4):3–272. Statistics Canada. The 10 leading causesof death, 2011. Available at: www.statcan.gc.ca/pub/82-625-x/2014001/article/11896-eng.htm. Accessed December 3, 20143. Public Health Agency of Canada. Streetyouth in Canada: findings from theenhanced surveillance of Canadianstreet youth, 1999-2003. 2006. Availableat: www.phac-aspc.gc.ca/std-mts/reports_06/pdf/street_youth_e.pdf.Accessed December 3, 20144. Roy E, Haley N, Leclerc P, Sochanski B,Boudreau JF, Boivin JF. Mortality ina cohort of street youth in Montreal.JAMA. 2004;292(5):569–5745. Greene JM, Ennett ST, Ringwalt CL.Substance use among runaway andhomeless youth in three nationalsamples. Am J Public Health. 1997;87(2):229–2356. Boivin JF, Roy E, Haley N, Galbaud du FortG. The health of street youth: a Canadianperspective. Can J Public Health. 2005;96(6):432–4377. Marshall BDL, Galea S, Wood E, Kerr T.Injection methamphetamine use isassociated with an increased risk ofattempted suicide: a prospective cohortstudy. Drug Alcohol Depend. 2011;119(1–2):134–1378. Bohnert AS, Roeder K, Ilgen MA.Unintentional overdose and suicideamong substance users: a review ofoverlap and risk factors. Drug AlcoholDepend. 2010;110(3):183–1929. Wilcox HC, Conner KR, Caine ED.Association of alcohol and drug usedisorders and completed suicide: anempirical review of cohort studies. DrugAlcohol Depend. 2004;76(suppl):S11–S1910. Rohde P, Noell J, Ochs L, Seeley JR.Depression, suicidal ideation and STD-related risk in homeless olderadolescents. J Adolesc. 2001;24(4):447–46011. Ompad DC, Ikeda RM, Shah N, et al.Childhood sexual abuse and age atinitiation of injection drug use. Am JPublic Health. 2005;95(4):703–70912. Dube SR, Felitti VJ, Dong M, Chapman DP,Giles WH, Anda RF. Childhood abuse,neglect, and household dysfunction andthe risk of illicit drug use: the adversePEDIATRICS Volume 136, number 3, September 2015 447childhood experiences study. Pediatrics.2003;111(3):564–57213. Bohnert ASB, Roeder KM, Ilgen MA.Suicide attempts and overdoses amongadults entering addictions treatment:comparing correlates in a U.S. NationalStudy. Drug Alcohol Depend. 2011;119(1–2):106–11214. Roy A. Risk factors for attempting suicidein heroin addicts. Suicide Life ThreatBehav. 2010;40(4):416–42015. Brown J, Cohen P, Johnson JG, SmailesEM. Childhood abuse and neglect:specificity of effects on adolescent andyoung adult depression and suicidality. JAm Acad Child Adolesc Psychiatry. 1999;38(12):1490–149616. Marshall BDL, Galea S, Wood E, Kerr T.Longitudinal associations between typesof childhood trauma and suicidalbehavior among substance users:a cohort study. Am J Public Health. 2013;103(9):e69–e7517. Feldmann J, Middleman AB. Homelessadolescents: common clinical concerns.Semin Pediatr Infect Dis. 2003;14(1):6–1118. Wood E, Stoltz JA, Montaner JSG, Kerr T.Evaluating methamphetamine use andrisks of injection initiation among streetyouth: the ARYS study. Harm Reduct J.2006;3:1819. Fischer B, Rehm J, Hall W. Cannabis usein Canada: the need for a ‘public health’approach. Can J Public Health. 2009;100(2):101–10320. DeMatteo D, Major C, Block B, et al.Toronto street youth and HIV/AIDS:prevalence, demographics, and risks.J Adolesc Health. 1999;25(5):358–36621. Kerr T, Marshall BD, Miller C, et al.Injection drug use among street-involvedyouth in a Canadian setting. BMC PublicHealth. 2009;9(171):17122. Radloff LS. The use of the Center forEpidemiologic Studies Depression scalein adolescents and young adults. J YouthAdolesc. 1991;20(2):149–16623. Radloff LS. The CES-D scale: a self reportdepression scale for research in thegeneral population. Appl Psychol Meas.1977;1(3):385–40124. Scher CD, Stein MB, Asmundson GJ,McCreary DR, Forde DR. The ChildhoodTrauma Questionnaire in a communitysample: psychometric properties andnormative data. J Trauma Stress. 2001;14(4):843–85725. Forde DR, Baron SW, Scher CD, Stein MB.Factor structure and reliability of thechildhood trauma questionnaire andprevalence estimates of trauma for maleand female street youth. J InterpersViolence. 2012;27(2):364–37926. Bernstein DP, Fink L. Childhood TraumaQuestionnaire: A Retrospective Self-report Manual. San Antonio, TX:Psychological Corporation; 199827. Bernstein DP, Stein JA, Newcomb MD,et al. Development and validation ofa brief screening version of theChildhood Trauma Questionnaire. ChildAbuse Negl. 2003;27(2):169–19028. Hallfors DD, Waller MW, Ford CA, HalpernCT, Brodish PH, Iritani B. Adolescentdepression and suicide risk: associationwith sex and drug behavior. Am J PrevMed. 2004;27(3):224–23129. Skinner R, McFaull S. Suicide amongchildren and adolescents in Canada:trends and sex differences, 1980-2008.CMAJ. 2012;184(9):1029–103430. Robins JM. Information recovery andbias adjustment in proportional hazardsregression analysis of randomized trialsusing surrogate markers. Proceedingsof the Biopharmaceutical Section,American Statistical Association. 1993:22–3331. Cain LE, Cole SR. Inverse probability-of-censoring weights for the correction oftime-varying noncompliance in the effectof randomized highly activeantiretroviral therapy on incident AIDSor death. Stat Med. 2009;28(12):1725–173832. Robins JM, Finkelstein DM. Correctingfor noncompliance and dependentcensoring in an AIDS clinical trial withinverse probability of censoringweighted (IPCW) log-rank tests.Biometrics. 2000;56(3):779–78833. Sobel ME. Asymptotic intervals forindirect effects in structural equationmodels. In: Leinhardt S, ed. SociologicalMethodology. San Francisco, CA: Jossey-Bass Publishers; 1982:290–31234. Lloyd JJ, Ricketts EP, Havens JR, et al. Therelationship between lifetime abuse andsuicidal ideation in a sample of injectiondrug users. J Psychoactive Drugs. 2007;39(2):159–16635. Kann L, Kinchen S, Shanklin SL, et al;Centers for Disease Control andPrevention. Youth risk behaviorsurveillance—United States, 2013. MMWRSurveill Summ. 2014;63(suppl 4):1–16836. Yoder KA. Comparing suicide attempters,suicide ideators, and nonsuicidalhomeless and runaway adolescents.Suicide Life Threat Behav. 1999;29(1):25–3637. Nock MK, Green JG, Hwang I, et al.Prevalence, correlates, and treatment oflifetime suicidal behavior amongadolescents: results from the NationalComorbidity Survey ReplicationAdolescent Supplement. JAMAPsychiatry. 2013;70(3):300–31038. Kessler RC, Berglund P, Borges G, NockM, Wang PS. Trends in suicide ideation,plans, gestures, and attempts in theUnited States, 1990-1992 to 2001-2003.JAMA. 2005;293(20):2487–249539. Kessler RC, Borges G, Walters EE.Prevalence of and risk factors forlifetime suicide attempts in the NationalComorbidity Survey. Arch Gen Psychiatry.1999;56(7):617–62640. Green JG, McLaughlin KA, Berglund PA,et al. Childhood adversities and adultpsychiatric disorders in the nationalcomorbidity survey replication I:associations with first onset of DSM-IVdisorders. Arch Gen Psychiatry. 2010;67(2):113–12341. Kelly K, Caputo T. Health and street/homeless youth. J Health Psychol. 2007;12(5):726–73642. Karabanow J. Getting off the street:exploring the processes of youngpeople’s street exits. Am Behav Sci. 2008;51(6):772–78843. Feng C, DeBeck K, Kerr T, Mathias S,Montaner J, Wood E. Homelessnessindependently predicts injection druguse initiation among street-involvedyouth in a Canadian setting. J AdolescHealth. 2013;52(4):499–50144. Fazel S, Khosla V, Doll H, Geddes J. Theprevalence of mental disorders amongthe homeless in Western countries:systematic review and meta-regressionanalysis. PLoS Med. 2008;5(12):e22545. Hwang SW. Homelessness and health.CMAJ. 2001;164(2):229–23346. Eynan R, Langley J, Tolomiczenko G, et al.The association between homelessness448 HADLAND et aland suicidal ideation and behaviors:results of a cross-sectional survey.Suicide Life Threat Behav. 2002;32(4):418–42747. Greene JM, Ringwalt CL. Youth andfamilial substance use’s association withsuicide attempts among runaway andhomeless youth. Subst Use Misuse. 1996;31(8):1041–105848. Kidd SA. “The walls were closing in, andwe were trapped”: a qualitative analysisof street youth suicide. Youth Soc. 2004;36(1):30–5549. McKenzie-Mohr S, Coates J, McLeod H.Responding to the needs of youth whoare homeless: calling for politicizedtrauma-informed intervention. ChildYouth Serv Rev. 2012;34(1):136–14350. Prescott L, Soares P, Konnath K, BassukE. A Long Journey Home: A Guide forCreating Trauma-Informed Services forMothers and Children ExperiencingHomelessness. Rockville, MD: SubstanceAbuse and Mental Health ServicesAdministration 2008. Available at: http://homeless.samhsa.gov/ResourceFiles/ALongJourneyHome.pdf. Accessed May28, 201551. Elliott DE, Bjelajac P, Fallot RD, MarkoffLS, Reed BG. Trauma‐informed ortrauma‐denied: principles andimplementation of trauma‐informedservices for women. J CommunityPsychol. 2005;33(4):461–47752. Goodman L, Saxe L, Harvey M.Homelessness as psychological trauma:broadening perspectives. Am Psychol.1991;46(11):1219–1225AWONDERFUL 800th BIRTHDAY: I was recently vacationing in England with mydaughter. I took her to some ofmy favoritemuseums, including theBritishMuseum,theTateModern, theVictoriaandAlbert,andtheBritishLibrary.TheBritishLibraryhas a fantastic collection, including Shakespeare folios, Da Vinci notebooks,a Gutenberg bible, and an ancient copy of Beowulf. However, I really wanted her tosee the Magna Carta.As reported inTheNewYorkTimes (Europe: June14, 2015), theMagnaCarta is 800years old this year. Revered by many, the Magna Carta is often considered thefoundation of English and American law. Interestingly, views of the importance oftheMagnaCarta in2015areradicallydifferent fromthatof1215orotheryears,andquite different inEnglandand theUnited States. TheMagnaCartawasanegotiatedpeace treatybetweenrebellious baronsangeredbyKing John’s tax levies, blackmail,and legal manipulations. Written in Latin (the language of the Church but not thecourt or the people), the document mostly deals with husbandry, land issues, andfeudal financial arrangements including inheritance. The agreement was betweentheeliteofmedieval societyandtheKing,butdidnotspecifically includecommoners.While King John did affix his seal in agreement at Runnymede, he immediatelyappealed to the Pope to annul the agreement, which the Pope did twomonths later.Edited and reissued several times over the next 100 years, the versions we mostassociate with the Magna Carta were most likely written in 1225 and 1297. In-terestingly, almost no part of the charter was actually honored in the years after1215.Only fourof theoriginal64provisionsarestill partofEnglish law.Nonetheless,a coreprinciplewas recorded– that theKingcouldnot imprisonpeoplewheneverhewanted for whatever reason he wanted. While it took centuries for due process tobecome part of enforceable English law, the concepts written in 1215 are stillapplicable today. Earlier this year, a provision in the charter was cited in a USSupreme Court decision on judicial integrity.While I amnot suremydaughterwas particularly impressed by the document itself,she and I could both appreciate its historical significance.Noted by WVR, MDPEDIATRICS Volume 136, number 3, September 2015 449


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items