Protective and risk factors of psychosocial wellbeing related to the reintegration of former child soldiers in Nepal

Copyrig

Adhikari et al.

Protective and risk factors ofpsychosocial wellbeing related to thereintegration of former child soldiersin NepalRamesh Prasad Adhikari, Brandon A. Kohrt, Nagendra Prasad Luitel,Nawaraj Upadhaya, Dristy Gurung &MarkJ.D. Jordans

This paper explores protective and risk factors

for mental health and psychosocial wellbeing

among 300 child solders (veri¢ed minors) through

a longitudinal study. Both the Hopkins Symptoms

Check list and the Posttraumatic Stress Disorder

Checklist (civilian version) were used to measure

mental health problems, while the Generalised

Estimating Equation was used to identify both

the protective and risk factors over time. Anxiety

and posttraumatic stress disorder decreased over a

nine month period, while depression prevalence

did not change. Social support, inter-caste marriage,

low caste and residence in far western geographic

regions were all associated with greater mental

health problems. Rehabilitation packages were not

associated with improved mental health, and former

child solders enrolled in vocational programmes

had greater posttraumatic stress disorder symptom

severity. The ¢ndings suggest that strong social

support is needed, as rehabilitation packages alone

may be insu⁄cient to improve mental health.

Keywords: child soldier, mental health,reintegration, risk and protective factors

IntroductionFormer child soldiers commonly su¡er frommental health problems as a result ofexposure to traumatic events during con£ictperiods (Derluyn et al., 2004). However,previous research (Kohrt et al., 2008) hasdemonstrated that the increased risk formental health problems is not only dueto traumatic events but that di⁄cultiesof reintegration also contribute. A study

ht © War Trauma Foundation. Unautho

conducted in Uganda showed that formerchild soldiers, in comparison to their nonrecruited counterparts, faced signicantlymore ‘‘emotional’’ and fewer ‘‘social and rela-

tional ’’challenges (Vindevogel et al., 2013).Post con£ict factors, such as educationaland economic deprivation, stigmatisationand discrimination have all been associa-ted with adverse outcomes (Blattman &Annan, 2010). Additionally, studies suggesteconomic and educational programmessupport mental health and psychosocialfunctioning. For example, returning to,and staying in, school was associated withhigher levels of con¢dence and pro socialbehaviours among youth in a SierraLeone study (Betancourt et al., 2010b).Post con£ict, reintegration of child soldiersis a priority because failed reintegrationis a threat to economic recovery, socialintegration, and peace (Annan et al., 2011).Successful reintegration is dependent onthe protective and risk factors presentin the children’s immediate environment,and within the broader national context.As child solders face di¡erent reintegra-tion problems than adults, this paperfocuses only on former child soldiers, andpresents the protective and risk factors forpsychosocial wellbeing during the reinte-gration process of former child soldiers.

Protective factorsPotential protective factors for successfulreintegration include: community sensitisa-tion, cleansing rituals, transitional periods

rized reproduction of this article is prohibited.367

Copyrig

Protective and risk factors of psychosocial wellbeing related to the reintegration of former child soldiers in Nepal,

Intervention 2014, Volume 12, Number 3, Page 367 - 378

in interim care centres, religious support,psychosocial counselling, family mediationand skills and vocational training (Kryger& Lindgren, 2011; Williamson, 2006).In Sierra Leone, higher levels of familyacceptance were linked to lower averagelevels of emotional distress among parti-cipants (Betancourt et al., 2010a). Suppor-tive parenting was associated with betterpsychosocial adjustment in one study ofUgandan child soldiers (Derluyn et al.,2004). Higher levels of social support wereassociated with increased adaptive andpro social behaviours and attitudes (Betan-court et al., 2010b).Former child soldiers need time and supportto move away from their past role and takea new role, identifying themselves as amember of a local community (Watson,2009). Generally, it is assumed that theinitial external support, in the formof huma-nitarian assistance, helps to improve therelationship with family and community(Kingma, 2000). Subsequent, long termassistance from di¡erent stakeholders suchas government, nongovernmental organis-ations (NGOs) and international NGOs(INGOs) can also contribute to successfulreintegration. Attending school and train-ing programmes are considered critical inassisting war a¡ected youth attain a senseof normalcy and safety in their everydaylives while also increasing future employ-ment opportunities (Betancourt & Khan,2008).

Background and context: veri¢edminors and late recruitsNepal su¡ered violent con£ict from 1996 to2006 that claimed the lives of more than13,000 people, while several thousands weresubjected to torture, intimidation, extortion,and abduction (Russell, 2012). The compre-hensive peace agreement, signed betweenthe government of Nepal and the Commu-nist Party of Nepal (Maoist) in November2006, paved the way for the United Nations’involvement in Nepal’s peace process. The

ht © War Trauma Foundation. Unautho368

veri¢cation of 32,250 Maoists combatantsregistered for the veri¢cation process wasconducted by the UN team who collectedpersonal military information and identitycards. Only 23, 610 combatants attendedthe individual interview session with theUN team. After assessing these combatantson criteria such as date of enrolment anddate of birth only 19,602 were veri¢ed asMaoist Army and the rest (4,008) wereveri¢ed as disquali¢ed (UNDP, 2011).Among those disquali¢ed,2,973wereminors(those born after 25 May 1988) and 1,035were ‘‘late recruits’’ (those who joined theMaoist army after the cease¢re of 25 May2006). As a result of this process, the term‘‘veri¢ed minors and late recruits’’ (VMLR) wascoined to refer to this speci¢c group ofcombatants.In principle, VMLR combatants shouldhave been discharged from the canton-ment (camps that process former warringfactions) immediately after the veri¢cationprocess, but this did not happen. Finally,after three years, the agreement wassigned, on 16 December 2009. At thebeginning of the formal discharge processin 2010, there were only 3,000 VMLRremaining in the cantonment. Over onethousand (1,008) VMLR had already leftas a result of the long duration betweenveri¢cation and release (Colekessian, 2009).The government of Nepal, in collabora-tion with UN agencies and NGOs, provi-ded a rehabilitation package for VMLRto support their transition back to civilianlife. The rehabilitation package consistedof four options: 1) formal educationalsupport which consisted of formal schoolenrolment and informal bridging courses;2) ¢nancing a micro enterprise, includingtraining in business oriented, short coursesand post training support to start up thebusiness; 3) vocational skill training toproduce skilled manpower, such aswelders, electricians, carpenters, etc.; and4) training and formal education in thehealth sector for work in paramedics, such

rized reproduction of this article is prohibited.

Copyrig

Adhikari et al.

as health assistant, nurse and lab technicians(UNDP, 2011).In addition, psychosocial support waso¡ered to the VMLR group in each careercentre, through regional psychosocialcounsellors. The regional counsellors alsoprovided information on di¡erent rehabili-tation packages and helped the combatantsto make informed decision on the choiceso¡ered to them. As of July 2012, about 2,231out of 4,008 combatants enrolled in therehabilitation package and 73% of themcompleted the package (UNIRP, 2012).The remaining 1,764 combatants did notenrol in the rehabilitation package due todissatisfaction with the content of thepackage and/or stigma associated with thelabel of being disquali¢ed (Colekessian,2009).

MethodsThe baseline study was conducted April^May 2011, whichcovered 24 outof 75 districtsof Nepal. This number of districts increasedconsiderably in the follow up study con-ducted during January^March 2012, when34 districtswere represented, due tomobilityof theVMLR group.

Study participants and samplingAs there were no up-to-date contact detailsavailable for the discharged VMLR group,a purposive sampling method was used toapproach VMLR who had either partici-pated in one of the rehabilitation packages,or were available in the district. At the base-line, 400 were interviewed, nearly 10% ofthe total 4,008 VMLR. At the follow-up310, or 77.5% of the initial interview group,were included.Out of 400 VMLR who participated inbaseline, 316 were former child soldiers. Ofthe baseline participants, 250 (79%) wereincluded in the follow-up study as someparticipants refused to participate further(N¼ 6), some were out of contact due tomigration (N¼ 59) and one was in prison.

ht © War Trauma Foundation. Unautho

InstrumentsPsychosocial and mental health problemswere assessed using standardised structuredquestionnaires, i.e. the Hopkins Symptoms

Checklist (HSCL-25) and PostTraumatic Stress

Disorder (PTSD) Checklist-civilian version

(PCL-C), which have been validated inNepal ‘‘against locally corresponding syndrome

and Composite International Diagnostic Inter-

views’’ (Thapa&Hau¡,2005).The validatedcut o¡ score points, in order to determinethe increased level of depression, anxiety,and PTSD, were score � 17.5 for anxiety(sensitivity¼ 0.77 and speci¢city¼ 0.58),score� 24.5 for depression (sensitivity¼ 0.870.87 and speci¢city¼ 0.60) in HSCL-25;and score � 50 for PTSD (sensitivity¼ 0.80and speci¢city¼ 0.80) in PCL-C.Additionally, a locally developed 11 itemquestionnaire was used for socio-economicreintegration. Of these, eight items wererelated to social integration, such as: com-munity acceptance; participation in socialactivities; trust; ignorance; support; careand support by family members, etc. Threeitems were related to economic reinte-gration, such as: standard of living com-pared to friends; perceptions of future andopportunities; and con¢dence of poten-tial economic improvement. Personal andhousehold information of each participantwere recorded through a questionnairethat included age, sex, education, region,caste/ethnicity, marital and family statusand source of income.

ProcedureThree teams, each consisting of one maleand one female researcher, were employedfor the baseline and follow-up data collec-tion. The ¢eld researchers received twoweeks of extensive training on study design,interviewing skills and research ethics.The eligible study participants were ident-i¢ed through the regional psychosocialcounsellors and representatives of part-ner organisations of UN Interagency Reha-bilitation Programme (UNIRP). Written

rized reproduction of this article is prohibited.369

Copyrig

Table 1. Background information(baseline)

Background characteristics N %

GenderMale 202 63.9Female 114 36.1

AgeLess than 20 61 19.320^22 215 68.023 24 40 12.7

Caste/ ethnicityBrahmin/Chhetri 129 40.8Janajati 142 44.9Dalit 45 14.2

Marital StatusUnmarried 161 50.9Married 155 49.1

Level of EducationBelow primary (illiterate/

informally literate)8 2.5

Primary 45 14.2Secondary 167 52.8School leaving

certi¢cate52 16.5

Intermediate level 44 13.9School enrolment

Yes 119 37.7No 197 62.3

Rehabilitation supportNone 85 26.9Education 68 21.5Training and formal

education in healthsector

36 11.4

Vocational training 45 14.2Micro enterprises 82 25.9

Development regionFar western region 84 26.6Midwestern region 56) 17.7Western region 20 6.3Central region 88 27.8Eastern region 68 21.5N 316 100.0

Protective and risk factors of psychosocial wellbeing related to the reintegration of former child soldiers in Nepal,

Intervention 2014, Volume 12, Number 3, Page 367 - 378

consent for the study was obtainedfrom all research participants. In the caseof children below age 18, consent wasobtained from both parents and children.Provision for the referral of participantsfor psychosocial support was set up incase of need. Strict adherence to con¢den-tiality was maintained.

Data analysisThe data was entered into the Statistical

Package for Social Science (SPSS) 16.0 version.A multivariate analysis with Generalized

Estimating Equation (GEE) was conductedto identify the supportive and risk factorsof psychosocial wellbeing over time for thoseparticipating in baseline and follow-upstudy.Through the GEE model, beta-coe⁄-cient, con¢dence interval and p value forindependent variables, such as age, gender,marital status, inter-caste marriage, geo-graphical location, caste/ethnicity, reha-bilitation package, level of education,schooling status, social support andeconomic opportunities were calculated.Possible ‘‘interaction e¡ects’’ between certainindependent variables were also included inthe model. Social and economic supportwas measured using the composite score,which included eight di¡erent questionsrelated to social support and three ques-tions related to economic status based ontheir distribution, correlation and the eigenvalues to reduce variance in the correlationmatrix. Because composite scores were notnormally distributed, standardised Z-scoreswere generated to use in the GEE models.Reintegration support was measured basedon the involvement status in the rehabilita-tion packages.

ResultsTable1presents the background informationof the respondents in baseline. Of the 316respondents in baseline, 202 (63.9%) weremales. The age of the respondents rangedfrom 16 to 23 years, with a mean age of 20.6years (standard deviation 1.4), with four

ht © War Trauma Foundation. Unauthorized reproduction of this article is prohibited.370

Copyrig

Adhikari et al.

below the 18 years of age. All participantswere child soldiers when they joined theMaoist party and continued until thepeace accord in November 2006. Themajority of the respondents belonged toJanajati (44.9%) group and had secondarylevel education (52.8%). In the Nepali castesystem, Brahmin/Chhetri are consideredhigher castes, Janajatis refer to indigenousethnic groups, while Dalits refer to agroup of people traditionally regarded asuntouchable (Panday et al., 2006).

Prevalence of anxiety, depression andPTSD over study periodTable 2 presents the prevalence rate ofanxiety, depression, and PTSD in thesample. At baseline,106 (33.5%) had preva-lence of anxiety above the cut o¡ point,which changed to 61 (24.4%) at follow-up.The prevalence of depression declinedfrom 110 (34.8%) at baseline to 75 (30%) inthe follow-up; and likewise, the prevalenceof PTSD changed from 75 (30%) at baselineto18 (7.2%) outof 250 in the follow-up.Thesechanges from baseline to follow-up weresigni¢cant for anxiety and PTSD, but notfor depression (seeTable 3).

Protective and risk factors ofpsychosocial wellbeing for formerchild soldiersThe GEEmodel (Table 3) shows poor socialsupport and inter-caste marriage as thestrongest risk factors for psychosocial well-being during the study period. There wassigni¢cant association of psychosocial well-being with development regions; respon-dents from eastern regions showed thelowest levels of psychosocial and mentalhealth problems, and respondents from farwestern regions showed the highest levels ofproblems. Dalits reported the highest levelsof depression and anxiety symptoms,whereas Janajati showed the best outcomes.Education, economic status, reintegration

ht © War Trauma Foundation. Unautho

support and gender had no signi¢cantimpact on their psychosocial wellbeing.Protective and risk factors for anxiety:Anxiety levels of the respondents reducedsigni¢cantly from baseline to follow-up.Social factor and inter-caste marriage wasthe strongest risk factors for high anxietylevels. The other contributing factors ofhigher levels of anxiety were caste/ethnicity(belonging toJanajati community was a sup-portive factor for anxiety while being aDalitwas a risk factor), marital status (being mar-ried was found to be a risk factor), and age(higher age was a supportive factor).Protective and risk factors for depres-sion:Depression scores did not change overtime. The scores were best predicted bysocial factors, caste/ethnicity, and region ofthe country during the study period. Bettersocial reception and belonging to Janjaticaste/ethnicity were found to be protectivefactors, while living in the far western regionwas associated with signi¢cantly higherdepression, compared to those in living theeast. There was no signi¢cant improvementfrom baseline to follow-up for depressionwhen controlling for factors in the model.Protective and risk factors for PTSD:For PTSD, time was a signi¢cant factor forreduction of symptom levels from baselineto follow-up. Social support (lower socialsupport was associated with higher PTSDsymptoms), inter-castemarriage (inter-castemarriage associated with higher PTSDsymptoms), and school enrolment (associ-ated with lower PTSD symptoms) were thestrongest predicting factors of PTSD. Otherfactors included development, region of thecountry (living in the far west was associatedwith higher PTSD compared to living inthe east) and caste/ethnicity (Janajati hadlower PTSD symptoms than Dalits).Interaction e¡ects: Time was evaluated(baseline vs. follow-up) by gender (femalevs. male). For depression, there was agender by time interaction e¡ect, with boysshowing greater depression reduction as agroup, from T1 to T2 [beta¼�2.28, 95%

rized reproduction of this article is prohibited.371

Copyrig

Table 2. Prevalence of psychosocial problems over time

Backgroundcharacteristics

Anxiety Depression PTSD Number

Gender 2011 2012 2011 2012 2011 2012 2011 2012

Male 68 (33.7) 36 (24.2) 76 (37.6) 45 (30.2) 34 (16.8) 10 (6.7) 202 149Female 38 (33.0) 25 (24.8) 34 (29.8) 30 (29.7) 9 (7.9) 8 (7.9) 114 101AgeBelow age 20 28 (45.9) 6 (21.4) 25 (41.0) 8 (28.6) 11 (18.0) 3 (10.7) 61 2820^22 65 (30.2) 45 (27.8) 71 (33.0) 52 (32.1) 26 (12.1) 12 (7.4) 215 16223^24 13 (32.5) 10 (16.7) 14 (35.0) 15 (25.0) 6 (15.0) 3 (5.0) 40 60Caste/ethnicityBrahmin/Chhetri 53 (41.1) 10 (28.8) 56 (43.4) 37 (33.3) 22 (17.1) 8 (7.2) 129 111Janajati 35 (24.6) 19 (18.8) 35 (24.6) 26 (25.7) 13 (9.2) 6 (5.9) 142 101Dalit 18 (40.0) 32 (26.3) 19 (42.2) 12 (31.6) 8 (17.8) 4 (10.5) 45 38Marital statusUnmarried 44 (27.3) 18 (18.0) 52 (32.3) 26 (26.0) 20 (12.4) 5 (5.0) 161 100Married 62 (40.0) 43 (28.7) 58 (37.4) 49 (32.7) 23 (14.8) 13 (8.7) 155 150Level of educationBelow primary 4 (50.0) 3 (42.9) 4 (50.0) 3 (42.9) 1 (12.5) 2 (28.6) 8 7Primary 17 (37.8) 13 (32.5) 16 (35.6) 17 (42.5) 7 (15.6) 5 (12.5) 45 40Secondary 53 (31.7) 22 (19.1) 53 (31.7) 32 (27.8) 19 (11.4) 5 (4.3) 167 115School leaving

certi¢cate21 (40.4) 14 (23.7) 23 (44.2) 16 (27.1) 10 (19.2) 4 (6.8) 52 59

Intermediate level 11 (25.00) 9 (31.0) 14 (31.8) 7 (24.1) 6 (13.6) 2 (6.9) 44 29RegionFar western region 42 (50.0) 26 (40.0) 51 (48.8) 33 (50.8) 21 (25.0) 7 (10.8) 84 65Midwestern region 21 (37.5) 17 (38.6) 23 (41.1) 17 (38.6) 10 (17.90 6 (13.6) 56 44Western region 3 (15.00) 3 (18.8) 5 (25.0) 5 (31.2) 1 (5.0) 0.0 20 16Central region 22 (25.0) 10 (13.9) 24 (27.3) 15 (20.8) 7 (8.0) 3 (4.2) 88 72Eastern region 18 (26.5) 5 (9.4) 17 (25.0) 5 (9.4) 4 (5.9) 2 (3.8) 68 53Rehabilitation supportNone 37 (43.5) 14 (36.8) 35 (41.2) 18 (47.4) 18 (21.2) 4 (10.5) 85 38Education 19 (27.9) 11 (18.3) 23 (33.8) 13 (21.7) 4 (5.9) 3 (5.0) 68 60Training and formal

education in healthsector

11 (30.6) 8 (23.5) 13 (36.1) 9 (26.5) 8 (22.2) 2 (5.9) 36 34

Vocational training 14 (31.1) 13 (31.7) 18 (40.0) 14 (34.1) 7 (15.6) 5 (12.2) 45 41Micro enterprises 25 (30.5) 15 (19.5) 21 (25.6) 21 (27.3) 6 (7.3) 4 (5.2) 82 77Total 106 (33.5) 61 (24.4) 110 (34.8) 75 (30.0) 43 (13.6) 18 (7.2) 316 250

Protective and risk factors of psychosocial wellbeing related to the reintegration of former child soldiers in Nepal,

Intervention 2014, Volume 12, Number 3, Page 367 - 378

CI (con¢dence interval) �4.30, �0.27,p¼ 0.03]. For anxiety, there was no genderby time interaction e¡ect. For PTSD, there

ht © War Trauma Foundation. Unautho372

was a gender by time e¡ect. PTSD symp-toms decreased from baseline to follow-up,but only among boys (beta¼�3.95, 95%

rized reproduction of this article is prohibited.

Copyright © War Trauma Foundation. Unauthorized reproduction of this article is prohibited.

Table3.

Protective/su

ppor

tive

andrisk

factorsof

anxiety,de

pression

andPTSD

Variable

Anx

iety

Depression

PTSD

Beta-

coe⁄

cient

95%

CI

p-value

Beta-

coe⁄

cient

95%

CI

p-value

Beta-

coe⁄

cient

95%

CI

p-value

Tim

e(baselinevs.follo

w-up)

�0.76

�1.49

^0.04

0.04

�0.40

�1.51^0.71

0.48

�2.47

�4.37

^0.57

0.01

Gender(m

alevs.fem

ale)

0.58

�0.57^1.71

0.32

0.30

�1.37

^1.98

0.72

1.64

�0.73

^4.02

0.17

Marita

lstatus(married

vs.u

nmarried)

1.69

0.64

^2.75

0.00

1.00

�0.68

^2.69

0.24

2.13

�0.24

^4.50

0.08

Inter-castemarriage

(novs.yes)

1.71

0.12^3.23

0.03

2.22

�0.21^4.65

0.07

4.10

0.98

^7.22

0.01

Scho

olingstatus

(currently

notinscho

olingVs.

currentlyat

scho

ol)

�0.87

�2.33

^0.58

0.24

�0.64

�2.85

^1.57

0.57

�4.03

�7.29

^0.78

0.01

Ages

�0.40

�0.74

^0.05

0.02

�0.47

�0.99

^0.04

0.07

�0.42

�1.15^1.21

0.26

Socialsupp

ort

�0.51

�0.64

^0.39

0.00

�0.97

�1.17^0.77

0.00

�1.48

�1.75^1.21

0.00

Econo

micop

portun

ities

0.09

�0.10^0.29

0.34

0.04

�0.25

^0.33

0.80

0.15

�0.27

^0.56

0.49

Caste/ethnicity

Bah

min/C

hhetri

�0.64

�2.30

^1.02

0.45

�1.43

�3.7

8^0.92

0.23

�2.59

�6.0^0.82

0.14

Janajati

�1.66

�3.25

^0.08

0.04

�2.58

�4.91^0.25

0.03

�3.32

�6.72

^0.09

0.06

Dalit[R

ef.]

Edu

cation

allevel

Interm

ediatelevel

�2.42

�6.73

^1.89

0.27

�3.20

�9.13^2.73

0.29

�4.84

�16.30^6.61

0.41

Scho

olleaving

certi¢cate(SLC)

�1.12

�5.27

^3.03

0.60

�2.34

�7.94

^3.25

0.41

�2.90

�14.16

^8.36

0.61

(continuedoverleaf

)

Adhikari et al.

373

Copyright ©

Variable

Anx

iety

Depression

PTSD

Beta-

coe⁄

cient

95%

CI

p-value

Beta-

coe⁄

cient

95%

CI

p-value

Beta-

coe⁄

cient

95%

CI

p-value

Second

ary

�2.77

�6.59

^1.05

0.15

�4.32

�9.43

^0.80

0.10

�6.38

�17.20^4.44

0.25

Primary

�1.33

�5.21^2.54

0.50

�1.92

�7.18^3.33

0.47

�3.34

�14.37^7.69

0.55

Below

prim

ary

Region

Farwestern

region

1.53

�0.27

^3.32

0.09

4.04

1.34

^6.75

0.00

4.09

0.62

^7.57

0.02

Midwestern

region

1.42

�0.66

^3.50

0.18

1.98

�0.87

^4.83

0.17

�0.18

�4.10^3.74

0.93

Western

region

0.73

�1.23

^2.70

0.46

3.19

�0.03

^6.41

0.05

�0.57

�4.28

^3.14

0.76

Centralregion

0.21

�1.10^1.53

0.75

0.66

�1.07

^2.40

0.45

0.43

�2.29

^3.14

0.76

Eastern

region

Rehabilitation

supp

ort

Micro

enterprises

�0.61

�2.28

^1.07

0.48

�0.14

�2.49

^2.21

0.90

�0.33

�3.38

^2.72

0.83

Vocationa

lskilltraining

0.28

�1.64

^2.20

0.78

1.92

�0.82

^4.65

0.17

4.24

0.55

^7.92

0.02

Training

andform

aleducationin

health

sector

�1.61

�4.56

^1.34

0.28

0.30

�3.83

^4.44

0.89

0.38

�6.16^5.40

0.90

Edu

cation

0.41

�3.08

^2.26

0.76

1.21

�2.63

^5.06

0.54

�2.68

8.09

^2.74

0.33

Non

eCI¼Con¢denceinterval.

Table3.

(Con

tinu

ed)

Protective and risk factors of psychosocial wellbeing related to the reintegration of former child soldiers in Nepal,

Intervention 2014, Volume 12, Number 3, Page 367 - 378

374

Copyrig

Adhikari et al.

CI �7.62, �0.28, p¼ 0.04).When this inter-action is included, it was found that PTSDworsened for girls as a group, fromT1 toT2.Gender impacts: Gender was evaluated(female vs. male) by type of reintegrationpackage (school, health programme, voca-tional skills training, micro-enterprise andnone). For depression, there was one signi¢-cant gender by package interaction e¡ect:boys using the health package was associ-ated with greater depression symptoms(beta¼ 8.44, 95% CI 2.67, 14.21, p¼ 0.004).For anxiety, there was no gender by packageinteraction e¡ect. For PTSD, an interactione¡ect was found for males using the healthpackage (beta¼11.31, 95% CI 2.72, 19.91,p¼ 0.01) andmales using vocational training(beta¼ 9.31, 95% CI 0.30, 18.31, p¼ 0.04),when compared to females and educationas the reference groups.

DiscussionAmong former child soldiers released fromcantonments in Nepal, changes in depres-sion, anxiety, and PTSD symptoms wereevaluated over a nine month period. It wasfound that depression symptoms did notchange during this period. In contrast, bothanxiety and PTSD symptoms reduced sig-ni¢cantly during the period. Demographicfactors were associated with di¡erences inmental health outcomes among former childsoldiers. Being Dalit was associated withworse mental health outcomes, while beingJanajati was associated with both better psy-chosocial and mental health outcomes.Thismay be due to social hierarchy and culturalnorms associated with di¡erent caste groupsin Nepal (Panday et al., 2006). A consistentrisk factor for all outcomes were the per-ceived lack of social support for inter-castemarriages, which is not looked at positivelyin Nepalese society and inter-caste couplesoften face social stigma, abuse from familymembers and rejection from the community(CESCR&HRTMCC,2013).The study alsodemonstrated that former child soldiers in

ht © War Trauma Foundation. Unautho

the far western parts of the country had sig-ni¢cantly poorer outcomes than those ineastern regions. This may also be due tohigher numbers of non participation inrehabilitation packages for former childsoldiers from far western regions, lackof incoming generating opportunities anddi⁄cult living conditions.For PTSD, participants in the vocationalskills training had greater PTSD symptomseverity compared to participants that didnot participate in rehabilitation packages.Because rehabilitation packages were notrandomised, we cannot determine whetherselection factors or aspects of the rehabilita-tion package itself predicted these poor out-comes, as compared to participants notreceiving any rehabilitation package. Itcould also be that participants with morePTSD symptoms were selected for this pro-gramme.When interactions were examined,the pattern of outcomes demonstratedimportant gender related issues. Althoughtherewas no change fromT1toT2 as a groupfor depression, therewas a signi¢cant genderinteraction with boys showing improvementin depression from T1 to T2, whereas girlsshowed worsening of depression during thenine months. Similarly, although the groupas whole improved for PTSD symptomsfromT1 toT2, this change was actually lim-ited to boys only and girls as a group showedworsening of PTSD symptoms during thenine months. In contrast, there was no sig-ni¢cant interaction for gender and timefor anxiety.In addition, there were signi¢cant inter-actions by type of rehabilitation package.The greater severity of symptoms for PTSD,based on the rehabilitation package, wasonly signi¢cant forboys, and it includedboththe vocational skills training and trainingand formal education on health rehabilita-tion package. So whereas boys in generalshowed improvement in PTSD, boys in theseprogrammes did not. This may either bedue to selection issues (e.g., boys selectedfor these programmes tended to have more

rized reproduction of this article is prohibited.375

Copyrig

Protective and risk factors of psychosocial wellbeing related to the reintegration of former child soldiers in Nepal,

Intervention 2014, Volume 12, Number 3, Page 367 - 378

PTSD), or these packages in some mannerexacerbated PTSD. Because rehabilitationpackages were not randomly assigned, wecannot distinguish between these twointerpretations. Interestingly, we also founda gender e¡ect for boys receiving the train-ing and formal education on health, withboys in this rehabilitation package showinggreater depression severity. In summary,the negative association of rehabilitationpackages with mental health outcomes waslimited to boys. Whereas girls as a grouptended to not show improvement in mentalhealth outcomes, over nine months, regard-less of the rehabilitation package type.These ¢ndings raise questions about thebene¢t of the rehabilitation packages thatwere o¡ered to former child soldiers tofacilitate reintegration into civilian life. Thecurrent study does not demonstrate abene¢t in terms of mental health outcomes,and in the case of boys there were negative¢ndings in training and formal educationon health or vocational skill training. Thiscan be attributed to the time factor, as dur-ing follow-up research former child soldierswere not involved in any income generat-ing activities and the bene¢ts of trainingwere not yet obvious. In contrast, social sup-port was a signi¢cant and consistent predic-tor of mental health for all outcomes andamong both boys and girls. The family andcommunity’s attitude towards those whoparticipated in the package was positive,which indirectly contributed to a better psy-chosocialwellbeing for former child soldiers.This suggests that programmes and activi-ties focusing on mobilising social supportare essential for mental health outcomes.Implications:The focus of these rehabilita-tion programmes operatedwith the assump-tion that former child soldiers werepreviously integrated back into their com-munities. However, some former childsoldiers had join armed groups becausethey felt excluded from society. For thosewho did not join voluntarily, powerfulstigma and discrimination that prevents

ht © War Trauma Foundation. Unautho376

reintegration may remain. For many adults,child soldiers are seen as threatening andmorally corrupt (Boyden, 2003). It is welldocumented that former child soldiersstruggle with community stigma uponreturn home (Burman & McKay, 2007;Denov & Maclure, 2007; Kohrt et al.,2010; Shakya, 2010). This is also supportedby our present ¢ndings, connecting lowsocial support with higher levels of mentalhealth problems.This study calls for a signi¢cant re-examin-ing of how best to support psychosocialwellbeing and mental health of formerchild soldiers during the post con£ict re-integration process. Social support, alongwith material support, may be the keyissue that needs to be increasingly addressed.In addition, further rigorous researchis needed to assess the impact of theserehabilitation packages of support.Limitations: First, the non-probabilitysampling method used in the study meansthat it may not be possible to generalise ¢nd-ings for all former child soldiers. Second,there may have been an over representationof respondents from particular geographicalareas and particular rehabilitation schemes.

ConclusionsDuring the process of reintegration, formerchild soldiers’mental health in the domainsof PTSD and anxiety improved, especiallyfor boys. However, depression symptomsdid not improve. It is important to note thatmental health of girls remained the same orworsened over time. Moreover, there wasno direct positive impact of receivingrehabilitation support to facilitate transi-tion to civilian life (provided by the govern-ment of Nepal, in collaboration with UNagencies and NGOs) on the psychosocialwellbeing and mental health among respon-dents. That said, the rehabilitation supportmight well be helpful for economic andeducational outcomes over time, whichshould be the subject of further study.Mental health and psychosocial support is

rized reproduction of this article is prohibited.

Copyrig

Adhikari et al.

especially important for girls, Dalit groups,participants residing in far western regions,and those who have had inter-caste mar-riage, as they have poorer mental healthoutcomes. Our ¢ndings suggest that improv-ing social support is a key area to target forpotential improvement of mental healthand psychosocial wellbeing.

AcknowledgementsWe would like to thank all the research partici-pants for their time and support.We are gratefulto Mr. Evan Rai, Child Protection O⁄cer, UNI-CEF, Kathmandu for his valuable support. Weare also thankful to our research sta¡ for theirsupport and cooperation. This study has beenmade possible with the ¢nancial support ofUNICEF.

ReferencesAnnan,J.,Blattman,C.,Mazurana,D.&Carlson,K. (2011). Civil War, Reintegration, and Genderin Northern Uganda. Journal of Con£ict Resolution,55(6), 877-908.

Annan, J., Brier, M. & Aryemo, F. (2009). From‘‘Rebel’’to‘‘Returnee’’: Daily life and reintegrationfor young soldiers in Northern Uganda. Journalof Adolescent Research, 24(6), 639-667.

Betancourt,T. S., Agnew-Blais, J., Gilman, S. E.,Williams,D.R.&Ellis, B.H. (2010a).Past horrors,present struggles:The role of stigma in the associ-ation between war experiences and psychosocialadjustment among former child soldiers in SierraLeone. Social Science andMedicine, 70(1),17-26.

Betancourt,T. S., Brennan,R.T.,Rubin-Smith, J.,Fitzmaurice,G.M.&Gilman,S.E. (2010b).SierraLeone’s former child soldiers: a longitudinal studyof risk, protective factors, andmental health.Jour-nal of theAmerican Academy of Child&Adolescent Psy-

chiatry, 49(6), 606-615.

Betancourt,T.S.&Khan,K.T. (2008).Thementalhealth of childrena¡ectedby armedcon£ict: Pro-tective processes and pathways to resilience. Inter-national Review of Psychiatry, 20(3), 317-328.

ht © War Trauma Foundation. Unautho

Blattman,C.&Annan,J. (2010).Theconsequencesof child soldiering.ReviewofEconomicsandStatistics,

92(4), 882-898.

Boothby, N. (2006).What happens when child sol-diers grow up? TheMozambique case study. Inter-vention, 4(3), 244-259.

Boyden, J. (2003). The Moral Development ofChild Soldiers: What Do Adults Have to Fear?Peace and Con£ict: Journal of Peace Psychology, 9(4),343-362.

Burman,M. E.&McKay, S. (2007).Marginaliza-tion of girl mothers during reintegration fromarmedgroups inSierraLeone. InternationalNursingReview, 54(4), 316-323.

Colekessian, A. (2009). Reintegrating Gender: AGendered Analysis of the Nepali RehabilitationProcess. In: F. Partida (Ed.),Gender, Peaceand Secur-ity SeriesWorking Paper. Santo, Dominican Repub-lic: United Nation International ResearchTraining.

Committee on Economic, Social and CulturalRights (CESCR) & Human Rights TreatyMonitoring Coordination Centre (HRTMCC)(2013). Economic, Social and Cultural Rights inNepal: A Civil Society Parallel Report. Nepal:CESCR&HRTMCC.

Denov,M. (2010). Coping with the traumaof war:Former child soldiers in post-con£ict SierraLeone. International SocialWork, 53(6),791-806.

Denov,M.&Maclure,R. (2007).Turningsandepi-phanies: Militarization, life histories, and themaking and unmaking of two child soldiers inSierra Leone. Journal ofYouth Studies, 10(2),243-261.

Derluyn, I., Broekaert, E., Schuyten,G.&deTem-merman,E. (2004).Post-traumatic stress in formerUgandan child soldiers. Lancet, 363(9412), 861-863.

Kingma, K. (2000). In Sub-Saharan Africa: The

Development and the Security Impacts. Basingstoke:Macmillan Press Ltd. p. 28.

Kohrt, B. A., Jordans, M. J. D.,Tol,W. A., Speck-man, R. A., Maharjan, S. M.,Worthman, C. M.

rized reproduction of this article is prohibited.377

Copyrig

Protective and risk factors of psychosocial wellbeing related to the reintegration of former child soldiers in Nepal,

Intervention 2014, Volume 12, Number 3, Page 367 - 378

& Komproe, I. H. (2008). Comparison of mentalhealthbetween former child soldiers and childrennever conscripted by armed groups in Nepal.JAMA, 300(6), 691-702.

Kohrt, B. A.,Tol,W. A., Pettigrew, J. & Karki, R.(2010). Children and Revolution: The MentalHealth and Psychosocial Wellbeing of ChildSoldiers in Nepal’s Maoist Army.. In: M. Singer& G. D. Hodge (Eds.), The War Machine and

Global Health (89^116). Lanham, Maryland:Altamira Press: Rowan & Little¢eld Publishers,Inc.

Kryger,L. S.&Lindgren,C.L. (2011). Fighting fora future: The potential for posttraumatic growthamong youths formerly associated with armedforces in Northern Uganda. Intervention, 9(1), 6-20.

Panday,T. R., Chemjong, D.,Mishra, S., Pokhrel,S. & Rawal, N. (2006). Forms and Patterns of SocialDiscrimination in Nepal. Nepal: United NationsOrganization for Education, Science and Culture(UNESCO) Jawalakhel, Lalitpur Kathmandu.

Russell, A. (2012). Transitional Justice and theTruthCommission inNepal 2012 SeniorHonorsPro-jects. Paper 272. Kingston, USA: University ofRhode Island.

Shakya, A. (2010). Experiences of children inarmed con£ict in Nepal. Children andYouth ServicesReview, 33(4), 557-563.

Thapa S. B. & Hau¡, E. (2005). Psychologicaldistressamongdisplacedpersonsduringanarmedconict in Nepal. Social Psychiatry and Psychiatric

Epidemiology, 40(8), 672^ 679.

United Nation Development Programme(UNDP) (2011). A project Document titled Veri-¢edminors and late recruits are supported intheirsocio-economic rehabilitation through individualrehabilitation option packages. UNDP, UnitedNations Children’s Fund (UNICEF)/United

ht © War Trauma Foundation. Unautho378

Nations Population Fund (UNFPA)/Inter-national Labour Organization (ILO)

UN Interagency Rehabilitation Programme(UNIRP) (2012). Rehabilitation update Newslet-ter, Oct 2012. UNIRP o⁄ce, Pulchowk, Lalitpur,Nepal.

Vindevogel, S., Schryver, M. D., Broekaert, E. &Derluyn, I. (2013). Challenges Faced by FormerChildSoldiers intheAftermathofWar inUganda.Journal of Adolescent Health, 52,757-764.

Watson, C. (2009). Socio-economic reintegrationof ex-combatants: What role for the EuropeanUnion? In C. Watson (Ed.), Reintegration Brie¢ng

Paper 1.1: International Alert.

Williamson, J. (2006).The disarmament, demobi-lization and reintegration of child soldiers: Socialandpsychologicaltransformation inSierraLeone.Intervention, 4(3),185-203.

Ramesh P. Adhikari, is a psychiatrist and

medical anthropologist at the Research Depart-

ment,Transcultural Psychosocial Organization

(TPO), Nepal

email: [email protected]

Brandon A. Kohrt, is a psychiatrist and medical

anthropologist at the Department of Psychiatry

and Behavioral Sciences Duke Global Health

Institute, Duke University, USA

Nagendra P. Luitel, is a research manager, at the

Research Department,TPO, Nepal

Nawaraj Upadhaya, is a project coordinator at

the Research Department,TPO, Nepal

DristyGurung, isaresearcho⁄cerattheResearch

Department,TPO, Nepal

MarkJ.D. Jordans, PhD, Head of Research at

Research and Development, HealthNet TPO,

theNetherlandsandSeniorLecturerattheCentre

for GlobalMental Health, Institute of Psychia-

try, King’s College London, UK

rized reproduction of this article is prohibited.