Adolescent Peer Experiences and Prospective Suicidal ...

Full Terms amp Conditions of access and use can be found athttpswwwtandfonlinecomactionjournalInformationjournalCode=hcap20

Journal of Clinical Child amp Adolescent Psychology

ISSN 1537-4416 (Print) 1537-4424 (Online) Journal homepage httpswwwtandfonlinecomloihcap20

Adolescent Peer Experiences and ProspectiveSuicidal Ideation The Protective Role of Online-Only Friendships

Maya Massing-Schaffer Jacqueline Nesi Eva H Telzer Kristen A Lindquist ampMitchell J Prinstein

To cite this article Maya Massing-Schaffer Jacqueline Nesi Eva H Telzer Kristen A Lindquistamp Mitchell J Prinstein (2020) Adolescent Peer Experiences and Prospective Suicidal Ideation TheProtective Role of Online-Only Friendships Journal of Clinical Child amp Adolescent Psychology DOI1010801537441620201750019

To link to this article httpsdoiorg1010801537441620201750019

Published online 23 Apr 2020

Submit your article to this journal

Article views 141

View related articles

View Crossmark data

Adolescent Peer Experiences and Prospective Suicidal Ideation The ProtectiveRole of Online-Only FriendshipsMaya Massing-Schaffer a Jacqueline Nesibc Eva H Telzera Kristen A Lindquista and Mitchell J Prinsteina

aDepartment of Psychology and Neuroscience University of North Carolina at Chapel Hill bDepartment of Psychiatry and Human BehaviorWarren Alpert Medical School of Brown University cBradley Hasbro Research Center Rhode Island Hospital

AbstractObjective This study examined characteristics of online-only friendships among suicidal and non-suicidal adolescents In addition the extent to which adolescentsrsquo online-only friendships mayoffer a protective function buffering the effects of peer stressors (ie friendship stress relationalvictimization) on prospective suicidal ideation was examinedMethod Adolescents aged 10 to 14 (n = 630) were assessed at baseline (Time 1) and one-yearfollow-up (Time 2) Measures of suicidal ideation sociometric relational victimization friendshipstress depressive symptoms online-only friendship status and quality and online-only friendshipquality compared to in-person friendship quality were obtained at Time 1 using sociometricprocedures and self-report questionnaires Self-report measures of suicidal ideation were collectedat Time 2Results Descriptive results suggested that online-only friendships are relatively common amongyouth (383) particularly for those experiencing suicidal ideation (463) Suicidal and non-suicidal adolescents reported comparable levels of intimate disclosure within their online-onlyfriendships Although adolescents without suicidal ideation endorsed more support from in-person friendships suicidal adolescents endorsed similar levels of support from their online-only and in-person friendships Moderation analyses indicated that the association betweenboth relational victimization and friendship stress and prospective suicidal ideation was attenu-ated among youth who reported having one or more online-only friendConclusion Online-only friendships are common and may offer protective benefits for youthparticularly those experiencing suicidal ideation Future studies should examine the specificmechanisms by which online-only friendships may confer this benefit

For many decades theories have suggested thatinterpersonal experiences are particularly relevantto suicidal ideation Past work suggests that inter-personal stressors often precipitate thoughts of sui-cide while supportive interpersonal ties may protectindividuals at risk (Gallagher amp Miller 2018 King ampMerchant 2008 Van Orden et al 2010) For exam-ple over 60 years ago Durkheim theorized thatpoor social integration leads to heightened risk forsuicide (Durkheim 1951) Linehanrsquos biosocial theorysuggests that suicidal ideation occurs in the contextof invalidating social environments (Linehan 1993)Joinerrsquos interpersonal theory of suicide (IPTS)(Joiner 2005 Van Orden et al 2010) also empha-sizes two social constructs relevant to suicidal idea-tion including thwarted belongingness (eg feelingisolated from onersquos community) and perceived bur-densomeness (eg feeling like a burden on lovedones) Conversely research indicates that supportive

interpersonal experiences such as high levels ofperceived social support or close friendship qualitymay serve as protective factors for individuals atrisk for suicide (Gallagher amp Miller 2018)

The role of interpersonal experiences in the risk foror protection against suicidal ideation may be espe-cially relevant for adolescents Note that rates of com-pleted suicide increase 17-fold in the transition frompreadolescence to adolescence (Centers for DiseaseControl and Prevention [CDC] 2014) and suicideremains the second leading cause of death among10ndash14 and 15ndash24 year olds (Boeninger et al 2010)Adolescentsrsquo peer relationships may play a key role inthe onset or maintenance of suicidal ideation Indeedadolescence is a period during which youth experiencean increase in peer-themed stressors (Rudolph 2008)heightened neural reactivity to interpersonal stress andincreased sensitivity to social reward (Somerville 2013)In prior work researchers have identified a number of

CONTACT Maya Massing-Schaffer mayamassingschafferuncedu Department of Psychology and Neuroscience University of North Carolina atChapel Hill Chapel Hill NC 27599

JOURNAL OF CLINICAL CHILD amp ADOLESCENT PSYCHOLOGYhttpsdoiorg1010801537441620201750019

copy 2020 Society of Clinical Child amp Adolescent Psychology

peer-related risk factors for adolescent suicidal ideationincluding relational victimization (eg Massing-Schaffer et al 2019) chronic close friendship stress(Pettit et al 2011) and perceived peer rejection(Arango et al 2016 Prinstein et al 2000) Data alsosuggest that high quality close friendships may mitigatesuicide risk in adolescence (Bonanno amp Hymel 2010Czyz et al 2012) However more work on the role ofpeer factors as predictors or protective factors for ado-lescent suicide is needed perhaps especially reflectingnew forms of peer interaction that have emerged in thelast decade

Contemporary research suggests that adolescentsrsquopeer experiences have transformed markedly in theirfrequency format and function (Nesi et al 2018a2018b) mostly due to rapid advances in technologyhowever this has not been reflected frequently inresearch on adolescent suicidality For example it isnow possible for adolescents to establish ldquoonline-onlyfriendshipsrdquo or friendships that occur entirely onlinewithout any in-person interaction (Nesi et al 2018a)According to recent data 57 of 13 to 17 year-oldshave met a new friend online and 29 report that theyhad made more than five friends online (Lenhart2015) Furthermore 77 of adolescents who had meta new friend online reported that they had never metthem in person indicating high rates of online-exclusive friendships within this age group (Lenhart2015) Despite the increasing prevalence of online-only friendships among todayrsquos adolescents few studieshave examined the relevance of these types of friend-ships for mental health outcomes particularly suicidalideation

Importantly a related body of work has begun toexamine the impact of the Internet more broadly (egonline message boards online crisis support exposureto online suicide-related content) on suicidality andself-harm Although some work has highlighted poten-tial consequences of the Internet for self-injury includ-ing prompting or reinforcing self-harm behaviors(Lewis amp Seko 2016) findings also support a numberof benefits particularly for non-suicidal self-injury(NSSI) In one thematic analysis participants acrossmultiple studies reported that online activities offeredprotection against NSSI through mitigation of socialisolation recovery encouragement emotional self-disclosure and curbing NSSI urges (Lewis amp Seko2016) Moreover data suggest that most young peoplewho self-injure try to seek help online indicating thatthe Intenet may promote help-seeking behaviorsamong youth experiencing suicidal and non-suicidalself-injury (Frost amp Casey 2016) Despite these impor-tant advances the association between risk for suicidal

ideation and adolescentsrsquo online activities specificallywithin online-only friendships rarely has been studiedempirically

This study thus offers two novel contributions Firstas an initial step toward understanding the potentialrisk or protective effects of online friendship interac-tions we provide descriptive data to characterize theprevalence and quality of online-only friendshipsamong youth with and without a history of suicidalideation Second we examine the extent to whichyouthsrsquo online-only friendships may offer a protectivefunction moderating the prospective associationbetween peer stressors (ie friendship stress relationalvictimization) and suicidal ideation one year later add-ing to a remarkable dearth of significant prospectivepredictors of suicidal ideation revealed in prior work(Franklin et al 2017)

Descriptive data regarding the quality of online-onlyfriendships may reveal a novel source of support forsuicidal youth Despite past research emphasizing therisks associated with online-only friendships for suicidalyouth (Marchant et al 2017) emerging data suggest thatyouth experiencing suicidal ideation may also formhigh-quality social interactions online As noted byNesi et al (2018a) the online environment providesa setting in which there are fewer nonverbal cues (egvisual auditory and context) and in which communica-tion is more asynchronous For teens who report diffi-culties in their in-person social relationships which iscommon among suicidal adolescents the increased con-trollability of online environments may create a safercontext for higher-quality social exchanges (ie seehyperpersonal communication theories) (Desjarlais ampWilloughby 2010 Tidwell amp Walther 2002 Walther1996) Indeed initial evidence suggests that online-onlyfriendships can be high in quality particularly in thecontext of high self-disclosure and longer relationshipduration (Chan amp Cheng 2004 Mesch amp Talmud2007) Unfortunately little is known about the compara-tive levels of intimate disclosure within online-onlyfriendships for suicidal and non-suicidal youth In addi-tion few researchers have examined how the quality ofadolescentsrsquo online-only friendships compares to thequality of their in-person friendships especially amongsuicidal and non-suicidal youth Thus to address theseimportant gaps the first aim of the current study is tooffer descriptive data on the quality of suicidal and non-suicidal adolescentsrsquo online-only friendships particularlycompared to that of their in-person friendships

A second goal of this study was to examine thepotential stress-buffering effect of online-only friend-ships on suicidal ideation over time According to thesocial compensation hypothesis proposed within the

2 M MASSING-SCHAFFER ET AL

computer-mediated communication literature (Ellisonet al 2007 McKenna amp Bargh 2000) individuals withpoorer in-person social lives may benefit more fromonline communication as online relationships mayprovide a compensatory social experience (Ybarraet al 2015) There are numerous mechanisms bywhich online-only friendships may protect against theeffects of peer stress on suicide risk First for youthwho feel isolated or experience in-person interpersonaldifficulties online forums can present an opportunityto feel connected with others potentially offering pro-tection against suicidal ideation (De Choudhury ampKiciman 2017 Marchant et al 2017) In additionprevious research has documented that adolescentsexperiencing suicidal ideation may receive social sup-port online from others who are experiencing similarconcerns (Marchant et al 2017) Moreover prior worksuggests that online interactions may contribute todecreased loneliness (Hood et al 2018) and depressivesymptoms especially among those who are alienatedoffline These benefits may be especially relevant in thecontext of high-quality online-only friendships includ-ing those characterized by high levels of intimate dis-closure and low levels of criticism Thus in the currentstudy an initial goal was to examine whether online-only friendships may mitigate the effects of offline peerstress on prospective suicidal ideation amongadolescents

Two forms of peer stress were examined based onprior literature revealing stressors particularly relevantfor suicidality yet rarely examined in longitudinal workon suicide First past work suggests that as comparedto other forms of peer stress relational forms of peervictimization are especially associated with internaliz-ing symptoms including suicidal ideation (Klomeket al 2008 Massing-Schaffer et al 2019 Tsypes ampGibb 2015) However most prior work has been lim-ited by cross-sectional associations and few prior stu-dies have examined moderators in the prospectiveassociation between victimization and suicide ideationSecond close friendship stress was examined to addressa dearth of prospective longitudinal studies in priorwork on this construct

In the present study we examine how online-onlyfriendships may be relevant to suicidal ideation ina sample of youth at the adolescent transitiona critical developmental vulnerability period associatedwith the greatest increases in suicidal ideation Asa preliminary exploration we compared the prevalenceand quality of online-only friendships to the quality ofin-person friendships and also examined the quality ofthese friendships between adolescents with and withoutsuicidal ideation It was hypothesized that suicidal

youth would be more likely than non-suicidal youthto report having online-only friendships It was alsohypothesized that adolescents experiencing suicidalideation would report higher or comparable levels ofintimate disclosure in their online-only friendshipscompared to non-suicidal youth Further we expectedthat adolescents perhaps especially those experiencingsuicidal ideation would report higher or comparablelevels of closeness similarity and supportiveness intheir online-only friendships compared to their in-person friendships

A second aim of this study was to examine whetherthe presence of online-only friendships wasa moderator of the prospective association betweenpeer-themed stress (eg relational victimizationfriendship stress) and suicidal ideation We hypothe-sized that having online-only friendships would bufferthe harmful effects of peer stressors on suicidal idea-tion after accounting for the effects of demographicrisk factors prior ideation and depressive symptoms

Method

Participants

Participants were 630 adolescents enrolled in Grades 6and 7 in three middle schools within a southeasternrural region of the US at the start of the study (490female M = 1179 years SD = 70 range 10 to 14) Thesample was racially and ethnically heterogeneous(358 Caucasian 254 HispanicLatinx 213African American 108 Multiracial 67 OtherRace) School records indicated that 667 to 721of students in these schools were classified as econom-ically disadvantaged (North Carolina School ReportCards [NCDPI] 2017) Additionally 695 of studentsin the district were eligible for free or reduced-pricelunch based on district reports

Procedures

Participants were recruited from three rural public mid-dle schools (N = 1385) for a study of peer relations andpsychological adjustment Letters of consent were mailedto all caregivers of students in 6th and 7th grade with anoption to grant or deny consent for their child to parti-cipate in this study Approximately 77 of families(n = 1059 families) returned signed forms 88(n = 935) of these gave consent for their child to partici-pate yielding a sample that represented 675 of thepopulation in this diverse low-income community

At Time 1 873 consented students attended datacollection sessions A proportion of students (n = 62)

JOURNAL OF CLINICAL CHILD amp ADOLESCENT PSYCHOLOGY 3

who initially consented to participate in the study didnot attend data collection sessions at Time 1 due toschool absences (n = 10) withdrawing from the studyafter parental consent (n = 16) withdrawing from theschool (n = 19) and unknown reasons (n = 17)Consented participants who did not attend data collec-tions sessions at Time 1 did not differ from consentedparticipants who participated in Time 1 data collectionon grade or schools attended Students in this districtwere randomly assigned to each of three middleschools In two schools shorter time periods wereavailable for data collection thus 243 (278) partici-pants did not complete a portion of survey measuresincluding those related to online-only friendship In thecurrent dataset online-only friendship items were com-pleted by 934 of students at School 1 745 at School2 and 516 at School 3 Other than school placementthe 243 adolescents who did not complete online-onlyfriendship items did not differ from the 630 adolescentswho were included in analyses below on any study ordemographic variables with the exception of race χ2

(4) = 2045 p lt 001 with White students more likely tohave completed online-only friendship measures andLatinx students less likely to have done so The finalsample for the current study included 630 students

At baseline assent and data were obtained using com-puter-assisted self-interviews (CASI) administered bytrained research staff in school Data were collected attwo time points one year apart Time 1 data were col-lected in Winter 2016 and Time 2 data were collected inWinter 2017 Of the 630 students who completed ques-tionnaires at Time 1 54 (86) did not participate at Time2 due tomoving (n = 25) declining participation (n = 21)or reason not listed (n = 8) Thus retention rates at Time2 within this low-income sample were adequate (914n = 576) The 54 participants who did complete Time 2questionnaires did not differ from those who wereretained at Time 2 on any demographic or study variableswith the exception that those who were missing reportedlower baseline levels of depressive symptoms t(7069) = minus 248 p = 015 All procedures were approvedby the university human subjects committee

Measures

Participants completed all measures at baseline(Time 1) At Time 1 and at one-year follow-up(Time 2) participants reported on suicidal ideation

Online-Only FriendshipAdapting an item from prior work differentiatingonline-only from in-person friendships (Smahel et al2012) adolescents were asked ldquoDo you have any online

friends you have NOT met in personrdquo Participantswere given two response choices ldquoyesrdquo or ldquonordquo result-ing in a dichotomous measure of online-only friend-ship Participants who endorsed this initial questionwere asked several follow-up questions about the plat-forms through which adolescents met their online-onlyfriends at the time of data collection (ie InstagramSnapchat Tumblr Facebook Twitter Vine Google+Kik or Whatsapp messaging apps Online pinboardsDiscussion boards Anonymous sharing or questionapps) (Lenhart 2015) as well as several additionalitems regarding the quality of their online-only friend-ships discussed below

Intimate disclosure with online only friends To assesslevels of intimate disclosure within online-only friend-ships adolescents responded to three items derivedfrom the Intimate Disclosure subscale of the Networkof Relationships Inventory (Furman amp Buhrmester1985) These items were only administered to thosewho reported having at least one online-only friendSpecifically adolescents rated how much they ldquotalkabout everythingrdquo ldquoshare secrets and private feelingsrdquoand ldquotalk about things [they] do not want others toknowrdquo with online-only friends using a scale from 1(none) to 5 (extremely much) The three items wereanalyzed using a mean score (Cronbachrsquos α = 85)

Comparative friendship quality As a preliminaryindex of adolescentsrsquo perceived comparison betweentheir friendship quality in online-only friendships to in-person friendships a brief three-item measure wasdeveloped These comparative friendship quality itemswere only administered to those who reported having atleast one online-only friend Before responding to theseitems adolescents were asked to consider their online-only friends (as defined above) and their ldquoin-personfriendsrdquo defined as ldquopeople who you know offlineeven if you also communicate with them onlinerdquoParticipants were asked to indicate how close they arehow similar they feel in terms of interests and feelingsand how much support they receive from online-onlyfriendships compared to in-person friendships Eachitem was rated on a similar 5-point Likert scalewhere lower values indicated more positive qualitiesattributed to online friends and higher values indicatedmore positive qualities attributed to in-person friendsFor example the closeness item was rated on a scalefrom 1 (Much closer to ONLINE friends) to 5 (Muchcloser to IN-PERSON friends) with 3 being ldquoEquallyclose to online and in-person friendsrdquo These threeitems were examined individually to evaluate uniqueassociations between each dimension of friendship


quality and suicidal ideation as well as to gatherdescriptive information on how online-only and in-person friendships compare in terms of closeness simi-larity and supportiveness

Relational VictimizationSociometric procedures were used to collect measures ofrelational victimization at Time 1 Adolescents were givenan alphabetized roster of students in their grade (in coun-terbalanced order) and asked to nominate an unlimitednumber of peers ldquowho get left out of activities ignored byothers because one of their friends is mad at them gos-siped about or have mean things said behind their backsrdquo(Grotpeter amp Crick 1996 Helms et al 2015) To furtherclarify this definition adolescents were informed thatfriendly or playful teasing as well as arguments betweenindividuals with similar strength or power do not qualifyas relational victimization (Salmivalli amp Peets 2018)A sum of relational victimization nominations was com-puted and standardized for each student within theirschool and grade Thus higher scores indicated higherlevels of relational victimization compared to same-gradepeers Sociometric nomination procedures have demon-strated strong reliability and validity in prior studies ofadolescent relational victimization (Crick amp Grotpeter1995) and peer status (Coie amp Dodge 1983)

Friendship StressFriendship stress was assessed at Time 1 using a newlydeveloped self-report measure adapted from theYouth Life Stress Interview (YLSI) (Rudolph ampFlynn 2007) The original semi-structured interviewwas designed to assess adolescentsrsquo stress levels acrossseveral domains including peers In the peer stressportion of the interview assessors use standardizedquestions to gather objective information aboutyouth friendship stress (eg friendship quality levelof trust support and closeness in friendships severityof conflict in friendships) This interview has demon-strated excellent reliability and validity (Rudolph ampFlynn 2007) The present measure was developedbased on the standardized questions used to assesschronic peer stress in the YLSI as well as on commonpeer stressors identified by adolescents during thisinterview in prior studies The resulting questionnaireincluded 11 items designed to assess adolescent friend-ship stress (eg ldquoA friend lied to yourdquo ldquoA friendstarted to date someone you had a crush onrdquo) Allquestions were rated on a 5-point scale ranging from1 (never) to 5 (very often) A mean of items wascomputed with higher values indicating higher levelsof friendship stress In support of its concurrent

validity higher scores on this measure were correlatedwith other measures of poor friendship qualityincluding friendship conflict (r = 18 p lt 001) andcriticism (r = 15 p lt 001) on the Network ofRelationships Inventory (Furman amp Buhrmester1985) In addition internal consistency of this mea-sure was good (Cronbachrsquos α = 084)

Depressive SymptomsDepressive symptoms were measured at Time 1 usingthe Short Mood and Feelings Questionnaire (MFQ)(Angold et al 1995) This self-report measure contains9-items designed to assess symptoms of depression inyouth ages 8 to 18 Items include statements such as ldquoIfelt miserable or unhappyrdquo and ldquoI didnrsquot enjoy anythingat allrdquo All items are scored on a 3-point scale from 0(not at all true) to 2 (mostly true) for the individualover the past two weeks Data were analyzed usinga mean score of all items with higher mean scoresindicating higher levels of depressive symptoms Priorresearch supports the reliability and validity of theMFQ (Daviss et al 2006) In the current study internalconsistency of this measure was excellent (Cronbachrsquosα = 092)

Suicidal IdeationSuicidal ideation was assessed at Times 1 and 2 usingthe Suicide Questionnaire (Heilbron amp Prinstein 2010)an 8-item measure assessing the frequency of passiveand active suicidal ideation in adolescents usinga 5-point Likert scale from 1 (never) to 5 (almostevery day) Participants were asked to report on thefrequency of suicidal ideation over the past year Thismeasure has demonstrated good psychometric proper-ties in prior research (Heilbron amp Prinstein 2010) andshowed excellent internal consistency in the presentstudy at Time 1 (Cronbachrsquos α = 094) and Time 2(Cronbachrsquos α = 094) In order to capture the fullrange of frequency and severity of suicidal ideation inthe sample analyses were performed treating suicidalideation as both a categorical variable (for descriptiveanalyses) and a continuous variable (for regressionanalyses) For descriptive analyses two groups werecreated to distinguish participants who had experiencedsuicidal ideation and those who had not Given thateven low levels of suicidal ideation including passiveideation are associated with heightened risk of inter-personal difficulties in youth (Arango et al 2016)participants were placed in the suicidal group if theyendorsed any suicidal ideation or at least one item onthis measure (ie summed values greater than 8) Forall other analyses a mean score of all items was


computed with higher scores indicating higher fre-quencies of suicidal ideation

Data Analyses

Preliminary analyses revealed that data were skewed forsuicidal ideation at Time 1 (skewness = 351) and Time 2(skewness = 313) Values were log-transformed resultingin reductions in skewness for suicidal ideation at Time 1(skewness = 231) and Time 2 (skewness = 220) Thesevalues were used to conduct Pearson and point-biserialcorrelations to examine bivariate associations among allstudy variables (see Table 1)

Given the skewness and overdispersion of suicidal idea-tion values (ie conditional variance of suicidal ideationgreater than conditional means at each value of online-onlyfriendship) a negative binomial regression analysis wasconducted using the original (ie not log-transformed)values of suicidal ideation This analysis was used to testthe hypothesis that participation in online-only friendshipswould moderate the longitudinal association between peerstressors (eg friendship stress relational victimization)and suicidal ideation In order to minimize multicollinear-ity and facilitate interpretation of results all continuouspredictor variables were mean centered Two interactionterms were computed (ie online-only friendshipx relational victimization and online-only friendshipx friendship stress) and included in a negative binomialregression model predicting Time 2 suicide ideation(Step 3) after controlling for prior ideation gender anddepressive symptoms (Step 1) and main effects of socio-metric relational victimization friendship stress andonline-only friendship (Step 2) Finally exploratory ana-lyses were conducted to test possible interaction effects ofgender by online-only friendships as well as three-wayinteractions of gender by online-only friendships by peerstressors

Results

Descriptive Statistics

Bivariate correlations among all study variables arepresented in Table 1 Significant positive associationswere found between intimate disclosure within online-only friendships and friendship stress Depressivesymptoms were negatively associated with comparativefriendship quality indicating that higher levels ofdepressive symptoms were associated with higher levelsof online friendships as compared to in-person friend-ships Higher levels of log-transformed suicidal ideationat Time 1 were associated with higher levels of Time 1depressive symptoms friendship stress and Time 2log-transformed suicidal ideation

Table 2 shows a comparison of study variables amongadolescents with and without online-only friends At Time1 383 of adolescents (n = 241) reported having at leastone online-only friendship Compared to adolescents with-out online-only friendships (n = 389) adolescents who hadat least one online-only friendship reported significantlygreater levels of depressive symptoms log-transformedTime 1 suicidal ideation and friendship stress In additionadolescents with at least one online-only friend were sig-nificantly older than those who did not have online-onlyfriendships Finally results revealed no significant differ-ences in the reported number of in-person friendshipsbetween those who did and did not have online-onlyfriendships t(628) = minus 527 p= 599

Prevalence and Nature of Online-Only Friendships

An initial goal of this study was to examine descriptivecharacteristics of online-only friendships among suicidaland non-suicidal youth Table 3 presents means andstandard deviations of online-only friendship variableswith comparisons by participants with and without

Table 1 Means standard deviations and bivariate correlations among study variablesVariable M (SD) 1 2 3 4 5 6 7 8 9 10

1 Age 1180 (070) ndash2 T1 Depressive Symptoms 037 (048) minus03 ndash3 T1 Suicide Ideation 127 (060) minus04 63 ndash4 T1 Online-Only Friendship ndash 09 14 09 ndash5 T1 OOF Intimate Disclosure 230 (105) 03 06 13 ndash ndash6 T1 Comparative Friendship Closeness 356 (139) 10 minus07 minus07 ndash minus08 ndash7 T1 Comparative Friendship Similarity 349 (124) 05 minus14 minus08 ndash minus24 53 ndash8 T1 Comparative Friendship Support 357 (128) 03 minus14 minus12 ndash minus15 53 58 ndash9 T1 Relational Victimization 004 (106) 05 17 minus02 06 minus03 minus03 minus00 minus07 ndash10 T1 Friendship Stress 217 (075) 04 38 25 09 19 minus06 minus08 01 05 ndash11 T2 Suicide Ideation 129 (063) minus02 31 31 07 minus07 00 06 02 09 13

p lt 05 p lt 01 p lt 001 OOF = online-only friendship T1 = Time 1 T2 = Time 2 Pearson correlations are reported for all continuous variablesPoint-biserial correlations are reported for online-only friendship Comparative Friendship Quality refers to comparison of online-only versus in-personfriendships Higher values indicate higher closeness similarity and supportiveness perceived within in-person friendships compared to online-onlyfriendship Values for Suicide Ideation at Times 1 and 2 are log transformed


suicidal ideation1 Given that online-only friendship qual-ity variables were only administered to those whoreported having at least one online-only friend thesecomparisons are limited to that subset of the total sampleThemost common social media platforms through whichyouth formed online-only friendships were Instagram(n = 142 589 of all participants with online-onlyfriends) Snapchat (n = 91 378) Messaging apps(n = 71 295) and Facebook (n = 65 270)

Compared to those with no suicidal ideation(n = 377 598) participants with suicidal ideation(n = 244 387) were significantly more likely tohave online-only friendships χ2 (1) = 1131 p = 001No significant differences in levels of intimate disclo-sure within online-only friendships between suicidaland non-suicidal youth were found However signifi-cant group differences were found in comparativefriendship quality (ie mean of the three comparativefriendship quality items) t(299) = 219 p = 030 Onaverage both groups reported higher quality in-personfriendships compared to online friendships (ie meanvalues greater than 3) However the suicidal ideationgroup reported significantly lower mean scores(M = 334 SD = 135) than the non-ideation group(M = 379 SD = 119) suggesting that adolescentswith suicidal ideation reported more comparable levels

of quality between their online-only and in-personfriendships than do adolescents without suicidal idea-tion This was particularly true for the supportivenessitem t(227) = 269 p = 008 Gender differences alsoemerged in the frequency of online-only friends withprevalence rates higher among boys (n = 136 424)compared to girls (n = 105 340) χ2 = 469 p = 03No gender differences were found in levels of intimatedisclosure within online-only friendships or compara-tive friendship quality

Online Only Friendship as a Moderator of theProspective Association between Peer Stress andSuicidal Ideation

A second goal of this study was to examine the presence ofonline-only friendship as a moderator of the longitudinalassociation between peer stressors (eg friendship stressrelational victimization) and suicidal ideation Results(Table 4) revealed a significant main effect of peer-reported relational victimization on suicidal ideationcontrolling for prior suicidal ideation gender depressivesymptoms friendship stress and both interaction effects(ie relational victimization by online-only friendshipfriendship stress by online-only friendship)

Table 2 Comparison of study variables among adolescents with and without online-only friendsTotal

(n= 630)Online-Only Friends

(n= 241)No Online-Only Friends

(n = 389)Variable M (SD) M (SD)N () M (SD)N() t (df)χ2

Age 1180 (070) 1188 (072) 1175 (069) minus215 (614)T1 Depressive Symptoms 037 (048) 045 (050) 031 (046) minus338 (46875)T1 Suicidal Ideation 127 (060) 134 (065) 123 (058) minus220 (46453)T1 Relational Victimization 004 (106) 011 (125) minus001 (092) minus138 (627)T1 Friendship Stress 217 (075) 225 (074) 212 (075) minus208 (627)T2 Suicidal Ideation 129 (063) 126 (060) 135 (069) minus177 (573)

p lt 05 p lt 01Values for Suicide Ideation at Times 1 and 2 are mean-centered and log transformed

Table 3 Comparison of online-only friendship variables between adolescents with and without suicidal ideationTotal

(N = 630)Suicidal Ideation

(n = 244)No Suicidal Ideation

(n = 377)N () N () N () χ2

Presence of Only-Only Friend 241 (383) (261) 113 (463) 124 (329) 11305

M (SD) M (SD) M (SD) t (df)Intimate Disclosure with OOFs 230 (105) 241 (104) 222 (106) minus145 (233)Comparative Friendship Quality DimensionsCloseness 356 (139) 345 (139) 370 (136) 142 (229)Similarity 349 (124) 338 (117) 361 (128) 139 (22499)Supportiveness 357 (128) 334 (135) 379 (119) 269 (227)

p lt05 p lt 01OOF = online-only friendship

1Supplemental analyses were conducted with groups recalculated using a score of 10 or higher (ie at the 75th percentile) on thesuicidal ideation measure Patterns of significant and nonsignificant results remained the same with the exception that youth withsuicidal ideation reported significantly lower scores on the ldquosimilarityrdquo variable t(df) = 229 (16334) p = 024


Regarding the moderating effects of online-onlyfriendship on relational victimization and prospectivesuicidal ideation analyses revealed a significant rela-tional victimization by online-only friendship interac-tion effect B = minus 068 SE = 024 Exp(B) = 0934 Thuscompared to the effect of relational victimization onsuicidal ideation for those with no online-only friendsthe effect of relational victimization on suicide for thosewith online-only friends decreases by a factor of(934ndash 1) x 100 = minus 66

Regarding the moderating effects of online-onlyfriendship on friendship stress and prospective suicidalideation results revealed a significant friendship stressby online-only friendship interaction effect B = minus 12SE = 004 Exp(B) = 0892 Thus compared to the effectof friendship stress on suicidal ideation for those withno online-only friends the effect of friendship stress onsuicidal ideation for those with online-only friendsdecreases by a factor of (892ndash 1) x 100 = minus 108

A series of sensitivity analyses were also conductedFirst as an alternative method of addressing skewnessin the suicidal ideation outcome variable a hierarchicallinear regression was conducted using log-transformedvalues for suicidal ideation Compared to the results ofthe primary negative binomial regression analyses theresults of the linear regression showed the same patternof significant and non-significant associationsAdditionally exploratory analyses were run testinginteraction effects of gender by online-only friendshipsas well as three-way interactions of gender by online-only friendships by peer stressors No significant effectsemerged To address the potential for Type I errors andfalse positives sensitivity analyses were conducted bytesting the model without covariates (ie excludingdepression and gender) Findings revealed the samepattern of results suggesting low likelihood of TypeI errors and false positives in the current study

Discussion

The current study examined online-only friendships a peerexperience uniquely afforded by the social media environ-ment as well as the possible role of online-only friends asa buffer of the effects of peer stress on adolescent suicidalideation Findings reveal that online-only friendships arerelatively common among adolescents and that this experi-ence may be significantly more common amongmales andyouth with suicidal ideation Furthermore findings suggestthat online-only friendships may offer protective benefitsfor youth as the association between important peer stres-sors (ie relational victimization friendship stress) andsuicidal ideation was attenuated among youth whoreported the presence of one or more online-only friendsUsing a longitudinal designwithin a large diverse sample ofadolescents results offer new evidence for the critical rele-vance of online peer experiences for understanding suiciderisk among adolescents

Findings revealed significant gender differences in thepresence of online-only friendships with adolescentmales reporting significantly more online-only friend-ships compared to females These results are consistentwith prior work indicating that males are more likely tomake online friends than females (Lenhart 2015) Thehigher rates of online-only friendships among males mayrelate to the medium through which males develop onlinefriends Recent data highlight the growing popularity ofonline gaming among males with 84 of adolescentmales endorsing online video game use in one study(Lenhart 2015) Moreover prior work suggests thatwhile males are more likely to make friends throughonline gaming females are more likely to make onlinefriends through social media platforms such as Instagram(Lenhart 2015) These differences may reflect uniquefunctions of online-only friendships for males andfemales For instance consistent with studies of genderdifferences in offine friendships (Rose amp Rudolph 2006)online-only friendships may promote shared activitiesand competitive experiences among males whereasthese friendships may provide opportunities for socialconversation and prosocial behaviors for femalesHowever in the present study no differences in intimatedisclosure within online-only friendships were foundamong males and females Further measures of onlinegaming were not available Therefore more work isneeded to understand the differing prevalence rates andpotential functions of online-only friendships amongadolescent males and females

Findings from this study also present initial data char-acterizing the quality of online-only friendships comparedto in-person friendships for suicidal and non-suicidalyouth Results suggest that suicidal and non-suicidal

Table 4 Final negative binomial regression model predictingsuicidal ideation at Time 2

Time 2 Suicidal Ideation

B (SE) Wald Exp(B) 95 CI

Time 1 Suicidal Ideation 001 (000) 1957 101 101 102Depressive Symptoms 016 (004) 1721 117 109 126Gender 001 (003) 023 101 096 107Online-Only Friendship (OOF) 003 (003) 086 103 097 109Relational Victimization 007 (002) 1463 107 103 110Friendship Stress 006 (002) 558 102 101 111OOF x RelationalVictimization

minus007 (002) 822 93 089 098

OOF x Friendship Stress minus012 (004) 944 89 083 096

p lt05 p lt01 p lt001 Time 1 suicidal ideation calculated as a totalscore All other variables mean centered with the exception of Genderand Online-Only Friendship Gender was coded as 0 for females and 1 formales Online-Only Friendship was coded as 1 for presence of at least oneonline-only friendship and 0 for no reported online-only friendshipsExp(B) refers to incidence rate ratios OOF = online-only friendship


adolescents reported comparable levels of intimate disclo-sure within their online-only friendships Further resultsindicate that adolescents with suicidal ideation reportedsimilar levels of support from their online-only and in-person friendships compared to adolescents without sui-cidal ideation who favored in-person friendships for sup-port These results offer a unique perspective compared toprior work which often cites the negative effects ofonline-only friendships on in-person friendships particu-larly for suicidal youth For instance proponents of theldquopoor-get-poorerrdquo hypothesis suggest that adolescentswith unstable in-person friendships ndash which oftenincludes suicidal youth according to past work ndash aremore likely to use online friendships to escape from andavoid problems with in-person friendships further bar-ring these youth from opportunities to develop high-quality and supportive in-person friendships (Armstronget al 2000 Kraut et al 2002 1998) However adolescentsexperiencing suicidal ideation in the present study indi-cated that in-person friendships offered similar levels ofsupport as online-only friendships and suicidal youth stillreported higher quality in-person friendships comparedto online-only friendships Although more research isneeded to compare the direct effects of adolescentsrsquoonline-only friendships on the quality of their in-personfriendships these findings suggest that online-only friend-ships may represent a relevant source of support foradolescents experiencing suicidal ideation

Furthermore results from this study suggest that foradolescents who had experienced relational victimiza-tion (according to peer report) and friendship stress(according to self-report) having at least one online-only friend moderated the negative effects of theseexperiences on prospective suicidal ideation The nega-tive effects of relational victimization and friendshipstress on future suicide risk were attenuated for thosewith online-only friends These findings are consistentwith multiple theories of suicide including the inter-personal theory of suicide (Joiner 2005 Van Ordenet al 2010) which emphasize the role of social experi-ences in conferring and protecting against suicide riskYet although theoretical and empirical work has high-lighted the critical role of in-person social support inbuffering against the negative effects of social stressors(Cohen amp Wills 1985) remarkably little research hasexamined the association between online forms ofsocial support and suicidal ideation especially amongadolescents

Some prior work has highlighted the potential risksassociated with online relationships particularly forvulnerable adolescents For example studies haveemphasized the potential for participation in ldquodeviantcommunitiesrdquo or those in which potentially harmful

behaviors (eg suicidality non-suicidal self-injury)become normalized (Lewis amp Seko 2016 Marchantet al 2017) Furthermore the anonymity of onlineenvironments presents risks for dangerous or illicitbehavior including the potential for unwanted solicita-tion which may worsen mental health concerns(Mitchell et al 2001) Although acknowledging theserisks is crucial the current study offers a much-neededadditional perspective First the same features of thesocial media environment that may facilitate theserisks ndash such as possibilities for anonymity lack ofinterpersonal cues and frequent availability (Massing-Schaffer amp Nesi 2020 Nesi et al 2018a) ndash may alsofacilitate the development of the online-only friend-ships that offer critical social support for vulnerableadolescents Indeed in contrast to prior work high-lighting the potential dangers of online-only friend-ships a growing body of literature suggests theirpotential benefits for suicidal youth For exampleonline-only friendships can offer increased social sup-port for youth who are marginalized in their offlinesocial environments For instance preliminary datasuggest that youth who are at-risk for suicide such asthose who are LGBTQ have specific interests or havemedical conditions can gain support from online-onlyfriendships who are going through similar experiences(Ybarra et al 2015) Studies have also demonstratedthat interaction with peers through online messageboards can also increase emotional support for youthwho struggle with psychiatric difficulties including sui-cidal ideation and self-injury (De Choudhury ampKiciman 2017 Marchant et al 2017) Thus resultsfrom this study add to these literatures by demonstrat-ing that for adolescents who may also feel isolated orwho experience stress in their in-person relationshipsonline-only friendships can offer protective effectsagainst the experience of suicidal ideation

Implications and Limitations

The current study offers an important extension ofprior literature by examining the role of online-onlyfriendships in contributing to suicide risk witha prospective longitudinal design in a large diversesample of adolescents Whereas prior work has oftenreported on youthsrsquo online activity few studies of sui-cide risk have examined friendships that take placeexclusively online This is particularly important givenrapid advances in social media in recent years whichhave transformed the social landscape such that nearlyall of adolescentsrsquo in-person friends also representldquoonline friendsrdquo Furthermore while prior studieshave documented the risks inherent in online


friendships for youth (Livingstone amp Smith 2014) thecurrent study suggests a more complex entanglement ofrisks and benefits for youth experiencing suicidal idea-tion Future research must adopt a nuanced perspec-tive which considers the need to adequately addressonline risks while maximizing access to online supportamong vulnerable youth

Findings suggest that online-only friendships may pro-tect against the negative effects of peer victimization forvulnerable youth However the specific mechanisms bywhich online-only friendships confer this benefit werenot examined For instance future research should exploreadditional qualities of these online friendships (ie lengthof friendship type and timing of social support received) tobetter characterize this protective effect More research isalso needed to test whether youth turn to online-onlyfriendships for support following victimization experi-ences or if adolescentsrsquo existing online-only friendshipsattenuate the negative impact of ongoing vicimization Inaddition this study advances prior suicide research byincorporating a peer sociometric nomination procedureto assess relational victimization However by usinga standard sociometric item it is not necessarily clearwhether relationship victimization occurred within thecontext of in-person or online-only friendships Given thefact that peer reports were based on nomination of schoolclassmates it is unlikely that adolescentsrsquo peers were report-ing on victimization by online-only friends In addition itis likely that the presence of online-only friends can protectagainst the negative effects of victimization that occurs bothonline and in-person given known protective roles ofonline and in-person social support in cyber- and in-person victimization (Cole et al 2017 Kowalski et al2014) However this possibility should be explored infuture work Further this study offers an exploratoryexamination of associations between online-only friend-ships and suicidal ideation within a community sampleand considers only those with suicidal ideation (rather thansuicidal behavior including attempts) Given the low fre-quency of suicide attempts over the follow-up period in thecurrent study we were underpowered to test the interac-tions of relational victimization and friendship stress byonline-only friendship in predicting this important out-come It will therefore be critical for future studies to lookat whether online-only friendship buffer the effect of offlinesocial stressors for suicide attempters perhaps especiallywithin clinical samples of adolescents who may be morevulnerable to risks associated with online-only friendshipsFinally although the present study provides interestingdata noting the possible benefits of online-only friendshipsthese effects are indeed small More work is thereforeneeded to expand on study findings and validate resultsfrom this work

Conclusion

Adolescentsrsquo peer relationships play an integral role inthe development maintenance and exacerbation ofsuicidal ideation and behaviors As youth increasinglyturn to digital media as a primary context for socialinteraction understanding the unique risk and protec-tive features of this context has become critical Thecurrent study finds that nearly half of youth with sui-cidal ideation report the presence of an online-onlyfriend and that these online friendships may be espe-cially important to the social development of youth atrisk for suicidal ideation Online-only friendshipsdespite their risks may also offer important protectivebenefits for vulnerable youth who have experiencedrelational victimization and friendship stress The cur-rent study highlights the need for a nuanced researchagenda considering both the risks and benefits ofonline friendships within the study of adolescent sui-cide risk

Disclosure statement

No potential conflict of interest was reported by the authorsDr Nesi was supported in part by grant PDF-010517 fromthe American Foundation for Suicide Prevention (AFSP)AFSP had no role in the design and conduct of the studycollection management analysis and interpretation of thedata and preparation review or approval of the manuscriptor decision to submit the manuscript for publication Thecontent is solely the responsibility of the authors and doesnot necessarily represent the official views of AFSP

ORCID

Maya Massing-Schaffer httporcidorg0000-0002-9663-3958

References

Angold A Costello E J Messer S C amp Pickles A (1995)Development of a short questionnaire for use in epidemio-logical studies of depression in children and adolescentsInternational Journal of Methods in Psychiatric Research 5(4) 237ndash249

Arango A Opperman K J Gipson P Y amp King C A(2016) Suicidal ideation and suicide attempts among youthwho report bully victimization bully perpetration andor lowsocial connectedness Journal of Adolescence 51 19ndash29httpsdoiorg101016jadolescence201605003

Armstrong L Phillips J G amp Saling L L (2000) Potentialdeterminents of heavier internet usage InternationalJouranl of Human Computer Studies 53(4) 113ndash122httpsdoiorg10172652328-2177201703001

Boeninger D K Masyn K E Feldman B J ampConger R D (2010) Sex differences in developmentaltrends of suicide ideation plans and attempts among


European American adolescents Suicide and Life-Threatening Behavior 40(5) 451ndash464 httpsdoiorg101521suli2010405451

Bonanno R A amp Hymel S (2010) Beyond hurt feelingsInvestigating why some victims of bullying are at greaterrisk for suicidal ideation Merrill-Palmer Quarterly 56(3)420ndash440 httpsdoiorg101353mpq00051

Centers for Disease Control and Prevention (2014) Youthrisk behavior surveillancemdashUnited States 2014 Morbidityand Mortality Weekly Report Surveillance Summary 55(SSndash5) 1ndash108

Chan D K-S amp Cheng G H-L (2004) A comparison ofoffline and online friendship qualities at different stages ofrelationship development Journal of Social and PersonalRelationships 21(3) 305ndash320 httpsdoi-orglibproxylibuncedu1011770265407504042834

Cohen S amp Wills T A (1985) Stress social support andthe buffering hypothesis Psychological Bulletin 98(2)310ndash357 httpsdoiorg1010370033-2909982310

Coie J D amp Dodge K A (1983) Continuities and changesin childrenrsquos social status A five-year longitudinal studyMerrill-Palmer Quarterly 29(3) 261ndash282

Cole D A Nick E A Zelkowitz R L Roeder K M ampSpinelli T (2017) Online social support for young peopleDoes it recapitulate in-person social support can it helpComputers in Human Behavior 68 456ndash464 httpsdoiorg101016jchb201611058

Crick N R amp Grotpeter J K (1995) Relational aggressiongender and social-psychological adjustment ChildDevelopment 66(3) 710ndash722 httpsdoiorg1023071131945

Czyz E K Liu Z amp King C A (2012) Social connected-ness and one-year trajectories among suicidal adolescentsfollowing psychiatric hospitalization Journal of ClinicalChild and Adolescent Psychology 41(2) 214ndash226 httpsdoiorg101080153744162012651998

Daviss W B Birmaher B Melhem N A Axelson D AMichaels S M amp Brent D A (2006) Criterion validity ofthe Mood and Feelings Questionnaire for depressive epi-sodes in clinic and non-clinic subjects Journal of ChildPsychology and Psychiatry 47(9) 927ndash934 httpsdoiorg101111j1469-7610200601646x

De Choudhury M amp Kiciman E (2017) The language ofsocial support in social media and its effect on suicidal idea-tion risk Proceedings of the Eleventh International AAAIConference on Weblogs and Social Media 2017 32ndash41

Desjarlais M amp Willoughby T (2010) A longitudinal studyof the relation between adolescent boys and girlsrsquo compu-ter use with friends and friendship quality Support for thesocial compensation or the rich-get-richer hypothesisComputers in Human Behavior 26(5) 896ndash905 httpsdoiorg101016jchb201002004

Durkheim E (1951) Suicide Free PressEllison N B Steinfield C amp Lampe C (2007) The benefits of

Facebook ldquofriendsrdquo Social capital and college studentsrsquo useof online social network sites Journal of Computer-MediatedCommunication 12(4) article 1 Retrieved June 2019 fromhttpsjcmcindianaeduvol12issue4ellisonhtml

Franklin J C Ribeiro J D Fox K R Bentley K HKleiman E M Huang X Musacchio K MJaroszewski A C Chang B P amp Nock M K (2017)Risk factors for suicidal thoughts and behaviors A

meta-analysis of 50 years of research PsychologicalBulletin 143(2) 187ndash232 httpsdoi-orglibproxylibuncedu101037bul0000084

Frost M amp Casey L (2016) Who seeks help online forself-injury Archive of Suicide Research 20(1) 69ndash79httpsdoi-orglibproxylibuncedu1010801381111820151004470

Furman W amp Buhrmester D (1985) Childrenrsquos percep-tions of the personal relationships in their social networksDevelopmental Psychology 21(6) 1016ndash1024 httpsdoiorg1010370012-16492161016

Gallagher M L amp Miller A B (2018) Suicidal thoughtsand behavior in children and adolescents An ecologicalmodel of resilience Adolescent Research Review 3(2)123ndash154 httpsdoiorg101007s40894-017-0066-z

Grotpeter J K amp Crick N R (1996) Relational aggressionovert aggression and friendship Child Development 67(5)2328ndash2338 httpsdoiorg1023071131626

Heilbron N amp Prinstein M J (2010) Adolescent peervictimization peer status suicidal ideation and nonsuici-dal self-injury Merrill-Palmer Quarterly 56(3) 388ndash419httpsdoiorg101353mpq00049

Helms S W Gallagher M Calhoun C D Choukas-Bradley S Dawson G C amp Prinstein M J (2015)Intrinsic religiosity buffers the longitudinal effects of peervictimization on adolescent depressive symptoms Journalof Clinical Child and Adolescent Psychology 44(3)471ndash479 httpsdoiorg101080153744162013865195

Hood M Creed P A amp Mills B J (2018) Loneliness andonline friendships in emerging adults Personality andIndividual Differences 133 96ndash102 httpsdoiorg101016jpaid201703045

Joiner T (2005) Why people die by suicide HarvardUniversity Press

King C A amp Merchant C R (2008) Social and interperso-nal factors relating to adolescent suicidality A review ofthe literature Archives of Suicide Research 12(3) 181ndash196httpsdoiorg10108013811110802101203

Klomek A B Marrocco F Kleinman M Schonfeld I Samp Gould M S (2008) Peer victimization depression andsuicidality in adolescents Suicide and Life-ThreateningBehavior 38(2) 166ndash180 httpsdoiorg101521suli2008382166

Kowalski R M Giumetti G W Schroeder A N ampLattanner M R (2014) Bullying in the digital ageA critical review and meta-analysis of cyberbullyingresearch among youth Psychological Bulletin 140(4)1073ndash1137 httpsdoiorg101037a0035618

Kraut R Kiesler S Boneva B Cummings J NHelgeson V amp Crawford A M (2002) Internet paradoxrevisited Journal of Social Issues 58(1) 49ndash74 httpsdoiorg1011111540-456000248

Kraut R Patterson M Lundmark V Kiesler SMukophadhyay T amp Scherlis W (1998) Internet paradoxA social technology that reduces social involvement andpsychological well-being American Psychologist 53(9)1017ndash1031 httpsdoiorg1010370003-066X5391017

Lenhart A (2015) Teens technology and friendships PewResearch Center httpwwwpewinternetorg20150806teens-technology-and-friendships

Lewis S P amp Seko Y (2016) A double-edged swordA review of benefits and risks of online nonsuicidal self-


injury activities Journal of Clinical Psychology 72(3)249ndash262 httpsdoiorg101002jclp22242

Linehan M (1993) Cognitive-behavioral treatment of border-line personality disorder Guilford press

Livingstone S amp Smith P K (2014) Annual researchreview Harms experienced by child users of online andmobile technologies The nature prevalence and manage-ment of sexual and aggressive risks in the digital ageJournal of Child Psychology and Psychiatry 55(6)635ndash654 httpsdoiorg101111jcpp12197

Marchant A Hawton K Stewart A Montgomery PSingaravelu V Lloyd K John A amp Purdy N (2017)A systematic review of the relationship between internetuse self-harm and suicidal behaviour in young people Thegood the bad and the unknown PloS One 12(8)e0181722 httpsdoiorg101371journalpone0181722

Massing-Schaffer M Helms S W Rudolph K DSlavich G M Hastings P D Giletta M Nock M Kamp Prinstein M J (2019) Preliminary associations amongrelational victimization targeted rejection and suicidalityin adolescents A prospective study Journal of ClinicalChild and Adolescent Psychology 48(2) 288ndash295 httpsdoiorg1010801537441620181469093

Massing-Schaffer M amp Nesi J (2020) Cybervictimization andsuicide risk in adolescence An integrative model of socialmedia and suicide theories Adolescent Research Review 549ndash65 httpsdoi-orglibproxylibuncedu1010801537441620181469

McKenna K Y A amp Bargh J A (2000) Plan 9 fromcyberspace The implications of the Internet for personalityand social psychology Personality and Social PsychologyRev i ew 4 (1 ) 57ndash75 h t tp s do i o rg 10 1207 S15327957PSPR0401_6

Mesch G S amp Talmud I (2007) Similarity and the qualityof online and offline social relationships among adoels-cents in Israel Journal of Research on Adolescence 17(2)455ndash466 httpsdoi-orglibproxylibuncedu101111j1532-7795200700529x

Mitchell K J Finkelhor D amp Wolak J (2001) Risk factorsfor and impact of online sexual solicitation for youthJAMA Journal of the American Medical Association 285(23) 3011ndash3014 httpsdoi-orglibproxylibuncedu101001jama285233011

Nesi J Choukas-Bradley S amp Prinstein M J (2018a)Transformation of adolescent peer relations in the socialmedia context Part 1mdasha theoretical framework and appli-cation to dyadic peer relationships Clinical Child andFamily Psychology Review 21(3) 267ndash294 httpsdoiorg101007s10567-018-0261-x

Nesi J Choukas-Bradley S amp Prinstein M J (2018b)Transformation of adolescent peer relations in the socialmedia context Part 2mdashapplication to peer group processesand future directions for research Clinical Child andFamily Psychology Review httpsdoiorg101007s10567-018-0262-9

North Carolina School Report Cards (2017) httpsncreportc a r d s ond emand s a s c om s r c s c hoo l s c hoo l =530336ampyear=2017amplang=english

Pettit J W Green K L Grover K E Schatte D J ampMorgan S T (2011) Domains of chronic stress and sui-cidal behaviors among inpatient adolescents Journal ofClinical Child and Adolescent Psychology 40(3) 494ndash499httpsdoiorg101080153744162011563466

Prinstein M J Boergers J Spirito A Little T D ampGrapentine W L (2000) Peer functioning family dys-function and psychological symptoms in a risk factormodel for adolescent inpatientsrsquo suicidal ideation severityJournal of Clinical Child Psychology 29(3) 392ndash405httpsdoiorg101207S15374424JCCP2903_10

Rose A J amp Rudolph K D (2006) A review of sex differ-ences in peer relationship processes Potential trade-offsfor the emotional and behavioral development of girls andboys Psychological Bulletin 132(1) 98ndash131 httpsdoiorg1010370033-2909132198

Rudolph K D (2008) Developmental influences on inter-personal stress generation in depressed youth Journal ofAbnormal Psychology 117(3) 673ndash679 httpsdoiorg1010370021-843X1173673

Rudolph K D amp Flynn M (2007) Childhood adversity andyouth depression Influence of gender and pubertal statusDevelopment and Psychopathology 19(2) 497ndash521 httpsdoiorg101017S0954579407070241

Salmivalli C amp Peets K (2018) Bullying and victimizationIn W M Bukowski B Laursen amp K H Rubin (Eds)Handbook of peer interactions relationships and groups(2nd ed pp 302ndash321) The Guilford Press

Smahel D Brown B B amp Blinka L (2012) Associationsbetween online friendship and Internet addiction amongadolescents and emerging adults Developmental Psychology48(2) 381ndash388 httpsdoiorg101037a0027025

Somerville L H (2013) The teenage brain Sensitivity to socialevalaution Current Directions in Psychological Science 22(2)121ndash127 httpsdoiorg1011770963721413476512

Tidwell L C amp Walther J B (2002) Computer-mediatedcommunication effects on disclosure impressions and inter-personal evaluations Getting to know one another a bit at atime Human Communication Research 28(3) 317ndash348httpsdoiorg101111j1468-29582002tb00811x

Tsypes A amp Gibb B E (2015) Peer victimization mediatesthe impact of maternal depression on risk for suicidalideation in girls but not boys A prospective studyJournal of Abnormal Child Psychology 43(8) 1439ndash1445httpsdoiorg101007s10802-015-0025-8

Van Orden K A Witte T K Cukrowicz K CBraithwaite S R Selby E A amp Joiner T E Jr (2010)The interpersonal theory of suicide Psychological Review117(2) 575ndash600 httpsdoiorg101037a0018697

Walther J B (1996) Computer-mediated communicationImpersonal interpersonal and hyperpersonal interactionCommunication Research 23(1) 3ndash43 httpsdoiorg101177009365096023001001

Ybarra M L Mitchell K J Palmer N A amp Reisner S L(2015) Online social support as a buffer against online andoffline peer and sexual victimization among US LGBT andnon-LGBT youth Child Abuse amp Neglect 39 123ndash136httpsdoiorg101016jchiabu201408006


Abstract

Method

Participants

Procedures

Measures

Online-Only Friendship

Relational Victimization

Friendship Stress

Depressive Symptoms

Suicidal Ideation

Data Analyses

Results



Online Only Friendship as aModerator of the Prospective Association between Peer Stress and Suicidal Ideation

Discussion


Conclusion


References

Adolescent Peer Experiences and Prospective Suicidal Ideation The ProtectiveRole of Online-Only FriendshipsMaya Massing-Schaffer a Jacqueline Nesibc Eva H Telzera Kristen A Lindquista and Mitchell J Prinsteina

aDepartment of Psychology and Neuroscience University of North Carolina at Chapel Hill bDepartment of Psychiatry and Human BehaviorWarren Alpert Medical School of Brown University cBradley Hasbro Research Center Rhode Island Hospital

AbstractObjective This study examined characteristics of online-only friendships among suicidal and non-suicidal adolescents In addition the extent to which adolescentsrsquo online-only friendships mayoffer a protective function buffering the effects of peer stressors (ie friendship stress relationalvictimization) on prospective suicidal ideation was examinedMethod Adolescents aged 10 to 14 (n = 630) were assessed at baseline (Time 1) and one-yearfollow-up (Time 2) Measures of suicidal ideation sociometric relational victimization friendshipstress depressive symptoms online-only friendship status and quality and online-only friendshipquality compared to in-person friendship quality were obtained at Time 1 using sociometricprocedures and self-report questionnaires Self-report measures of suicidal ideation were collectedat Time 2Results Descriptive results suggested that online-only friendships are relatively common amongyouth (383) particularly for those experiencing suicidal ideation (463) Suicidal and non-suicidal adolescents reported comparable levels of intimate disclosure within their online-onlyfriendships Although adolescents without suicidal ideation endorsed more support from in-person friendships suicidal adolescents endorsed similar levels of support from their online-only and in-person friendships Moderation analyses indicated that the association betweenboth relational victimization and friendship stress and prospective suicidal ideation was attenu-ated among youth who reported having one or more online-only friendConclusion Online-only friendships are common and may offer protective benefits for youthparticularly those experiencing suicidal ideation Future studies should examine the specificmechanisms by which online-only friendships may confer this benefit

For many decades theories have suggested thatinterpersonal experiences are particularly relevantto suicidal ideation Past work suggests that inter-personal stressors often precipitate thoughts of sui-cide while supportive interpersonal ties may protectindividuals at risk (Gallagher amp Miller 2018 King ampMerchant 2008 Van Orden et al 2010) For exam-ple over 60 years ago Durkheim theorized thatpoor social integration leads to heightened risk forsuicide (Durkheim 1951) Linehanrsquos biosocial theorysuggests that suicidal ideation occurs in the contextof invalidating social environments (Linehan 1993)Joinerrsquos interpersonal theory of suicide (IPTS)(Joiner 2005 Van Orden et al 2010) also empha-sizes two social constructs relevant to suicidal idea-tion including thwarted belongingness (eg feelingisolated from onersquos community) and perceived bur-densomeness (eg feeling like a burden on lovedones) Conversely research indicates that supportive

interpersonal experiences such as high levels ofperceived social support or close friendship qualitymay serve as protective factors for individuals atrisk for suicide (Gallagher amp Miller 2018)

The role of interpersonal experiences in the risk foror protection against suicidal ideation may be espe-cially relevant for adolescents Note that rates of com-pleted suicide increase 17-fold in the transition frompreadolescence to adolescence (Centers for DiseaseControl and Prevention [CDC] 2014) and suicideremains the second leading cause of death among10ndash14 and 15ndash24 year olds (Boeninger et al 2010)Adolescentsrsquo peer relationships may play a key role inthe onset or maintenance of suicidal ideation Indeedadolescence is a period during which youth experiencean increase in peer-themed stressors (Rudolph 2008)heightened neural reactivity to interpersonal stress andincreased sensitivity to social reward (Somerville 2013)In prior work researchers have identified a number of

CONTACT Maya Massing-Schaffer mayamassingschafferuncedu Department of Psychology and Neuroscience University of North Carolina atChapel Hill Chapel Hill NC 27599

JOURNAL OF CLINICAL CHILD amp ADOLESCENT PSYCHOLOGYhttpsdoiorg1010801537441620201750019

copy 2020 Society of Clinical Child amp Adolescent Psychology














Method

Participants


Procedures







Measures















Data Analyses



Results
























(n = 377)N () N () N () χ2









Discussion








minus007 (002) 822 93 089 098













Conclusion




ORCID


References
































































Abstract

Method

Participants

Procedures

Measures



Friendship Stress

Depressive Symptoms

Suicidal Ideation

Data Analyses

Results




Discussion


Conclusion


References














Method

Participants


Procedures







Measures















Data Analyses



Results
























(n = 377)N () N () N () χ2









Discussion








minus007 (002) 822 93 089 098













Conclusion




ORCID


References
































































Abstract

Method

Participants

Procedures

Measures



Friendship Stress

Depressive Symptoms

Suicidal Ideation

Data Analyses

Results




Discussion


Conclusion


References




Measures















Data Analyses



Results
























(n = 377)N () N () N () χ2









Discussion








minus007 (002) 822 93 089 098













Conclusion




ORCID


References
































































Abstract

Method

Participants

Procedures

Measures



Friendship Stress

Depressive Symptoms

Suicidal Ideation

Data Analyses

Results




Discussion


Conclusion


References









Data Analyses



Results
























(n = 377)N () N () N () χ2









Discussion








minus007 (002) 822 93 089 098













Conclusion




ORCID


References
































































Abstract

Method

Participants

Procedures

Measures



Friendship Stress

Depressive Symptoms

Suicidal Ideation

Data Analyses

Results




Discussion


Conclusion


References















(n = 377)N () N () N () χ2









Discussion








minus007 (002) 822 93 089 098













Conclusion




ORCID


References
































































Abstract

Method

Participants

Procedures

Measures



Friendship Stress

Depressive Symptoms

Suicidal Ideation

Data Analyses

Results




Discussion


Conclusion


References




Discussion








minus007 (002) 822 93 089 098













Conclusion




ORCID


References
































































Abstract

Method

Participants

Procedures

Measures



Friendship Stress

Depressive Symptoms

Suicidal Ideation

Data Analyses

Results




Discussion


Conclusion


References










Conclusion




ORCID


References
































































Abstract

Method

Participants

Procedures

Measures



Friendship Stress

Depressive Symptoms

Suicidal Ideation

Data Analyses

Results




Discussion


Conclusion


References



























































Abstract

Method

Participants

Procedures

Measures



Friendship Stress

Depressive Symptoms

Suicidal Ideation

Data Analyses

Results




Discussion


Conclusion


References

Adolescent Peer Experiences and Prospective Suicidal ...

Documents

Transcript of Adolescent Peer Experiences and Prospective Suicidal ...