ARTICLES A Prospective Study of the Effects of Optimism on Adolescent Helth Risks AUTHORS: George C. Ptton, MD,,b,c Michelle M. Tollit, EdPsych,,b Helen Romniuk, PhD,,b,c,d Susn H. Spence, PhD, e Jennie Sheffield, PhD, f nd Michel G. Swyer, PhD g,h Centre for Adolescent Helth, Royl Children s Hospitl, Prkville, Victori, Austrli; d Clinicl Epidemiology nd Biosttistics Unit, b Murdoch Childrens Reserch Institute, Prkville, Victori, Austrli; c Deprtment of Peditrics, University of Melbourne, Victori, Austrli; e Griffith Helth Institute, Griffith University, Queenslnd, Austrli; f School of Psychology, University of Queenslnd, Queenslnd, Austrli; g Discipline of Peditrics, University of Adelide, Adelide, Austrli; nd h Reserch nd Evlution Unit, Children, Youth nd Women s Helth Service, Adelide, Austrli KEY WORDS optimism, psychologicl style, depression, nxiety, substnce use, ntisocil behviors, dolescence ABBREVIATIONS GEE generlized estimting equtions CI confidence intervl www.peditrics.org/cgi/doi/10.1542/peds.2010-0748 doi:10.1542/peds.2010-0748 Accepted for publiction Nov 11, 2010 Address correspondence to George C. Ptton, MD, Centre for Adolescent Helth, 2 Gtehouse St, Prkville 3052, Austrli. E- mil: george.ptton@rch.org.u PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275). Copyright 2011 by the Americn Acdemy of Peditrics FINANCIAL DISCLOSURE: The uthors hve indicted they hve no finncil reltionships relevnt to this rticle to disclose. WHAT S KNOWN ON THIS SUBJECT: Optimism lter in life is protective ginst rnge of helth problems. It hs commonly been focus in dolescent mentl helth promotion. Crosssectionl studies suggest protective effect ginst dolescent helth risks, but prospective studies hve been lcking. WHAT THIS STUDY ADDS: Optimism is somewht protective ginst dolescent helth risks; the strongest effect ws seen ginst the onset of new depressive symptoms. Its protective effect ginst hevier substnce use nd ntisocil behvior ws modest nd only for the highest ctegories compred to the lowest. Promoting optimism long with other positive spects of psychologicl nd emotionl style hs role in mentl helth promotion tht is likely to be enhnced if n intervention lso ddresses risk nd protective fctors in n dolescent s socil context. bstrct CONTEXT: The promotion of optimism hs been widely dvocted for children nd dolescents, but epidemiologic dt to support this pproch re scnt. METHODS: This ws 3-wve longitudinl study of helth nd socil development in younger dolescents from 3 Austrlin sttes. The 5634 student prticipnts, initilly ged 12 to 14 yers, were ssessed for optimistic thinking style, emotionl problems, substnce use, nd ntisocil behviors. RESULTS: Cross-sectionl ssocitions between optimism nd ech of the study outcomes were strongly protective but tended to differ ccording to gender in extent. In prospective nlyses of the onset of new cses of ech study outcome, protective ssocitions were weker. Those in the highest optimism qurtile hd risks for depressive symptoms tht were reduced by lmost hlf (odds rtio: 0.54 [95% confidence intervl: 0.42 0.70]) compred with those in the lowest ctegory. No effect ws seen in prevention of nxiety symptoms fter djustment for other spects of psychologicl style. In predicting the onset of hevy substnce use nd ntisocil behvior, high optimism hd modest protective effects. CONCLUSIONS: Optimistic thinking style is somewht protective ginst dolescent helth risks; the clerest effects re seen ginst depressive symptoms. Promoting optimism long with other spects of psychologicl nd emotionl style hs role in mentl helth promotion tht is likely to be enhnced if n intervention lso ddresses risk nd protective fctors in n dolescent s socil context. Peditrics 2011;127:000 PEDIATRICS Volume 127, Number 2, Februry 2011 1 Downloded from www.ppublictions.org/news by guest on Mrch 5, 2019
Optimism hs long been linked to cdemic, occuptionl, nd politicl success. 1,2 Its role in both mentl nd physicl helth hs lso ttrcted ttention. 3 In lter life it predicts good globl helth sttus, 4 lower rtes llcuse mortlity, 5,6 crdiovsculr deths, 7,8 nd depressive symptoms. 9 Vrious explntions hve been put forwrd. Its potentil role in buffering ginst the negtive emotionl consequence of socil dversity 1,10 hs led to view of optimism s n index of resilience. 11 Optimists my lso be more likely to dopt lifestyles tht promote helth or fvorble progression through disese. 12,13 Were such n ssocition with helthy lifestyle to begin erly in life, the cumultive effects might be profound. 10,14 For these resons, optimism hs often fetured prominently in mentl helth promotion in children nd dolescents. 15,16 Yet, filure to find consistent positive outcomes from recent interventions rises question bout how protective it my be in younger people. 15 17 Indeed, optimism hs been linked to some helth hzrds in dolescents, becuse optimists my not void risks tht they perceive to be unlikely to ffect them. 18 Unwrrnted optimism bout the personl helth risks hs been found in young smokers 19 nd drug users. 20 To dte, most studies tht hve exmined the effects of optimism on dolescent helth hve been cross-sectionl nd generlly considered single helth outcomes. 21 One recent prospective study of more extreme pessimism found tht perceived likelihood of erly deth predicted rnge of helth nd behviorl outcomes in young dulthood. 22 In our study, we prospectively exmined the ssocition between optimistic thinking styles nd the commoner emotionl problems nd helth risk behviors of younger dolescents. METHODS Ethics pprovls were obtined from the pproprite bodies in ech stte. Prticiption required written prentl consent. The study ws designed s cluster-rndomized tril of n intervention to improve individul coping nd socil skills. 23 The outcomes were reported recently with no difference in outcomes between the intervention nd control groups. 24 Procedure nd Smple Schools were recruited through n expression-of-interest process; n invittion ws extended to ech school in ech stte with t lest 100 8thgrde students. The 105 schools tht responded were strtified ccording to sector (government, Ctholic, nd independent) nd whether they were metropolitn. Twenty-five school pirs (50 schools in totl) were selected with mtching on socioeconomic sttus, enrollment size, nd eduction sector to enroll typicl rnge of schools in ech stte. All but 2 of the selected schools prticipted. Prticipting schools were in Queenslnd (n 18), South Austrli (n 16), nd Victori (n 16). Of the 50 schools, 42 were mixed-gender nd 8 were single-gender schools. Thirty-four schools (68%) were government funded nd 16 (32%) were privtely funded. Thirty-eight (76%) were metropolitn nd 12 (24%) were nonmetropolitn. Students completed the bseline ssessment in term 2 (My through June) 2003. The cdemic yer in Austrli begins in Februry nd ends in erly December. Follow-up ssessments took plce in ll schools in term 4 ech yer to follow the cycle of curriculum intervention (October through December) in 2004 nd 2005 when the students were in 9th nd 10th grde. Mesures Depressive Symptoms The 20-item Centre for Epidemiologic Studies Depression Scle (CES-D) ws used to ssess depressive symptoms. 25,26 Respondents rted their experience of ech symptom in the previous week on 4-point scle from rrely or none of the time ( 1 dy) to most or ll of the time (5 7 dys). Summed scores could rnge from 0 to 60; higher scores indicted more depressive symptoms. A cutoff score of 16 identified dolescents with mild or greter levels of depression. 27 Anxiety Symptoms A short 8-item scle dpted from the Spence Children s Anxiety Scle 28 ws used to ssess nxiety symptoms. Prticipnts were sked to indicte how often they experienced ech symptom by using 4-point scle (never, sometimes, often, or lwys). Totl scores rnged from 0 to 24; higher scores reflected higher levels of nxiety symptoms. Internl consistency ws high (Cronbch s 0.89). A cutoff point of 12 ws used to define high nxiety, which reflected those who scored in the top 15% t bseline ssessment. Substnce Use Substnce use ws mesured by selfreport of frequency of tobcco, lcohol, nd cnnbis use. Tobcco use in the previous week ws defined on 3 levels: nonsmokers nd ex-smokers who reported hving stopped for more thn 1 month; current smoking on 3 dys in the previous week; nd smoking on t lest 3 dys in the previous week. Alcohol use ws defined on 3 levels: no use in the previous week; drinking 3 dys in the pst week; nd drinking 3 dys in the previous week nd/or binge-drinking ( 5 stndrd drinks in row) on t lest 2 occsions in the previous 2 weeks. Cnnbis use ws defined s no use in the previous 6 2 PATTON et l Downloded from www.ppublictions.org/news by guest on Mrch 5, 2019
ARTICLES months; use t lest once in the previous 6 months but less thn weekly; or t lest weekly use. Any substnce use ws defined s hving used lcohol, tobcco, or cnnbis t lest t the intermedite level. Hevy substnce use ws s ny lcohol, tobcco, or cnnbis use t the highest level defined bove. Antisocil Behvior Antisocil behvior ws ssessed with 5 items from the Self-reported Erly Delinquency Scle tht cover getting into physicl fights, driving cr without permission, running wy from home, being suspended from school, nd doing grffiti in public plce in the previous 6 months. 29 Prticipnts responded on 3-point scle (not t ll, once, or more thn once), which provided totl score tht rnged from 0 to 10. Internl consistency ws cceptble (Cronbch s 0.79). Any ntisocil behvior referred to t lest 1 instnce of ny of the 5 behviors in the previous 6 months, nd frequent referred to either 2 or more instnces of the sme behvior or 2 or more different ntisocil behviors on t lest 1 occsion. Optimistic Thinking Style A 12-item scle ws developed to ssess students thinking style. The structure of the scle ws designed to reflect Beck s cognitive trid, which reltes to positive thoughts bout the self, the world, nd the future. 30 The mesure incorported tendency for the children to perceive themselves in fvorble light. 1 Prticipnts rted the experience of specific positive thoughts over the previous week. Explortory fctor nlysis using bseline dt reveled single fctor (eigenvlue 7.11) tht explined 59.25% of the vrince with ll items loded strongly ( 0.6). Cronbch s for the totl scle ws 0.94. In this study, optimistic thinking style ws ctegorized on 4 levels to llow evlution of ny dose-response reltionships in ssocitions: very low (0 17), low (18 23), high (24 29), nd very high (30 36), which corresponded to the qurtilevlue cutoff points t bseline. This ctegoriztion llowed explortion of ssocitions with the study outcomes without the ssumption of linerity. Interpersonl Competence Interpersonl competence ws ssessed through modified version of the Adolescent Interpersonl Competence Questionnire. 31 The modified scle included 15 items tht ssessed inititing reltionships, providing emotionl support, sserting influence, selfdisclosure, nd conflict resolution. Cronbch s for the scle ws 0.89. Scores were coded on 3 levels: low ( 25), intermedite (25 to 32), nd high (32 45), which corresponded to the tertile scores t bseline. Negtive Coping Style Coping style ws mesured by using the Coping Actions Scle, which consists of sttements tht ssess coping skills tken from the Socil Problem Solving Inventory-Revised 32 nd the Selfreporting Coping Scle. 33 The negtivecoping-strtegies dimension included 8 items tht ssessed the tendency to rect to or del with problems in n voidnt or unconstructive mnner. Cronbch s for the scle ws 0.88. These scores were coded on 3 levels: low ( 10), intermedite (10 to 15), nd high (15 32), which corresponded to the tertile scores t bseline. Life events over the previous 12 months were evluted by using n dpttion of the List of Thretening Experiences Questionnire, brief mesure of 12 events with considerble long-term contextul thret. 34,35 The events included serious illness or injury in prticipnt or close reltive, deth of close friend or reltive, brek-up of stedy reltionship, suspension from school, problems with the police, serious conflict with close friend, nd mjor finncil problems for the prticipnt or fmily. The vrible ws ctegorized s none, 1 or 2, or more thn 2 life events. Anlysis For the summry sttistics for the overll prevlence rtes of depression, nxiety, substnce use, nd ntisocil behvior, we took into ccount the clustering of students in schools by using robust SEs. The reltionship between concurrent optimism nd study outcomes cross the 3 wves were nlyzed by using logistic generlized estimting equtions (GEE). 36 Models were djusted for wve, demogrphic vribles of interest (child s ge, gender, country of birth, prentl mritl sttus), intervention sttus, nd other fctors known to be relted to the outcomes (negtive coping style, interpersonl competence, nd recent life events). Interctions between optimism nd gender, optimism nd wve, nd optimism nd life events were tested by using the Wld test. Gender ws focus for interction testing, becuse gender differences in the prevlence of the outcomes rised possibility of different cusl processes. 37 Trnsitionl logistic GEE regression models were used to exmine ssocitions between optimism 12 months erlier nd the onset of the outcomes studied. Those who hd scored bove the relevnt cutoff point t the previous wve were excluded from these models. Initil models controlled for wve, demogrphic vribles of interest, intervention sttus, nd, where pproprite, outcome mesures t the previous wve. Fully djusted models lso controlled for previous negtive coping style nd interpersonl competence nd concurrent recent life events. Interctions between optimism nd gender, life events, nd intervention sttus were tested. For depressive PEDIATRICS Volume 127, Number 2, Februry 2011 3 Downloded from www.ppublictions.org/news by guest on Mrch 5, 2019
symptoms, the popultion-preventble frction ws clculted to estimte proportion of cses of depression potentilly prevented by hving n optimism level higher thn the lowest ctegory. 38 All estimtes of prevlence nd ssocition re presented with 95% confidence intervls (CIs). Dt nlysis ws undertken by using Stt 11. 39 RESULTS Within the individul schools, 64% of students in the trgeted yer level completed the survey t bseline. Of these 5634 dolescents, 92% hd complete demogrphic vribles nd t lest 1 outcome of interest t wve 1, 85% t wve 2, nd 75% t wve 3. Thirty-six percent of the smple ws from the stte of Victori, 35% from Queenslnd, nd 29% from South Austrli. The men ge of the smple t the outset in 2003 ws 13.1 yers (SD: 0.5). The smple ws 53% femle. Seventy-five percent of the smple cme from fmilies with both prents living t home, nd 93% of the smple ws Austrlin born, which is consistent with ntionl popultion estimtes. 40 Bseline rtes of high depressive symptoms (Centre for Epidemiologic Studies Depression Scle scores 16) were 29% in boys nd 39% in girls. Similr rtes were found t wves 2 nd 3 for boys (27% nd 28%, respectively) nd girls (43% nd 43%, respectively). High nxiety symptoms ( 12) were reported by 11% of the boys nd 19% of the girls t bseline. These rtes remined consistent over subsequent wves. Rtes of ny substnce use incresed cross the wves for boys (21%, 35%, nd 45%) nd girls (17%, 30%, nd 41%). Rtes of hevy substnce use lso incresed for boys (8%, 17%, nd 22%) nd girls (5%, 12%, nd 18%). Antisocil behvior ws reported by just less thn hlf of the boys nd pproximtely one-third of the girls t ech wve (boys: 45%, 45%, nd 41%; girls: 29%, 32%, nd 28%) t ech wve. At wve 1, 19% of boys nd 11% of girls reported frequent ntisocil behvior; these rtes were similr for the succeeding wves. At bseline, 20% of boys hd very low optimism, 25% hd low optimism, 29% hd high optimism, nd 25% hd very high optimism, wheres 25% of girls hd ech level of optimism. For boys these rtes remined similr cross wves, but for girls the rtes of low optimism incresed (wve 3: 30% very low optimism, 26% low optimism, 24% high optimism, 20% very high optimism). Cross-sectionl Associtions of Optimism With Helth Risks Tble 1 shows the prevlence of the study outcomes ccording to gender cross the 4 optimism levels t bseline. At bseline, dolescents who reported higher levels of optimism tended to report lower levels of depression, nxiety, substnce use, nd ntisocil behvior. This pttern ws repeted t wves 2 nd 3 (dt not shown). Tble 2 shows concurrent ssocitions between study outcomes nd optimism fter controlling for demogrphic nd other fctors known to be relted to the outcomes. Significnt interctions were found between optimism nd gender for ll the outcomes except nxiety. Optimism hd strong protective ssocition for depression, which differed between boys nd girls. Compred with girls with very low levels of optimism, boys with the sme level of optimism were pproximtely hlf s likely to be depressed. The risk of depression for boys nd girls decresed s the level of optimism incresed, but boys hd lower risk of depression for ech level of optimism. Optimism hd protective ssocition with nxiety. Boys were less likely to TABLE 1 Prevlence of Study Outcomes t Wve 1 in 5634 8th-Grde (12- to 13-Yer-Old) Secondry School Students Ctegorized According to Gender nd Concurrent Optimism Level Optimism Level Emotionl Problems Substnce Use Antisocil Behvior Depression Anxiety Any b Hevy c Any Frequent Boys (N 2635) Very low 59 (53 64) 22 (18 26) 30 (25 35) 13 (9 17) 62 (58 67) 32 (27 36) Low 31 (27 36) 11 (8 14) 17 (13 21) 7 (5 9) 49 (45 53) 19 (15 23) High 18 (15 21) 7 (4 10) 10 (8 13) 6 (4 8) 40 (36 44) 15 (12 18) Very high 15 (12 17) 6 (4 8) 10 (7 12) 5 (3 8) 34 (30 39) 12 (8 16) Overll 29 (26 32) 11 (9 13) 21 (18 24) 8 (6 10) 45 (43 48) 19 (17 21) Girls (N 2972) Very low 76 (72 80) 36 (32 40) 32 (27 37) 11 (8 14) 49 (44 54) 23 (19 26) Low 43 (39 47) 19 (15 22) 23 (19 27) 5 (3 8) 32 (28 35) 10 (8 12) High 25 (22 28) 13 (10 15) 17 (15 20) 2 (1 3) 22 (18 25) 7 (5 9) Very high 16 (12 19) 9 (7 12) 15 (12 18) 2 (1 4) 17 (14 20) 5 (3 7) Overll 39 (37 42) 19 (17 20) 17 (14 19) 5 (4 7) 29 (27 32) 11 (9 13) Prevlence estimtes re presented with 95% CIs with djustment for clustering within schools. b Any substnce use ws defined s using tobcco or lcohol in the previous week or cnnbis in the previous 6 months. c Hevy substnce use refers to using tobcco t lest most dys, binge-drinking ( 5 units in row) in the previous week, or weekly cnnbis use. 4 PATTON et l Downloded from www.ppublictions.org/news by guest on Mrch 5, 2019
ARTICLES TABLE 2 Cross-sectionl Associtions of Study Outcomes With Optimism Across 3 Annul Wves of Dt Collection in 5607 Secondry School Students Initilly in 8th Grde Emotionl Problems, OR (95% CI) Substnce Use, OR (95% CI) Antisocil Behvior, OR (95% CI) Depression Anxiety b Any c Hevy d Any e Frequent f Optimism level Very low Girls 1.0 1.0 1.0 1.0 1.0 1.0 Boys 0.44 (0.37 0.54) 1.22 (1.00 1.49) 1.35 (1.05 1.74) 1.66 (1.41 1.95) 1.65 (1.35 2.02) Low Girls 0.28 (0.24 0.33) 0.67 (0.56 0.80) 0.56 (0.45 0.70) 0.49 (0.51 0.67) 0.47 (0.40 0.56) 0.57 (0.51 0.64) Boys 0.17 (0.14 0.21) 0.97 (0.80 1.19) 0.90 (0.74 1.10) 1.36 (1.15 1.61) 1.09 (0.90 1.33) High Girls 0.13 (0.11 0.15) 0.53 (0.44 0.64) 0.46 (0.37 0.57) 0.35 (0.30 0.40) 0.32 (0.26 0.40) 0.41 (0.36 0.47) Boys 0.07 (0.06 0.09) 0.74 (0.63 0.86) 0.73 (0.60 0.89) 1.06 (0.91 1.23) 0.88 (0.76 1.02) Very high Girls 0.07 (0.05 0.09) 0.37 (0.30 0.46) 0.32 (0.22 0.46) 0.34 (0.29 0.40) 0.32 (0.23 0.45) 0.41 (0.33 0.50) Boys 0.06 (0.05 0.08) 0.70 (0.57 0.84) 0.77 (0.60 0.98) 0.86 (0.73 1.00) 0.91 (0.73 1.13) Associtions were estimted by using GEE with djustment for ge, gender, country of birth, prentl mritl sttus, recent life events, negtive coping style, nd interpersonl competence. OR indictes odds rtio. Sttisticl significnce of min effects for optimism, gender, nd their interction ws.001 for ech. b No interction between optimism nd gender. c Sttisticl significnce of min effects of optimism, gender, nd their interction ws.001,.05, nd.004, respectively. d Sttisticl significnce of min effects of optimism, gender, nd their interction ws.001,.02, nd.04, respectively. e Sttisticl significnce of min effects for optimism, gender, nd their interction ws ll.001. f Sttisticl significnce of min effects of optimism, gender nd their interction ws.001,.001, nd.004, respectively. report nxiety, but ssocition with optimism did not differ with gender. Optimism, gender, nd the interction between optimism nd gender were ll found to be relted to ny substnce use. The risk of ny substnce use declines s optimism increses; girls were less likely to use thn boys, especilly for very high levels of optimism. A similr result ws found for hevy substnce use. For ny ntisocil behvior, there gin were ssocitions with optimism, gender, nd the interction between them. Boys were more likely to report ntisocil behviors. The risk of ntisocil behvior decresed s optimism incresed; the reduction in risk differed slightly between boys nd girls. Similr effects were lso found for frequent ntisocil behvior. These reltionships did not vry over time. Optimistic Thinking Style nd the Onset of New Problems We conducted prospective nlyses to exmine optimism s predictor of the onset of emotionl problems, substnce use, nd ntisocil behviors 12 months lter. Depressive nd Anxiety Symptoms In predicting new depressive symptoms, optimism hd protective effects in the prtilly djusted model, nd there ws trend to incresing protection with higher levels of optimism (Tble 3). In the fully djusted model, optimism hd similr protective effect, nd there ws some evidence for possible vrition between genders; boys hd lower levels of risk of depression t ny given level of optimism. Multiple life events were ssocited with n lmost fourfold increse in risks for depressive symptoms. Estimtes of the popultion-preventble frction suggested tht up to 32% more incident cses of depressive symptoms would hve occurred if those with higher levels of optimism hd not been exposed to its protective effects. The risks for nxiety were lower in ll 3 higher-optimism groups compred with the very-lowoptimism group in the prtilly djusted model. In the fully djusted model, the protective effect of optimism ws much reduced fter djustment for interpersonl competence, negtive coping style, nd reported life events. There ws no interction between optimism nd gender in either the prtilly or fully djusted model. There ws no interction found between optimism nd life events in the prediction of either new depressive symptoms (P.74) or nxiety symptoms (P.51). Substnce Use Incident substnce use ws defined on 2 different levels: ny nd hevy substnce use (Tble 4). Level of optimism ws protective ginst ny substnce use, nd its effect vried between boys nd girls. In the prtilly djusted model both boys nd girls with the high nd very high levels of optimism hd reduced risk of new substnce use. In the fully djusted model, the protective effect of high nd very high levels of optimism remined only for girls. Those with the highest level of optimism, compred with those with the lowest level, hd lmost hlf the risk for the onset of hevy substnce use in the prtilly djusted model. PEDIATRICS Volume 127, Number 2, Februry 2011 5 Downloded from www.ppublictions.org/news by guest on Mrch 5, 2019
TABLE 3 Prospective Associtions Between Optimism nd the Onset of Emotionl Problems 12 Months Lter Depressive Symptoms (N 3307), OR (95% CI) Prtil Adjustment,b This protective effect ws substntilly reduced in the fully djusted model fter djustment for interpersonl competence. There ws no interction between optimism nd gender in either model. Antisocil Behvior Antisocil behvior ws defined on 2 levels (ny nd frequent); with the Full Adjustment c,d Anxiety Symptoms (N 4191), OR (95% CI) Prtil Adjustment,e Full Adjustment c,f Optimism level Very low 1.00 1.00 1.00 1.00 Low 0.82 (0.65 1.03) 0.78 (0.62 0.98) 0.62 (0.49 0.79) 0.71 (0.56 0.92) High 0.68 (0.54 0.86) 0.65 (0.51 0.83) 0.66 (0.54 0.81) 0.82 (0.65 1.04) Very high 0.55 (0.43 0.72) 0.54 (0.42 0.70) 0.69 (0.55 0.89) 0.92 (0.70 1.21) Life events None NA 1.00 NA 1.00 1 2 1.66 (1.41 1.94) 1.44 (1.14 1.81) 3 4.35 (3.70 5.11) 3.87 (2.94 5.11) Negtive problem-solving Low 1.0 1.0 Intermedite NA 1.17 (1.00 1.36) NA 1.24 (1.01 1.52) High 1.51 (1.24 1.84) 1.74 (1.40 2.15) Interpersonl competence Low NA 1.00 NA 1.00 Intermedite 0.91 (0.74 1.10) 0.76 (0.59 0.97) High 0.85 (0.65 1.11) 0.70 (0.53 0.92) Odds rtios (95% CIs) were estimted by using GEE. OR indictes odds rtio; NA, not pplicble. Prtilly djusted models included wve, intervention, ge, prentl mritl sttus, nd erlier subthreshold symptoms. b Sttisticl significnce of min effects of optimism, gender, nd their interction ws.001,.03, nd.12 respectively. c Fully djusted models included life events, negtive coping style, nd interpersonl competence t the previous wve. d Sttisticl significnce of min effects of optimism, gender, nd their interction ws.002,.07, nd.05. e Sttisticl significnce of min effects of optimism nd gender in the prtilly djusted model ws.001 for ech. f Sttisticl significnce of min effects of optimism nd gender in the fully djusted model ws.02 nd.001 respectively. models we exmined the trnsition to frequent ntisocil behvior, djusted for ny ntisocil behvior t the previous wve (Tble 5). Protective effects of optimism in predicting ny nd frequent ntisocil behvior were reduced fter fully djusting for negtive life events, negtive coping style, nd interpersonl competence; ech covrite reduced the ssocition. No interction ws found between gender nd level of optimism in predicting either outcome. DISCUSSION In the cross-sectionl nlysis, optimistic thinking style hd strong protective ssocitions with dolescent emotionl nd conduct problems nd substnce use. In prospective nlysis these ssocitions were weker. In predicting new depressive symptoms, the highest optimism ctegory hd lmost hlf the risk of the lowest optimism ctegory. Without the protective effect of higher levels of optimism, incidence rtes for depressive symptoms might hve been 32% higher thn those observed. For nxiety symptoms, modest reductions in risk tht lrgely disppered fter djustment for other psychologicl ttributes were found. In predicting the onset of substnce use, optimistic thinking style hd protective effect for girls but not boys. For frequent ntisocil behvior nd hevier substnce use, high optimism hd modest protective ssocitions in both genders. By lter life, optimism seems stble, 35 lthough socil nd culturl context in erly life re importnt influences. 41 In our study, optimism seemed less st- TABLE 4 Prospective Assocition Between Optimism nd the Onset of New Substnce Use Optimism Level Any Substnce Use (N 4040), OR (95% CI) Hevy Substnce Use (N 4636), OR (95% CI) Prtil Adjustment,b Full Adjustment c,d Prtil Adjustment e,f Full Girls Boys Girls Boys Adjustment e,g Very low 1.00 0.80 (0.63 1.02) 1.00 0.85 (0.66 1.09) 1.00 1.00 Low 0.78 (0.63 0.97) 0.88 (0.67 1.15) 0.80 (0.64 1.00) 1.00 (0.73 1.35) 0.87 (0.71 1.06) 0.94 (0.77 1.17) High 0.66 (0.52 0.84) 0.77 (0.59 1.00) 0.69 (0.54 0.88) 0.91 (0.69 1.21) 0.69 (0.56 0.84) 0.79 (0.64 0.97) Very high 0.45 (0.35 0.58) 0.71 (0.54 0.93) 0.50 (0.38 0.66) 0.81 (0.60 1.08) 0.59 (0.48 0.72) 0.74 (0.59 0.93) Odds rtios (95% CIs) were estimted by using GEE. OR indictes odds rtio. Prtil djustment for wve, intervention sttus, ge, country of birth, nd prentl mritl sttus. b Sttisticl significnce of min effects of optimism, gender, nd their interction in the prtilly djusted model ws.001,.07, nd.001, respectively. c Fully djusted models lso djusted for recent life events, negtive coping style, nd interpersonl competence t the previous wve. d Sttisticl significnce of min effects of optimism, gender, nd their interction in the fully djusted model ws.001,.20, nd.002, respectively. e Prtil djustment for wve, intervention sttus, ge, country of birth, prentl mritl sttus, nd ny substnce use t the previous wve. f Sttisticl significnce of min effects of optimism nd gender in the prtilly djusted model ws.001 for ech. g Sttisticl significnce of min effects of optimism nd gender in the fully djusted model ws.02 nd.001, respectively. 6 PATTON et l Downloded from www.ppublictions.org/news by guest on Mrch 5, 2019
ARTICLES TABLE 5 Prospective Assocition Between Optimism nd the Onset of Antisocil Behvior Optimism Level Any Antisocil Behvior (N 3189), OR (95% CI) Frequent Antisocil Behvior (N 4268), OR (95% CI) Prtil Adjustment,b ble becuse young girls showed decline over the 2 yers of follow-up. Whether this decrese reflects pubertl development or the influence of socil context in erly dolescence is uncertin. This study hd notble strengths in size, high retention rtes, nd coverge of relevnt ge group for the onset of emotionl nd behviorl problems. However, limittions should be noted. The smple ws chosen for mounting cluster-rndomized tril. In prt, becuse ctive consent ws required, initil prticiption rtes were only fir. It is possible, therefore, tht the distribution of optimism nd the study outcomes my differ somewht from truly representtive smple, which could mke modest difference in the ssocitions found but seems unlikely to ccount for the min findings. The mesure of optimistic thinking style ws broder thn some used in erlier studies in tht it incorported current ttitudes to the self nd others together with future-oriented elements. This scle hd excellent internl consistency, nd subnlyses in which only the future-oriented items Full Adjustment c,d Prtil Adjustment e,f Full Adjustment c,g Very low 1.00 1.00 1.00 1.00 Low 0.88 (0.72 1.07) 1.01 (0.81 1.25) 0.73 (0.60 0.89) 0.84 (0.70 1.01) High 0.72 (0.60 0.87) 0.86 (0.70 1.05) 0.68 (0.54 0.85) 0.83 (0.64 1.06) Very high 0.58 (0.46 0.74) 0.74 (0.57 0.97) 0.60 (0.47 0.76) 0.74 (0.55 0.98) Odds rtios (95% CIs) were estimted by using GEE. OR indictes odds rtio. Prtil djustment for wve, intervention sttus, ge, country of birth, nd prentl mritl sttus. e Prtil djustment for wve, intervention sttus, ge, country of birth, prentl mritl sttus, nd ny ntisocil behvior t the previous wve. c Fully djusted models include life events, negtive coping style, nd interpersonl competence t the previous wve. b Sttisticl significnce of min effects of optimism nd gender in the prtilly djusted model ws.001 for ech. d Sttisticl significnce of min effects of optimism nd gender in the fully djusted model ws.05 nd.001, respectively. f Sttisticl significnce of min effects of optimism nd gender in the prtilly djusted model ws.001 for ech. g Sttisticl significnce of min effects of optimism nd gender in the fully djusted model ws.12 nd.001, respectively. were used produced similr findings to those of the complete scle. The explntory style of optimists my be responsible for the protective effect ginst depressive symptoms. An optimist generlly sserts tht the good things hppening to me will continue to hppen in the future nd they re of my doing; bd things cn hppen occsionlly by chnce but re unlikely to recur. 42 A possibility tht such ttitudes my buffer ginst socil dversity ws not borne out in tht optimism s effects were independent of life events. 43 Optimism neither predicted life events nor hd ny interction with life events in predicting depressive symptoms or other study outcomes. 44 The results of other work hve suggested tht optimists tend to use more problem-focused coping strtegies thn pessimists. 45 However, lesser tendency to dopt negtive coping strtegies did not explin optimism s protective effects ginst depression. A gender difference in the effects of optimism ws pprent in regrd to substnce use, for which optimism ws protective for girls but did little to reduce risks for boys. Erlier reports hve suggested tht dispositionl optimism my led to n underestimtion of both lter helth risks nd person s cpcity to control prticulr behvior. 46 Why this risk reltionship should differ for boys nd girls is less cler but could be relted to evolutionry pressures on dolescent boys to tke risks to improve their socil sttus. 47 CONCLUSIONS The presence of protective effects cross rnge of dolescent helth outcomes provides support for promoting n optimistic thinking style in dolescents. Yet, focus on optimism lone, without ddressing other spects of cognitive, interpersonl, nd emotionl style, seems unlikely to hve lrge effects. 15 17 Preventive interventions round dolescent mentl helth nd behvior should lso ddress other spects of psychologicl functioning nd, given the lrge effects of life events in this nd other studies, lso consider the socil context in the intervention strtegy. 48,49 ACKNOWLEDGMENTS Funding for the study design nd dt collection ws provided by beyondblue (the Austrlin ntionl depression inititive) nd the Ntionl Helth nd Medicl Reserch Council. Dr Ptton is supported by Senior Principl Reserch Fellowship from the Ntionl Helth nd Medicl Reserch Council. The reserch ws principlly supported with funding from beyondblue with dditionl support in the preprtion of the rticle from Austrlin Helth Mngement. PEDIATRICS Volume 127, Number 2, Februry 2011 7 Downloded from www.ppublictions.org/news by guest on Mrch 5, 2019
REFERENCES 1. Tylor SE, Brown JD. Illusion nd well-being: socil psychologicl perspective on mentl helth. Psychol Bull. 1988;103(2): 193 210 2. Peterson C. The future of optimism. Am Psychol. 2000;55(1):44 55 3. Scheier MF, Crver CS. Optimism, coping, nd helth: ssessment nd implictions of generlized outcome expectncies. Helth Psychol. 1985;4(3):219 247 4. Mrut T, Collign R, Mlinchoc M, Offord K. Optimism-pessimism ssessed in the 1960s nd self-reported helth sttus 30 yers lter. Myo Clin Proc. 2002;77(8):748 753 5. Gilty E, Geleijnse JM, Zitmn F, Hoekstr T, Schouten EG. Dispositionl optimism nd ll-cuse nd crdiovsculr mortlity in prospective cohort of elderly Dutch men nd women. Arch Gen Psychitry. 2004; 61(11):1126 1135 6. Dnner DD, Snowdon DA, Friesen WV. Positive emotions in erly life nd longevity: findings from the Nun Study. J Pers Soc Psychol. 2001;80(5):804 813 7. Kubznsky L, Sprrow D, Vokons P, Kwchi I. Is the glss hlf empty or hlf full? A prospective study of optimism nd coronry hert disese in the Normtive Aging Study. Psychosom Med. 2001;63(6):910 916 8. Nvrro VM, Cstellno JM, Grci-Glino D, Ten-Sempere M. Neuroendocrine fctors in the initition of puberty: the emergent role of kisspeptin. Rev Endocr Metb Disord. 2007;8(1):11 20 9. Gilty E, Zitmn F, Kromhout D. Dispositionl optimism nd the risk of depressive symptoms during 15 yers of follow-up: the Zutphen Elderly Study. J Affect Dis. 2006;91(1): 45 52 10. Adler NE. Helth disprities: wht s optimism got to do with it? J Adolesc Helth. 2007;40(2):106 107 11. Gllo LC, Mtthews KA. Understnding the ssocitions between socioeconomic sttus nd physicl helth: do negtive emotions ply role? Psychol Bull. 2003;129(1): 10 51 12. Ironson G, Hywrd H. Do positive psychosocil fctors predict disese progression in HIV-1? A review of the evidence. Psychosom Med. 2008;70(5):546 554 13. Milm JE, Richrdson JL, Mrks G, Kemper CA, McCuthchn AJ. The roles of dispositionl optimism nd pessimism in HIV disese progression. Psychol Helth. 2004; 19(2):167 181 14. Gilty E, Geleijnse JM, Zitmn F, Buijsse B, Kromhout D. Lifestyle nd dietry correltes of dispositionl optimism in men: the Zutphen Elderly Study. J Psychosom Res. 2007;63(5):483 490 15. Jycox LH, Reivich KJ, Gillhm J, Seligmn ME. Prevention of depressive symptoms in school children. Behv Res Ther. 1994;32(8): 801 816 16. Roberts C, Kne R, Thomson H, Bishop B, Hrt B. The prevention of depressive symptoms in rurl school children: rndomized controlled tril. J Consult Clin Psychol. 2003; 71(3):622 628 17. Gillhm JE, Reivich KJ, Freres DR, et l. School-bsed prevention of depressive symptoms: rndomized controlled study of the effectiveness nd specificity of the Penn Resiliency Progrm. J Consult Clin Psychol. 2007;75(1):9 19 18. Weinstein ND. Optimistic bises bout personl risks. Science. 1989;246(4935): 1232 1233 19. Reppucci JD, Revenson TA, Aber M, Reppucci ND. Unrelistic optimism mong dolescent smokers nd non-smokers. J Prim Prev. 1991;11(3):227 236 20. Schmid H. Swiss dolescent drug users nd nonusers optimism bout their future. J Appl Soc Psychol. 1998;28(20):1889 1902 21. Jmieson P. Unrelistic ftlism in US youth ges 14 to 22: prevlence nd chrcteristics. J Adolesc Helth. 2007;42(2):154 160 22. Borowsky IW, Irelnd M, Resnick MD. Helth sttus nd behviorl outcomes for youth who nticipte high likelihood of deth. Peditrics. 2009;124(1). Avilble t: www. peditrics.org/cgi/content/full/124/1/e81 23. beyondblue. Schools reserch inititive. Avilble t: www.beyondblue.org.u/index. spx?link_id 4.64. Accessed December 7, 2010 24. Swyer MG, Pfeiffer S, Spence SH, et l. School-bsed prevention of depression: rndomised controlled study of the beyondblue schools reserch inititive. J Child Psychol Psychitry. 2010;51(2):199 209 25. Rdloff LS. The CES-D scle: self-report depression scle for reserch in the generl popultion. Appl Psychol Mes. 1977;1(3): 385 401 26. Rdloff LS. The use of the Centre for Epidemiologic Studies Depression Scle in dolescents nd young dults. J Youth Adolesc. 1991;20(2):149 166 27. Brnes GE, Prosen H. Depression in Cndin generl prctice ttenders. Cn J Psychitry. 1984;29(1):2 10 28. Spence SH, Brrett P, Turner CM. Psychometric properties of the Spence Children s Anxiety Scle with young dolescents. J Anxiety Disord. 2008;17(6):605 625 29. Moffitt TE, Silv PA. Self-reported delinquency: results from n instrument for New Zelnd. AustNZJCriminol. 1988;21(Dec): 227 240 30. Beck AT. Cognitive Therpy nd Emotionl Disorders. New York, NY: Interntionl Universities Press; 1976 31. Buhrmester D. Intimcy of friendship, interpersonl competence, nd djustment during predolescence nd dolescence. Child Dev. 1990;61(4):1101 1111 32. D Zurill TJ, Nezu AM. Mnul for the Socil Problem-Solving Inventory Revised (SPRI-R). North Tonwnd, NY: Multi-Helth Systems; 2009 33. Cusey DL, Dubow EG. Development of selfreport coping mesure for elementry school children. J Clin Child Adolesc Psychol. 1992;21(1):47 59 34. Brugh TS, Crgg D. The List of Thretening Experiences: the relibility nd vlidity of brief life events questionnire. Act Psychitr Scnd. 1990;82(1):77 81 35. Brugh TS, Bebbington PE, Tennnt C, Hurry J. The List of Thretening Experiences: subset of 12 life event ctegories with considerble long term contextul thret. Psychol Med. 1985;15(1):189 194 36. Ling KY, Zeger SL. Longitudinl dt nlysis using generlized liner models. Biometrik. 1986;73(1):13 22 37. Gll SL, Abbott-Chpmn J, Ptton GC, Dwyer T, Venn A. Intergenertionl eductionl mobility is ssocited with crdiovsculr disese risk behviours in cohort of young Austrlin dults: the Childhood Determinnts of Adult Helth (CDAH) Study. BMC Public Helth. 2010;10:55 38. Rockhill B, Newmn B, Weinberg C. Use nd misuse of popultion ttributble frctions [published correction ppers in Am J Public Helth. 2008;98(12):2119]. Am J Public Helth. 1998;88(1):15 19 39. Stt [computer progrm]. Relese 7.0. College Sttion, TX: Stt Corp; 2001 40. Austrlin Bureu of Sttistics. Fmily Chrcteristics. Cnberr, Austrli: Austrlin Bureu of Sttistics; 2003 41. Dukntite D, Bergmn L. Childhood roots of women s subjective well-being: the role of optimism. Eur Psychol. 2005;10(4): 287 297 42. Villnt GE. Mentl helth. Am J Psychitry. 2003;160(8):1373 1384 43. vn der Velden P, Kleber R, Fournier M, Grieveink L, Drogendijk A, Gersons B. The 8 PATTON et l Downloded from www.ppublictions.org/news by guest on Mrch 5, 2019
ARTICLES ssocition between dispositionl optimism nd mentl helth problems mong disster victims nd comprison group: prospective study. J Affect Dis. 2007; 102(1 3):35 45 44. Li JCL. Dispositionl optimism buffers the impct of dily hssles on mentl helth in Chinese dolescents. Pers Individ Diff. 2009; 47(4):247 249 45. Scheier M, Crver C, Bridges M. Distinguising optimism from neuroticism (nd trit nxiety, self-mstery, nd selfesteem): reevlution of the life orienttion test. J Pers Soc Psychol. 1994;67(6): 1063 1078 46. Howrd T, Glenn A. The costs nd benefits of optimistic explntions nd dispositionl optimism. J Pers. 1987;55(2):376 393 47. Dhl RE. Biologicl, developmentl, nd neurobiologicl fctors relevnt to dolescent driving risks. Am J Prev Med. 2008;35(3 suppl):s278 S284 48. Ptton GC, Bond L, Crlin JB, et l. Promoting socil inclusion in schools: group-rndomized tril of effects on student helth risk behvior nd wellbeing. Am J Public Helth. 2006;96(9): 1582 1587 49. Merry S, McDowell H, Hetrick S, Bir J, Muller N. Psychologicl nd/or eductionl interventions for the prevention of depression in children nd dolescents. Cochrne Dtbse Syst Rev. 2004;(1):CD003380 PEDIATRICS Volume 127, Number 2, Februry 2011 9 Downloded from www.ppublictions.org/news by guest on Mrch 5, 2019
A Prospective Study of the Effects of Optimism on Adolescent Helth Risks George C. Ptton, Michelle M. Tollit, Helen Romniuk, Susn H. Spence, Jennie Sheffield nd Michel G. Swyer Peditrics originlly published online Jnury 10, 2011; Updted Informtion & Services Permissions & Licensing Reprints including high resolution figures, cn be found t: http://peditrics.ppublictions.org/content/erly/2011/01/10/peds.2 010-0748 Informtion bout reproducing this rticle in prts (figures, tbles) or in its entirety cn be found online t: http://www.ppublictions.org/site/misc/permissions.xhtml Informtion bout ordering reprints cn be found online: http://www.ppublictions.org/site/misc/reprints.xhtml Downloded from www.ppublictions.org/news by guest on Mrch 5, 2019
A Prospective Study of the Effects of Optimism on Adolescent Helth Risks George C. Ptton, Michelle M. Tollit, Helen Romniuk, Susn H. Spence, Jennie Sheffield nd Michel G. Swyer Peditrics originlly published online Jnury 10, 2011; The online version of this rticle, long with updted informtion nd services, is locted on the World Wide Web t: http://peditrics.ppublictions.org/content/erly/2011/01/10/peds.2010-0748 Peditrics is the officil journl of the Americn Acdemy of Peditrics. A monthly publiction, it hs been published continuously since 1948. Peditrics is owned, published, nd trdemrked by the Americn Acdemy of Peditrics, 141 Northwest Point Boulevrd, Elk Grove Villge, Illinois, 60007. Copyright 2011 by the Americn Acdemy of Peditrics. All rights reserved. Print ISSN: 1073-0397. Downloded from www.ppublictions.org/news by guest on Mrch 5, 2019