SS Prevalence and Development of Psychiatric Disorders in Childhood and Adolescence E, Jane Costello, PhD; Sarah Mustilo, PhD; Alaattin Erkanli, PhD; Gordon Keeler, MS; Adrian Angold, MRCPsych Background: This longitudinal community study as- sessed the prevalence and development of psychiatric dis- ‘orders from age 9 through Lo years and examined ho- molypic and heterotypic continuity Metheds: 4 representative population sample of 1420 children aged 9 to 13 years at intake Were assessed an- nally for DSM-IV disorders until age 16 years. Results: Although 3-month prevalence of any disorder averaged 13.3% (05% confidence interval [CI], 11-7%%- 15.0%), during the study period 36.7% of participants 31% of girls and 42% of boys) had at least 1 psychiatric disorder. Some disorders (social anxiety, panic, depres- sion, and substance abuse) increased in prevalence, whereas others, including separation anxiety disorder and attention-deficivhyperactivity disorder (ADHD), decreased. Lagged analyses showed that children with a history of psychiatric disorder were 3 times more likely than those with no previous disorder to have a diagnosis at any subsequent wave (odds ratio, 3.7; 05% Cl, 2.0- 4.9; P<.001). Risk from a previous diagnosis was high among both girls and boys, but it was significantly higher among girls. Continuity of the same disorder (homotypic) was significant forall disorders except spe- cific phobias. Continuity from one diagnosis to another (heterotypic) was significant from depression to anxiety and anxiety to depression, from ADHD to oppositional defiant disorder, and from anxiety and conduct disorder to substance abuise. Almost all the heterotypic continuity was seen in girls, Conclusions: The risk of having at least 1 psychiatric disorder by age 16 years is much higher than point es- mates would suggest. Concurrent comorbidity and ho- motypic and heterotypie continuity are more marked in girls than in boys, Arch Gen Psychiatry. 2003;60:837-844 ‘TUDIES THAT follow the same subjects as they grow up are the best source of informa- tion about the prevalence and ‘causes of continuity and dis. or an underlying disorder that has diffe ent manifestations at different ages. The clinical and research implications of homo- \ypic and heterotypic continuity are quite different From the Departments of Psychiatry and Behavioral Sciences (Drs Costello, ‘Mastill, and Angold and (Mr Keeler) and Biostatistics «and Bioinformatics (Dr Erkanl), Duke University ‘Medical School, Durkan, NC. continuity of psychiatric disorders. A re- view’ ofthe few studies that coverboth child and adolescent psychiatric disorders” showed that between 23% and 619% of chi dren with adiagnosisat one wave had adi agnosis, although not necessarily the same fone, at a subsequent wave. This suggests quite high level of continuity. However, few studies have the power to distinguish between homotypic continuity (the samedi- agnosis at different assessments) and het- crotypic continuity (continuity of disorder buta different diagnosis). Homotypic con- tinulty is evidence fora disorder that has a similar manifestation across the age range of the study, whereas heterotypic continu- ity suggests an underlying vulnerability 10 psychiatric illness that may expose chil- dren to different disorders at different ages (©2003 American Med Downloaded From: http:/jamanetwork.com/ on 11/26/2016 In this article we address the follow= ing questions about continuity for the age range 9 to 16 years: What is the prevs- lence of DSM-IV disorders at different ages? (Becatise DSM-IV is inconsistent in whether impairment is required for a di agnosis, we include the prevalence of lous emotional disturbance (SED), the erm used by the federal government” for psychiatric disorder accompanied by si nificant impairment in the child’s fune- tioning.") Does the prevalence of these dis- orders increase or decrease as children grow up? What are the patterns of comor- bidity? Is there continuity of disorder across this developmental period? Which disorders predict which other disorders? And are there significant sex dilferences fm any of these estimates? 1 Association, All rights reserved.‘Table 1. Participant Age and Date of Interviews by Cohort conat age.y Ties 2883100941080 71099) x ° "1 0 2 5 w oy 4 2 2 M & rt a 83 “4 a a a5 6 a 85 46 6 ce 85 a Response at, % oz ms sao 70 moa ® 752 a7 “Financ consis preertad us rom ntarewing any of cohort A dring 1987 and he ost hal of 1908 Because subj ware randomly selected and itrvawed around tata, subjects intr ewad gurng the second hal of 1098 wee a anda sample of he whos coor (es The Great Smoky Mountains Stady (GSMS) is longitudinal study ofthe devclopment of pychinte disorder andthe need formental health services?" Thre cohort of children, aged 9° Lean 1 yeaa at as, were eerie from U1 counties in westem North Carolina A multistage sampling design was ted, vith weighting tht returned the weighted mpl ots crigial size Potenial participants were randomly selected from the population of some 20000 cidren using house- hod eal probably acclerated cohort design." This means that each cohort reaches given age in diferent year, thus ontoling for cohor effects The inital random sample of $007 ycked 3800 scrccning questionnaires (05%) consisting mainly ofthe externalizing (behavioral) problem sae of the Child behavioe Checklist completed bys parent (sual the mothe) .by telephone or in person. Allee scoring above A predetermined cutll pont the top 25% of the otal cores) pls a1 in 10 random sample ofthe rest, were rere for Ute imerviews ‘About 8 of area residents andthe sample are African “American and fever than Isare Hispanic: American indians male up only abost 3% ofthe popslation ofthe stay aren, ‘which ir overwhelmingly whitey bat were erupted Gros hoolrecordsto constitute 25% ol the study sample This was donc by using he sme screning procedure bit reerking al ‘American Indan children respective of screen score. Of he $56 American indian children dented, screening question: naites were obinined from 96%, and 81% (2390) partic pated inthe stay Although race was include inal asalyes, Ep comeliscrm ae awn article bout racial or elas Similarities or difrences, which ar reported clewhere =" ‘Table 1 presents the study desgh andthe numbers of checrtionsuleach wate Fandlogectatsinn prevented Dir interviewing the youngest cohort from January 1097 through Je 1998 Becadne seca were randomly selected, the co tovt members interviewed between july nd December 1998 areatandom sample ofthe whole cohort Daa were coleted Gn I cohort at ages 9 and 10 year, 2 cohorts tages 11, 12, Sn 13 years anal cohorts sages 15, and {9 year In the results presented inthe curren study, the data fom the interviews of 8-and 10-year-old are combined to provide a tore reliable estimate MEASURES The Child and Adolescent Psychiatric Assessment (CAPA) is an interviewer-based interview." The goal of interviews us ing this format, such as the adult Present State Examination” (©2003 American Medical Ass Downloaded From: http:/jamanetwork.com/ on 11/26/2016 or Schedles for Clinical Assessment in Neuropsychiatry. is tocombine the advantages of clinical interviews with those of highly structured “epidemiologic” interview methods. A de- tailed glossary provides the operational rues foridentifying clink cally significant symptoms, ‘With the CAPA, parent and child are interviewed sepa- rately by different interviewers, In this article, with the excep- tion of attention deficivbyperactivity disorder (ADHD) symp- toms, about which only the parent was interviewed, we counted symptom as present if twas reported by cither the parent or the child or both, as is standard elinieal practice ‘The time frame of the CAPA for determining the pres- ence of most psychiatric symptoms the past 3 months. Inthe case ofa few rare and severe acts, sch as fresetting oF a=- saul, lifetime frame of reference is used, as required by DSM 1V. Two-week test-retest reliability of CAPA diagnoses in chil dren aged 10 through 18 years is comparable to that of other highly structured child psychiatric interviews. PROCEDURE Interviewers were residents ofthe study area and had at least buchelors-level degrees. They received 1 month of training, con- stant monitoring of quality contro, and instruction by Depart tment of Social Services stall in North Carolina’ requirements for reporting abuse or neglect. All suspected cases were re- ferred to the appropriate agency Interviews usually took place at home. The parent and child signed informed consent forms, and each was paid $10. DATA MANAGEMENT AND ANALYSIS Scoring programs for the CAPA, writen in SAS statistical sol ware (SAS institute Inc, Cary, NC), combined information about the date of onset, duration, and intensity of each symp- {om to create diagnoses according tothe fourth edition of the DsaLv GENERATING POPULATION PREVALENCE [AND VARIANCE ESTIMATES To generate population prevalence estimates from a multi- stage sampling design'® subjects were assigned a weight versely proportional to their probability of selection. Correla- tion matrices were introduced to aecount for within-subject correlations. We used general estimation equations (GEES)” toaccount forboth the sampling design and within-subject cor- relation. In GEE, subject is introduced asa cluster (class) vari able, and the sampling weights are introduced asa scale vec- lation, All rights reserved.‘Table 2. Three-Month Prevalence of Selected Dsdr-v Psychiatric Disorders® Tela +" fs oys ago (w=0079 (oss) (= 2009)_(n= 3008) Roydogosst 183105 a a7 1s 158 (NTi50) (54244 ca0t62) 2-130) riz (135185) Swiousamotoral GB 4636 er 56 79 dsubanet §— G89) 453) 2749) (233) 379 6595) ‘ny behav 7 574 63 4585 ‘earcer 032 azz aren usa as5n coi) Canductdsorde 27272 28 2 42 2139 a4 1334) (1749) 729 arse) Opposition ar at 19 3 2 un eft csorder (22-22) (1528) (1426) oan (1629 atta) 0H 0922, 14 03 3 15 siz) 1289 728) 0105) 209 021) sue 2 0 ° 14 20 28 (1331) (0108 0207) 0728 BETH GAt05 1329 (19.40) ‘yan th 45602 20 ‘arse (1831) @875) (1545) (514 (1039 935) (164) @RS1) BOL (13.31) Aoydepessie = 225 to 0k 25 raat 28 16 ‘arse (1630) @340) (1035) (0298 (1353) (145) 235) 1853) (184g) (10-28) ‘Abrvaions- ADHD, tao dehy deorder SUDs, abetance vedere: “Daa ate gin asthe pacantags of paras (5 candace itn). {One ermare of sapaaton ae deordegnaaled arty order snp oh tare epee, ipa ore, ADHD, conduct dearer oppestorl dora postraumate ses: der, ojchoss, of absassie conpueve dso ‘Sofeus enol duane is dtined a any dagrets accompanied by si tor that multiplies the subjects wave to wave correlation matrix. ‘We lso used the robust variance estimate (ie, sandwich types limates), together with sampling weights, to adjust the standard ‘errors ofthe parameter estimates to account forthe multiphase sammpling design, The use of maltiwave data with the appropri ste sample weights thus capitalized on the multiple observation points over time, while controlling forthe elfect on variance es- Uimates of repeated measures. Identical results can be generated svithSUDAAN" using primary sampling unitsas the subject iden- ‘fiers along with sampling weight, To obtain the prevalence of| ‘ach ofthe various peychiatric disorders we fited logistic regre=- sion models, again using GEE, with no predictors but including the intercept only. We used the Delta method (Taylor series ex- pansion) to generate overall and time-averaged prevalence est- nates and standard errors, corrected for within-subject correla. tions and design ellects. Statistical teste such as comparison of prevalence between sexes used the same approach ‘Lagged analyses were used to answer the questions about hhomotypie and heteroypic continsity. Twosets of analyses were done, one looking atthe ellect of a diagnosis in the year i= meditely preceding any annual interview and one looking at the elect ofa diagnosis at any interview before the current one Although the effete were stronger when the I-yeatlag was sed, results are more generalizable using the entire preceding pe- Flod, and s the later are reported here. Logistic regression analy ses regressed this year's diagnosis on all past years, with other pst and present psychiatric disorders entered as covariates (es THREE-MONTH PREVALENCE (OF PSYCHIATRIC DISORDER Table 2 presents the 3-month prevalence of a range of DSM-IV diagnoses by age and sex. A fuller version of the table can be found on our Web site (hutp://www.devepi (©2003 American Medical Ass Downloaded From: http:/jamanetwork.com/ on 11/26/2016 tia, pai soe agoraphobia, aor depression dst, depression nat rood ness, buems mana tchellomana, nse, reopen, SUD, at unin enpairmant & ey Das . Gee SD a. Bo Ay Dagss i . “ease g Tyoe-month prevalence of ay disorder and sarous emotional turbance (SED) by age ad sex. smeduke-edu; Web Table 1). In the ease of depressive disorders, data are also presented on depressive disor- der not otherwise specified, which includes the DSM-IV experimental category of minor depression. The overall prevalence of any disorder was highest in 9- to 10-year olds, falling to its lowest level in 12-year-olds and then rising slowly (Figure). Twelve years was the age at which many of the disorders of childhood (ie, ADHD, separa- tion anxiety disorder [SAD], enuresis and encopresis, and motor and verbal ic disorders) had almost disappeared, especially in boys, while those of adolescence and adiult- hhood had not yet developed. Childhood disorders also decreased in girls across the same period but were less frequent to start with, lation, All rights reserved.‘Table 3, Predicted Cumulative Preval of Psychiatric Disorders by Age 16 Years” Diagn Tua Bor ‘ry doer Bren sos esa ‘Any erationl disorder = 150(17) W74(17) 130(18) ‘ny arty disorder Bois) ris) 7708) ‘ny depressive sore 5 (11) 117 (12) 73010) ‘ry bahar disorder zo(t7) 161(12) 20922) ‘Condt order goi2) 3a) 41.8) Oppositoraldetnt disorder 113(10) 0.1(10) 134 (10) 30H 4) 1403) ToC) sue 1228) 10148) 14317), ‘obrevalons: ADHD, ateniondeiUhyparactiy ded, ‘Us subetane tee ordre, ‘Datare ven asthe percentage of parcipans(S) The move into adolescence was marked by a rise in rates of depression and social phobia in girls that was not ‘seen in boys. In middle adolescence the increase in sub- stance use disorders (SUDs) in both sexes was dra- matic, and there was also a modest increase in panic and generalized anxiety disorder (GAD). The Figure also shows that SED increased as a pro- portion of psychiatric morbidity across adolescence, es- pecially in boys. Atage 0 and 10 years only 20% of boys ‘and 31% of girls with a psychiatric disorder qualified as significantly impaired by SED. By age 16 years, 79% of diagnosed boys and 58% of diagnosed girls had SED. ONSET OF DISORDERS BETWEEN AGES 9 AND 16 YEARS We estimated the cumulative prevalence of disorders across the age range of the study; that is, the accumula tion of new cases in previously unaffected children. (Note that the individual may have had an episode before en- ty into the study.) Table 3 presents the predicted cu- rulative prevalence by age 16 years. By then, 36.7% of children had met DSM-IV criteria for 1 or more disor- ders. The estimate for boy's greatly exceeded that for girls, ‘asa result ofa much higher cumulative prevalence of con- duet disorder (CD) and ADHD. Girls accumulated more ‘eases of depression and anxiety disorders, CONCURRENT COMORBIDITY Concurrent comorbidity refers to the co-occurrence of 2 or more diagnosesat the time of measurement," in this ‘case, within the same 3-month period (A.A.,A.E.,EJ-C., and Helen Egger, MD, unpublished research, 2001). Table 4 presents the major types of concurrent comor~ bidity by sex. In bivariate analyses, 25 of 30 comparisons showed significant comorbidity. This fell to 16 of 30 after con- trolling for other comorbidities. There was significant co- morbidity among the behavioral disorders and between anxiety and depression, as expected. The strong bivari- ate association between anxiety disorders and the behav- loral disorders fell markedly and in most cases became nonsignificant when controlling for other forms of co- (©2003 American Med Downloaded From: http:/jamanetwork.com/ on 11/26/2016 morbidity. This was mainly because of the strong asso- ciation between oppositional defiant disorder (ODD) and depression, on the one hand, and depression and anxi- ety, on the other, There were 2 marked sex differences in patterns of concurrent comorbidity, both involving depression, AL- tercontrolling for other comorbidity, depression was co- morbid with CD in girls but not boys. Conversely, de- pression was comorbid with SUD in boys but not girls. HOMOTYPIC AND HETEROTYPIC CONTINUITY Children with a history of psychiatric disorder were 3 limes more likely than those with no previous disorder to have a diagnosis at any subsequent wave of data col- lection (28.0% vs 9.3%; odds ratio [OR], 3.7; 959% con- fidence interval [Cl], 2.9-4.0; P<.001). Risk from a pre vious diagnosis was high in girls (OR, 5.2; 05% C1, 3.4- 7.9; P<001) and boys (OR, 29; 95% Cl, 2.0-4.1; P<.001), but it was significantly higher in glels (interaction OR, 1.8; 95% Cl, 1.0-3.7; P=.03). Within this general pic- ture, there were marked differences in continuity from one diagnosis to the same one (homotypic continuity) or to another (heterotypic continuity). Looking first at homotypic continuity, this was sig- nificant for every diagnosis except for specific phobias. The details can be found on our Web site (hup/Avww devepi me-duke.edu; Web Table 2). The disorders showing the highest level of continuity were panic disorders, psycho- sis, verbal tics, encopresis and enuresis, and SUDs. Girls showed more continuity than boys, even though they had {ewer psychiatric disorders. Prediction from past episodes \vas significantly higher in girls for depression, GAD, so- cial phobia, and specific phobia. On the other hand, en- copresis showed continuity only in boys. There were no significant sex differences in continuity for the behavioral disorders or SUDs, despite their greater Irequency in boys. ‘We next tested whether there were cases in which the presence of one disorder earlier in life increased the risk ofa different diagnosis later, net of the ellects of con- current comorbidity. Models included both homotypic and heterotypic continuity and then tested whether add- {ing a term for concurrent comorbidity affected het \ypie continuity, Results ate summarized in Fable 5. There was strong heterotypic continuity from de pression toanxiety, which was not greatly alfected by the high level of concurrent comorbidity between the 2 dis orders. Prediction from anxiety to depression also oc- curred, and remained significant, although with a lower OR when controlling for concurrent comorbidity. An ety predicted later SUDs, but depression did not. Among, the disruptive behavior disorders, ADHD showed a mod- est but significant prediction to later ODD, even con- trolling for the high level of concurrent comorbidity be- oven the two. Conduct disorder predicted both ADHD and SUD im the bivariate model, but the very high level of concurrent comorbidity explained this. Separate analyses by sex showed that heterotypic con- Linuity was much more common in girls. Anxiety pre- dicted depression, and vice versa, even controlling for concurrent comorbidity, but this was not the case for boys. Similarly, the link between past anxiety or CD and eui 1 Association, All rights reserved.‘Table 4, Concurent Comorbdity*t ‘ay Amity Any Depressive Diagnosis Disorder oisrser AoHO conic leareer atin ier sue ‘ry arty ior 21093) SS RO15H] 25TH 33/75] 05/0219) MA@ses| 240299 10/0799 —os(eses) oz (mony ‘ny depressive soror 28.4 (142.55. So(34165) 50123-1129) 207 (88488) 99(31-315), 749 (18.8002) z2qnes2) —a7jez24) 10.7 68478)| 10.4(27-403)) sono 7727-222) 33(07-151) S7(1876) 87 (46-164) eo(isznng 0.1 (2.00240) 19(1033) — oaserza, Conduct erder 35(1359) 213(00505)) 13 (140-2199) (63.1734 03 (0012) weersens 30 (11130) so(sco) 9728-199) Oppositoraldetantcsorder 7.7(38457H| 151(74306)) 79.0G21-1015)) 441 (157-1244) 4917-1025 24(ose7) 1.122220) 03 (100405.0) 20.4 (08-20.0), 19 (07-40) sue Oaioots BSI 03 l0oL24) — ZaAlOI-s0ay) 2414-524 ic 20019) Ne 0.7 (127735) 04 (0002.7) Abbreviations: ADHD, atenton “Dat ae qn as ods at (5% contance ara). Data fo gis ae {aus in rogular ype ae spl vate estimates 2nd values in bola unger consiaraton (9, arty and depression) conting far comorbid of dprasson with oppsional dean dor P05 Bpeot ee vhyparactiy ded, NC, rade woud not canvrge:SUDe, substance ue deoders eset bal the aganal ool amp) cls: boys, above the dagonal yp af cored freer poss emorbiiesivolng the 2 dagnses ‘Table 5. Homolyple and Heterotypie Continuity With and Without Controls for Comorbidity” Freitag Fast Depression Past Armety Parl conduet Disorder Past O00 Past ADO ——_—Past SUDS Depression Toeatag S07 S05 Contlingforcomortdiy 42(2vasyy 27(18528 ‘sty STR2U59 24 (108.79 Contlingorcomortdty 28(12654 20(r2eayt Conduct derder 2921.95 Contzling fr comer 103 (9324.75 ono s7@zeng 20/1108 Contzling for comer arerang 21,1142 0H 22(io4st 107 (3.222.395 Conlin for comer 1800039) 9.0 (4021295 sue 2oii2as¢ 27012651 213 (s7239§ Conlin for comer 2ouian 17 06-47), 77 (883.95 ‘Abbravalons: ADHD, atenton-deiiuhypractny dsorder, 00D, opost *Datgare gun asthe odds a (95% on ‘ypandesiss homes contin, PaO peat secon rent SUDs (both of which persisted when controlled for ‘concurrent comorbidity) were specific to gels. The ink between past ADHD and current ODD was also specific to girls but appeared to be mainly the result of current ‘comorbidity ee “Although there has never been a representative popula tion survey of child psychiatric disorders in the US com- parable with the National Comorbidity Survey of 15-10 54-year-olds,* or the recent British national survey of 5- to 15-year-olds,” researchers have patched together a pic- ture of the prevalence of psychiatric disorders in Ameri- ‘ean children of different ethnic groups and ages. This te search has been reviewed several times,"*" and a ‘consensus has been reached that at any given time 1 child ©2003 American Medi Downloaded From: http:/jamanetwork.com/ on 11/26/2016 orl datant order, SUDe, substance ue order rental). Empty as inate hat coniuity withoui controls fr comerbdty was not sigicant alae in 5 will have a psychiatric disorder, This was also the conclusion that we reached on the basis ofthe first wave of data from GSMS, published in 1999," Analyses of further waves from the same data set, presented here, lead us to revise this picture. The 1-in-5 estimate was true of the youngest children, but as they grew up the 3-month prevalence fell toa low of 8.3% at age 12 years, before beginning to rise again in adoles- cence. This is consistent with the one other (cross- sectional) study that provides annual data by age group. Earlier epidemiologic studies of children and ado- lescents in the 1980s and eatly 1990s tended to generate very varied, and sometimes very high, rates of some dis orders, in particular anxiety disorders.” ADHD,” and in some cases depression." However, recent studies have tended toward greater agreement and, on the whole, lower population estimates." Association, AI rights reserved,METHODOLOGICAL CONSIDERATIONS The GSMS sample comes from a small (11-county) and predominantly rural area of the southeastern United Sates. There were no Asian American subjects, and only 8% were [African American, Thus, does not represent the Ameri- ‘can population. A recent study of another part of North Carolina, a largely rural area including a large percent- age of African American youth, found remarkably simi lar rates of disorder in white and African American chil- dren.” A recent British population study of youth aged 5 to 15 years produced quite similar results; for ex- ample, the prevalence of any anxiety disorder was 3.8% (GSMS, 24%); depression, 0.9% (GSMS, 2.2%); ADHD, 1.4% (GSMS, 0.0%); ODD, 2.0% (GSMS, 2.79%); and tic disorders, 0.1% (GSMS, 0.1%). Despite the overlapping cohorts design, which used 3 age cohorts to permit age X cohort comparisons, there sno escaping the fact that in the later data waves of GSMS the children's mean age is greater than it was in the ear= lier waves and the participants have been study subjects for longer. Thus, lower prevalence relative to eross- sectional studies may reflect (1) a tendency for subjects to report fewer symptoms in later interviews (ie, “symp- tom attenuation"); (2) cohort differences, with the youngest cohort having more symptoms and the oldest ‘cohort the fewest, controlling for age; and (3) differen- tial dropout of children with and without psychiatric disorders. Another type of methodological challenge lies in the fact that the CAPA uses a 3-month window to ceslimate prevalence. If none of these methodological challenges is upheld, then we need to look for substan- live reasons for the differences between this and earlier studies. Symptom Auenuation Ina test-retest reliability study of 77 children aged 10 to 18 years using the CAPA," there was a very modest amount of attenuation between the first and second in- terviews; the difference was significant only for CD symp- toms. The sample, however, had no 10-year-old, so at- tenuation remains a possbiliy forthe youngest children. Cohort Dilferences Analyses of symptom and diagnostic counts by cohort in GSMS (unpublished data) showed that the cohort dif ferences were nonsignificant Differential Dropout The response rates for the different waves of the study are reported in Table 1. OF 7944 possible interviews dur- ing the study period, 6675 (84%) were completed. Four subjects died, and 5.9% of participants only completed, single interview; these participants were no more or less likely than those who completed multiple interviews to have a diagnosis. Subjects who refused to be inter- viewed at one wave frequently returned later. This sug- {gests that biased nonresponse is not a major reason for the observed prevalence rates, (©2003 American Med Downloaded From: http:/jamanetwork.com/ on 11/26/2016 Three-Month Reference Period Because the CAPA inquires only about the 3-month riod preceding the annual interview, our estimate of cu- mulative prevalence may underestimate the number of cases occurring since the last interview, and thus the “bur- den of disease” seross childhood and adolescence There is no question that the CAPA underesti- mates cumulative prevalence, It misses cases that began and ended in the © months between interviews, includ- ing those for which some symptoms were present dur- ing the 3-month window, but not enough to reach the diagnostic threshold. Many psychiatric interviews adopt a 6-month,"" 12-month,"* of even lifetime” time frame for estimating the prevalence of psychiatric disorders. However, the evidence from studies using longer time frames suggests that a lot of forgetting occurs; for ex- ample, in the National Comorbidity Survey the lifetime rate for 15-10 54-year-olds (48.0%) was less than double the 12-month rate (29.5%)."* Clearly, both approaches underestimate the burden of psychiatric disorder. The CAPA uses a 3-month time frame because of the strong, psychometric evidence that memory for the type of phenomena sought in psychiatric interviews is highly unreliable further back than 3 months.*°°* Ideally, study subjects would be interviewed every 3 months, but the burden imposed on subjects would introduce another set of biases. Substantive Reasons Controlling for wave, the 9- and 10-year-olds had si nificantly more symptoms and diagnoses than any other age. This was particularly true of the younger boys. As noted earlier, 9-and 10-year-old boys had very high rates of enuresis (0.5%), ADHD (3.6%), ti disorders (5.5%), and SAD (4.1%), all ofthem disorders of childhood that had fallen by more than 50% by age 11 years and disap- peared by age 16 years, Excluding these diagnoses, the 3-month prevalence of psychiatric disorder for 9-and 10- year-old boys would fall to 8.3% (05% CI, 5.2%-12.0%). In summary, bias is always a risk in survey re- search designed (0 estimate the population prevalence of a disorder.” We have guarded against it to some ex- tent by choosing an interview that shows minimal at- tenuation effects, maintaining a good response rate over 8 years of data collection, and adopting appropriate sta- Uistical controls for the sampling design and the use of correlated data seross waves. The result is an estimate of the 3-month prevalence of psychiatric disorder that fs lower than our wave 1 estimate largely because of the fall in disorders of childhood: enuresis and encoprests, ADHD, tic disorders, and SAD. PREVALENCE OVER TIME: THE BURDEN OF DISEASE Multiple waves of data show that although only 13.3% of children, on average, had a diagnosis at any measure- ‘ment point, almost 3 times this number had 1 or more disorders over the period of the study. This means that single-wave, cross-sectional studies are likely to under- 1 Association, All rights reserved.cestimate the burden of disease" in the general popula tion of children over time COMOREIDITY AND DEVELOPMENTAL CONTINUITY As with other cross-sectional and longitudinal” stud- les, we found considerable concurrent comorbidity: 25.5% of children with a diagnosis had 2 or more. It appears that much of the association between CD/ODD and ADHD isin fact carried by ODD rather than CD.” In both sexes, the high level of association between CD and ADHD fell noticeably when other diagnoses were included in the model, whereas the association between ADHD and ODD did not. Concurrent comorbidity between anxiety and depression was, as expected, strong in both sexes. Co- morbidity between depression and ODD wasalso strong, ‘but comorbidity between depression and CD was sig- nificant only for girls, when other types of comorbidity were controlled, The data presented here on continuity of disorder across time sulfer from the same shortcomings as do the ‘cumulative prevalence figures: we have only a series of snapshots on which to base our estimates. Given that, the degree of homotypic continuity is remarkable. In bi- variate analyses, every DSM-IV diagnosis examined showed significant homotypic continuity with the ex: ‘ception of specific phobias. Apart from encopresis, which rarely occurred in girls, continuity was higher in girls than in boys. Compared with the level of homotypic continuity, there were few cases of heterotypic continuity. Hetero- lypie continuity was much stronger in girls than in boys. This suggests that the DSM-IV taxonomy may fit boys! developmental patterns better than those of girls. There isno evidence that boys with an emotional disorder were at increased risk of a behavioral disorder, or vice versa, ‘whereas girls with anxiety disorders had Increased risk for later SUDs. (See Kaplow et al” for a more detailed ‘examination of this issue.) Robins and Price's predic~ tion™ that later SUDs will be predicted by CDs holds here for gitls but not for boys. This sample is too young to test the prediction of Zoccolillo et al” that CD in slrls is likely co lead to later depression and somatiza- tion disorders. ‘An important issue that exceeded the scope of this anticleis developmental continuity among the anxiety dis- ‘orders and between specific anxiety disorders and other diagnoses. Continuity was much higher for some anxi- ‘ety disorders (panic disorder and posttraumatic stress dis- order) than for others (specific phobia and GAD) (hup: ‘iwww.devepi.me duke-edu; Web Table 2). Much more ‘work needs to be done on comorbidity and heterotypic continuity among these disorders and with depression. In summary, data on a representative population of children and adolescents growing up in the 1990s show that at any time 1 in 6 will have a psychiatric disorder and at least 1 in 3 will have 1 or mote psychiatric disor- ders by age 16 years. As children grow older, psychiat- He disorders are more and more likely to be accompa- nied by significant functional impairment. Once children, particularly girls, develop a psychiatric disorder their (©2003 American Med Downloaded From: http:/jamanetwork.com/ on 11/26/2016 chances of continuing to have one, or of developing an- other episode after remission, are much higher than those of their unaffected peers, By mid-adolescence, although some disorders of childhood have disappeared, impair- ing adult disorders such as depression, panic disorder, and SUDs are becoming the most prevalent problems Much more work on the childhood antecedents of these disorders is needed if prevention programs are to be el- fective Submitted for publication September 4, 2002; final revi- sion received January 13, 2003; accepted January 22, 2003. This study was supported by grants DA11301 and MHB7761 and Independent Scientist Award MHO1 167 from the National Institutes of Health, Bethesda, Md, and Fac- ulty Scholar awards from the William T. Grant Founda- tion, New York, NY (Drs Costello and Angold). Corresponding author and reprints: E. Jane Costello, PhD, Center for Developmental Epidemiology, Depart. ment of Psychiatry and Behavioral Sciences, Duke Univer- sity Medical School, DUMC Box 3454, Durham, NC 27710 (e-mail: jeostell@psych.me.duke-ed) (Es 4. Costa El, Angad A, Onvapmental pceiay. ica, Cohen -DeslpmentaPychopatloy Val Ne York WY-Jon Wly& Sas 15:23:56, sha, MeGoe RO, rion SE, nderson J, Waton LA Wan, Siva PA Rab A, Ory, Mekoeu DH Dress na supe ol yeh tte: roves anf ocd characte, Arch Gon shy TERA a2 Anderson J, Wile , Mee, Sha PA. OSH cords in pate ent hn: rts ina le sare rom te geal papuaon. Ach Sen Psyeniany. O874438-77, Ge ecan Mi, Wars, Pride F Sa PA Kay. DS ds fersinalrgesanglctacolescane, JA ad Gi Adless Peay 1000 acre Fergsson OM, Howond Lynsey MT Pravnes and coma oft ‘Adagroses naire coor ot 1 year ois. J Am Acad hs Ades Ps hay 1093.2 127-1138 Velez Cbs, Cher. langual anit sect kts of thle psyhopaolagy.JAn esd Chil Ado Pech 1HBZ861- tt Chen Cohn, KasnS, Vee Harmar Johan iM, rok, Sain LAr pei sty of ocr nace an ado lesen, ae- an ged sectic prone J Ch Pah sen 1085 est 67 Rainer, Gicona AM, aot ES, Pai, Frat AK. Prana y= hate dors na communty popaon of le sdesens. J Am Acad (hd Adal ayy. 1985328977, Levinson PH, Rode, Kin ON, Ses JA. Nua cous of adolescent a iordpesvdioraer cocinty ina Your acutood Am Acad Cn Tes Psy. 1000383688, 58 Fedral git 20425108) nino Cost gad A Assessing tunciaalimaiment nd saci apeation fr child marca sercas esearch review meas, Jet ath Sr 100188108, Casto, Angad, ums, StnglDK Twseé DL Ean, Worthman Th ret rot Mourne Sly of You: goal, dene, mathe, nthe evans of DSUILR disorders. Ach Ge sj chat. 195531720135 Burs Cost, Ago Tweed 0, SurlD Famer EZ Ean Chi- ‘he's mentale use cas Sere sectors. Hath A (Mod, 4147-158, Ang Prncegst Mh, CoxA HarngtonR iment Rute M. The Cid tan Adlon Psa Asessnet (APR) Psy Me. 185, 25:730- 78 Catala Angol A Marc Fan Livni and post uma ste the dvlopmart tn maser for chien an odolans. Psychol ded sone 2751208, 1 Association, All rights reserved.16. Argo Cosa. Atesttst realy sty fit oparted pyc sympoms ad dagases ug the Child nd Adlescer Paya Asses te CAPAC) jal Me. 100525 75-70. ogo, Eta Conall, tar Pacis, ela and asaya ‘tengo yom one inci sn acs pychopatlgy. Chi Paya Pec. 19657 67-6 Shout PE. Hewan SC. Desig fo phase prewace sues trae dsr (re, Bomar 16545540 55, Seale. arr ode ort stuf dvlpmental polos. Psychl Bal es 6492 107. inka, Kuper LL. Morgner Emile seh Princes Quanta Mero ew York MY Van star Ris 1982 ‘Aceach TH Ebook 5 Behave problems ancompdenes rpoted by preset noma an dsturbedchlren ape ou tough seen. Monogr Soo Cid De WET 4582, Catal Famer, Ang Bun BEA. Payee derdets mg ‘aan nanan ie youn palais te Great Soy Moras Su ‘hn Pub aah. 07 872780, FedranFS, Cosel Anal Famer EU, Ean A elope of st strc se nd pyc comorbity man pierce uy of hit and ‘Amarcn an young ssesce te Get Say Moun Study. Drag Ala Depend, 19874307 Cart E Farar EZ god A Same ace detect hia: Whe ad ‘Anaad cen in th Agpabchian Meunis :CoenP Rais, Somos, el Whore and Wher Histor and Gograph Aspects of Psyoptotgy. Maa, W: Lanne aun Associ: 108270298, ‘Anol AFsher PW tence sed aris Sha, Lucas Rc {ue de Diagnose Asesmnt in Chil on sce Pjchopathaloy Now Yar Gator Pros: 0084-54 Wing oopr Sari Pres Sate xinatn (PSE). Wg J, Cooper Sarria ade. he Metsurmant and lsat of ayer ‘Symptoms Canbigs, Egan: Cambios Urey Pres, 74-2 Wing J bor Bagh, ute J Caner Eel, ls Reger, Sarr SCALE Schule for in eseenent i Maupsyhiy. Ach Gen Peyehiny 100047380508, ‘Anecan seh Asocison. Dagnastc and Sasa! Mana Mental ‘Dard, ut Eation Washowon, DC Anan Psychiat Asscizon 1 Zag SL, ing KY, Alu FS Molla ate pra et rang equation approach. anes. TORE 109-1050. Sand oson RE LaVangeLM Whesss 0 Boye VE Wilms RL SUDAAN Safar te Sista Anais ofCoreled Oat: Sunt Mees and ‘bem Agonthms Used i SUDAA, 1995 Raseeh Tange Pat HE: AT naan Angol Cosel, aA Comorbiiy, J Chi sya Pei 1988, “05747 Kesler BO, McGonagle HA Zhao, Neon C8 Muphs Mt Etismans, Witchen HU Kander KS Lista 2-nach rove tOSHEI-Apschiaie gs ‘rds nt Und Sate esha hata amity Sure, Ach Gon Pye 100451818. NetaurH Gata , Goodman R Fors. The Mena Hath of Cian and ‘Adtscensin Grex iain London, Egan asa Sates 108, Cast Developments incl psychic een. J An Aad Cit Adlai. H6028 9968. Caso Ang! A Eenlgy Marc Ami Disord in Ch- tena doecnts Mow York W-Gallord Pres 06 100-124 Caso 1, rgulé A. Delopmontl epiderlgy 2 ranaveck or doy ing sjhapahoopy- I Lens Sumer Aes. Habock of evelgpmen- 1 Pjehpataloy Zn ew Yr, MY: Huser earl Perum Pas 2o0037-3, rannbuy MA Fidman RM, Sve SE. To epieiiogy of hihood py ‘hie ors. Am Aad Cn dase Pei, 1990207682 ‘gol, Earl A, Farmer EZ Fara J, Bums Bel, Cost Pye dsardar,imparmen and sere sein Afizan Amercan and Whos youh Art Gan Pei. 200258298 90, Casto, Cosel Al, rack C Burs B,D MK Bet, ais 2 a 2 a 4 28 2% a. Ey Ey o a 2 a et 8 8 a Py oe (nernnyrep) SRC GEN VCHATRYVVOLE, AOC ET ©. 4 2 2 “4 6 6 a. 8 2 0 5 2 ® en 6 6 7, %e 0 S.Pychawicdordrsin peda pinay cr pres and isco Ach en Pein. OBESE 7-116 Br HR Cain 6, Fubo-Sipe Gould MS, be Sosnan M Weedbuy ,Hugas-Gldman 8 Pagan A, Sancher-Lcay A Mesto ML Estima ot the prove fchidhodmaadjstertinaeormanty suneyinPueo Re: the ie of cond measures. Ac Gen Pacha 98B45 120-126. Leinsohn PU Hops, Roberts RE. Sey JR, dens JA. Adoscan py ‘hopablogy, prersene and iene of dpessian a ote DSWLIE ‘ordain igh sal suse J Abram Poet 10H 12 12144, ‘Skater, PU, Duan Dai M, Pace Schwab Ste M, Li hay BB, Bourdon Jensen, rH Cain, ager Ta WMH dagoasic iri shel for chien (1SC 23): esriginacepbiy, prev Tacs, an parrmaesin th MEA tu. J Ar cad hs dls Poh ry 100538 365-277 mean PS, Shaler Ra Darina, Bid HR, Deana BB, bi- ‘Sipe Goodman, Rens JE tenant he Osgrastcitrve Sehed efor Chiltan (1S0 21}: sex, age a 10 elonshps. Paper pres at $n Homa eig of tho Aarne of Chad oct Poe thay Oetaer 192 Washing, 0. Pei eeanP Sha D rR, Canna, Odean Riche Rape Fr Pil ves, Can J Goodman abi, Bourdon K.Kteutton fel psyiaicdagost iteration: magtule and raicosip os ‘eri. Paper resend a Anu Meaig ofthe Soi tr Ressrch nC nd Asescet Pychoptlogy Jl 102; San Francia, Cl Paani Ropar Jnsn Lucas Faber, Be, Bourdon K.Sctwab- Stone M, Rube My aves M, Olan M.forar-ased etermnants ‘symptom atanusoninsracured hil pychaiceaeews Aba his Paya. 198771728. uray, Lopez, Te Global ure of sas anes, Suton Wd Heal Ogi: 196 ‘Shae. Fisher PW, Seba StneM, Pace, Wik J. Te Dag Inne Sched fr Chie (DISC 2 1 Roki, Mé atonal stat of Ma anh 06. ‘ht FierP, uc, Davie Tha Diagnoses Sead Ue or Chien Version 23 (16023) description aro, prance ‘ns, anpereancs inte MECA sty Am Aca Chl Ades Pye. “0063536587 Line Zr Lavish, Sack Lite coma aang ane sty orders and between ane srr an oer mental dards ae Toscan Aniy Disord 100711377308, Su AK Bacal apse in Demographic ques NewYork, NY Asi Pali ing House 173. Bld OF, Mui J. Chie’ elopsnt of ae reall mary: remen- bing onion Aon Ci De. 10885 70-123. vow 0, Lint MS. grat betas qusionae dat and moda = exten for azar of cal. Ar Eide 198012528328 Palomar 8, White SH. Chana ees acl by chien and te 4 hil er Bea. 108621207 240, ‘Ang Costa Tovardesashing an empl basis oh agoss ‘foposnal tant dear. Am Acad Chi Adolesc 165 ‘ease Rotman, GrelndS. Made penis Pita Pe: Lpinat- Raven 008 Conio Gi Bra HR Sout PE Rubio Spc, Bao M Marz, Seman ,Gueara LM. Th preaencs ot spe psi dors in Pus Re Ata en Pct: 1987 $4727. plow J, Gran Pl Angol, Cotsia El. Th prospec rion eveen simension of ant ard he tisiono adler aoa se. Jl Cs Payal 201 3016328. otis LN, Price RK Aut saree predicted by ciood conduct prot Tans sus tte UA Epiemilog Caterers Propet Pye, saniseriet, Zeca, Pls, uirton,Ruter The atconeot hood arcuet ‘ior impeatons lor dein aft petsoraly dare and conducts de. Payee 1222071 5, (©2003 American Medical Association, All rights reserved. Downloaded From: http:/jamanetwork.com/ on 11/26/2016