Original Paper

Psychopathology 2011;44:125–132 Received: October 15, 2009 DOI: 10.1159/000320893 Accepted after revision: September 2, 2010 Published online: January 12, 2011

Psychometric Properties of the Young Rating Scale for the Identification of Mania Symptoms in Spanish Children and Adolescents with Attention Deficit/Hyperactivity Disorder

a, b b a a Eduardo Serrano Lourdes Ezpeleta José A. Alda José L. Matalí a, c Luis San a Department of Child and Adolescent Psychiatry and , Hospital Sant Joan de Déu, and b Unitat d’Epidemiologia i de Diagnòstic en Psicopatologia del Desenvolupament, Universitat Autònoma de c Barcelona, and Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Barcelona , Spain

Key Words acteristic curve analysis showed good diagnostic efficiency Attention deficit/hyperactivity disorder ؒ Mania, detection ؒ in differentiating mania in ADHD (area under the curve of -Children and adolescents ؒ 0.90). Conclusions: The Spanish version of the YMRS is a val id and reliable instrument for detecting and quantifying the symptoms of mania in children and adolescents with ADHD. Abstract The results provide further knowledge about the frequent Background: Diagnosing mania in children is difficult, due association between ADHD and manic symptoms in chil- to the high comorbidity and symptom overlap with atten- dren. Copyright © 2011 S. Karger AG, Basel tion deficit/hyperactivity disorder (ADHD). The detection of manic symptoms in ADHD has important implications for prognosis and choice of treatment. Our objective was to study the utility of the Young Mania Rating Scale (YMRS) for Introduction discriminating mania in Spanish children with ADHD. Meth- od: One hundred children and adolescents with ADHD be- In recent decades, it has been recognized that bipolar tween 8 and 17 years of age were evaluated with a structured disorder (BPD) could have onset at young ages [1–3] . diagnostic interview (Diagnostic Interview for Children and There is a consensus among researchers and clinicians Adolescents-IV), the YMRS, the Parent-Young Mania Rating about the difficulty of diagnosing pediatric mania. This Scale (P-YMRS), the Child Mania Rating Scale-Parent Version difficulty is primarily due to 3 factors: (1) the clinical pre- (CMRS-P) and the Children’s Global Assessment Scale. Re- sentation of mania in children, (2) the overlapping of sults: The YMRS showed a 1-dimensional structure with symptoms of mania with attention deficit/hyperactivity good internal consistency and test-retest reliability. The disorder (ADHD) and (3) its high comorbidity with YMRS was associated with the P-YMRS and the CMRS-P. The ADHD. scores obtained with the YMRS differentiated between The clinical presentation of BPD in prepubertal chil- ADHD with and without mania. The receiver operating char- dren and in early adolescence is a matter of debate and

© 2011 S. Karger AG, Basel Eduardo Serrano 0254–4962/11/0442–0125$38.00/0 Department of Child and Adolescent Psychiatry and Psychology Fax +41 61 306 12 34 Hospital Sant Joan de Déu, Passeig Sant Joan de Déu, 2 E-Mail [email protected] Accessible online at: ES–08950 Esplugues del Llobregat, Barcelona (Spain) www.karger.com www.karger.com/psp Tel. +34 93 280 40 00, ext. 4350, Fax +34 93 280 63 49, E-Mail eserrano @ hsjdbcn.org shows differences with respect to classic manic-depres- multidimensional and more complex than the classic eu- sive symptoms in adults. The symptoms of euphoria and phoric-dysphoric dichotomy [21] . inflated self-esteem, and the episodic course are rare in The difficulty in discriminating between BPD and children [4, 5]. Affective phenomenology is often mixed, ADHD leads to the diagnosis of false negatives (not rec- and dysphoria with ‘affective storms’ and severe out- ognizing BPD if there is ADHD) or false positives (diag- bursts of anger are much more frequent than euphoria. nosing BPD when only ADHD is present). The impor- The course is often chronic, discontinuous and erratic tance of differentiating symptoms of mania in children [6–8] . The National Institute for Mental Health research and adolescents with ADHD and making a proper diag- roundtable on prepubertal BPD [9] reached an agreement nosis is not only a nosological matter, but has significant that pediatric BPD can present as ‘narrow’ (fitting the implications for prognosis and choice of treatment [22– classic definition of BPD-I or BPD-II described in DSM- 24] . IV [10] ) or ‘broad’ (basically BPD not otherwise specified Therefore, it is essential to have specific instruments that includes children who do not meet DSM criteria but with strong psychometric properties to assess pediatric who are still severely impaired by symptoms of mood in- mania and enable the clinician to quantify the presence stability). Other authors have suggested a phenotypic sys- and severity of manic symptoms. We currently have few tem of juvenile mania consisting of a narrow phenotype, instruments that specifically assess juvenile mania. The 2 intermediate phenotypes and a broad phenotype, ac- most widely studied in BPD, and developed to be used on cording to the presence of distinct episodes and hallmark adults, is the Young Mania Rating Scale (YMRS) [25] . The symptoms [11] . YMRS is an 11-item scale designed to be completed by the Comorbidity studies show the frequent co-occurrence clinician. The scale assesses the intensity of manic symp- of ADHD and BPD. In a sample of young people with toms based on a clinical interview with the patient and ADHD, Biederman et al. [12] found that 23% had BPD, takes into account the subjective comments of the inter- which is similar to the 22% reported by Butler et al. [13] viewee and the clinician’s own observation. The instru- and the 20% found by Wozniak et al. [8] . In Europe, Di- ment has shown adequate psychometric properties in ler et al. [14] found in an outpatient sample (7–13 years of adult inpatients [25] . Subsequently, studies have been age) with ADHD that 8.2% also had BPD. conducted to validate its use with children and adoles- There is a considerable debate about the difference in cents, which have also shown satisfactory psychometric pediatric BPD prevalence between the USA and Europe. properties [26–28] . The scale used on prepubertal chil- Epidemiologic studies in clinical samples in Europe re- dren (6–12 years old) differentiated BPD from ADHD, ported a prevalence of mania ranging from 0% in the UK while the scores from hyperactivity-specific instruments [15] to 4% in Spain [16] . In US outpatient samples it rang- (for parents and teachers) did not differ between groups es from 5.9 to 19.6% [17–20] . These differences may be [26] . In a later study also conducted on prepubertal chil- explained by the variations in methodology (range of age dren, the YMRS showed adequate internal consistency considered, rating scale used, person interviewed), ICD- (␣ = 0.80) and convergent validity with another measure 10 and DSM-IV differences in diagnostic criteria for BPD of mania (r = 0.83; p ! 0.0001) [27]. Youngstrom et al. [28] (ICD-10 requires at least 2 episodes of significantly dis- reported a single-factor structure and good internal con- turbed mood and activity), clinician bias against BPD sistency (␣ = 0.91) in children and adolescents (5–17 years (preference to classify such clinical presentations as se- of age). vere ADHD, conduct, depressive and personality disor- Studies for validating the YMRS in children have fo- ders instead of BPD), an overdiagnosis of the disorder cused on the scale’s effectiveness in discriminating BPD in the USA and/or a true higher prevalence of pediatric from ADHD or another disorder. Until now, no study has BPD in the USA [16] . assessed the utility of the YMRS in discriminating mania Another aspect that makes diagnosis difficult is the in children with ADHD. We do not currently have any frequent overlap of BPD and ADHD symptoms (distract- validated scale for the Spanish population that specifi- ibility, high levels of motor activity, racing thoughts, rap- cally assesses pediatric mania. The objective of this study id or accelerated speech, impulsivity, aggression and ir- is to analyze the reliability and validity of the Spanish ver- ritability) [12]. Despite considerable symptom overlap, sion of the YMRS for children and adolescents with mania can be distinguished from ADHD by the presence ADHD and thus evaluate the utility of the scale to detect of core symptoms of mania such as euphoria and grandi- and quantify symptoms of mania in ADHD. osity [17] . Nonetheless, the structure of mania may be

126 Psychopathology 2011;44:125–132 Serrano /Ezpeleta /Alda /Matalí /San

Table 1. Sample description and comparison of scores on mania and functioning scales

Total A DHD with mania (n = 14) ADHD p (n = 100) without mania BPD (DSM) BPD-NOS BPSD (n = 86) (n = 8) (n = 6) (n = 14)

Median age, years 11 13.50 12 13 11 0.082 (range = 8–17) (range = 8–17) (range = 8–17) (range = 8–17) (range = 8–17) Sex Boys 84 (84) 7 (87.5) 6 (100) 13 (92.9) 71 (82.6) 0.458* Girls 16 (16) 1 (12.5) 0 1 (7.1) 15 (17.4) ADHD type Combined 75 (75) 7 (87.5) 5 (83.3) 12 (85.7) 63 (73.3) 0.508* Hyperactivity 9 (9) 0 1 (16.7) 1 (7.1) 8 (9.3) 1* Inattentive 16 (16) 1 (12.5) 0 1 (7.1) 15 (17.4) 0.458* Comorbidity Defiant disorder 71 (71) 8 (100) 6 (100) 14 (100) 57 (63.3) 0.001* Conduct disorder 15 (15) 4 (50) 5 (83.3) 9 (64.3) 6 (7) 0.001* CMRS-P 15 [0–53] 26 [3–53] 16.50 [10–31] 17 [3–53] 14 [0–43] 0.045 P-YMRS 12 [0–52] 18 [6–52] 17.50 [10–25] 18 [6–52] 11 [0–32] 0.017 YMRS 3 [0–33] 15 [3–33] 6.50 [5–27] 12 [3–33] 3 [0–16] 0.001 CGAS 55 [21–70] 40 [21–50] 50 [30–60] 41 [21–60] 59 [40–70] 0.001

Fig ures are numbers of cases with percentages in parentheses or medians with ranges in square brackets unless otherwise indi- cated. BPD (DSM) = (DSM); BPD-NOS = bipolar disorder not otherwise specified; BPSD = bipolar spectrum disorder; ADHD = attention deficit/hyperactivity disorder; CGAS = Children Global Assessment Scale; CMRS-P = Child Mania Rating Scale- Parent Version; P-YMRS = Parent-Young Mania Rating Scale; YMRS = Young Mania Rating Scale. p value: significance of Mann- Whitney test for the comparison of ADHD without mania and BPSD. * ␹2 test (Fisher’s exact test).

Methods (87.5%) had ADHD combined type and 1 case (12.5%) had inat- tentive type. With regard to BPD-NOS, 5 cases (83.3%) had com- Subjects bined type and 1 case (16.7%) had hyperactivity-impulsive type. A total of 100 children and adolescents between 8 and 17 years Therefore, in BPSD we found 12 cases (85.8%) with ADHD com- old (mean = 11.03; SD = 2.61) consecutively referred to an outpa- bined type, 1 case (7.1%) with ADHD hyperactivity-impulsive tient child and adolescent psychiatry and psychology unit of 3 type and 1 case (7.1%) with ADHD inattentive type, respectively. hospitals (2 of public health network and 1 private) in Barcelona The study excluded the subjects with a diagnosis of mental retar- (Spain) participated in the study. dation, pervasive developmental disorder, schizophrenia or se- To be part of the study, the children had to have received a vere neurological disorders. principal diagnosis of ADHD according to DSM-IV criteria [10] assessed with the Diagnostic Interview for Children and Adoles- Materials or Instruments cents-IV (DICA-IV) [29] . Table 1 shows the demographic and Diagnostic Interview for Children and Adolescents-IV clinical characteristics of the sample. The distribution of socio- Children and adolescents were diagnosed using the DICA-IV economic level was: 30% of the parents had a high professional [29], a semistructured diagnostic interview that assesses a wide level (professionals and executives), 35% had an average profes- range of psychological disorders in children and adolescents sional level (skilled workers), 32% had a low professional level (un- based on diagnostic criteria from the DSM-IV [10] . The Spanish skilled workers and manual laborers) and 3% were unemployed. version of the DICA-IV has been adapted and validated for the Seventy-five subjects (75%) had combined type ADHD, 16 Spanish population with satisfactory psychometric properties (16%) had predominately inattentive-type ADHD, and 9 (9%) had [30, 31]. The following sections were used: ADHD, oppositional hyperactive-impulsive-type ADHD. Seventy-one (71%) of the defiant disorder, conduct disorder and mania. The diagnoses children and adolescents with ADHD met the criteria for opposi- were established by child and adolescent clinical psychologists tional defiant disorder and 15 (15%) for conduct disorder. Eight combining the information provided by parents and children/ad- cases (8%) fulfilled DSM-IV criteria for BPD (DSM), 6 (6%) for olescents for each symptom assessed. The symptom was consid- bipolar disorder not otherwise specified (BPD-NOS) (see defini- ered present if the parents or child/adolescent referred it in the tion below) and 14 (14%) for bipolar spectrum disorder (BPSD) interview. For this study, the same clinician administered the (BPD + BPD-NOS) (table 1 ). Of the 8 cases with BPD (DSM), 7 DICA-IV to the parents and to the child/adolescent, taking into

Psychometric Properties of the YMRS in Psychopathology 2011;44:125–132 127 Spanish Children with ADHD Table 2. Comparison of means by diagnostic group, standard deviation (SD) and factor loading of the YMRS

YMRS items Total ADHD ADHD with mania (n = 14) ADHD Factor (n = 100) without mania without mania loading BPD (DSM) BPD-NOS BPSD (n = 86) vs. BPSD (n = 8) (n = 6) n = 14) p

1. Elevated mood 0.2480.63 0.0880.31 1.3881.06 181.26 1.2181.12 0.001 0.85 2. Increased motor activity/energy 0.9481.07 0.8080.99 1.8781.24 1.6781.21 1.7981.18 0.001 0.73 3. Sexual interest 0.1780.57 0.0680.23 1.2581.38 0.3380.81 0.8681.23 0.001 0.55 4. Sleep 0.3880.83 0.2480.59 1.2581.75 1.1781.16 1.2181.47 0.002 0.50 5. Irritability 0.7081.21 0.4780.95 2.2581.28 282.19 2.1481.65 0.001 0.53 6. Speech (rate and amount) 0.9881.34 0.8681.17 2.2582.25 181.67 1.7182.05 0.160 0.58 7. Language/thought disorder 0.2780.60 0.2180.51 0.8881.12 0.3380.51 0.6480.92 0.016 0.47 8. Thought content 0.2480.81 0.0780.36 1.7581.98 0.6781.03 1.2981.68 0.001 0.65 9. Disruptive/aggressive behavior 0.1080.52 0.0780.48 0.2580.70 0.3380.81 0.2980.72 0.093 0.37 10. Appearance 0.0780.38 0.0580.26 0.0080.00 0.5081.22 0.2180.80 0.726 0.30 11. Insight 0.8281.29 0.5781.04 2.6381.59 281.78 2.3681.64 0.001 0.57 Total score 4.9585.85 3.5083.35 1689.78 1188.87 13.8689.40 0.001 –

S ignificant differences are indicated in italics.

consideration independently the information provided by each son/daughter, in his/her mood or in his/her thoughts. Each ques- one. tion has 5 possible answers, from 0 (no symptoms/normal behav- In addition to the DSM diagnosis of BPD, BPD-NOS was also ior) to 4 (extreme deviation). As in the YMRS, questions 5, 6, 8 defined according to the criteria set out by Birmaher et al. [32] , and 9 have a double weight for calculating the total score. Each such as the presence of relevant BPD symptoms that did not fulfill one of the 11 questions on the P-YMRS occurs in the same order the DSM-IV criteria for BPD. Specifically, subjects needed to have and assesses the same symptom as the items described in the (1) (a) elated mood and 2 associated DSM-IV symptoms or (b) ir- YMRS. The aim of the P-YMRS is to have a standardized scale for ritable mood and 3 associated DSM-IV symptoms; (2) a change in parents used to assess the presence and severity of manic symp- the level of functioning; (3) duration of a minimum of 4 h within toms in children and adolescents aged 5–17 years. The original a 24-hour period, and (4) at least 4 cumulative lifetime days meet- version of the P-YMRS showed acceptable internal consistency ing the criteria. (Cronbach’s ␣ of 0.72) and good discriminant validity between BPD and unipolar disorder, disruptive behavior disorder and any Young Mania Rating Scale other diagnosis [34] . The P-YMRS has not been validated in Spain. This is an instrument developed by Young et al. [25] to be com- To conduct the study, the P-YMRS was translated from English to pleted by the clinician while conducting the interview on the pa- Catalan and from Catalan to English with permission of the au- tient (15–30 min). It is an 11-item scale (see table 2 for item con- thors. Each translation was done by an official translator in his/ tents). The clinician rates the severity of the symptoms from 0 (no her native language in conjunction with consultation to 3 clinical symptoms/normal behavior) to 4 (extreme deviation) based on experts. The decision to translate one of the scales into Catalan the subjective information provided by the patient about the last was based on the interest of the authors to have standardized ma- 48 h and the clinical observation of behavior during the interview. nia assessment instruments in a language that is the mother According to the developer of the questionnaire, items 5, 6, 8 and tongue of a significant part of the studied population. The Catalan 9 have a double weight for calculating the total score [25] . version of the P-YMRS showed adequate internal consistency Validation studies with pediatric outpatients obtained inter- (Cronbach’s ␣ of 0.71) in this sample. nal consistency (Cronbach’s ␣ ) between 0.80 and 0.91, and con- vergent validity with another measure of mania of r = 0.83 (p ! Child Mania Rating Scale-Parent Version 0.0001) [26–28] . The YMRS has become the most widely used This is a questionnaire for parents developed by Pavuluri et al. measure in the research of BPD in adults and children. [35] to screen for child mania. It contains 21 items and is based on The Spanish version, adapted for adults by Colom et al. [33] , DSM-IV symptoms for manic episode. Each question has 4 pos- was used to validate the YMRS in children and adolescents with sible answers on a Likert scale of 0 to 3, where 0 is never/rarely , 1 ADHD. is sometimes , 2 is often and 3 is very often. The original version of the Child Mania Rating Scale-Parent Version (CMRS-P) showed Parent-Young Mania Rating Scale good psychometric properties (Cronbach’s ␣ of 0.96) and conver- Developed by Gracious et al. [34] , this is the parent version of gent validity with the YMRS (r = 0.78, p ! 0.001), and differenti- the YMRS. It consists of 11 questions in which parents report on ated between children with mania and children with ADHD [35] . possible changes that they have observed in the behavior of their The scale is designed to be answered by parents and it takes about

128 Psychopathology 2011;44:125–132 Serrano /Ezpeleta /Alda /Matalí /San

10–15 min. There is currently no version of the scale adapted for which indicate a significant contribution of each item in the Spanish population. For this study, the scale was translated the factor. All factor loadings were 6 0.30. into Spanish with permission of the authors. The Spanish version was then backtranslated into English. The translations were done by official translators and it was reviewed by 3 clinical experts. In R e l i a b i l i t y this sample, the CMRS-P obtained good internal consistency Internal Consistency. Cronbach’s ␣ coefficient for the (Cronbach’s ␣ of 0.91). total scale was 0.83 for the entire sample. Test-Retest Reliability. The 1-week test-retest reliability Children’s Global Assessment Scale ! This scale, developed by Shaffer et al. [36], was used to assess in 32 patients was 0.95 (p 0.001). the degree of functional impairment of the child at the time of the examination. Children’s Global Assessment Scale (CGAS) scores C o n v e r g e n t V a l i d i t y range from 1 (maximum impairment) to 100 (normal function- Convergent validity was studied by correlating the 1 ing). Scores 70 reflect normal functioning. The psychometric scale with the P-YMRS and the CMRS-P, which are 2 properties of the scale adapted for the Spanish population have been studied with satisfactory results in terms of reliability (test- scales that aim to measure the same construct (manic retest and inter-interviewers) and validity (since it differentiated symptoms). The YMRS was significantly related to the P- groups of children with and without pathology and subjects with YMRS (B = 0.66; 95% CI = 0.37–0.95) and to the CMRS- a different number of disorders) [37] . P (B = 0.64; 95% CI = 0.27–1.01). The association of the YMRS with a functional impairment measure, the CGAS, P r o c e d u r e Once written consent to participate in the study had been ob- was also studied through linear regression, controlling tained from parents and children (1 12 years old) and verbal con- for sex and age. The results showed a significant associa- sent had been obtained from children (! 12 years old), parents and tion between both scales (B = –1.07; 95% CI = –1.34 to children and adolescents were interviewed separately with the –0.80). For every point of increase on the YMRS, the DICA-IV. After the interview the parents answered the P-YMRS functioning decreased 1.07 points. and the CMRS-P and the clinicians completed the YMRS and the CGAS. The YMRS was completed again after 1 week in 32 pa- tients to evaluate the test-retest reliability. P r e d i c t i v e V a l i d i t y Logistic regression analysis, controlling for sex and Statistical Analysis age, indicated that the score on the scale is significantly The data were analyzed with SPSS 17.0. The factor structure associated with the diagnosis of ADHD + BPD (DSM) was analyzed through confirmatory factor analysis with AMOS 17. The unweighted least squares method of estimation was used, (OR = 1.24; 95% CI = 1.09–1.4); the likelihood of diagnos- regarding nonnormality and sample size, but this method of esti- ing ADHD + BPD (DSM) versus ADHD without mania mation does not provide significance tests. In order to make ad- was multiplied by 1.24. The same analysis showed a sig- justment possible items 5, 6, 8 and 9 were recoded into a 5-point nificant association between the YMRS score and the di- Likert-type scale [38] . Internal consistency was calculated using agnosis of ADHD + BPSD (OR = 1.45; 95% CI = 1.17–1.8); Cronbach’s ␣ and test-retest reliability was analyzed by Pearson correlation. External validity was studied by relating the total for each point of increase on the scale, the likelihood of YMRS score with the total scores of the mania scales (P-YMRS and diagnosing ADHD + BPSD versus ADHD without mania CMRS-P) and functional impairment scale (CGAS) with linear was multiplied by 1.45. regression analysis, controlling for sex and age. The relationship between the total score on the scale and the ADHD with mania Discriminative Validity diagnosis was analyzed by logistic regression, also controlling for sex and age. Lastly, receiver operating characteristic (ROC) curves The means of the items of the YMRS and the total were analyzed, and the area under the curve was generated in or- score for the ADHD + BPSD group were significantly der to assess the ability of the YMRS to discriminate between di- greater than those of the ADHD without mania group in agnostic groups (ADHD with and without mania). all of the items except speech (rate and amount), disrup- tive/aggressive behavior and appearance ( table 2 ). The ROC curve analysis showed areas under the curve R e s u l t s between 0.90 and 0.88 to differentiate ADHD + BPD (DSM) and ADHD + BPSD from ADHD without mania, Factorial Structure respectively. A cutoff point between 6 and 7 achieved a The confirmatory factor analysis showed that the scale specificity of 0.78 and a sensitivity of 0.87 for ADHD + is 1-dimensional (SRMR = 0.117; GFI = 0.966; eigenval- BPD (DSM). A cutoff point between 5 and 6 achieved a ue = 4.206); a single factor explained 38.24% of the vari- specificity of 0.76 and a sensitivity of 0.78 to discriminate ance. Table 2 shows the standardized factor loadings, ADHD + BPSD ( fig. 1 and 2 ).

Psychometric Properties of the YMRS in Psychopathology 2011;44:125–132 129 Spanish Children with ADHD ROC curve ROC curve 1.0 1.0

0.9 0.9

0.8 0.8

0.7 0.7

0.6 0.6

0.5 0.5 Sensitivity 0.4 Sensitivity 0.4

0.3 0.3

0.2 0.2

0.1 0.1

0 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1 – specificity 1 – specificity

Fig. 1. Diagnostic accuracy of YMRS to differentiate ADHD + Fig. 2. Diagnostic accuracy of YMRS to differentiate ADHD + BPD (DSM) from ADHD without mania. BPSD from ADHD without mania.

Discussion increased motor activity/energy, sexual interest, sleep, ir- ritability, language/thought disorder, thought content and The results obtained from validating the YMRS sug- insight. Geller et al. [6] found that grandiosity, elated gest that the Spanish version of the YMRS is a valid and mood, hypersexuality, flight of ideas and decreased need reliable instrument for detecting and quantifying the for sleep differentiated children with BPD from ADHD. symptoms of mania in children and adolescents with It is worth noting that children with BPD were included ADHD. The YMRS used on children and adolescents in the study if they had grandiosity and/or elation. with ADHD shows satisfactory psychometric properties On the other hand, speech (rate and amount) , disrup- and differentiates between ADHD patients with and tive/aggressive behavior and appearance were not dis- without mania. criminatory. The appearance of children, unlike that of The factor analysis of the scale indicates that it is 1-di- adults, is not directly related to psychopathology because mensional; a single factor explains the largest percentage the way children dress greatly depends on their parents. of the variance, which indicates that it assesses a single This item also had the lowest factor loading in the factor construct (mania). analysis. Speech (rate and amount) and disruptive/aggres- In addition to good reliability (internal consistency sive behavior were the manic symptoms that neither dis- and test-retest reliability), the validity of the Spanish ver- criminated between ADHD and ADHD + BPSD. Disrup- sion of the YMRS for children with ADHD is shown by tive/aggressive behavior also occurs in ADHD when it is (1) the significant association with other measures of ma- associated with oppositional defiant and conduct disor- nia, the P-YMRS and the CMRS-P, (2) the significant as- der. However, the association between conduct problems sociation between the YMRS and functional impairment and BPD in adolescents should be noted. In our sample, measured with the CGAS, (3) the diagnostic effectiveness we found 64% of conduct disorder in children and ado- shown in the ROC curve analysis, and (4) the significant lescents with ADHD + BPSD. These results are similar to association between the total score on the scale and the 69% of lifetime comorbidity of BPD with conduct disor- diagnosis of ADHD + BPD. der found for Kovacs and Pollok [39] . Endrass et al. [40] The manic symptoms of the YMRS that best differen- found that children and adolescents showing behavior tiated BPD in children with ADHD were elevated mood, such as repeated running away from home and physical

130 Psychopathology 2011;44:125–132 Serrano /Ezpeleta /Alda /Matalí /San

fighting were 2.6–3.5 times more likely to experience a NOS progress to meet full criteria for BPD-I or BPD-II bipolar II disorder than those without any indication of within 2 years. These results provide preliminary valida- conduct problems. Longitudinal studies have confirmed tion for the definition of BPD-NOS. impaired social functioning as a predictor for mood dis- Other authors are skeptical of the new phenotypes de- orders and BPD in particular [41–42] . fined by US researchers and affirm that pediatric BPD Finally, the item speech (rate and amount) also overlaps phenotypes remain hypothetical [48] . with the rapid or accelerated speech in children with In any case, the results of this study provide further ADHD [12] . knowledge about the association between ADHD and Psychometric properties obtained in the Spanish ver- manic symptoms in children and adolescents in a non- sion of the YMRS for children and adolescents with Anglo-Saxon sample. Given the high prevalence of ADHD are similar in factorial structure, internal consis- ADHD, which is between 3 and 5% in school-age children tency, test-retest reliability and validity data to the valida- [10] , a co-occurrence of 14% of BPSD represents a consid- tion results for the American population of the original erable number of comorbid cases. Keeping in mind the version of the YMRS for child and adolescent outpatients worse prognosis of ADHD associated with mania [22– [26–28] . 24] , proper detection of comorbid cases with BPD (aided The validity of the diagnosis of mania (and BPD-NOS) by specific scales for assessing pediatric mania) will de- in children and adolescents has been under debate, espe- termine the choice of treatment, on both a pharmaco- cially in Europe. Youngstrom et al. [43] reviewed the clas- logical and psychosocial level. sic diagnostic validators applied to pediatric BPD. They noted that studies of phenomenology could inform diag- L i m i t a t i o n s nostic progress in children with BPD. In this way, elated The study poses some limitations arising mainly from mood is highly specific to BPD (45–87%) [32, 44–47] and the limited number of participants with comorbidity irritability (present in 1 80%) [32, 44, 46] appears to be the with BPSD. The sample size of the ADHD with mania complement of elated mood. The genetic component of group should be increased in subsequent studies. Anoth- diagnostic validation appears important as the presence er limitation was the fact that the same clinician com- of bipolar illness in the parents of such children is an im- pleted DICA-IV interviews with parents and children for portant diagnostic marker. The course of illness tends to the diagnosis, which could have biases the assessment. involve mixed episodes with rapidly fluctuating mood However, in other interviews, as the Schedule for Affec- states, and such an early onset of bipolar illness may pre- tive Disorders and Schizophrenia for School-Age Chil- dict worse prognosis into adulthood. With respect to dren, the assessment of parents and children is made by treatment response, children with BPD treated with an- the same interviewer. tidepressants and/or amphetamines may lead to mania or mixed states. There is an emerging consensus about mor- phological and functional changes associated with pedi- Acknowledgements atric BPD. We want to thank Dr. E. Penelo for her advice on confirma- With regard to BPD-NOS, Birmaher et al. [32] found tory factor analysis. This study was supported, in part, by grant that 25% of the youths initially meeting criteria for BPD- PSI2009-07542 of the Ministry of Science and Innovation (Spain).

References

1 Akistal HS, Downs J, Jordan P: Affective dis- 3 Geller B, Fox LW, Clark K: Rate and predic- 6 Geller B, Zimerman B, Williams M, Bolhof- orders in referred children and younger sib- tors of prepubertal bipolarity during follow ner K, Craney JL, Delbello MP: Diagnostic lings of manic-depressives: mode of onset up of 6- to 12-year-old depressed children. characteristics of 93 cases of a pre-pubertal

and prospective course. Arch Gen Psychia- Am J Psychiatry 1994; 33: 461–468. and early adolescent bipolar disorder pheno-

try 1985; 42: 996–1003. 4 Carlson AG, Kelly KL: Manic symptoms in type by gender, puberty and comorbid atten- 2 Duffy A, Alda M, Kutcher S, Fusee C, Grof psychiatrically hospitalized children – what tion deficit disorder. J Child Adolesc Psycho-

P: Psychiatric symptoms and syndromes do they mean? J Affect Disord 1998; 51: 123– pharmacol 2000; 10: 157–164. among adolescent children of parents with 135. 7 Pavuluri MN, Birmaher B, Naylor MW: Pe- lithium-responsive or lithium-nonrespon- 5 Geller B, Luby J: Child and adolescent bipo- diatric bipolar disorder: a review of the past

sive bipolar disorder. Am J Psychiatry 1998; lar disorder: a review of the past 10 years. J 10 years. J Am Acad Child Adolesc 2005; 44:

155: 431–433. Am Acad Child Adolesc 1997; 36: 1168–1176. 846–871.

Psychometric Properties of the YMRS in Psychopathology 2011;44:125–132 131 Spanish Children with ADHD 8 Wozniak , Biederman J, Kiely K, Ablon JS, 21 Akiskal S, Azorin JM, Hantouche EG: Pro- 35 Pavuluri MN, Henry DB, Devineni B, Car- Faraone SV, Mundy E, Mennin D: Mania- posed multidimensional structure of mania: bray JA, Birmaher B: Child Mania Rating like symptoms suggestive of childhood-on- beyond the euphoric-dysphoric dichotomy. J Scale (CMRS): development, reliability and

set bipolar disorder in clinically referred Affect Disord 2003; 73: 7–18. validity. J Am Acad Child Adolesc 2006; 45:

children. J Am Acad Child Adolesc 1995; 34: 22 Biederman J, Faraone SV, Chu MP, Wozniak 550–560. 867–876. J: Further evidence of a bi-directional over- 36 Shaffer D, Gould MS, Brasic J, Ambrosini P, 9 National Institute for Mental Health: Na- lap between juvenile mania and conduct dis- Fisher P, Bird H, Aluwahlia S: A children’s tional Institute for Mental Health research order in children. J Am Acad Child Adolesc global assessment scale (CGAS). Arch Gen

roundtable on pre-pubertal bipolar disorder. 1999; 38: 468–476. Psychiatry 1983; 40: 1228–1231.

J Am Acad Child Adolesc 2001; 40: 871–878. 23 Biederman J, Mick E, Faraone SV, Van Patten 37 Ezpeleta L, Granero R, Osa N de la: Evalu- 10 American Psychiatric Association: Diagnos- S, Burback M, Wozniak JA: Prospective fol- ación del deterioro en niños y adolescentes a tic and Statistical Manual of Mental Disor- low-up study of pediatric bipolar disorder in través de la Children’s Global Assessment

der, ed 4. Washington, American Psychiatric boys with attention-deficit hyperactivity Scale (CGAS). Rev Psiquiatr Infant Juv 1999;

Association, 1994. disorder. J Affect Disord 2004; 82: 25–34. 1: 18–26. 11 Leibenluft E, Charney DS, Towbin KE, 24 Geller B, Tilman R, Craney JL, Bolhofner K: 38 Jöreskog KG (2005): Structural equation Bhangoo RK, Pine DS: Defining clinical phe- Four-year prospective outcome and natural modeling with ordinal variables using LIS- notypes of juvenile mania. Am J Psychiatry history of mania in children with a prepu- REL. http://www.ssicentral.com/lisrel/tech-

2003; 160: 430–437. bertal and early adolescent bipolar disorder docs/ordinal.pdf (accessed November 16.

12 Biederman J, Faraone S, Mick E, Wozniak J, phenotype. Arch Gen Psychiatry 2004;61: 2007). Chen L, Ouellette C, Marrs A, Moore P, Gar- 459–467. 39 Kovacs M, Pollock M: Bipolar disorder and cia J, Mennin D, Lelon E: Attention-deficit 25 Young RC, Biggs JT, Ziegler VE, Meyer DA: comorbid conduct disorder in childhood hyperactivity disorder and juvenile mania: A rating scale for mania: reliability, validity, and adolescence. J Am Acad Child Adolesc

an overlooked comorbidity? J Am Acad and sensibility. Br J Psychiatry 1978; 133: Psychiatry 1995; 34: 715–723.

Child Adolesc 1996; 35: 997–1008. 429–435. 40 Endrass J, Vetter S, Gamma A, Gallo WT, 13 Butler FS, Arredondo DE, McCloskey V: Af- 26 Fristad MA, Weller EB, Weller RA: The ma- Rossegger A, Urbaniok F, Angst J: Are be- fective comorbidity in children and adoles- nia rating scale; can it be used in children? A havioral problems in childhood and adoles-

cents with attention deficit hyperactivity preliminary report. Am J Psychiatry 1992; 31: cence associated with bipolar disorder in

disorder. Ann Clin Psychiatry 1995; 7: 51–55. 252–257. early adulthood? Eur Arch Psychiatry Clin

14 Diler RS, Uguz S, Seydaoglu G, Erol N, Avci 27 Fristad MA, Weller RA, Weller EB: The ma- Neurosci 2007; 257: 217–221. A: Differentiating bipolar disorder in Turk- nia rating scale (MRS): further reliability 41 Carlson GA, Weintraub S: Childhood behav- ish prepubertal children with attention-def- and validity studies with children. Ann Clin ior problems and bipolar disorder – relation-

icit hyperactivity disorder. Bipolar Disord Psychiatry 1995; 7: 127–132. ship or coincidence? J Affect Disord 1993; 28:

2007; 9: 243–251. 28 Youngstrom EA, Danielson CK, Findling 143–153. 15 Meltzer H, Gatward R, Goodman R, Ford T: RL, Gracious BL, Calabrese JR: Factor struc- 42 Lewinsohn PM, Klein DN, Seeley JR: Bipolar Mental Health of Children and Adolescents ture of the Young Mania Rating Scale for use disorders in a community sample of older in Great Britain. London, Stationery Office, with youths ages to 5 to 17 years. J Clin Child adolescents: prevalence, phenomenology,

2000. Adolesc Psychol 2002; 31: 567–572. comorbidity, and course. J Am Acad Child

16 Soutullo CA, Chang KD, Diez-Suarez A, 29 Reich W: Diagnostic Interview for Children Adolesc Psychiatry 1995; 34: 454–463. Figueroa-Quintana A, Escamilla-Canales I, and Adolescents (DICA). J Am Acad Child 43 Youngstrom EA, Birmaher B, Findling RL:

Rapado-Castro M, Ortuno F: Bipolar disor- Adolesc 2000; 39: 59–66. Pediatric bipolar disorder: validity, phenom- der in children and adolescents: internation- 30 Ezpeleta L, de la Osa N, Júdez J, Doménech enology, and recommendations for diagno-

al perspective on epidemiology and phenom- JM, Navarro JB, Losilla JM: Diagnostic agree- sis. Bipolar Disord 2008; 10: 194–214.

enology. Bipolar Disord 2005; 7: 497–506. ment between clinician and the Diagnostic 44 Kowatch RA, Youngstrom EA, Danielyan A, 17 Geller B, Zimerman B, Williams M, Delbello Interview for Children and Adolescents – DI- Findling RL: Review and meta-analysis of MP, Bolhofner K, Craney JL, Frazier J, CA-R in a Spanish outpatient sample. J Child the phenomenology and clinical characteris-

Beringer L, Nickelsburg MJ: DSM-IV mania Psychol Psychiatry 1997; 38: 431–440. tics of mania in children and adolescents. Bi-

symptoms in a prepubertal and early adoles- 31 Ezpeleta L, Osa N de la, Domenech JM, Na- polar Disord 2005; 7: 483–496. cent bipolar disorder phenotype compared varro JB, Losilla JM: Fiabilidad test-retest 45 Findling RL, Gracious BL, McNamara NK, to attention-deficit hyperactive and normal de la adaptación española de la Diagnostic Youngstrom EA, Demeter CA, Branicky LA, controls. J Child Adolesc Psychopharmacol Interview for Children and Adolescents – Calabrese JR: Rapid, continuous cycling and

2002; 12: 11–25. DICA-R. Psicothema 1997; 9: 529–539. psychiatric co-morbidity in pediatric bipolar

18 Biederman J, Faraone SV, Wozniak J, Mick E, 32 Birmaher B, Axelson D, Strober M, Gill MK, I disorder. Bipolar Disord 2001; 3: 202–210. Kwon A, Cayton GA, Clark SV: Clinical cor- Valeri S, Chiappetta L: Clinical course of 46 Axelson D, Birmaher B, Strober M, Gill MK, relates of bipolar disorder in a large, referred children and adolescents with bipolar spec- Valeri S, Chiappetta L, Ryan N, Leonard H,

sample of children and adolescents. J Psychi- trum disorders. Arch Gen Psychiatry 2006; Hunt J, Iyengar S, Bridge J, Keller M: Phe-

atr Res 2005; 39: 611–622. 63: 175–183. nomenology of children and adolescents 19 Hunt JI, Dyl J, Armstrong L, Litvin E, Sheer- 33 Colom F, Vieta E, Martínez-Arán A, Garcia- with bipolar spectrum disorders. Arch Gen

an T, Spirito A: Frequency of manic symp- Garcia M, Reinares M, Torrent C, Goikolea Psychiatry 2006; 63: 1139–1148. toms and bipolar disorder in psychiatrically JM, Banús S, Salamero M: Versión española 47 Geller B, Williams M, Zimerman B, Frazier hospitalized adolescents using the K-SADS de una escala de la manía: validez y fiabilidad J, Beringer L, Warner KL: Prepubertal and

Mania Rating Scale. J Child Adolesc Psycho- de la Escala de Young. Med Clin (Barc) 2002; early adolescent bipolarity differentiate from

pharmacol 2005; 15: 918–930. 119: 366–371. ADHD by manic symptoms, grandiose delu- 20 Youngstrom E, Youngstrom JK, Starr M: Bi- 34 Gracious BL, Youngstrom EA, Findling RL, sions, ultra-rapid or ultradian cycling. J Af-

polar diagnoses in community mental Calabrese JR: Discriminative validity of a fect Disord 1998; 51: 81–91. health: Achenbach child behavior checklist parent version of the Young Mania Rating 48 Parry P, Furber G, Allison S: The Paediatric

profiles and patterns of comorbidity. Biol Scale. J Am Acad Child Adolesc 2002;41: Bipolar Hypothesis: the view from Australia

Psychiatry 2005; 58: 569–575. 1350–1359. and New Zealand. Child Adolesc Ment

Health 2009; 3: 140–147.

132 Psychopathology 2011;44:125–132 Serrano /Ezpeleta /Alda /Matalí /San