Ciência & Saúde Coletiva ISSN: 1413-8123 [email protected] Associação Brasileira de Pós-Graduação em Saúde Coletiva Brasil

Wilkinson, Paul Conceptualization about internalizing problems in children and adolescents Ciência & Saúde Coletiva, vol. 14, núm. 2, abril, 2009, pp. 373-381 Associação Brasileira de Pós-Graduação em Saúde Coletiva Rio de Janeiro, Brasil

Available in: http://www.redalyc.org/articulo.oa?id=63013532007

How to cite Complete issue Scientific Information System More information about this article Network of Scientific Journals from Latin America, the Caribbean, Spain and Portugal Journal's homepage in redalyc.org Non-profit academic project, developed under the open access initiative 373 ARTIGO ARTICLE Internalização, Análise fatorial de

Esta revisão discute o conceito de trans-

tornos internalizantes, descrevendo os dois princi- tornos internalizantes, descrevendo os dois Será pais tipos deste problema: depressão e ansiedade. tam- discutido o quanto eles têm em comum, mas ilustrar bém as principais diferenças entre Para eles. se- estas características em comum e as diferenças, aná- rão usados dados de estudos modernos usando e prog- lise fatorial de sintomas, etiologia, tratamento de nóstico. Na conclusão, será feita uma tentativa responder a questão onde os problemas internali- zantes deveriam ser inseridos nos esquemas futuros de classificação diagnóstica. Palavras-chave Resumo sintomas, Classificação diagnóstica Internalizing disorders, Symptom fac- disorders, Internalizing

This review will discuss the concept of Conceptualization about internalizing problems internalizing about Conceptualization adolescents in children and e adolescentes internalizantes em crianças O conceito de problemas internalizing disorders. It will describe the two main internalizing disorders. It will describe the two anxi- types of internalizing disorder: depressive and much in ety disorders. It will discuss how they have The common, but that there are also key differences. of symp- review will use data from modern studies treatment analysis, and prog- factor aetiology, tom differen- nosis to illustrate the commonalities and in- ces. It will conclude by trying to answer where ternalizing disorders should be placed in future di- agnostic classification schemes. words Key tor analysis, Diagnostic classification Abstract 1 Department of Psychiatry, 1 University of Cambridge. Road 18b Trumpington Cambridge CB2 8AH UK. [email protected] Paul Wilkinson 374 Wilkinson P The first attempt to classify child psychopatholo- Depressive disorders gy was made by Hewitt and Jenkins in 19461. They carried out a factor analysis of symptoms and Depressive disorders are about more than feel- linked factors with social situation. This led to the ing sad. They are syndromes of persistent emo- differentiation of “emotional disturbance” and two tional, biological and psychological symptoms, classes of disruptive disorders, based both on in- accompanied by impaired social functioning. The tercorrelations between symptoms within each fac- most important of the depressive disorders is ma- tor and different patterns of psychosocial variables jor . in the three factors. Since then, there has developed For a DSM-IV (Diagnostic and Statistical Man- infinitely greater complexity to psychiatric diagno- ual of Mental Disorders, fourth edition4) diagno- sis. Yet a similar split between emotional (or “in- sis of major depression, at least five depressive ternalizing”, where the patients feel distress inside symptoms must have been present most of the themselves) and behavioural (or “externalizing”, time for two weeks and represent a change from where the patients cause distress to people external previous functioning. At least one of the symp- to themselves) problems is still often made. toms must be: This article will look in detail at the conceptual- . Depressed or irritable mood ization of internalizing disorders in children and . Markedly diminished interest or pleasure in adolescents. It will explain the main disorders, look- almost all activities (anhedonia). ing at what separates the disorders, but also at Other possible symptoms are: what these disorders have in common, validating . Decreased or increased weight or appetite Hewitt and Jenkins’ original classification. It will . Increased or decreased sleep discuss the aetiology of these disorders, and how . Psychomotor agitation (fidgetiness) or retar- modern scientific technology has been able to dem- dation (slowed down speech/movements) onstrate differences and commonalities within this . Reduced energy group of disorders. It will finish by discussing . Worthlessness or excessive guilt whether the concept of internalizing disorders (and . Reduced concentration or indecisiveness the sub-classification of internalizing disorders) is . Recurrent thoughts of death or suicidal ide- helpful and valid. ation. In addition, there must be significant distress or impairment in functioning (such as at school, Classification of internalising disorders with friends or with the family). Symptoms must not be due to a medical condition, medication, il- The two main emotions suffered by patients with licit substances or bereavement. Onset is rare be- emotional disorders are sad (or depressed) mood fore adolescence. and worry (or ). These emotions form the However, here we have the first of our prob- basis of the two main types of internalizing disor- lems with the conceptualisation of internalising ders: depressive and anxiety disorders. The third disorders. As clinicians, it is convenient to be able edition of the Diagnostic and Statistical Manual of to say whether or not people have a diagnosis, a Mental Disorders (DSM-III,1980) reflected this, categorical approach. This makes it easy at a su- and was the first diagnostic classification that made perficial level to decide what treatment we should a major separation between “affective” and “anxi- use, and whether somebody is eligible for treat- ety” disorders. ment at all. However, human biology and psychol- This article will focus on these two main groups ogy is not so simple. In the same way as blood of disorders. It will not include bipolar disorder pressure lies on a continuum, and at one end of and obsessive-compulsive disorder. The phenom- this, the doctor and patient must decide whether enology, biology and genetics of these disorders to treat “high” blood pressure, depressive symp- demonstrate that while these disorders have much toms and the associated social dysfunction lie upon in common with internalizing disorders, they share a continuum. The American Diagnostic and Sta- many features of, respectively, schizophrenia2 and tistical Manual of Mental Disorders, 4th edition tic disorders3. This overlap demonstrates part of (DSM-IV)5 defines a threshold of five symptoms. the difficulty of a simple split of psychiatric disor- But this is fairly arbitrary, and is not based on any ders into a small number of discrete groups. empirical data. It has been shown that there is a gradual reduction in social functioning with more symptoms in children and adolescents6 and in adults7 and that older adolescents with subthresh- 375 Ciência & Saúde Coletiva, 14(2):373-381, 2009

old depressive symptoms are as likely as those with tion optimally in a fight-or-flight dangerous situ- the full depressive syndrome to develop a subse- ation. Sometimes anxiety is maladaptive, and it is quent depressive disorder8. So it is not helpful to here that we see anxiety disorders. People may have clinicians to use a simple yes/no approach to treat- excess anxiety around situations that they need to ment. Instead it is better to consider the number of be in to function normally (such as school or symptoms, the social context, the level of dysfunc- shops). This anxiety can be very distressing and tion, the duration of symptoms and the views of can be so overwhelming that it affects normal func- the patient on treatment. The more symptoms, the tion in these arenas (such as stopping children from more likely it is that we should use treatments. concentrating on their schoolwork). In extreme Clinicians do not feel it right to ignore the pa- cases, anxiety can be so overwhelming that people tients who are clearly suffering but only have four cannot face the feared situation, and avoid it com- depressive symptoms. In some healthcare systems, pletely. This avoidance leads to a great feeling of it is not even possible to get treatment without a relief and anxiety levels fall. Sadly, this benefit is “diagnosis”. Other depressive disorders have since greatly reinforcing, and makes it even harder for been described. Dysthymic disorder is a more chron- that person to face the feared situation on the next ic, less severe condition, consisting of depressed occasion. Anxiety is often accompanied by physi- mood for most of the time, with at least two other cal symptoms, such as heart racing, stomach pains depressive symptoms, lasting at least one year, and and sweating. These are mediated by the sympa- in which criteria for a major depressive episode are thetic nervous system/adrenergic response, that not met for the first year. Many children with dys- would be useful in a fight-flight dangerous situa- thymia later develop major depression. tion, but which often makes feelings even worse in DSM-IV discussed “minor depression” as 2-4 the non-dangerous anxiety producing situation. depressive symptoms, including depressed/irrita- As with depressive disorders, anxiety disorders ble mood or anhedonnia, over a two week period, lie on a spectrum, from the common feeling of but concluded there was not sufficient evidence to slight anxiety at the start of a new school year, to include minor depression as a formal diagnosis. children unable to leave their house for many years Instead, depressive disorder, not otherwise speci- because of fear of what people may think of them. fied, was included, to refer to people with depres- We do not have a simple symptom count to help sive problems, but not enough criteria for the full us divide people into those with and without a dis- diagnosis. Minor depression has been shown to order (some may say this is a blessing!). Instead, have financial costs intermediate between those of diagnosis is determined by the level of distress and major depression and no depression in adults9. how impaired the patient is by the anxiety, in par- Randomised controlled trials have demonstrat- ticular whether it leads to avoidance. Of course, ed that the standard treatments for depression, normal development must be taken into account. and psychological therapies, are Great anxiety and protest at being left by the mother effective for minor depression10 and dysthymia11,12 in a shop would be normal, indeed adaptive, for 15 in adults. This again demonstrates the dimension- month olds. It is not for a 15 year old. al nature of depression, and the inappropriateness Anxiety disorders can be further subclassified of using the cut-off of 5 symptoms for major de- by the stimuli which the patient is anxious about. pression in deciding upon suitability for treatment. Some anxiety disorders present at earlier ages than Adjustment disorder is emotional and/or be- others, reflecting the normal children have havioural symptoms within 3 months of an iden- at different ages. Many children have more than tified stressor. Caseness is determined by either one anxiety disorder at the same time, in particular marked distress or social dysfunction in excess of generalized anxiety disorder is often found along- what would be expected from such a stressor. One side another anxiety disorder13. subtype is “with depressed mood”. Such a diagno- Specific phobic disorder is anxiety around a spe- sis cannot be made if full criteria for major depres- cific object (e.g., spiders) or situation (e.g., flying), sion are met. and is present whenever the person encounters the specific object. Onset is often in early childhood. Anxiety disorders Separation anxiety disorder is anxiety at sepa- ration from caregivers, either when the child leaves Anxiety can be useful and adaptive, warning us the caregiver (e.g., to go to school) or when the that we should avoid somewhere dangerous. The caregiver leaves the child at home when they go simultaneous adrenergic response (high heart rate, out. It sometimes leads to school refusal. Onset fast breathing, sweating) would help our body func- peaks in late childhood. 376 Wilkinson P Social phobia (sometimes called social anxiety Comorbidity may occur for several reasons. It disorder) is anxiety in social situations due to fear may be an artifact of imperfect diagnostic systems. of other people´s evaluations and reactions and Some symptoms are present in depressive and anx- the possible resultant embarrassment or humilia- iety disorders, such as insomnia, poor concentra- tion. It can be generalized, present in most social tion and fatigue, therefore not many symptoms settings, or non-generalized, such as only present are needed of a second disorder to make it a full at parties, or when speaking or eating in front of comorbid disorder. However, non-overlapping others. The generalized form may have earlier age symptoms are also present for both disorders in of onset, a worse prognosis and more psychopa- comorbid patients, and so this can only be a par- thology in parents14. Onset is often in adolescence tial explanation at best. or adulthood. Secondly, one disorder may “cause” the other. Agoraphobia is anxiety about being away from For example, a chronic anxiety disorder with its home especially in crowds, where they cannot leave distress and the effects on socializing with friends easily, such as in shops or on buses. It is often co- and going to school may make a person sad, then morbid with panic disorder, and the main focus of depressed. Anxiety disorders precede depression in the anxiety may then be worry that the person will two thirds of cases where both are present17,18. have a panic attack in public and not be able to Thirdly, common predisposing factors may in- escape. Onset is often in adolescence or adulthood. crease the risk of both disorders. An adult twin study Panic disorder is the repeated experience of sud- demonstrated that the liability to major depression den, unprovoked panic attacks, with intense anxiety and generalized anxiety disorder is influenced by and physical symptoms. It sometimes leads to ago- the same genetic factors, and that different envi- raphobia. Onset is often in adolescence or adulthood. ronmental factors must therefore determine which Generalized anxiety disorder is multiple wor- disorder an individual develops19. This shared ge- ries about many aspects of life, rather than about netic liability is partially mediated by the personal- specific stimuli, with resultant distress and reduced ity trait of neuroticism20. It has been demonstrated functioning. Worry is the main symptom, and in adults that the life events of humiliation, bereave- avoidance is not common. It is often found with ment and respondent-initiated separation predict- other psychiatric disorders. When found only with- ed major depression but not generalized anxiety, in a depressive episode, the diagnosis cannot be while dangerous life events predicted generalized given, as these symptoms are also part of depres- anxiety but not depression. Life events that predict- sion. Onset is often in adolescence or adulthood. ed mixed major depression and generalized anxiety were the sum of those that predicted the individual disorders21. A smaller study in children also dem- Comorbidity between depressive onstrated that loss events, family and friendship and anxiety disorders problems and schoolwork stress were significantly associated with high depressive, but not anxiety, It is common for people to have more than one symptoms; threat events were associated with high psychiatric disorder at the same time, a phenome- anxiety, but not depressive, symptoms22. However non called “comorbidity”. It has been demonstrat- it has been demonstrated that one genetic factor ed that in people with one psychiatric disorder, the underlies phobic and panic disorders while a sepa- presence of a second disorder is more than can be rate factor may underly depressive and generalized accounted for by the prevalence of the second dis- anxiety disorders23. order in the healthy population and chance15. Fourthly, comorbid depression and anxiety Comorbidity is certainly a feature of anxiety may be a different syndrome to each of the indi- and depressive disorders. In a large meta-analysis vidual diagnoses. Tyrer in particular has argued of community epidemiology studies, 32% of chil- for the existence of a “cothymia”, a disorder with dren/adolescents with major depression also had both anxiety and depression present at a syndro- an anxiety disorder and 24% of adolescents with mal level24. In a meta-analysis of studies of adults, an anxiety disorder had major depression16. The such “cothymia” has worse prognosis than either odds ratio for the other disorder to be present in disorder alone 25. In adolescents, comorbid depres- probands with one disorder compared with sion has been shown to worsen the outcome of probands without that disorder was 8.2. However, anxiety disorders 26. It is possible that the worse comorbidity is not specific to internalizing disor- prognosis of combined depression is because risk ders: there is also great comorbidity between inter- factors for both are present, and so this is simply a nalizing and externalizing disorders. more severe illness, rather than a different illness. 377 Ciência & Saúde Coletiva, 14(2):373-381, 2009

However, there may be differences at a biological Treatment level. High salivary cortisol has been demonstrat- ed in the evening in depressed people in several There are three main areas to consider when studies27, but in one of these groups, evening corti- making a treatment plan: biological, psychological sol level was similar to that of healthy controls in and social. Social treatment, the improving of a depressed adolescents with comorbid panic or patient´s environment, should always be guided by phobic disorder28. the problems in that specific environment, more than by the diagnosis. Therefore I shall not consid- er it further in this discussion on whether separa- Depressive and anxiety disorders – tion of anxiety and depressive disorders is helpful. two truly different disorders? The most basic level of psychological treatment is psychoeducation and non-specific supportive lis- DSM-III separated depressive and anxiety disor- tening. Again, this is not diagnosis-specific. The ders. Was this the correct thing to do? The main most widely used specific psychological therapy, distinction between the disorders is the core emo- with the greatest evidence-base, for both anxiety31, tion – depressed mood vs. anxiety. Yet the second- 32 and depressive33,34 disorders is cognitive-behav- ary symptoms may be seen in both types of disor- ioural therapy (CBT). However, the techniques used der. And there is great overlap, with many people in CBT are rather different in the two disorders. In with one disorder also having the other. There may anxiety disorders, the focus is on facing the feared be a common genetic basis underlying the two dis- situations, moving up a hierarchy of more difficult orders. Are we best keeping this distinction? Or and anxiety provoking situations, while dealing would we be better changing to a common depres- with any inappropriate and maladaptive thoughts sive/anxiety disorders classification, with patients about the anxiety-provoking stimuli. In depres- lying on a depressed-both-anxiety spectrum? Or sion, the focus is on encouraging taking part in should we separate the classification further into pleasurable activities, rather than ruminating about depression only, anxiety only, and mixed anxiety- problems, and challenging inappropriate, mal- depression? adaptive and negative thoughts about the self, the One way to answer this question is to use more world and the future. Meta-analysis has demon- modern statistical techniques than Hewitt and Jen- strated that remission rates in depressed adoles- kins to investigate which symptoms cluster in very cents randomized to CBT are 48% and to placebo large samples of children with psychiatric distur- 34%35; the respective figures for anxiety disorders bance. Such studies have also shown a split be- are 57% and 35%32. Different statistical methods tween internalizing and externalizing disorders29. make it not possible to compare relative effective- They have also demonstrated two correlated sub- ness of the two treatments. factors within internalising disorders: “fear”, in- The other specific psychological therapy with cluding simple phobia, social phobia, panic disor- proven effectiveness against paediatric depression der and separation anxiety disorder; and “distress/ is interpersonal therapy36, 37. It was specifically de- misery”, including depressive disorders and gener- signed to treat depression, and looks at the rela- alized anxiety disorder. A study of young adults tionship between affect and interpersonal relation- showed that there is an underlying latent “inter- ships, and tries to improve relationships, thus im- nalising” trait responsible for some of the variance proving mood. While effective against depression in anxiety and depressive disorders but also addi- in many adults and child studies, it did not dem- tional disorder-specific traits responsible for some onstrate greater effectiveness against anxiety dis- of the variance30. orders than supportive therapy, in the only study The best way to answer this question is to think to date (with adults)38. about what would help us as we see our patients. The name of the medication class “antidepres- There are three main ways a diagnosis is useful. It sants” suggests they are for treating depression. should guide us towards specific treatments. It Research has demonstrated one , the should help us to advise our patients on what their selective serotonin re-uptake inhibitor (SSRI) flu- prognosis is likely to be. It helps communication oxetine, to be more effective than placebo in treat- (to patients and fellow professionals) if we can use ing childhood and adolescent depression39-41. How- a word or two to sum up the main problem(s) of ever, SSRIs are also effective against anxiety disor- a patient. I shall consider these three uses of a diag- ders in children and adolescents42-45. The most re- nosis to help us to decide on the most appropriate cent meta-analysis46 of SSRI and other new genera- classification system. tion antidepressants showed that for depression, 378 Wilkinson P 61% of adolescents responded to antidepressants prognosis will be like if they receive optimum treat- and 50% to placebo; while for non-OCD anxiety ments in clinic. The most recent meta-analysis of disorders, 69% responded to antidepressants and psychological treatments for depression showed 39% to placebo. Effect sizes using continuous mea- that 50% of those randomised to psychological sures also demonstrated greater drug-placebo dif- treatments “responded” after a course of treat- ferences for anxiety disorders (0.69) than for de- ment35. Meta-analysis of CBT for anxiety disor- pressive disorders (0.20). 95% confidence intervals ders demonstrated remission rate of 57%32. The for drug-placebo differences did not overlap be- most recent meta-analysis of SSRI and other new tween the two disorders in either recovery rate or generation antidepressants showed that for 61% continuous measures analyses, suggesting that the of depressed adolescents and 69% of adolescents antidepressant-placebo difference is significantly with non-OCD anxiety disorders responded to greater for anxiety disorders than depressive disor- antidepressants46. Again, caution must be taken in ders. Some caution should be made when compar- comparing studies due to different methodologies ing the results of different studies, as differences and definitions of response. may be due to different methodologies and remis- A way to compare outcome that avoids the sion criteria. In addition, meta-analyses did not dif- confounds of different methodologies between ferentiate between the different anxiety disorders. studies is to look at those with co-morbid anxiety In conclusion, the same pharmacological treat- and depressive disorders in the same study, and ment is effective for anxiety and depressive disor- compare outcome of each disorder. Anxiety pre- ders, while different psychological treatments are cedes depression in 2/3 of cases where both occur effective. This partly reflects genetic findings that together, and often persists after remission of de- similar genes may underly the two types of disor- pression17,18, suggesting anxiety disorders have der (in particular generalized anxiety disorder), worse prognosis. reflecting similar biological vulnerability, while dif- People with internalizing disorders have high ferent psychosocial events lead to the two disor- recurrence rates of both the initial disorder and ders19. However, antidepressants may be relatively other internalizing disorders49. When followed up more effective compared with placebo for anxiety over 7 years, continuity of all internalizing disor- than depressive disorders. ders was mediated by the common latent “inter- nalising disorder” trait. There was also disorder- Prognosis specific continuity for depressive and phobic dis- orders, but not panic and generalized anxiety dis- It is difficult to compare the prognosis of dif- orders 30. ferent disorders, as the prognosis in a sample very In summary, people with anxiety disorders may much depends on the severity of illness of individ- have longer time to recovery than those with de- uals in that sample. Different studies, with differ- pressive illnesses if left untreated; however, those ent disorders, differ greatly in recruitment, inclu- with anxiety disorders may be more likely to re- sion and exclusion criteria. Outcome is also affect- cover if treated. Long-term follow up suggests that ed by treatment in clinic samples, and cohort ef- while recurrence of the index illness is high, there is fects are important: later studies may be at times also a greater incidence of other internalizing dis- when better treatments are available, appearing to orders than people who have not had any inter- demonstrate a better prognosis for the disorder. A nalizing disorder. sample of depressed adolescents recruited in the early 1990s and followed up prospectively demon- Communication strated that community-ascertained cases had a median time to full remission of 3 months for com- It may be tempting at this point to think that munity-recruited cases and two years for clinic- all internalizing disorders lie on a spectrum from recruited cases47. A sample of adolescents asked no disorder to very ill; and that as internalizing retrospectively about past and present anxiety dis- disorders may overlap, then we can say that peo- orders in the late 1980s, who were mainly recruited ple lie on the “internalising disorders spectrum”. So from the community, had a median time of recov- why not just assign people scores on the different ery of 8 years48. Of course, a major confound with dimensions? While this may be scientifically quite prospective studies is that ascertainment may lead pure, it is of little help when talking to real people! to treatment. It helps people to know whether or not they have Of more interest to our patients than sponta- an illness, and to give a name to that illness. More- neous recovery in community cases is what their over, this name should bear some relation to what 379 Ciência & Saúde Coletiva, 14(2):373-381, 2009

the illness feels like. People know what “depressed” (in particular genetic), separate life events lead to and “anxious” moods are, and so depressive and the different disorders; anxiety disorders make sense to them. People can 2. Outcome appears different. Despite the lim- also understand that they have both types of dis- itations of comparing the results of different stud- order, and being given a dual diagnosis of an anx- ies, it appears that: iety and a depressive disorder may make more sense a. People take longer to recover from anxiety than “cothymia”. While fellow academic psychia- disorders than depressive disorders if untreated in trists may understand a dimensional system of clas- the community, or are treated with inactive place- sification, many professionals we speak with about bo in clinic; the children we are trying to help have very differ- b. Conversely, there is a slightly better response ent trainings, for example as teachers or social to active treatment (whether pharmacological or workers. Again, a straightforward and logically- psychological) among those with anxiety disor- named classification system will make communi- ders; cation far easier. While severity lies on a spectrum, c. In people with both anxiety and depressive we can talk about disorders as being mild, moder- disorders, the anxiety disorder has a worse prog- ate and severe, which people can understand. DSM- nosis; IV even gives us a list non-major depressive disor- d. While there is non-specific risk of future in- ders. While using symptom counts to assign a di- ternalizing disorders, there is also some specific risk agnosis can be fairly arbitrary, deciding whether of recurrence of the index disorder. somebody is significantly impaired, and so has an 3. Different forms of psychological treatment “illness”, has greater face validity. are needed for the different disorders; 4. People understand anxiety and depression as different emotions, which may of course both Conclusion: the future of conceptualization be present. We need a very good reason to go against about internalizing problems people’s understanding and say they are part of in children and adolescents the same disorder. Both types of disorder may be present in the Internalizing and externalizing disorders were pro- same individual at the same time. Rather than call posed as separate entities 60 years ago. Is this sep- this a different disorder, I would argue that the pa- aration still valid today? The answer is a resound- tient has a real problem of two different illnesses, ing yes. Research has demonstrated anxiety and which may worsen each other and worsen overall depressive disorders to have much in common. prognosis. Treatment of both disorders is neces- Modern factor analytic studies have demonstrat- sary, particularly if psychological treatment is used. ed that symptoms of these disorders cluster to- Of course, no classification system is perfect. gether, and separately to those of externalizing dis- In particular, generalized anxiety disorder poses orders. There is great overlap between individual some problems: it may share more genetic liability symptoms amongst diagnoses and between diag- with major depression, yet outcome may be more noses amongst individuals. There may be shared similar to anxiety disorders. genetic liability to internalizing disorders, particu- Current classifications are currently being re- larly depression and generalized anxiety disorders. vised, so that we shall soon benefit from the im- Internalising disorders all respond to similar treat- proved 5th version of the Diagnostic and Statistical ments, in particular selective serotonin re-uptake Manual of Mental Disorders, and the 11th version inhibitor antidepressants and different types of of the International Classification of Diseases. My cognitive-behavioural therapy. People who devel- own opinion is that depression and anxiety disor- op one internalizing disorder have a high risk of ders should be brought together into the same chap- developing any internalizing disorder in future. ter, as they have much in common, in particular What is less clear cut is classification within the when they are compared against other types of dis- internalizing disorders. With so much in common, orders, such as externalizing and psychotic disor- should anxiety disorders and depressive disorders ders. I still think Hewitt and Jenkins were correct be separated? I would argue that they should, for back in 1946. However, the distinction between the several reasons. diagnoses should still remain. In particular depres- 1. While there are common aetiological factors sive and anxiety disorders should remain separate. 380 Wilkinson P References

1. Hewitt LE, Jenkins RJ. Fundamental Patterns of Mal- 16. Angold A, Costello EJ, Erkanli A. Comorbidity. J Child adjustment: the Dynamics of Their Origin. Springfield, Psychol Psychiatry 1999; 40(1):57-87. Illinois: State of Illinois Publications; 1946. 17. Kovacs M, Gatsonis C, Paulauskas SL, Richards C. 2. Craddock N, Owen MJ. Rethinking psychosis: the Depressive disorders in childhood. IV. A longitudi- disadvantages of a dichotomous classification now nal study of comorbidity with and risk for anxiety outweigh the advantages. World Psychiatry 2007; disorders. Arch Gen Psychiatry 1989; 46(9):776-782. 6(2):20-7. 18. Alpert JE, Maddocks A, Rosenbaum JF, Fava M. Child- 3. Leckman JL, Cohen DJ. Tic Disorders. In: Rutter M, hood psychopathology retrospectively assessed among Taylor E, editors. Child and Adolescent Psychiatry. 4th adults with early onset major depression. J Affect Dis- ed. Oxford: Blackwell; 2002. ord 1994; 31(3):165-171. 4. American Psychiatric Association. Diagnostic and sta- 19. Kendler KS, Neale MC, Kessler RC, Heath AC, Eaves tistical manual of mental disorders (4th edition). Wash- LJ. Major depression and generalized anxiety disor- ington, D. C.: American Psychiatric Association; 1994. der. Same genes, (partly) different environments? Arch 5. American Psychiatric Association. Diagnostic and sta- Gen Psychiatry 1992; 49(9):716-722. tistical manual of mental disorders. 4th edition. Wash- 20. Kendler KS, Gardner CO, Gatz M, Pedersen NL. The ington, D. C.: American Psychiatric Association; 1994. sources of co-morbidity between major depression 6. Pickles A, Rowe R, Simonoff E, Foley D, Rutter M, and generalized anxiety disorder in a Swedish na- Silberg J. Child psychiatric symptoms and psychoso- tional twin sample. Psychol Med 2007; 37(3):453-462. cial impairment: relationship and prognostic signifi- 21. Kendler KS, Hettema JM, Butera F, Gardner CO, Pres- cance. Br J Psychiatry 2001; 179:230-235. cott CA. Life event dimensions of loss, humiliation, 7. Cuijpers P, de Graaf R, van Dorsselaer S. Minor de- entrapment, and danger in the prediction of onsets of pression: risk profiles, functional disability, health major depression and generalized anxiety. Arch Gen care use and risk of developing major depression. J Psychiatry 2003; 60(8):789-796. Affect Disord 2004; 79(1-3):71-79. 22. Eley TC, Stevenson J. Specific life events and chronic 8. Fergusson DM, Horwood LJ, Ridder EM, Beautrais experiences differentially associated with depression AL. Subthreshold depression in adolescence and and anxiety in young twins. J Abnorm Child Psychol mental health outcomes in adulthood. Arch Gen Psy- 2000; 28(4):383-394. chiatry 2005; 62(1):66-72. 23. Kendler KS, Walters EE, Neale MC, Kessler RC, Heath 9. Cuijpers P, Smit F, Oostenbrink J, de Graaf R, Ten AC, Eaves LJ. The structure of the genetic and envi- Have M, Beekman A. Economic costs of minor de- ronmental risk factors for six major psychiatric disor- pression: a population-based study. Acta Psychiatr ders in women. Phobia, generalized anxiety disorder, Scand 2007; 115(3):229-236. panic disorder, bulimia, major depression, and alco- 10. Fogel J. Recognising Minor Depression. Medscape Psy- holism. Arch Gen Psychiatry 1995; 52(5):374-383. chiatry and Mental Health. [serial on the Internet] 24. Tyrer P. The case for cothymia: mixed anxiety and 2006;11(2). Available from: http://www.medscape.com/ depression as a single diagnosis. Br J Psychiatry viewarticle/528985_1 2001;179:191-193. 11. Dunner DL. Treatment of dysthymic disorder. De- 25. Emmanuel J, Simmonds S, Tyrer P. Systematic review press Anxiety 1998; 8(Suppl 1):54-58. of the outcome of anxiety and depressive disorders. 12. Lima MS, Moncrieff J. Drugs versus placebo for dys- Br J Psychiatry Suppl 1998; 34:35-41. thymia. Cochrane Database Syst Rev 2000; 26. Lewinsohn PM, Rohde P, Seeley JR. Adolescent psy- (4):CD001130. chopathology: III. The clinical consequences of co- 13. Klein RG, Pine DS. Anxiety Disorders. In: Rutter M, morbidity. J Am Acad Child Adolesc Psychiatry 1995; Taylor E, editors. Child and Adolescent Psychiatry. Ox- 34(4):510-519. ford: Blackwell; 2002. 27. Goodyer IM, Herbert J, Altham PM, Pearson J, Sech- 14. Wittchen HU, Stein MB, Kessler RC. Social fears and er SM, Shiers HM. Adrenal secretion during major social phobia in a community sample of adolescents depression in 8- to 16-year-olds, I. Altered diurnal and young adults: prevalence, risk factors and co- rhythms in salivary cortisol and dehydroepiandros- morbidity. Psychol Med 1999; 29(2):309-323. terone (DHEA) at presentation. Psychol Med 1996; 15. Caron C, Rutter M. Comorbidity in child psychopa- 26(2):245-256. thology: concepts, issues and research strategies. J 28. Herbert J, Goodyer IM, Altham PM, Pearson J, Sech- Child Psychol Psychiatry 1991; 32(7):1063-1080. er SM, Shiers HM. Adrenal secretion and major de- pression in 8- to 16-year-olds, II. Influence of co- morbidity at presentation. Psychol Med 1996; 26(2):257-263. 381 Ciência & Saúde Coletiva, 14(2):373-381, 2009

29. Clark LA, Watson D. Distress and fear disorders: an 41. March J, Silva S, Petrycki S, Curry J, Wells K, Fair- alternative empirically based taxonomy of the “mood” bank J, Burns B, Domino M, McNulty S, Vitiello B, and “anxiety” disorders. Br J Psychiatry 2006; 189:481- Severe J; Treatment for Adolescents With Depression 483. Study (TADS) Team. Fluoxetine, cognitive-behavior- 30. Fergusson DM, Horwood LJ, Boden JM. Structure of al therapy, and their combination for adolescents with internalising symptoms in early adulthood. Br J Psy- depression: Treatment for Adolescents With Depres- chiatry 2006;189:540-546. sion Study (TADS) randomized controlled trial. Jour- 31. NICE NIfHaCE. Anxiety (amended). Management of nal of the American Medical Association 2004; anxiety (panic disorder, with or without agoraphobia, 292(7):807-820. and generalised anxiety disorder) in adults in primary, 42. Rynn MA, Siqueland L, Rickels K. Placebo-controlled secondary and community care. London: NICE; 2007. trial of sertraline in the treatment of children with 32. Cartwright-Hatton S, Roberts C, Chitsabesan P, generalized anxiety disorder. Am J Psychiatry 2001; Fothergill C, Harrington R. Systematic review of the 158(12):2008-2014. efficacy of cognitive behaviour therapies for child- 43. Wagner KD, Berard R, Stein MB, Wetherhold E, Car- hood and adolescent anxiety disorders. Br J Clin Psy- penter DJ, Perera P, Gee M, Davy K, Machin A. A chol 2004; 43(Pt 4):421-436. multicenter, randomized, double-blind, placebo-con- 33. NICE. Depression in Children and Young People: iden- trolled trial of paroxetine in children and adolescents tification and management in primary, community and with social anxiety disorder. Arch Gen Psychiatry 2004; secondary care. London: British Psychological Soci- 61(11):1153-1162. ety, Royal College of Psychiatrists; 2005. 44. PPASG. Fluvoxamine for the treatment of anxiety dis- 34. Weisz JR, McCarty CA, Valeri SM. Effects of psycho- orders in children and adolescents. The Research therapy for depression in children and adolescents: a Unit on Pediatric Psychopharmacology Anxiety Study meta-analysis. Psychol Bull 2006; 132(1):132-149. Group. N Engl J Med 2001; 344(17):1279-1285. 35. Watanabe N, Hunot V, Omori IM, Churchill R, Fu- 45. Birmaher B, Axelson DA, Monk K, Kalas C, Clark DB, rukawa TA. Psychotherapy for depression among chil- Ehmann M, Bridge J, Heo J, Brent DA. Fluoxetine for dren and adolescents: a systematic review. Acta Psy- the treatment of childhood anxiety disorders. J Am chiatr Scand 2007; 116(2):84-95. Acad Child Adolesc Psychiatry 2003; 42(4):415-423. 36. Mufson L, Dorta KP, Wickramaratne P, Nomura Y, 46. Bridge JA, Iyengar S, Salary CB, Barbe RP, Birmaher Olfson M, Weissman MM. A randomized effective- B, Pincus HA, Ren L, Brent DA. Clinical response ness trial of interpersonal psychotherapy for depressed and risk for reported suicidal ideation and adolescents. Arch Gen Psychiatry 2004; 61(6):577-584. attempts in pediatric antidepressant treatment: a meta- 37. Mufson L, Weissman MM, Moreau D, Garfinkel R. analysis of randomized controlled trials. Jama 2007; Efficacy of interpersonal psychotherapy for depressed 297(15):1683-1696. adolescents. Arch Gen Psychiatry 1999; 56(6):573-579. 47. Dunn V, Goodyer IM. Longitudinal investigation into 38. Lipsitz JD, Gur M, Vermes D, Petkova E, Cheng J, childhood- and adolescence-onset depression: psy- Miller N, Laino J, Liebowitz MR, Fyer AJ. A random- chiatric outcome in early adulthood. Br J Psychiatry ized trial of interpersonal therapy versus supportive 2006; 188:216-222. therapy for social anxiety disorder. Depress Anxiety 48. Keller MB, Lavori PW, Wunder J, Beardslee WR, 2008; 25(6):542-553. Schwartz CE, Roth J. Chronic course of anxiety disor- 39. Emslie GJ, Rush AJ, Weinberg WA, Kowatch RA, ders in children and adolescents. J Am Acad Child Hughes CW, Carmody T, Rintelmann JW: A double- Adolesc Psychiatry 1992; 31(4):595-599. blind, randomized, placebo-controlled study of flu- 49. Pine DS, Cohen P, Gurley D, Brook J, Ma Y. The risk oxetine in depressed children and adolescents. Arch for early-adulthood anxiety and depressive disorders Gen Psychiatry 1997; 54:1031–1037. in adolescents with anxiety and depressive disorders. 40. Emslie GJ, Heiligenstein JH, Wagner KD, Hoog SL, Arch Gen Psychiatry 1998; 55(1):56-64. Ernest DE, Brown E, Nilsson M, Jacobson JG. Fluox- etine for acute treatment of depression in children and adolescents: a placebo-controlled, randomized clinical trial. J Am Acad Child Adolesc Psychiatry 2002; Artigo apresentado em 05/11/2007 41(10):1205-1215. Aprovado em 16/06/2008