Our young

Few of us recognise that Australia was the first country to estimate the prevalence of mental disorders in children and adolescents (Sawyer et al., 2001). Practising clinicians sometimes need to be reminded that such survey data refer to the general population, not the much smaller fraction of young persons whom they may encounter in their work. The latter are far from the whole picture. And it is the whole picture that is needed to guide service policy.

Now a second nation-wide study has been completed, its main findings being a large part of this September issue, starting with a constructive Editorial by Merry (this issue). The methods are exemplary (Hafekost et al., this issue). This writer knows how such advanced epidemiological work can be achieved only by the harmonious collaboration of a well-constructed multidisciplinary team held together over some years. Note the pains taken to examine a sample as nationally representative as practicable, totalling 6310 parents and almost 3000 young persons. Impressively, the interviewers attempted contact in the respondents’ homes up to six times before giving up. A striking finding was that these young people agreed to take part much more often (almost 90%) than did their parents (55%). The latter figure is unfortunate for interpretation of the parent-derived data.

In 4- to 17-year-olds, the disorders with the highest 1-year prevalence, in descending order, were attention deficit hyperactivity disorder (ADHD), anxiety disorders and major depressive disorder (Lawrence et al., this issue). If the rate for ADHD of over 7% is close to the truth, what questions follow? For a start, is the disorder really so common, has it increased over time and should seven children in every 100 be treated for it? For all forms of morbidity, the distribution across our population shows some intriguing correlates: our young have better mental health when their parents do not use English as their main language; and parents are often unaware that their child has a major depressive disorder, itself often associated with suicidal thoughts or acts of self-harm. So what do these findings mean? For operational reasons and to avoid overload, the survey does not include estimates on drug abuse or eating disorders, and that has to be accepted as realistic. But what about recreational Internet use as a promising independent variable?

On the matter of suicidal behaviour and acts of self-harm ‘without suicidal intent’, the two papers (Zubrick et al., this issue [a], this issue [b]) may tax some readers’ understanding. Here, we enter an area that is hard to classify neatly. Suicidal behaviour and thoughts are not binary, in the sense of being present or absent. To separate suicidal behaviour into two categories (with and without suicidal intent) approximates clinical reality rather poorly. Erwin Stengel (1965) proposed that suicidal mental states are phenomenologically multifaceted, calling it ‘The Mask of Janus’, where the person looks to both life and death at the same time. So survey data such as we have here, from self-report on a tablet, usually in the young person’s home, are necessarily only part of the picture. The respondents who reported having carried out an act of self-harm, purportedly ‘without suicidal intent’ (Zubrick et al., this issue [b]), are precisely in this category. Caution is also needed for lifetime estimates of self-harm, said here to be as high as 11% in the 12- to 17-year-olds. This is because lifetime estimates invariably use the question ‘Have you ever …?’ which is open to reporting error of serious magnitude. At an epidemiology forum where lifetime estimates in the Epidemiologic Catchment Area (ECA) were being discussed, Ernest Gruenberg referred to it derisively: ‘Ah yes, the McCarthy question’, as used in Central Intelligence Agency (CIA) interrogations about membership of the Communist Party. If anything, 11% is an underestimate, but who knows? Regarding the prevalence of attempted suicide, 8% of the adolescents said they had done so in the previous 12 months. But only 1/10th of them (0.8%) reported they had needed medical treatment (presumably admission). Here is a matter that could be improved by operational data, not very costly to acquire: the national admission rate for self-poisoning and self-injury

The findings on service use deserve our close attention (Johnson et al., this issue). They will again prove invaluable in further efforts to improve access for troubled young people. Here, this writer is very conscious of something we just do not know. It is probably a very good thing that our young can get help more readily. They go to general practitioners (GPs), psychologists and counsellors. The Better Access programme has allowed more needful people to get professional help. But when a troubled teenager goes to a GP or is referred to a psychologist or counsellor, the key questions we should be asking ourselves are as follows: What actually happens between the two people involved? and What is the outcome some months or years later? An even more uncomfortable question is to ask whether having more mental health services has meaningfully reduced the overall prevalence of morbidity in our young. This very question is currently much under consideration (e.g. Jorm et al., 2016). Even if the rate were shown to have dropped in the last 10 or more years, that change may well be due to other unknown and unmeasured factors. One component of service use was not included in the survey: prescribed medication. For this, we have the landmark study by Karanges et al. (2014) based on Australian national data, but this of course does not tell us what the present sample were prescribed or actually consumed. In summary, this Second National Survey is a credit to our resources to conduct epidemiological research of outstanding value.

Who would think of bones and mental health being linked? Williams et al. (this issue) do just this, opening our minds to several associations with much relevance for clinical practice. Inevitably, they prompt us to bear in mind that omnipresent third set of variables that can account for some of the association. The phenomenology of mental disorders seems to have gone out of fashion. In teaching medical students and registrars the fundamentals of psychiatry, the contemporary syllabus now carries material that comes more from the consultants’ contemporary preoccupations rather than the fundamental building blocks of good clinical psychiatry. The study of Schneider’s First Rank Symptoms by Toh et al. (this issue) is therefore all the more welcome in these pages. The paper on placebos by Darragh et al. (this issue) is an example of another basic issue in medicine and psychiatry.

In a similar vein, the placebo response is becoming less of a mystery, as its expanding literature testifies. Arandjelovic et al. (this issue) ask us to look at mental health services in Asia, encouraging a greater contribution from Australia, and indeed New Zealand. Suppose we invert this, asking whether psychiatry and mental health services in Asia can enrich us here more than we appreciate. This could extend from social values or the organisation of services to biological research. What an appropriate proposition to be examined in an international forum. The masterly meta-analysis of international case control data by Oo et al. (this issue) goes some way to confirm the risk for both major depression and alcohol dependence in persons having the S allele of the 5HTTLPR polymorphism. Such studies are important in encouraging continuing efforts in the fascinating field of genetics and mental disorders.