Abstract
Adolescence is a transitional stage of human development, during which the individual undergoes marked physiological, psychological and social change in the process of growing from a child into an adult. In addition to the developmental disturbances of adolescence, many serious mental disorders may have their onset at this time, including schizophrenia, mood disorders, anorexia nervosa, drug and alcohol abuse and personality disorders. Early recognition of psychiatric disorders in adolescents and prompt initiation of adequate treatment are of great importance to the outcome of treatment, to mental health in adulthood [1, 2] and to the health service costs.
One way to improve the early recognition of adolescent psychiatric disorders is to check the mental status of all adolescents during general medical examinations at schools and in the general health services [3, 4]. General practitioners are in a key position to do this, because young people consult them about their health problems [5]. However, the recognition of mental disorders may be difficult, since even normal adolescents are susceptible to occasional mental and psychosomatic symptoms and the clinical picture of adolescent mental disorders is often non-specific at first. Indeed, the evidence suggests that general practitioners feel that they do not have adequate training in recognizing or intervening in adolescent mental problems [5–8].
An awareness of the factors associated with psychiatric disorders among adolescents may improve the recognition of mental disorders. It is important that doctors are able to distinguish between adolescents who are at high risk and those at low risk [9]. Childhood behavioural disorders, learning difficulties, conflicts in human relations, traumas, and various sociodemographic factors have been identified in population surveys as risk factors for adolescent mental disorders [10–14]. However, these variables are not of great help in the consultation situation, since detecting these risk factors necessitates a vast anamnestic interview and extensive clinical knowledge.
The purpose of our study was to investigate whether it is possible for all practitioners to identify detectable signals that would be clinically helpful in improving the recognition of mental disorders among those adolescents who have mental symptoms.
Method
The study sample consisted of adolescents seeking outpatient psychiatric assessment in three specialist psychiatric adolescent clinics during 1 year (n = 239); 173 adolescents were willing to participate in this study. Inadequate data were received from 9, so the final sample consisted of 164 subjects. Sixty-six adolescents (28%) refused to participate in the study. The mean age of the dropout group (18.1 ± 2.7 years) did not differ statistically (p = 0.05) from that of the study group (17.7 ± 2.3 years), but there were more boys (54%) among dropouts than among study subjects (28%). All adolescents were white Caucasians. Only 55% of study group (54% of dropouts) had a referral from primary care doctors or school nurses; the others came of their own or their parents’ accord.
Seventy-one per cent of those studied were at school, 7% had dropped out of school, 6% were at work and 16% were unemployed. Over one-third (38%) lived with both parents and 23% with one parent; 35% had moved away from home and 3% were in foster homes or homes for young people. For cultural and sociopolitical reasons the number of adolescents living alone in Finland may be high compared with the proportions found in many other countries. However, the adolescents living alone were older (mean age 19.2 ± 0.8 years) than the other study subjects.
Background information (onset of puberty, appearance of secondary sexual signs, education, work status, attitude to himself/herself, attitude to the future and plans for the future) was collected by questionnaire, which was completed during the interview with the help of the interviewer, if needed.
Psychiatric diagnosis was made using the Structured Clinical Interview (SCID-I) for the axis I DSM-III-R diagnoses [15, 16]. The SCID is a semistructured interview method that allows for considerable variation in interview style, depth of probing and clinical judgement as to whether a patient's description of a particular behaviour meets a relevant diagnostic criterion. The SCID has mostly been used with adults, but also with adolescents [17, 18].
In this study the Conduct Disorder items from the SCID-II (interview for evaluating 12 personality disorders) were added to the SCID-I interview, but other SCID-II disorders were not evaluated as it is not possible to assess them in a cross-sectional study [19]. The SCID interview does not diagnose attention deficit hyperactive disorder or oppositional defiant disorder, and for this reason the investigators made a separate assessment on the basis of medical files to determine whether the diagnostic criteria of these disorders were fulfilled.
Four experienced adolescent psychiatrists trained in the use of the SCID-I and II performed the clinical psychiatric assessments. An assessment was completed in one to three interviews of 45 min each depending on the collaborative capacity of the adolescents.
The psychiatrists assessed psychosocial functioning capacity using the Global Assessment of Functioning Scale (GAF) [19]. The GAF score describes psychological, social and occupational functioning on a hypothetical continuum of mental health–illness. Previous research shows that a score of 61 discriminates those with mental disorders very well from those without [20].
The self-rated 21-item Beck Depression Inventory (Beck) was used to assess cognitive, behavioural, affective and somatic components of depression [21]. On the basis of previous studies [22], the score of 11 was used as a cut-off in multivariate analysis.
The conscious self-image of the adolescents was assessed using the 130-item version of the Offer Self-Image Questionnaire (OSIQ). These 130 items form 12 scales [23]. We used the total score of the OSIQ and the sum scores of scales S2 (emotional tone), S7 (family relationships) and S10 (emotional health). The scales S2, S7 and S10 were assessed separately, because earlier studies have shown that these scales in particular differ between disturbed and normal adolescents [23, 24]. A high score indicates a negative self-image both in the total score and for each scale.
The data were analyzed statistically with the paired t-test for continuous variables and χ2 test for class variables. Logistic regression analysis was used to identify the independent variables associated with psychiatric diagnosis.
Results
On the basis of the SCID interview, 76% of the adolescents were diagnosed as having a psychiatric disorder, with 44% having more than one disorder. The most common were depressive disorders (Table 1).
Psychiatric diagnosis (DSM-III-R) according to SCID interview
The adolescents who had psychiatric disorders were slightly older (mean = 17.7 ± 1.9 years, 95% CI = 7.5–18.3, range = 13.3–22.8) than the others (mean = 17.1 ± 2.3 years, 95% CI = 16.5–17.7, range = 13.4–21.0). There was no difference in the age of onset of puberty between those with psychiatric disorders and the others. The average age of menarche was 12 years (12 years ± 0.9 years, range = 9–15), and the average age of first ejaculation was 13 years (13 years ± 1 years, range = 11–16).
Of those with psychiatric disorders, 50% had had symptoms for more than 1 year (mean = 23.9 ± 22.5 months), and of those without any psychiatric diagnoses, 28% had had behavioural and emotional symptoms for more than 1 year (mean = 15.7 ± 16.1 months, p = 0.01).
There were no statistically significant differences regarding social relationships between the two groups (diagnosis or not). Over 90% of the adolescents had peer relationships and 82% had dated. Altogether, 46% of those who had psychiatric disorders and 36% of those without disorders stated that their relationship with their mother was problematic or non-existent (p = NS), while 58% of those with psychiatric disorders and 49% of the others reported having problems with their fathers (p = NS).
Current smoking among the subjects was common, although no statistical differences were found between the groups. Alcohol consumption was also common, and only 18% reported that they did not consume alcohol at all. Adolescents who had psychiatric disorders used alcohol in order to get drunk more frequently than others. Illicit drug use was also more common among adolescents who had psychiatric disorders (Table 2).
Smoking and use of alcohol, medicines and drugs in adolescents with and without a psychiatric diagnosis
The GAF score for adolescents who had psychiatric disorders was lower (mean = 60.2, 95% CI = 58.5–61.8) than that for others (mean = 67.6, 95% CI = 65.2–70.2).
Adolescents who had one or more psychiatric disorders according to SCID interview felt themselves more depressive than adolescents without any psychiatric diagnoses (Beck score) and they also had a more pessimistic attitude towards the future. The self-image (OSIQ score) of the adolescents with psychiatric disorders was more negative. A statistically significant difference was found on the OSIQ scales describing mood (S2) and emotional health (S10) between adolescents who had psychiatric disorders and those without (Table 3).
Self-image (OSIQ), depressiveness (Beck), self-destruction, perceived hopelessness and attitude towards the future in adolescents with and without a psychiatric diagnosis
The psychiatrists found that 17 adolescents felt chronic hopelessness, and all of these were diagnosed as having a psychiatric disorder (depressive disorders, n = 16 and one anorexia nervosa). Twenty subjects (12%) had previously attempted suicide, and all had psychiatric disorders (depressive disorders, n = 19 and one conduct disorder). Suicidal ideation was found to occur in both groups, but more frequently among those who had psychiatric disorders. Only the adolescents who had psychiatric disorders had continuous suicidal ideas or suicidal plans (Table 3).
More of the adolescents with psychiatric disorders than others had abandoned a hobby important to them (50% vs 26%, p = 0.007). A pessimistic attitude towards the future was noted in the adolescents’ views about setting up a family: 31% of those who had psychiatric disorders compared with 13% in the group of those without disorders did not consider it possible (p = 0.04). One-quarter (26%) of those with disorders compared with 8% of the others had no plans for the future (p = 0.02).
Independent factors for psychiatric diagnosis were assessed by using logistic regression analysis (the Enter method). The following variables were included in the model: Beck score (0–11 = 0, < 11 = 1), attitude towards the future (confident = 0, uncertain or pessimistic or unwilling to think = 1), giving up a hobby important to oneself (not having given up = 0, given up = 1), alcohol consumption (no consumption at all or moderate consumption = 0, getting drunk at least once a week = 1), suicidal ideation (no ideation at all or occasional ideas = 0, continuous suicidal ideation or plans = 1), GAF score (< 60 = 0, < – 60 = 1). The OSIQ score was not included in the model, because it correlates highly with the Beck score (r = 0.55). The independent factors proved to be a GAF score < 60 (OR = 3.0, 95% CI = 1.2–7.5), and an uncertain or pessimistic attitude towards the future (OR = 9.1, 95% CI = 1.1–72.9).
Discussion
The most important variables associated with psychiatric diagnosis were low psychosocial functioning, an uncertain or pessimistic attitude towards the future, depressive symptoms, suicidal thoughts, previous suicide attempts, hopelessness and a lack of plans for the future.
Naturally these results will depend on the sample. The high frequency of depressive disorders has an effect on our results. However, depressive patients are seen in primary care and in outpatient clinics and so our results may help to distinguish adolescents who have psychiatric disorders from those who merely have mental symptoms.
To summarize, from among the factors studied, the independent factors for psychiatric diagnosis proved to be an uncertain or pessimistic attitude towards the future and a low GAF score. It is important to note that all adolescents had sought psychiatric help. Low psychosocial functioning has also been found previously to be a significant and discriminating factor between psychiatric symptoms and disorders [20, 25]. However, adolescents’ attitudes towards the future have not previously been reported to be associated with psychiatric diagnosis.
It is well known that adolescents’ trust in their abilities and attitudes towards the future are closely related to their self-perception. Furthermore, it has been shown that the self-image of adolescents with psychiatric disorders is more negative on average than that of others [24]; moreover, the self-image of adolescents who seek help from their parents is more positive than that of those who do not [26, 27].
A suicide attempt is a clear signal of a psychiatric disorder. Psychological autopsy studies indicate that most of the adolescents who commit suicide have had psychiatric disorders [28]. Hopelessness in adolescents, whether associated with suicidal thoughts or not, is a signal of a severe psychic cul-de-sac; depression, self-destruction and the prodromal phase of psychosis [29, 30]. Our results suggest that an adolescent who appears hopeless or who talks about hopelessness should be assessed carefully. Questions about future plans and hobbies may be helpful in this respect.
Most of our study subjects had conflicts with their parents but this did not discriminate the adolescents who had psychiatric disorders from those who did not. Our results indicate that continual conflicts with parents cannot be used for screening adolescent psychiatric disorders, although previous studies among the general population and psychiatric patients suggest that conflicts in family relations and family functioning on a wider scale may be related to adolescent psychiatric symptoms [14, 31, 32]. Dissimilarity in the results with the previous studies may be due partly to the fact that our subjects were outpatients.
We found that among help-seeking adolescents there were twice as many smokers as among Finnish adolescents in general [33]. Brown et al. [34] and Kandel et al. [12] found that smoking was related only to the diagnoses of drug abuse, drug dependence and behavioural disorder. The distribution of diagnoses of our data may explain the difference between our results and those of Brown et al. It is advisable, however, that the doctors who see adolescents should ask them about smoking habits and alcohol consumption, as these may indicate a psychiatric disorder or other problems in life.
The adolescents in our study had had symptoms for over 12 months on average before the psychiatric assessment. According to previous studies, as many as 80% of adolescents suffering from mental problems do not seek help at all, and will never be referred for psychiatric assessment [35–37] and many adolescents’ problems are difficult to recognize even by their parents [37]. Therefore, whenever possible, the adolescents should be asked whether they have any psychic symptoms in the course of general medical examinations.
Our study subjects consisted of a clinical sample of adolescents who attended specialized outpatient clinics, but it is true that half of them came without a referral from a doctor. One-quarter did not have any psychiatric disorders, and consequently our data are suitable for the study of factors associated with mental illness among adolescents with mental symptoms [9]. The associates for psychiatric diagnosis found in our sample dealing with treatment-seeking adolescents may not be the same as those in the general population, but the factors that we found could be useful for doctors.
The SCID-I diagnostic interview that we used together with selected parts of SCID-II items can be regarded as reliable and valid. The SCID is based on the criteria of DSM-III-R, but permits additional questions to be asked. The Beck scale has been shown to assess depression well [23] and the OSIQ to predict normal adolescent development [38, 39]. The fact that the clinical assessment was made by experienced adolescent psychiatrists, and that a structured clinical diagnostic interview and structured scales were used, improves the validity of our study.
Recognizing a psychiatric disorder in adolescents is always a challenge for the clinician. Suicidal thoughts and suicide attempts are clear, well-known risk factors for adolescent psychiatric disorders. The clinician should also pay attention to adolescents’ psychosocial functioning and to their attitudes towards the future and towards themselves. Smoking and substance abuse, specifically alcohol consumption, also need to be investigated during anamnestic interviews.