Validity of the 12-item General Health Questionnaire (GHQ-12) in detecting depressive and anxiety disorders among high school students

https://doi.org/10.1016/j.psychres.2010.10.010Get rights and content

Abstract

Despite the common use of the 12-item General Health Questionnaire (GHQ-12) with adolescents, there is limited data supporting its validity with this population. The aims of the study were to investigate the psychometric properties of the GHQ-12 among high school students, to validate the GHQ-12 against the gold standard of a diagnostic interview, and to suggest a threshold score for detecting depressive and anxiety disorders. Six hundred and fifty-four high school students from years 10 to 12 (ages 15–18) completed the GHQ-12 (Likert scored) and the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders-IV-Test Revision (DSM-IV-TR). Receiver operating characteristic (ROC) curves were plotted. The mean GHQ-12 score for the total sample was 9.9 (S.D. = 5.4). Results from the ROC curve indicated that the GHQ-12 performed better than chance at identifying depressive and anxiety disorders (area under the curve (AUC) = 0.781). A GHQ-12 threshold score of 9/10 for males and 10/11 for females was found to be optimal. Given the significant proportion of mental illness among high school students, there may be a need to introduce screening for mental illnesses as part of the school curriculum. This can assist with the early identification and enable low stigma preventive intervention within the school environment.

Introduction

Mental health problems, in particular depression, are common among young people, with the peak period of onset occurring during adolescence. The 12-month prevalence of mental illness in young people is in the range of 16–30%, with depressive illness being among the most common (Andrews et al., 2001). Mental disorder in young people is associated with a range of negative outcomes including recurrent illness, relationship difficulties, substance misuse and poor educational and vocational outcomes (McGorry et al., 2007). Early identification and treatment of such disorders can lead to improved outcomes, yet many young people experiencing mental health problems do not seek help (Rickwood et al., 2007).

Schools have been identified as an appropriate setting for the identification of young people with mental health problems (U.S. Department of Health and Human Services, 1999). Screening school students for mental disorder is one way of identifying people at risk without relying upon help-seeking (Shaffer et al., 2004, Cuijpers et al., 2006, Nemeroff et al., 2008). However critical to the success of any screening program is the availability of a valid and reliable screening tool that can be easily administered in a school setting (Pena and Caine, 2006).

The General Health Questionnaire (GHQ; Goldberg and Williams, 1988) is a measure designed to assess current mental wellbeing by assessing normal ‘healthy’ functioning and the appearance of new, distressing symptoms. It was originally developed as a 60-item screening instrument from which shorter versions were subsequently developed (e.g., GHQ-12 and GHQ-28; Goldberg and Williams, 1988). The GHQ has been used extensively with adults, assessing psychological distress in relation to a variety of stressors, such as traumatic events (Dyregrov et al., 1996), job satisfaction (Virtanen et al., 2002), and poor hearing (Fellinger et al., 2007).

Numerous studies have assessed the validity of the GHQ for use with adults (Tarnopolsky et al., 1979, Pan and Goldberg, 1990), the most extensive of which was conducted in 15 centres worldwide, covering both developed and developing countries (Goldberg et al., 1997). This cross-cultural validation was conducted with 5438 patients as part of a World Health Organization project on mental illness. Goldberg and colleagues found that the GHQ, and in particular the 12-item version, performed well and was remarkably robust.

Although the GHQ was designed for use in adult populations, numerous studies have used the GHQ with adolescent populations. For example, it has been used to assess psychological distress in adolescents following parental loss (death or separation) (Raphael et al., 1990), in school settings (Robinson et al., 2010) and in the context of unemployment (Stafford et al., 1980).

However, despite its use in these studies, there is little evidence regarding the validity of the GHQ in adolescents. A review conducted by Tait et al. (2002) identified only four studies that validated the GHQ solely with adolescent populations and a further four that included both adolescents and young adults. The study groups used varied: three of these studies were carried out in schools (Mann et al., 1983, Shek, 1989, Winefield et al., 1989), two in universities (Chan and Chan, 1983, Radovanovic and Eric, 1983), two in the community (Banks, 1983, Seva et al., 1992) and one in the army (Politi et al., 1994). For example, Banks assessed the validity of the GHQ by comparison with a Structured Clinical Interview, the Present State Examination (Wing et al., 1974). The sample comprised 200 17-year-old school leavers recruited through a careers service. Using the GHQ scoring method (0–0–1–1), it was found that a cut-off score of 2/3 on the GHQ-12 yielded a sensitivity of 71.4% and a specificity of 79.8% (Banks, 1983).

Threshold scores on the GHQ-12 can vary not only by study sample, but also by the scoring method used. For example, Politi and colleagues used the Likert scoring method (0–1–2–3) and recommended a cut-off score of 8/9 on the GHQ-12. These researchers assessed the validity of the GHQ-12 among a sample of 252 military recruits that had passed a physical and psychiatric examination into the army and 111 people who did not. The entire sample comprised males aged 18 years of age. Comparing the results from the GHQ-12 against a psychiatrists' rating (International Classification of Diseases 9th Revision (ICD-9) criteria), it was found that a cut-off score of 8/9 yielded a sensitivity of 68% and a specificity of 59% (Politi et al., 1994). However, the generalisability of these findings to other young persons is unclear as the sample comprised a highly homogenous sample that may have exaggerated emotional distress to avoid military service.

A further study examining the psychometric properties and validity of the GHQ-12 with adolescents was conducted by Tait et al. (2003). These researchers surveyed 336 students from grades 7 to 10, aged 11–15 years. The students were from single sex and mixed schools from both the government and private sectors in Western Australia. It was reported that the GHQ-12 showed good internal consistency, and scores for both males and females increased with age. Using the Likert scoring method, threshold scores of 13/14 for males and 18/19 for females appeared optimal.

However the authors concluded that while the GHQ-12 appeared to be a valid index of psychological wellbeing in this population, the criterion measure employed by this study was sub-optimal. The authors selected students scoring within the top 3–4% on depression and anxiety measures. The GHQ scores were then compared to this combined category. Acknowledging this limitation, the authors identified the need for future research to validate the GHQ-12 against a ‘gold standard’ diagnostic interview to identify the optimal threshold score to use in a high school setting (Tait et al., 2003). Future research is also required that evaluates this measure among students from a broader age range including those from years 10 to 12 (ages 15–18), which are arguably the most stressful years of high school.

Thus the primary aims of the current study were to examine the psychometric properties and validity of the GHQ-12 among high school students from years 10 to 12; to assess the sensitivity and specificity of GHQ-12 against a gold standard diagnostic interview, and to identify an appropriate threshold score for detecting likely depressive and anxiety disorders.

Section snippets

Participants and setting

During 2005, Year 10 secondary students were invited to participate in a study investigating emotional experiences among adolescents. Students were recruited from 34 secondary schools in the northern and western suburbs of Melbourne, Australia, that consented to participate from a total of 60 secondary schools approached (20 government, five Catholic and nine independent schools, participation rates of 71%, 29% and 60% respectively). Of 4797 students initially invited, 946 (19.7%) agreed to

Sample characteristics

Of the 654 participants that took part in the research assessment, 297 were male (45.4%) and 357 (54.6%) were female. The mean age was 16.59 years (S.D. = 0.58 years; range: 15.02–18.39 years). The majority of students were at school full-time with either part-time/casual work (n = 347, 53.1%) or no work (n = 300, 45.9%). The majority of participants also lived in a house with their family of origin (n = 623, 95.4%), reported experiencing no significant medical history in the 6 months prior to interview (n =

Discussion

The aim of the study was to assess the validity of the GHQ-12 for detection of depressive and anxiety disorders in high school students aged between 15 and 18 years and to identify an appropriate threshold score for identifying those who met the diagnostic criteria for depression and anxiety according to the Structured Clinical Interview for DSM-IV-TR (SCID-IV). This is the first study to date that has assessed the validity of the GHQ-12 against a diagnostic interview with adolescents aged

Acknowledgments

This research was funded by the Colonial Foundation and an NHMRC Program Grant. The authors gratefully acknowledge the statistical assistance provided by Hok Pan Yuen, Statistical Consultant at Orygen Youth Health Research Centre. The authors also acknowledge the participation of the students and the schools in this research project.

References (49)

  • M. Banks

    Validation of the General Health Questionnaire in a young community sample

    Psychological Medicine

    (1983)
  • D. Chan et al.

    Reliability, validity and the structure of the General Health Questionnaire in a Chinese context

    Psychological Medicine

    (1983)
  • N. Craddock et al.

    The beginning of the end for the Kraepelinian dichotomy

    The British Journal of Psychiatry

    (2005)
  • P. Cuijpers et al.

    Screening and early psychological intervention for depression in schools: systematic review and meta-analysis

    European Child & Adolescent Psychiatry

    (2006)
  • M. Dolan et al.

    Violence risk prediction. Clinical and actuarial measures and the role of the psychopathy checklist

    British Journal of Psychiatry

    (2000)
  • A. Dyregrov et al.

    Voluntary and professional disaster-workers: similarities and differences in reactions

    Journal of Traumatic Stress

    (1996)
  • J. Fellinger et al.

    Mental distress and quality of life in the hard of hearing

    Acta Psychiatrica Scandinavica

    (2007)
  • M.B. First et al.

    Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition

    (2002)
  • D.J. French et al.

    Measurement invariance in the General Health Questionnaire-12 in young Australian adolescents

    European Child & Adolescent Psychiatry

    (2004)
  • D. Goldberg et al.

    A User's Guide to the General Health Questionnaire

    (1988)
  • D. Goldberg et al.

    The validity of two versions of the GHQ in the WHO study of mental illness in general health care

    Psychological Medicine

    (1997)
  • M.S. Gould et al.

    Evaluating iatrogenic risk of youth suicide screening programs: a randomised controlled trial

    Journal of the American Medical Association

    (2005)
  • R. Kendell et al.

    Distinguishing between the validity and utility of psychiatric diagnoses

    The American Journal of Psychiatry

    (2003)
  • P.M. Lewinsohn et al.

    Adolescent psychopathology: I. Prevalence and incidence of depression and other DSM-III-R disorders in high school students

    Journal of Abnormal Psychology

    (1993)
  • Cited by (0)

    View full text