Brief report
The performance of the Alcohol Use Disorder Identification Test (AUDIT) in detecting alcohol abuse and dependence in a population of depressed or anxious persons

https://doi.org/10.1016/j.jad.2010.04.019Get rights and content

Abstract

Background

Alcohol use disorders are highly prevalent but often remain unrecognized among depressed and/or anxious persons. This study examines the performance of the Alcohol Use Disorder Identification Test (AUDIT) in detecting alcohol abuse and dependence in this high-risk group and compares it to that in healthy controls.

Methods

Data from the Netherlands Study of Depression and Anxiety (NESDA) were used, including 1756 persons with a past-year depressive and/or anxiety disorder and 648 persons without a lifetime depressive and anxiety disorder. The performance of the AUDIT was compared against the gold standard of a CIDI-based diagnosis of past-year alcohol abuse or dependence by means of sensitivity, specificity and areas under receiver operating characteristic curves (AUCs).

Results

The AUDIT accurately detected alcohol dependence in depressed and/or anxious men (AUC = 0.89) and women (AUC= = 0.88), with detected cut-off points of ≥ 9 and ≥ 6, respectively, comparable to that in healthy controls (men: AUC = 0.89; women: AUC = 0.94). However, the overall accuracy in detecting alcohol abuse was limited in depressed/anxious men (AUC = 0.74) and women (AUC = 0.78) and no adequate cut-off points with both acceptable sensitivity and specificity could be identified.

Limitations

Persons with a primary diagnosis of an addiction disorder were excluded and therefore the sample may not be fully representative of the most severely addicted patients.

Conclusions

These findings confirm the accuracy of the AUDIT in detecting alcohol dependence, but not alcohol abuse, in depressed and/or anxious persons. Screening for alcohol dependence in this high-risk group could improve identification of persons suffering from this impairing comorbid condition.

Introduction

Alcohol use disorders (AUDs; alcohol abuse [AA] or alcohol dependence [AD]) tend to worsen the course of depressive and anxiety disorders and may severely impede treatment (Barkow et al., 2003, Bijl and Ravelli, 2000, Bruce et al., 2005, Buckner et al., 2008, McDermut et al., 2000, Vuorilehto et al., 2009). Although AUDs are highly prevalent in the general population, and especially among persons with depressive and/or anxiety disorders (Boschloo et al., submitted for publication; Burns and Teesson, 2002, de Graaf et al., 2003, Hasin et al., 2007, Merikangas et al., 1998, Pirkola et al., 2005), AUDs often remain unrecognized (Cleary et al., 1988, Rydon et al., 1992). Adequate screening for AUDs may help to identify those depressed and/or anxious patients suffering from this impairing comorbid condition and, consequently, could improve effectiveness of health care.

The Composite International Diagnostic Interview (CIDI) (World Health Organization, 1997) is a widely-used diagnostic instrument for the assessment of AUDs in both clinical and research settings. As a time-efficient screening instrument, the Alcohol Use Disorder Identification Test (AUDIT) (Babor et al., 1992) has been developed to detect hazardous and harmful alcohol use. A growing body of research evidence supports the validity of the AUDIT in screening for AUDs in various settings and in diverse populations (Allen et al., 1997, Reinert and Allen, 2002, Reinert and Allen, 2007), but empirical evidence regarding the diagnostic accuracy in depressed or anxious patients is lacking. Most studies also fail to distinguish between the performances of the AUDIT in detecting AA versus AD.

Since test performances often vary across populations (Ransohoff and Feinstein, 1978), the AUDIT may be less accurate in depressed and/or anxious persons due to social desirable answering or symptom overlap. Therefore, we examined and compared the performance of the AUDIT against the gold standard of a CIDI-based diagnosis of past-year AA or AD in a large sample of persons with a past-year depressive and/or anxiety disorder versus persons without a lifetime depressive and anxiety disorder.

Section snippets

Study sample

Baseline data of the Netherlands Study of Depression and Anxiety (NESDA) were used, comprising 2981 adults (18–65 years). Of that number, 19% were recruited from the community, 54% from primary care settings and 27% from specialized outpatient mental health care settings. Exclusion criteria were insufficient command of the Dutch language and a known primary clinical diagnosis of bipolar disorder, obsessive compulsive disorder, addiction disorder or psychotic disorder. A detailed description of

Sample characteristics

Mean age was 41.3 (SD = 13.0) years and 65.8% were women. In the past year 4.4% had AA and 7.5% AD. Table 1 presents gender-stratified sample characteristics comparing non-depressed/non-anxious versus depressed and/or anxious persons.

Screening characteristics of the AUDIT

Based on the sensitivity and specificity across the full range of cut-off points, Receiver Operating Characteristic (ROC) curves for the AUDIT in detecting AA (blue lines) and AD (black lines) are presented in Fig. 1. Separate lines are shown for men and women with

Discussion

The present study shows that the AUDIT accurately detects CIDI-based alcohol dependence in depressed/anxious men (AUC = 0.89) and women (AUC = 0.88), comparable to its performance in healthy controls. The standard cut-off point of 8 for men was adequate (sensitivity: 0.88; specificity: 0.76), but specificity could be further improved to 0.81 without losing sensitivity by choosing a cut-off point of 9. The identified optimal cut-off point of 6 with a sensitivity of 0.85 and a specificity of 0.80 for

Role of funding source

The infrastructure for the NESDA study (www.nesda.nl) is funded through the Geestkracht program of the Netherlands Organization for Health Research and Development (Zon-Mw, grant number 10-000-1002) and is supported by participating universities and mental health care organizations (VU University Medical Center, GGZ inGeest, Arkin, Leiden University Medical Center, GGZ Rivierduinen, University Medical Center Groningen, Lentis, GGZ Friesland, GGZ Drenthe, Scientific Institute for Quality of

Conflict of interest

None of the authors report competing interest.

Acknowledgements

We thank Dr. Adriaan W. Hoogendoorn for his assistance with statistical analyses.

References (29)

  • C. Bouza et al.

    Efficacy and safety of naltrexone and acamprosate in the treatment of alchol dependence: a systematic review

    Addiction

    (2004)
  • Boschloo, L., Vogelzangs, N., Smit, J. H., van den Brink, W., Veltman, D. J., Beekman, A. T. F., and Penninx, B. W. J....
  • S.E. Bruce et al.

    Influence of psychiatric comorbidity on recovery and recurrence in generalized anxiety disorder social phobia, and panic disorder: A 12-year prospective study

    Am. J. Psychiatry

    (2005)
  • J.D. Buckner et al.

    Implications of comorbid alcohol dependence among individuals with social anxiety disorder

    Depress. Anxiety

    (2008)
  • Cited by (37)

    • Psychological risk factors and the course of depression and anxiety disorders: A review of 15 years NESDA research

      2021, Journal of Affective Disorders
      Citation Excerpt :

      In multivariate models, disorder groups were marked by high neuroticism levels (Noteboom et al., 2016) and low extraversion as reflected in low positive affect and low sociability (Spinhoven et al., 2014). Neuroticism showed substantial correlations with many of the more specific psychological risk factors, such external locus of control (LoC), hopelessness, rumination, worry, experiential avoidance, and fearful avoidance (r = .65 to .80); and all these specific risk factors also showed substantial correlations with cognitive reactivity (r > .50, Kruijt et al., 2013) and low extraversion (r ≤ -.45, see Boschloo et al., 2010; Drost et al., 2014; Glashouwer et al., 2011, 2012; Hovenkamp-Hermelink et al., 2019; Kok et al., 2017; Spinhoven, Elzinga, et al., 2015; Spinhoven, Drost, et al., 2016; Struijs, Lamers, Spinhoven, et al., 2018; Struijs et al., 2020). These concepts are thus all closely related.

    • AUDIT and AUDIT-C as screening instruments for alcohol problem use in adolescents

      2018, Drug and Alcohol Dependence
      Citation Excerpt :

      Comorbidity of AUDs with other mental disorders is common (Pirkola et al., 2000). Validity studies of the AUDIT and AUDIT-C among psychiatric patients have indicated mostly acceptable screening properties for alcohol problems (Boschloo et al., 2010; Dawson et al., 2005; Reinert and Allen, 2007). Alcohol use in adolescence is associated with disruptive disorders, poor impulse-control, and conduct disorders but also with depression and depressive symptoms (Masten et al., 2008).

    • Prevalence, correlates and treatment status of alcohol use disorders in psychiatric patients in China

      2017, General Hospital Psychiatry
      Citation Excerpt :

      The Alcohol Use Disorders Identification Test (AUDIT) is a structured and standardized screening instrument developed by the WHO in 1982, which consists of 10 items about recent alcohol use, alcohol dependence symptoms, and alcohol-related problems [18]. Studies conducted in many countries and a variety of settings [19–24] suggest that the AUDIT provides valid and reliable detection of unhealthy alcohol use when the cutoff is 8 and a score of > 13–20 has generally been associated with alcohol dependence. However, the recommended cutoff points of the AUDIT in Chinese version for unhealthy alcohol use and alcohol dependence were 7 and 16, respectively [25].

    • Differential association of somatic and cognitive symptoms of depression and anxiety with inflammation: Findings from the netherlands study of depression and anxiety (NESDA)

      2013, Psychoneuroendocrinology
      Citation Excerpt :

      Markers of an unhealthy lifestyle were considered as covariates, because they have been linked to both psychopathology and inflammation. Smoking status was categorized as nonsmoker, former, and current smoker and alcohol intake was measured with the Alcohol Use Disorders Identification Test (Saunders et al., 1993; Boschloo et al., 2010) and defined as <1 glass per week, 1–14 glasses per week and >14 glasses per week. Body Mass Index (BMI) was determined as measured weight in kilograms divided by the square of the measured height in meters.

    View all citing articles on Scopus
    View full text