Tools for Assessing Generalized Anxiety Disorder

March 1, 2006

Generalized anxiety disorder (GAD) is a chronic, impairing and highly comorbid psychiatric condition. A small but sufficient group of empirically supported instruments to assess the severity of GAD are now available.

Generalized anxiety disorder (GAD) is a chronic, impairing and highly comorbid psychiatric condition afflicting an estimated 2.1% to 3.1% of the U.S. population during any given 12-month period (Grant et al., 2005; Kessler et al., 2005). Although historically neglected relative to other anxiety disorders, recent years have witnessed increasing attention and interest in the nature and treatment of GAD. A number of factors have contributed to these developments including elimination of GAD's status as a residual category in the DSM-III-R and several empirically based refinements in the diagnostic criteria for GAD. Among the most consequential of these revisions has been the designation of uncontrollable worry (apprehensive expectation) as the hallmark feature of the disorder and the specification of six key associated symptoms, primarily centering on motor tension (e.g., muscle tension, aches or soreness, restlessness) and vigilance (e.g., feeling keyed up or on edge, difficulty concentrating).

As currently defined in the DSM-IV-TR, a diagnosis of GAD involves excessive anxiety and worry about a number of events or activities, which occur more days than not for at least six months. The worrying is difficult to control, and the individual experiences three or more associated symptoms, including restlessness, fatigue, difficulty concentrating, irritability, muscle tension and sleep disturbance. The focus of worry is not on a situation that could be more parsimoniously subsumed under another anxiety disorder (e.g., fear of embarrassment or humiliation in social situations, as in social anxiety disorder). Additionally, there must be evidence of clinically significant functional impairment.

Given its prevalence and associated impairment, the significant burden imposed on health care resources, accurate assessment of GAD and its severity by mental health and primary care clinicians is an increasingly important goal. Reliable diagnosis and assessment of disorder severity can guide the nature, frequency and duration of therapeutic interventions. Moreover, accurate assessment of initial disorder severity provides a benchmark from which ongoing evaluation of treatment effectiveness can proceed.

Assessing Severity in GAD

Assessing the severity of GAD can be done quantitatively (e.g., a 0-10 rating scale), qualitatively (e.g., categorical classifications such as mild, moderate or severe), or with some combination of these approaches. It should apply to relevant domains (e.g., worry, associated symptoms, subjective distress, impairment in role-functioning). In the recent National Comorbidity Survey-Replication study, severity of 12-month GAD cases was evaluated among a large representative community sample (Kessler et al., 2005). Findings revealed that 32.3% of individuals with GAD were classified as serious, 44.6% of cases as moderate and 23.1% as mild, with severity defined by the consequences of GAD in several domains of functioning. Serious cases were defined by: a recent suicide attempt; substantial work limitation; substance dependence with serious role impairment; or 30 or more days out of role in the year. Moderate cases were defined by: a suicide gesture, plan or ideation; substance dependence without serious role impairment; and moderate work limitation or moderate role impairment. Disorders not meeting the criteria for serious or moderate severity were classified as mild.

A complementary way to assess the severity of GAD is to focus on the specific symptoms constituting the disorder. A number of psychometrically sound instruments designed toward this end have been reported in the anxiety disorder and GAD research literature (Table 1). (For more comprehensive reviews see Roemer and Medaglia [2001] and Turk et al. [2004]).

Clinician Measures

The Hamilton Anxiety Rating Scale (HAM-A) is a widely used interview measure designed to assess anxiety. Although it predates current conceptualizations of GAD, it assesses several of the associated symptoms of GAD. The HAM-A features both psychic and somatic anxiety subscales. The psychic subscale, which is comprised of items that address the more subjective cognitive and affective components of anxious experience (e.g., anxious mood, tension, fears, difficulty concentrating), is particularly useful in assessing the severity of GAD. In contrast, the somatic subscale emphasizes features of GAD that are somewhat less typical, including autonomic arousal, respiratory and cardiovascular symptoms. The HAM-A has become an industry standard in clinical trials of pharmacotherapy and psychotherapy for GAD. Full-scale pretreatment scores of 18 to 20 or greater are a common index of severity required for entry into a clinical trial. A 40% to 50% reduction in the HAM-A total score (i.e., a full-scale score in the range of 8 to10 or less) is a typical criterion for defining treatment response (Gelenberg et al., 2000; Rickels et al., 2005). One notable limitation in using the HAM-A to assess GAD severity is that it does not evaluate excessive or difficult-to-control worry, which is recognized as the central feature of GAD in current conceptualizations (e.g., DSM-IV-TR).

Recently, Shear and colleagues (in press) have developed the Generalized Anxiety Disorder Severity Scale (GADSS). The GADSS was explicitly designed to be a specific measure of GAD severity, and it facilitates probing of typical domains of worry (e.g., future, health, finances, relationships) and detection of the six associated symptoms defining GAD. Severity is assessed by ratings of frequency and distress due to worrying, the associated symptoms of GAD, and impairment/interference in social and work functioning. Preliminary evaluation has shown that the GADSS has good reliability, validity and treatment sensitivity.

Self-Report Measures

Self-report measures have the advantages of brevity, ease of administration and scoring, and a decreased demand on human resources, which are often limited in high-volume outpatient clinics and private practice settings. The Beck Anxiety Inventory (BAI) and the Trait Anxiety subscale of the State-Trait Anxiety Inventory (STAI-T) are among the most widely used self-report measures of anxiety, although each has limitations. The BAI is biased toward somatic and panic-like symptoms of anxiety, which have been shown to be less characteristic of GAD, whereas the STAI-T appears to confound anxiety and depression and may be more a measure of general negative affectivity than anxiety. Despite this limitation, the STAI-T has demonstrated sensitivity to treatment change in studies of cognitive-behavioral therapy for GAD.

The Penn State Worry Questionnaire (PSWQ) measures a general trait-like tendency to worry excessively. As such, it is an especially useful instrument to assess the severity of pathological worry characteristic of GAD. It has been shown to discriminate among the anxiety disorders, as individuals with GAD score significantly higher on the PSWQ than any other anxiety disorder group, and it is also sensitive to change in studies of cognitive-behavioral therapy for GAD (Molina and Borkovec, 1994). The Generalized Anxiety Disorder Questionnaire-IV (GADQ-IV) is a brief measure designed to screen for the presence of GAD, based on DSM-IV diagnostic criteria.

Other Features

Several other dimensions of individual functioning also warrant attention during assessment, as these may significantly impact and/or contribute to overall severity of GAD (Table 2). Among these, people with GAD often have maladaptive beliefs about worry, including the assumption that worry serves a protective function or that worry will prevent negative and feared outcomes from occurring. Asking a patient to describe their beliefs about the effects of worry may identify targets to address in treatment.

Emotional functioning is another important area to examine, as it has been shown that people with GAD have difficulty identifying their emotions or may consciously attempt to avoid experiencing emotions by using worry as a cognitive avoidance strategy. In such cases, interventions aimed at improving emotion regulation may be of benefit. While overt avoidance may not be as common in GAD as in phobic disorders, subtle behavioral avoidance is often present and may serve as a maintaining factor.

Careful probing during the clinical evaluation and between-session patient self-monitoring can help to identify areas of avoidance to target in treatment. Interpersonal functioning may also be impaired, and social or intimate relationships are often a major focus of patient's excessive worrying. Fears of disapproval, negative evaluation and appearing incompetent are common in patient's clinical presentation and need to be addressed.

Finally, careful assessment of comorbid Axis I and Axis II psychopathology is crucial, as GAD has been found to be a highly comorbid disorder with 30-day and lifetime comorbidity rates estimated at 66.3% and 90.4%, respectively (Wittchen et al., 1994). Comorbid psychopathology has been shown to be associated with increased severity of GAD. Among the most common co-occurring conditions are other anxiety disorders, depressive disorders and substance use disorders.

Concluding Comments

As public awareness and recognition of GAD continues to grow, clinicians will likely evaluate and treat these patients with increasing frequency. Accurate detection, diagnosis and assessment of disorder severity can inform the clinician's interventions and improve treatment effectiveness. A small but sufficient group of empirically supported instruments to assess the severity of GAD are now available and should be judiciously combined with one's clinical acumen.

Dr. Belzer is a research fellow in affective, anxiety and related disorders, Department of Psychiatry, Columbia University College of Physicians and Surgeons and the Anxiety Disorders Clinic of New York State Psychiatric Institute. He has indicated he has nothing to declare.




Gelenberg AJ, Lydiard RB, Rudolph RL et al. (2000), Efficacy of venlafaxine extended-release capsules in nondepressed outpatients with generalized anxiety disorder: A 6-month randomized controlled trial. JAMA 283(23):3082-3088.


Grant BF, Hasin DS, Stinson FS et al. (2005), Prevalence, correlates, co-morbidity, and comparative disability of DSM-IV generalized anxiety disorder in the USA: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Psychol Med 35(12):1747-1759.


Kessler RC, Chiu WT, Demler O et al. (2005), Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. [Published erratum Arch Gen Psychiatry 62(7):709. Merikangas, KR (added).] Arch Gen Psychiatry 62(7):617-627 [see comment].


Molina S, Borkovec TD (1994), The Penn State Worry Questionnaire: psychometric properties and associated characteristics. In: Worrying: Perspectives on Theory, Assessment and Treatment, Davey GCL, Tallis F, eds. New York: Wiley, pp265-283.


Rickels K, Pollack MH, Feltner DE et al. (2005), Pregabalin for treatment of generalized anxiety disorder: a 4-week, multicenter, double-blind, placebo-controlled trial of pregabalin and alprazolam. Arch Gen Psychiatry 62(9):1022-1030.


Roemer L, Medaglia E (2001), Generalized anxiety disorder: a brief overview and guide to assessment. In: Practitioner's Guide to Empirically Based Measures of Anxiety, Antony MM, Orsillo SM, Roemer L, eds. New York: Kluwer Academic/ Plenum Publishers, pp189-195.


Shear K, Belnap BH, Mazumdar S et al. (in press), The Generalized Anxiety Disorder Severity Scale (GADSS): a preliminary validation study. Depress Anxiety.


Turk CL, Heimberg RG, Mennin DS (2004), Assessment. In: Generalized Anxiety Disorder: Advances in Research and Practice, Heimberg RG, Turk CL, Mennin DS, eds. New York: Guilford Press, pp219-247.


Wittchen HU, Zhao S, Kessler RC, Eaton WW (1994), DSM-III-R generalized anxiety disorder in the National Comorbidity Survey. Arch Gen Psychiatry 51(5):355-364.