RUNNING TITLE: Geriatric Anxiety Inventory
Development and Validation of the Geriatric Anxiety Inventory
Nancy A. Pachana
School of Psychology
University of Queensland
Gerard J. Byrne
Associate Professor, Discipline of Psychiatry, School of Medicine
University of Queensland
Helen Siddle
Consultant Psychiatrist, Princess Alexandra Hospital, Brisbane
Natasha Koloski
Emma Harley
School of Psychology
University of Queensland
Elizabeth Arnold
Research Nurse, Discipline of Psychiatry, School of Medicine
University of Queensland
Corresponding Author:
Nancy A Pachana, Ph.D.
School of Psychology
University of Queensland
Brisbane, QLD 4072
Australia
Tel: +617 3365-6832
Fax: +617 3365-4466
e-mail: npachana@psy.uq.edu.au
Word Count = ~3,000
Paper in submission to International Psychogeriatrics
RUNNING TITLE: Geriatric Anxiety Inventory
ABSTRACT
Background:
Anxiety symptoms and anxiety disorders are highly prevalent among older people
although infrequently the subject of systematic research in this age group. One
important limitation is the lack of a widely accepted instrument to measure
dimensional anxiety in both normal older people and older people with mental health
problems seen in various settings. Accordingly, we undertook the development and
initial testing of a short scale to measure anxiety in older people.
Methods:
We generated a large number of potential items de novo and by reference to existing
anxiety scales and then reduced the number of items to 60 through consultation with a
reference group consisting of psychologists, psychiatrists and normal older people. We
then tested the psychometric properties of these 60 items in 502 normal older people
and 46 patients attending a psychogeriatric service. We were able to reduce the number
of items to 20. We chose a one week perspective and a dichotomous response scale.
Results:
Cronbach’s alpha for the GAI-20 was 0.91 among normal older people and 0.93 in the
psychogeriatric sample. Concurrent validity with a variety of other measures was
demonstrated in both the normal sample and the psychogeriatric sample. Inter-rater
and test-retest reliability were found to be excellent. Receiver operating characteristic
(ROC) analysis indicated a cut point of 10/11 for the detection of Generalised Anxiety
Disorder in the psychogeriatric sample with 83% of patients correctly classified with a
specificity of 84% and a sensitivity of 75%.
Conclusions:
The Geriatric Anxiety Inventory (GAI) is a new 20-item self-report or nurse-
administered scale that measures dimensional anxiety in older people. We have
demonstrated that it has sound psychometric properties. Initial clinical testing indicates
that it is able to discriminate between those with and without any anxiety disorder and
between those with and without DSM-IV Generalised Anxiety Disorder.
RUNNING TITLE: Geriatric Anxiety Inventory
Key words: Anxiety; Anxiety disorder; Aged; Aged, 80 and over; Generalised Anxiety
Disorder; Psychological Test;
RUNNING TITLE: Geriatric Anxiety Inventory
DEVELOPMENT AND VALIDATION OF THE
GERIATRIC ANXIETY INVENTORY
Introduction
The prevalence of anxiety symptoms and anxiety disorders has been reported to
decline with advancing age (Flint, 1994; Henderson et al., 1998). Despite this decline,
anxiety in its various forms remains one of the most common psychiatric problems
experienced by older people (Australian Bureau of Statistics, 1998). However, while
anxiety is a common problem in older adults, anxiety symptoms and anxiety disorders
remain both under-recognised and under-treated by health professionals (Scogin,
1998). This is despite the fact that anxiety disorders contribute to significant morbidity,
loss of functioning, and lower quality of life (Blazer et al., 1991).
Anxiety disorders are more prevalent in older adults with chronic general
medical conditions and are also highly co-morbid with depressive disorders (Beekman
et al., 2000; Lenze et al., 2001). Anxiety disorders, however, remain less well studied
in older adults than other disorders such as depression and dementia. An accurate
picture of the true prevalence and incidence of anxiety disorders remains elusive
(Krasucki et al. 1998). This may be due in part to methodological factors such as the
use of diagnostic criteria and instruments not validated for use with older adults
(Fuentes & Cox, 1997) and to response bias during epidemiological surveys (Jorm,
2000). Diagnostic difficulties, including problems of recognising age-specific
symptoms, distinguishing symptoms of chronic physical disorders from the symptoms
of anxiety, and the influence of age-related psychosocial issues on presentations of
anxiety in later life have been increasingly discussed in the literature (Palmer et al.,
1997).
There is little agreement as to a “classic” presentation of the common form of
generalised anxiety in later life. Shapiro et al (1999) suggest both cognitive and
affective components are fundamental, not only to a discussion of core aspects of the
presentation of pathological anxiety, but also in terms of distinguishing anxiety from
other conditions such as depression. There is also debate as to the degree to which
anxiety may be distinguished from more commonplace occurrences like worry. In fact,
as anxiety serves a necessary, indeed adaptive, function in the everyday lives of older
people, its absence is neither natural nor desirable. However, it is unclear at which
RUNNING TITLE: Geriatric Anxiety Inventory
point anxiety should be considered abnormal and maladaptive in older people. To
further investigate these and a variety of other issues, a valid and reliable measure of
anxiety in older people is needed.
Many instruments have been developed to measure the symptoms, distress
levels and characteristics of the various anxiety disorders; the vast majority of these
have been developed in and for young adult populations. Yet the importance of age-
congruent norms is critical (Owens et al., 2000). A few such instruments [e.g. the Beck
Anxiety Inventory (BAI; Beck et al., 1988)] have normative data for older populations.
Some have been modified for use with older adults (e.g. the Adult Manifest Anxiety
Scale – Elderly Version; Lowe & Reynolds, 2000). And a very few anxiety measures
[e.g. the Short Anxiety Screening Test (SAST; Sinoff et al., 1999)] have been
specifically designed for use with older adult populations. Instruments to measure
anxiety levels can be constructed as clinician-rated or observational in nature (e.g.
Hamilton Anxiety Scale; Hamilton, 1959; Maier et al., 1988) or can be designed as
self-report measures (e.g. the State-Trait Anxiety Inventory; Spielberger et al., 1970;
and the Padua Inventory; Sanavio, 1988).
However, many of these instruments, even those designed specifically for older
adult populations, have shortcomings in terms of clinical and/or psychometric utility.
These deficiencies fall into three main categories: a) many inventories (e.g. Hospital
Anxiety and Depression Scale; Zigmond & Snaith, 1983) are found to be poor in
detecting anxiety in older samples (Davies et al., 1993); b) many inventories (e.g. Beck
Anxiety Inventory) are less suitable for older adults with mild cognitive deficits (e.g.
wording of items and/or response sets too long or complex); and c) somatic items in
some inventories (e.g. Goldberg and Bridges Scale; Goldberg et al.1988) fail to reflect
the somatic nature of some older adults’ manifestations of anxiety disorders (Turnbull,
1989) while resulting in too great an overlap with somatic symptoms of normal ageing,
co-morbid medical conditions or medication side effects (e.g. shortness of breath in
chronic obstructive pulmonary disorder or cardiac failure, conditions that are relatively
prevalent in later life).
Some authors have argued that late-life anxiety scales should be able to
measure the symptoms of anxiety in older adults with and without co-morbid
depression, and in the latter situation, distinguish depression from anxiety (Beck &
