Aging & Mental Health

ISSN: 1360-7863 (Print) 1364-6915 (Online) Journal homepage: https://www.tandfonline.com/loi/camh20

Assessment of anxiety in older adults: A

systematic review of commonly used measures

Zoé Therrien & John Hunsley

To cite this article: Zoé Therrien & John Hunsley (2012) Assessment of anxiety in older adults:
A systematic review of commonly used measures, Aging & Mental Health, 16:1, 1-16, DOI:
10.1080/13607863.2011.602960

To link to this article: https://doi.org/10.1080/13607863.2011.602960

Published online: 15 Aug 2011.

Submit your article to this journal

Article views: 3498

View related articles

Citing articles: 21 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=camh20
Aging & Mental Health
Vol. 16, No. 1, January 2012, 1–16

Assessment of anxiety in older adults: A systematic review of commonly used measures

Zoé Therrien* and John Hunsley
School of Psychology, University of Ottawa, Ottawa, Canada
(Received 22 March 2011; final version received 17 June 2011)

Objectives: The authors set out to systematically review the research literature in order to identify the anxiety
measures most commonly used in the assessment of older adults. Once identified, the literature was reviewed to
determine the extent to which these instruments had age-relevant norms and psychometric data supporting their
use with older adults.
Method: Literature searches were conducted in PsycINFO and PubMed to identify research articles in which
anxiety measures were completed by older adults. After screening for suitability, a total of 213 articles were
reviewed to determine the most commonly used anxiety measures with older adults to examine the psychometric
properties of these instruments and to evaluate whether the instruments are appropriate for use with older adults.
Results: A total of 91 different anxiety measures were used in the 213 included articles. Twelve anxiety measures
were most commonly used in the literature and of those three were specifically developed for older adults.
Conclusions: Of the most commonly used measures, the majority lacked sufficient evidence to warrant their use
with older adults. Based on psychometric evidence, three measures (Beck Anxiety Inventory, Penn State Worry
Questionnaire, and Geriatric Mental Status Examination) showed psychometric properties sufficient to justify
the use of these instruments when assessing anxiety in older adults. In addition, two measures developed
specifically for older adults (Worry Scale and Geriatric Anxiety Inventory) were also found to be appropriate for
use with older adults.
Keywords: older adults; anxiety; assessment; psychometric properties; systematic review

Introduction that so little effort has been put toward identifying

Over the past decade, there has been a growing and promoting evidence-based assessment (Hunsley &
emphasis on the importance of evidence-based mental Mash, 2007). Indeed, without scientifically sound
health practice (American Psychological Association assessment, clinicians cannot clearly evaluate a
Presidential Task Force on Evidence-Based Practice, patient’s level of functioning; without this information,
2006; Institute of Medicine, 2001). According to the the development of solid case formulations (including
Institute of Medicine (2001), evidence-based practice diagnostic considerations) is not possible, and without
in health care services integrates information derived high quality case formulations, it becomes difficult to
from the best research evidence, clinical expertise, and make informed treatment choices. Moreover, in
the patient’s values when contemplating health care research contexts, assessment measures are used to
services for a patient. This framework is based on the select participants and evaluate treatment; thus, they
acknowledgment that positive outcomes are less likely are critical for developing evidence-based treatment
if the chosen service does not have a research base that (Cohen et al., 2008). In sum, whether assessment
measures are used for diagnostic purposes, to select
shows the potential to improve the client’s functioning
research participants, to establish case conceptualiza-
(Forman, Fagley, Steiner, & Schneider, 2009). The
tion, to inform the choice of a treatment plan, or to
evidence-based practice movement has been endorsed
monitor treatment outcome, the choice of measures is
through reports released by several scientific organi-
central to the quality of services provided. Fortunately,
zations in the field of health and mental health.
some recent efforts have been made toward the
Much of the efforts to evaluate and disseminate
development of general guidelines to be used in
evidence-based mental health practices have focused
selecting the best instruments. These efforts emphasize,
on intervention services. For example, a growing body
in particular, the importance of solid psychometric
of research has concentrated on the identification of properties, appropriate norms, and evidence of clini-
efficacious psychological interventions (Ayers, Sorrell, cal utility (Holmbeck & Devine, 2009; Hunsley &
Thorp, & Wetherell, 2007) and the establishment of Mash, 2008).
guidelines for determining whether an intervention can Although research on evidence-based assessment
be considered evidence-based (Chambless & Hollon, is increasing, research on the assessment of older adults
1998). As assessment is a necessary component of is much less developed (Ayers et al., 2007). With the
health care services that should inform treatment growing number of older adults in the general popu-
selection and implementation, it seems incongruous lation, there is also a concomitant rise in the number of

*Corresponding author. Email: zther093@uottawa.ca

ISSN 1360–7863 print/ISSN 1364–6915 online

ß 2012 Taylor & Francis
http://dx.doi.org/10.1080/13607863.2011.602960
http://www.tandfonline.com
2 Z. Therrien and J. Hunsley

older adults who require mental health services. As affective disorder (Regier et al., 1988). With regard to
diagnosis and treatment selection is informed by the older adult population, one study using a commu-
assessment data, it is necessary to have measures that nity-based sample of Canadian adults aged 55 and
are appropriate for an older population, but the lack of older found that depression was the most common
research evidence for the psychometric quality of many comorbid disorder among those with anxiety disorders,
of these measures makes it challenging to choose an with 23% of those with an anxiety disorder also
appropriate measure for use with older adults meeting criteria for major depressive disorder
(Edelstein et al., 2001). (Cairney, Corna, Veldhuizen, Herrmann, & Streiner,
The empirical literature on anxiety prevalence 2008). Studies of depressed older adults also indi-
suggests that it has become a widespread problem in cate that approximately half of these individuals meet
late life. With prevalence estimates ranging from 1.2% the criteria for an anxiety disorder (Beekman et al.,
to 15% in community samples of older adults (Bryant, 2000).
Jackson, & Ames, 2009; Wolitsky-Taylor, Castriotta, The detection of anxiety in older adults is also
Lenze, Stanley, & Craske, 2010), and from 1% to 28% complicated by the high frequency of medical disorders
in clinical samples of older adults (Bryant et al., 2009), in this age group. With studies suggesting that between
it is more common than depression. Furthermore, 80% and 86% of adults aged 65 and older have at least
among hospitalized geriatric patients, the prevalence one significant medical condition (Dawson, Kline,
of anxiety disorders is estimated to be as high as 43%, Wiancko, & Wells, 1986; Haley, 1996; Naughton,
and known to remain high after discharge (Kvaal, Feely, & Bennett, 2007), many symptoms of anxiety
Macijauskiene, Engedal, & Laake, 2001). Moreover, may be overlooked or wrongly attributed to a medical
the prevalence of older adults with anxiety symptoms illness. Older adults may be more likely to attribute
that do not meet criteria for an anxiety disorder has physical symptoms related to anxiety (including muscle
been found to range from 15% to 52.3% in community tension, hypervigilance, and difficulties related to
samples and 15–56% in clinical samples (Bryant et al., sleep) to a medical problem than to anxiety (Gurian
2009). There is significant impairment and a lower level & Miner, 1991). In turn, many physical conditions
of quality of life among anxious individuals such as cardiovascular disease, respiratory disease,
(Mendlowicz & Stein, 2000) and, according to de hyperthyroidism, and pulmonary difficulties can
Beurs et al. (1999), older adults who present some involve anxiety symptoms, making it difficult to
anxiety symptoms are as negatively affected in their establish the underlying cause of these symptoms
quality of life as those who meet criteria for an anxiety (Alwahhabi, 2003; Kogan et al., 2000). Additionally,
disorder. anxiety symptoms can occur as side effects of medica-
Unfortunately, research on anxiety disorders in tion that is being used to treat a medical condition.
older adults has not grown at the same rate as research In assessing elderly patients with anxiety complaints
on anxiety disorders in younger populations (Dennis, or unexplained physical symptoms, clinicians need to
Boddington, & Funnell, 2007). One consequence of obtain enough information to make an appropriate
this is that relatively little is known about the evidence- diagnosis or, at the least, to describe possible differen-
based assessment of anxiety in older adults. In both tial diagnoses. Moreover, current anxiety diagnostic
clinical and research contexts, self-report appears to be criteria and measures, most of which were developed
the dominant assessment method for gathering infor- originally for use with much younger persons, are
mation on the experience of anxiety (Alwahhabi, 2003; weighted heavily with somatic items, making it difficult
Dennis et al., 2007). For self-report measures to to distinguish between medical and psychological
provide valid information, information on additional causes of anxiety in this population. Methods for
factors such as the frequent comorbidity of mental and assessing anxiety in older adults can be enhanced
physical health problems (Cully et al., 2006; Wolitsky- through consideration of unique aspects of anxiety in
Taylor et al., 2010) and the use of multiple medications this population, including age-relevant aspects of
in older populations must be obtained and incorpo- physical and mental health status.
rated into assessment decisions. Anxiety disorders are also frequently comorbid
The assessment and diagnosis of late life anxiety is with cognitive decline and dementia among the elderly
especially challenging, as symptoms of anxiety can be (Seignourel, Kunik, Snow, Wilson, & Stanley, 2008).
confused with some aspects of the normal aging This comorbidity is partially due to the fact that there
process (Lenze & Wetherell, 2009) as well as with may be a specific relation with anxiety and several
medical conditions and comorbid mental disorders types of cognitive impairments (Wolizky-Taylor et al.,
(Kogan, Edelstein, & McKee, 2000). It is well- 2010). Cognitive decline is an important factor to
established that anxiety and depression are frequently consider when assessing anxiety in this age group
comorbid in younger adults (de Graaf, Bijl, Spijker, because it may affect the presentation of the symptoms
Beekman, & Vollebergh, 2003; Kessler, et al., 2005). In as well as the ability to communicate them to a
the Epidemiological Catchment Area study of people clinician. What seems to be symptoms of anxiety in
aged 18–54, 20% of individuals who received a older adults (e.g., agitation) may in reality be the result
diagnosis of any anxiety disorder in the past 6 of the challenges associated with memory impairment.
months also received a diagnosis of some type of Finally, older adults or clinicians may misattribute
 Aging & Mental Health 3

anxiety symptoms (e.g., fatigue, difficulty concentrat- combined the following terms anxiety, anxiety disor-
ing) to normal aging processes. It has also been ders, generalized anxiety disorder AND geriatric assess-
suggested that the previous cohort of older adults ment, geriatric patients, geriatric psychiatry, geriatric,
might be less comfortable than current generations in gerontology. The key words were searched as article
discussing emotions and, therefore, more likely to keywords, titles, and abstracts. The search was limited
minimize their symptoms (Pachana, 2008). to older adults by selecting ‘Age 65 and older’ in the
In sum, as older adults experience anxious symp- Age Group section of the search. The search was
toms in a different way than younger adults do, and restricted to articles published between January 1960
as there are many factors that can complicate the use and August 2009, in English or French (as these were
of self-report measures with older adults, it is critical the languages understood by the authors). Articles
that the anxiety measures used with older adults be were obtained through our university library network
demonstrated to be scientifically sound. In recent or, for the articles that were unavailable in print or
years, researchers have reviewed anxiety measures electronically at our university, through inter-library
often used with older adults (Edelstein et al., 2008; loans. If an article was unavailable through both these
Kogan et al., 2000). Although such reviews provide methods, a search was conducted at the National
an important first step in evaluating the scientific Archives of Canada.
merits of available instruments, they do have signif-
icant limitations. Most notably, neither Edelstein
et al. (2008) nor Kogan et al. (2000) indicated the
bases for selecting the instruments they chose to Search results
review. Furthermore, only limited psychometric infor- The search yielded a total of 1427 articles (785 in
mation on selected instruments was presented in these PubMed and 642 in PsycINFO). After reviewing the
reviews, thus making it difficult for readers to titles and abstracts, 592 unique articles were retained at
determine the scientific adequacy of these instruments this point for further examination to determine if they
for clinical or research purposes. fully met our inclusion criteria. As part of this review,
With this background in mind, we set out to only studies presenting original data were included in
systematically review the research literature in order to our final set of retained articles. Articles were excluded
identify the anxiety measures most commonly used in because they did not have a sample that included only
the assessment of older adults. Once identified, our older adults (49.2%), did not use an anxiety measure
intent was to review the literature to determine the (8.5%), did not have a sample of older adults and did
extent to which these instruments had age-relevant not have an anxiety measure (7.1%), were not empir-
norms and psychometric data (reliability, convergent ical studies (20.0%), did not use original data (6.3%),
validity, discriminant validity, and treatment sensitiv- were not in French or English (5.2%), or were
ity) supporting their use with older adults. By summa- unavailable in print or electronic form within our
rizing this research, we provide critical up-to-date university library network, through inter-library loans
information regarding the status of evidence-based or through the National Archives of Canada (3.6%).
instruments for assessing anxiety in older adults. As
indicated, this information is crucial in developing and
providing evidence-based mental health services to
older adults. Data coding
For all the 592 selected articles, the full article was
obtained and reviewed to assess the fit with our
inclusion criteria. After this detailed review, a total
Method of 213 articles were retained. Articles were excluded
Study eligibility and search strategy because, after thorough review, they were found to not
The electronic databases PubMed and PsycINFO were meet the inclusion criteria which required the articles
searched for published journal articles that included to be empirical studies that employed an anxiety
some type of anxiety measure used in a population of measure with adults aged 65 and older. The details of
older adults. Studies meeting the following criteria the 213 retained articles were coded to summarize key
were selected: (1) an empirical study, (2) used an aspects of the study and the anxiety measures used in
anxiety measure with at least one participant, and (3) the study. For each anxiety measure, the rater noted if
only included adults aged 65 years and above in the the authors indicated that the measure was appropri-
sample. ate for older adults, if it was designed for older adults,
To ensure the broadest possible search of the and if relevant norms are available. Participant vari-
databases, two separate searches combining different ables (age range of the sample, the mean age and
key words and study criteria were carried out in each standard deviation of the sample, proportion of men
database. The first search included the following terms: and women included in the sample), and sample
anxiety, anxiety disorder, generalized anxiety disorder information (selection of participants, research setting)
AND assessment, measurement. The second search were also coded.
4 Z. Therrien and J. Hunsley

Results Questionnaire (GHQ, Goldberg et al., 1978; eight

General characteristics of the studies studies), the Beck Anxiety Inventory (BAI; Beck,
Epstein, Brown, & Steer, 1988; 8 studies), the Brief
Of the retained articles, close to half (45%) were in
Symptom Inventory (BSI; Derogatis & Spencer, 1982;
journals that specialize in research on older adults. The
seven studies), the Penn State Worry Questionnaire
majority of articles meeting the inclusion criteria (83%)
(PSWQ; Meyer, Miller, Metzger, & Borkovec, 1990;
were published after 1997. The age of the participants
six studies), and the Symptom Checklist-90-Revised
in the retained articles ranged from 65 to 102 years.
(SCL-90-R; Derogatis, 1994; six studies).
The samples reported in the majority of studies
Only one of the most commonly used measures
consisted of older adults recruited from within their
(GMSE) was created specifically for older adults. In
communities (56%). With respect to study recruitment
the past decade, considerable effort has been made to
procedures, participants were also recruited from
develop anxiety measures specifically suited for older
medical settings (38%), mental health settings (9%),
adults. As such measures are relatively new and might
and residential settings (11%). It is worth noting that
not yet be frequently used in research on older adults,
the recruitment total does not add to 100% as some
we decided to also include in our detailed review the
studies recruited participants in more than one setting
most commonly employed measures created specifi-
(e.g., use of both a clinical and community sample in a
cally for older adults. Accordingly, we relaxed our
study). In approximately half of the studies, the intent
criteria to consider such measures. Based on our review
was to conduct research on normal, community-
of the literature, there were two instruments that have
dwelling older adults; samples were also recruited
been used with sufficient frequency to allow for some
because of having a specific medical disorder (26%),
replication of psychometric findings: the Worry Scale
mental disorder (15%), or both (1%).
(WS; Wisocki, Handen, & Morse, 1986; five studies)
and the Geriatric Anxiety Inventory (GAI; Pachana,
Byrne, Siddle, Koloski, Harley, & Arnold, 2007; four
Most commonly used measures of anxiety studies).
A total of 91 different anxiety measures were used in
the 213 included articles. However, the majority of
these measures were used in only one or two studies. Psychometric evaluation of measures
Most (89%) of the anxiety measures were developed All these measures are standardized self-report ques-
for use with younger adults but used with older aged tionnaires, clinician-administered rating scales, or
samples; few measures (16%) were created with older standardized interviews that have been used to evalu-
adults’ specific experiences or needs in mind. Of the ate anxiety in a variety of settings. The measures and
measures created specifically for older adults, only one their psychometric properties are described in detail
(Geriatric Mental State Examination, GMSE; below. To give a fuller sense of the scientific adequacy
Copeland et al., 1976) was commonly used in the of each measure, we describe the nature of the
studies we examined (14 times). Study authors rarely measures and provide information about the availabil-
mentioned whether measures were appropriate for ity of age-appropriate norms. Additionally, we
older adults (24%), and even fewer reported whether describe both reliability (internal consistency, test–
specific age-relevant norms were available to aid in the retest reliability, and/or inter-rater reliability, as
interpretation of the anxiety data (21%). appropriate) and validity (convergent validity, discrim-
In our review of the scientific status of anxiety inant validity, treatment sensitivity) evidence based on
measures used with older adults, we focused on data from samples of older adults. The information
measures that had been used frequently in the litera- regarding psychometric properties was obtained from
ture, thus allowing for independent replication of the studies identified in the literature search as well as
findings and the availability of data from multiple from the instrument manuals.
types of samples. With this in mind, and in consider-
ation of the measure frequency distribution we found
in the 213 studies, we decided to concentrate on Stait Trait Anxiety Inventory
instruments that had been used in six or more studies. The STAI (Spielberger et al., 1983) is a 40-item self-
Using this criterion, the most commonly employed report questionnaire, derived from the Minnesota
measures to evaluate anxiety in older adults were the Multiphasic Personality Inventory, that is designed
State Trait Anxiety Inventory (STAI; Spielberger, to measure and differentiate between anxiety as a trait
Gorsush, Lushene, Vagg, & Jacobs, 1983; 35 studies), and as a state. The two scales consist of 20 items
the Hospital Anxiety and Depression Scale (HADS; answered on a four-item scale and provide a score
Zigmund & Snaith, 1983; 26 studies), the GMSE ranging from 20 to 80, with higher scores indicating
(Copeland et al., 1976; 14 studies), the Hamilton higher levels of anxiety. A cut-off score of 39/40 for
Anxiety Rating Scale (HARS; Hamilton, 1959; 13 both single scales is normally used to identify clinically
studies), the Goldberg Anxiety and Depression Scale significant symptoms of anxiety. However, optimal
(GADS; Goldberg, Bridges, Duncan-Jones, & cut-off scores for older adults have been found to range
Grayson, 1988; eight studies), the General Health between 44 and 55 (Himmelfarb & Murrell, 1984;
 Aging & Mental Health 5

Kvall, Ulstein, Nordhus, & Engedal, 2005). Hospital Anxiety and Depression Scale
Means and standard deviations are available from The HADS (Zigmond & Snaith, 1983) is a 14-item
the original authors for adults aged 50–69, but there self-report questionnaire developed to evaluate the
are no normative data reported specifically for older presence and the severity of anxiety and depression in
adults. non-psychiatric outpatients. Items referring to symp-
The scale was developed for young and middle- toms that may have physical causes such as insomnia
aged adults, but further research has examined its and dizziness were excluded from the scale during its
psychometric properties with older adults. Good development. It is, therefore, unbiased by comorbid
internal consistency has been shown for both the trait medical conditions. The measure is divided into a
and state scales in older psychiatric outpatients seven-item anxiety subscale (HADS-A) and a seven-
( ¼ 0.92–0.94 for the state version and ¼ 0.88–0.90 item depression subscale (HADS-D). The items are
for the trait version; Kabacoff, Segal, Hersen, & Van rated on a four-point scale and summed to provide a
Hasselt, 1997; Stanley, Beck, & Zebb, 1996; Stanley, score ranging from 0 to 21 for anxiety and for
Novy, Bourland, Beck, & Averill, 2001) and in depression. There are no fixed cut-off scores for the
community-dwelling older adults ( ¼ 0.79–0.90 for HADS but, in their original study, Zigmond and
the trait version and 0.85 for the state version; Snaith (1983) recommended two cut-off scores: 7/8 for
Himmelfarb & Murrell, 1984; Stanley et al., 1996). possible anxiety and depression and 10/11 for probable
Test–retest reliability has been found to be good for the anxiety or depression. A cut-off score of 14/15 was
trait scale (r ¼ 0.58–0.84) and appropriately lower for later added for severe anxiety or depression, but no
the state scale (r ¼ 0.51–0.62; Stanley et al., 1996, empirical data support this score (Snaith & Zigmond,
2001). Unfortunately, there is only limited evidence of 1994). The cut-off scores have not been validated with
concurrent validity, as demonstrated by moderate older adults but there is some evidence that use of the
correlations between the STAI-T and other measures 7/8 cut-off score will correctly identify the majority of
of anxiety (r ¼ 0.33–0.57; Kabacoff et al., 1997; Stanley anxious older adults (Dennis et al., 2007; Haworth,
et al., 1996, 2001); slightly lower correlations were Moniz-Cook, Clark, & Wang, 2007). No norms are
found between the STAI-S and other measures of available for older adults.
anxiety (r ¼ 0.15–0.52; Kabacoff et al., 1996; Stanley The HADS was originally developed for medical
et al., 1996, 2001). Both scales are substantially outpatients aged between 16 and 65, but further
correlated with measures of depression (r ¼ 0.41–0.70; research has since been conducted to validate its use
Stanley et al., 1996, 2001), indicating only limited with older adults. Even though not originally designed
discriminant validity. There is also a concern that the for use with psychiatric outpatients, the internal
STAI is lengthy and easily misinterpreted by older consistency of the HADS is high in samples of such
adults (Dennis et al., 2007). individuals ( ¼ 0.73–0.80; Flint & Rifat, 2002;
Few studies identified in our review used the STAI Wetherell, Birchler, Ramsdell, & Unutzer, 2007). It
to assess the effects of mental health treatments for has also been found to be high in samples of older
older adults. In a sample of older adults receiving medical inpatients ( ¼ 0.75–0.84; Bryant et al., 2009;
cognitive-behavior therapy (CBT) for GAD, the STAI Johnston, Pollard, & Hennessey, 2000; Yu, Lee, Woo, &
was used before and after initiating treatment (5–20 Hui, 2007) and in community samples ( ¼ 0.84–0.85;
weeks after the pretreatment assessment; Stanley et al., Spinhoven et al., 1997). Although we found no study
2001). Post hoc paired comparisons of scores on the that reported evidence of test–retest reliability, the
STAI revealed a significant reduction in anxiety anxiety scale of the HADS has been found to show no
symptoms from pre- to post-treatment. Another significant changes when administered at a two month
study (Stanley et al., 2003) examined the efficacy of a interval in a population of inpatients in a geriatric
CBT treatment relative to a minimal contact control in hospital (Bryant et al., 2009). The HADS-A has
a sample of older adults with GAD. All participants demonstrated only limited evidence of concurrent
completed measures of anxiety (STAI and HARS) and validity, with moderate correlations with other psycho-
worry (WS and PSWQ) before and after initiating logical distress measures such as the BSI (r ¼ 0.54;
treatment. An analysis of simple effects demonstrated Wetherell & Aréan, 2007), the HARS (r ¼ 0.57; Dennis
significant improvements on the STAI as well as on the et al., 2007), and observer ratings of global anxiety
HARS and PSWQ for the CBT group but not for the (r ¼ 0.28; Kenn, Wood, Kucyj, Wattis, & Cunane,
control group. Thus, evidence from two studies 1987). It has shown moderate evidence of discriminant
suggests that the STAI can be sensitive to treatment validity with depression measures (r ¼ 0.26–0.47), but
effects. Although promising, evidence of the instru- moderate correlations between the anxiety and depres-
ment’s treatment sensitivity with other disorders and sion subscale (r ¼ 0.43–0.73) could result in high rates of
other treatments is needed. Overall, the studies exam- misclassification (Davies, Burn, Mckenzie, Brotherwell,
ining the psychometric properties of the STAI have & Wattis, 1993; Dennis et al., 2007; Flint & Rifat, 2002;
yielded mixed results and, therefore, it should be used Johnson et al., 1995; Spinhoven et al., 1997).
with some caution when assessing anxiety in A study conducted by Yu et al. (2007) examined
older adults. the effects of relaxation and exercise training on
6 Z. Therrien and J. Hunsley

psychological outcomes in older patients with heart treatments for older adults. In sum, the GMSE is a
failure. Participants completed the HADS at baseline useful tool to assess the mental health of older adults in
and at the 12th week of treatment. Analyses revealed medical settings as it excludes the effects of physical
that older adults who participated in the relaxation or illness. It has been subjected to many reliability and
exercise therapy showed lower levels of anxiety on the validation studies and is often used when examining
HADS compared to the control sample. Although the validity of other instruments used with older adults
supportive of the instrument’s treatment sensitivity, (Mottram, Wilson, & Copeland, 2000). Although
additional research studies assessing the effects of evidence regarding the validity of the anxiety-related
mental health treatments with the HADS are needed clusters and the treatment sensitivity of the impairment
in the geriatric population. is required, data on the correspondence with DSM
Overall, although the HADS excludes somatic diagnostic criteria suggest that it is likely to be useful in
symptoms and shows high internal consistency, the assessing clinically significant anxiety in older adults.
correlations between the anxiety and depression scale
suggests that it may be most useful as an overall
indicator of distress. These factors, when combined Hamilton Anxiety Rating Scale
with the lack of evidence for clinical cut scores or The HARS (Hamilton, 1959) is a 14-item clinician-
norms relevant to older adults, suggest that this administered rating scale developed to assess the
frequently used measure is not a good option for severity of anxiety symptoms in adults. Seven of
assessing anxiety in older adults (Bryant et al., 2009). the items address psychic/cognitive anxiety and the
remaining seven items somatic anxiety. The items are
rated on a five-point scale and summed to provide a
Geriatric Mental State Examination score ranging from 0 to 56. A score of 17 or less
The GMSE (Copeland et al., 1976) is a semi-structured represents mild anxiety, a score between 18 to 24 mild
clinical interview designed as a mental health assess- to moderate anxiety, and a score of 25 and above
ment for older adults. The original version of the GMSE moderate to severe anxiety. The cut-off scores have not
had 541 items, but it has since been shortened so that been validated with older adults and there are no
it can be administered in 20–50 min. The data can be published norms for older adults (Kogan, Edelstein, &
analyzed by the computerized system AGECAT McKee, 2000; Scheikh, 1991).
(Automated Geriatric Examination or Computer Although the HARS was developed for young and
Assisted Taxomany) to obtain a suggested psychiatric middle-aged adults, there is some recent support for its
diagnosis. Symptoms are grouped into eight syndrome use with older adults. Adequate internal consistency
clusters: organicity, schizophrenia and related para- has been shown in samples of older adults diagnosed
noia, mania, depression, hypochondrias, phobias, and with generalized anxiety disorder ( ¼ 0.77–0.86; Beck,
obsessional and anxiety neurosis. A diagnostic confi- Stanley, & Zebb, 1999; Diefenbach et al., 2001;
dence level is provided for each syndrome, ranging from Schuurmans et al., 2009). High inter-rater reliability
0 (no symptoms) to 5 (very severely affected). A level of has been shown with community samples and with
three or more in a cluster represents a diagnostic case older adults diagnosed with generalized anxiety disor-
(Copeland, Dewey, & Griffiths-Jones, 1986). der (r ¼ 0.81–0.95; Lenze et al., 2009; Stanley et al.,
The GMSE was developed and normed for older 2009; Wetherell, Gatz, & Craske, 2003). The HARS
adults, and has become one of the most widely used showed limited concurrent validity with the STAI-T
comprehensive structured mental health assessments (r ¼ 0.23; Diefenbach et al., 2001) and the BAI
for older adults (Copeland et al., 2002). High inter- (r ¼ 0.47; Morin et al., 1999). The scale has been
rater reliability for the general scale has been found shown to differentiate older adults with generalized
in samples of community-dwelling older adults, older anxiety disorder from those with no anxiety disorders
medical patients, and older psychiatric inpatients due (Beck, Stanley, & Zebb, 1996; Edelstein et al., 2008).
to its well-established procedures and structured However, the scale correlates considerably with the
approach to the administration of questionnaire Hamilton Depression Rating Scale in samples of older
( ¼ 0.73–0.80; Ames, Flynn, Tuckwell, & Harrigan, adults, raising concerns about its discriminant validity
1994; Copeland et al., 1975; Turrina et al., 1991). Test– (r ¼ 0.72–0.92; Beck et al., 1996, 1999; Diefenbach
retest reliability has been somewhat inconsistent, with r et al., 2001). Importantly, the usefulness of the HARS
values ranging from 0.49 to 0.75 (Copeland et al., 1976; with older adults has been questioned due to the heavy
Henderson, Duncan-Jones, & Finlay-Jones, 1983). emphasis placed on somatic symptoms (e.g., tension)
Evidence of concurrent validity in a variety of cultures that are common experiences in aging individuals
has been shown with high correlations between the (Kogan et al., 2000; Skopp et al., 2006).
GMSE and DSM diagnostic criteria (r ¼ 0.76–0.78, Several studies identified in our review used the
Ames et al., 1994; Copeland et al., 1999). We were HARS to assess the effects of psychological and
unable to find any study in which discriminant validity pharmacological treatments for anxiety in older
of the anxiety-related clusters was examined. There adults. In a randomized controlled trial comparing
were no studies identified in our review in which the sertraline and CBT for the treatment of late-life
GMSE was used to assess the effects of mental health anxiety, the HARS showed moderate to large effect
 Aging & Mental Health 7

sizes for the sertraline group and small to moderate anxiety measures, with  values of 0.13–0.28 indicat-
effect sizes for the CBT group at both post-treatment ing poor concurrent validity (Koloski, Smith, Pachana,
and one year follow-up (Schuurmans et al., 2009). & Dobson, 2008). Moderate correlations have been
Similar results were found in a recent randomized found between the GADS anxiety subscale and the
controlled trial comparing the effect of escitalopram GAI in a sample of older adults (Pachana et al., 2007).
with a placebo (Lenze et al., 2009). However, studies Finally, there is also evidence that the anxiety and
examining the effect of CBT in samples of older adults depression subscales of the measure are highly corre-
with GAD have yielded mixed results. Two studies lated in both the general population and in samples of
(Stanley et al., 2003; Wetherell et al., 2003) found that older adults, which suggests poor discriminant validity
CBT participants had improved significantly on the (Christensen et al., 1999; Huber et al., 1999; Koloski
HARS as well as on other anxiety measures, whereas et al., 2008). There were no studies identified in our
another study (Stanley et al., 2009) found that changes review in which the GADS was used to assess the
on the HARS were not significantly different between effects of mental health treatments for older adults. As
the CBT and control group. a result, there is no evidence with respect to the
Overall, the HARS has shown high reliability in treatment sensitivity of the GADS. The inclusion of
samples of older adults and some evidence of treatment somatic symptoms (e.g., waking early and headaches)
sensitivity with psychopharmacological interventions. can overestimate the prevalence of anxiety and depres-
However, more studies are needed to establish its sion and result in classification errors. Overall, the
validity and treatment sensitivity across treatments in psychometric evidence available thus far on the GADS
diverse groups of older adults. Given the lack of age- is limited and provides little support for its use with
relevant norms and cut-off scores, and concerns about older adults.
discriminant validity, this rating scale is not an optimal
choice for use with older adults.
Beck Anxiety Inventory
The BAI (Beck et al., 1988) is a 21-item self-report
Goldberg Anxiety and Depression Scale questionnaire designed to measure the severity of
The GADS (Goldberg et al., 1988) is an 18-item self- anxiety and to distinguish anxiety from depression.
report questionnaire that measures symptoms of The items are rated on a four-point scale and summed
depression and anxiety experienced in the past to provide a score ranging from 0 to 63, with higher
month. The items are rated on yes (1) or no (0) scores representing higher levels of anxiety. According
answers and are summed to provide a score ranging to the manual (Beck & Steer, 1990), the score can be
from 0 to 9 for the depression subscale and for the interpreted as follows: 0–9 (normal anxiety), 10–18
anxiety subscale. According to Goldberg (1988), (mild to moderate anxiety), 19–29 (moderate to severe
patients with anxiety scores of 5 or more or with anxiety), and 30–63 (severe anxiety). Based on the
depression scores of 2 or more have a 50% chance of a information reported in the manual, it is unclear how
clinically important disturbance, and the probability of those cutoffs were derived and there is no mention of
a significant disturbance increases substantially with whether different cut scores should be used with older
higher scores. The anxiety cut-off score has not been populations. In subsequent research conducted with
validated with older adults and there are no normative older adults, no single BAI cutoff proved to be optimal
data available for older populations. due to the tradeoffs between sensitivity and specificity
The GADS was not developed for use with older (Kabacoff et al., 1997).
adults and only a few studies have examined its The BAI was developed and normed with samples
psychometric properties with this population. The of psychiatric adult outpatients. However, since its
GADS showed good internal consistency in a sample development, there have been several studies that
of adults aged 18–79 where no substantial differences evaluated its use with older populations. The internal
were noted when the alpha was calculated separately consistency of data collected with the BAI is high in
for different age groups ( ¼ 0.82; Christensen et al., samples of older adult medical outpatients ( ¼ 0.91–
1999). Another study, conducted with older medical 0.92; Diefenbach, Tolin, Meunier, & Gilliam, 2009;
inpatients, reported good internal consistency Wetherell & Aréan, 1997), older adult psychiatric
( ¼ 0.82; Huber, Mulligan, Mackinnon, Nebuloni- outpatients ( ¼ 0.81–0.93; Kabacoff et al., 1997;
French, & Michel, 1999). Although this shows initial Wetherell & Gatz, 2005), and in community samples
evidence of reliability, no other published studies have ( ¼ 0.87–0.89, Morin et al., 1999; Wetherell & Gatz,
examined reliability indices specifically in samples of 2005). The BAI showed adequate test–retest reliability
older adults. Furthermore, we were unable to find any (r ¼ 0.64–0.75) in samples of older adults (Beck et al.,
reports of test–retest reliability among samples of older 1988; Diefenbach et al., 2009). Moderate correlations
adults. In the general population, the GADS correctly between the BAI and other anxiety measures show
identified over 80% of adult patients with anxiety evidence of concurrent validity, with correlations
disorders (Goldberg et al., 1988). However, when it ranging from 0.29 to 0.63 (Dennis et al., 2007;
comes to older adults, poor agreement has been found Diefenbach et al., 2009; Kabacoff et al., 1997;
between the anxiety subscale of the GADS and other Wetherell & Gatz, 2005). However, despite efforts in
8 Z. Therrien and J. Hunsley

the development of the BAI to disentangle symptoms mental disorders, very little is known about its use with
of anxiety and depression, relatively high correlations older adults. The internal consistency of the measure is
(r ¼ 0.56–0.65) between the BAI and depression mea- high in samples of older community-dwelling adults
sures show only limited evidence of discriminant when using either the 12- or 28-item version ( ¼ 0.75–
validity (Wetherell & Aréan, 1997). 0.90; Boey & Chiu, 1998; Cheung, 2002; Clarke &
In a study of older adults with GAD, participants Clarkson, 2009; Costa et al., 2006; Malakouti,
were randomly assigned to a CBT group, a discussion Fatollahi, Mirabzadeh, & Zandi, 2007), in cognitively
group, or a waiting list, and were assessed before and impaired older adults when using the 12-item version
after treatment (Wetherell et al., 2003). At post- ( ¼ 0.81, Costa et al., 2006) and in older medical
treatment, mean effects for time among GAD partic- outpatients when using either the 12- or 30-item scale
ipants were significant for other psychological distress ( ¼ 0.82–0.92; Dale, Saevareid, & Soderhamn, 2009;
measures, but not for except the BAI and the HDRS, Thygesen, Saevareid, Lindstrom, Nygaard, & Engedal,
suggesting that the BAI may not be sensitive to 2008). We were unable to locate any study reporting
treatment change. Overall, the initial evidence of evidence of test–retest reliability for older adults. Few
psychometric properties as well as the simplicity of studies have examined the validity of the GHQ with
the BAI make it a useful tool to detect the presence older adults. There were no studies identified in our
of anxiety in older adults. However, because of (1) review in which the GHQ was used to assess the effects
potential confounds with depressive symptoms and (2) of mental health treatments for older adults. As a
high somatic item content (13 of the 21 items are result, there is no evidence with respect to the
related to somatic symptoms), the BAI should be used treatment sensitivity of the GHQ. Although most
with caution, especially with samples recruited from studies found that the GHQ could differentiate
medical settings. Additionally, more evidence is between older adults with and without mental disorder
required before it is deemed acceptable for use in (Costa et al., 2006; Malakouti et al., 2007; Mowry &
evaluating treatment effects. Burvill, 1990; Seva, Sarasola, Merino, & Magallon,
1991), one found that it does not differentiate mentally
ill patients from those with somatic illness (Malakouti
General Health Questionnaire et al., 2007). In light of the limited psychometric
The GHQ (Goldberg, 1978) is a self-report question- information available on the GHQ when used with
naire designed to evaluate the presence of minor, non- older adults, the GHQ should be used with consider-
psychotic psychiatric disorders in community setting. able caution when assessing anxiety symptoms in older
The original questionnaire consists of 60 items, but adults.
shorter versions of 30, 28, 20, and 12 items have also
been developed. The 28-item version is most commonly
used in the general population but the shorter 12-item Brief Symptom Inventory
version has been found to be more appropriate with The BSI (Derogatis & Spencer, 1982) is a 53-item self-
older populations (Clarke & Clarkson, 2009). The report questionnaire designed to assess the psycholog-
GHQ incorporates four scales: somatic symptoms, ical distress of medical and psychiatric patients. It is a
anxiety and insomnia, social dysfunction, and severe brief form of the SCL-90-R and covers nine symptom
depression. The items are rated on a four-point scale dimensions (somatization, obsessive-compulsive, inter-
(‘not at all,’ ‘no more than usual,’ ‘rather more than personal sensitivity, depression, anxiety, hostility,
usual,’ and ‘much more than usual’) and are summed phobic anxiety, paranoid ideation, and psychotism)
to provide a score ranging from 0 to 84, with higher and three global indices of distress (Global Severity
scores representing higher levels of distress. When Index (GSI), Positive Symptom Distress Index, and
using the full 60-item scale, a cutoff of 23/24 for the Positive Symptom Total). The global indices measure
total scale is suggested. However, Goldberg and Hillier the level of symptomatology, intensity of symptoms,
(1979) suggest using an alternative binary scoring and number of reported symptoms. The items are rated
method in which the two least symptomatic answers on a five-point scale ranging from 0 (not at all) to 4
(‘not at all’ and ‘no more than usual’) are given a score (extremely). The overall score of the BSI is referred to
of 0 and the two most symptomatic answers (‘rather as the GSI. Scores on the GSI ranges from 0 to 72 and
more than usual’ and ‘much more than usual’) a score each subscale has a score ranging between 0 and 24.
of 1. A total score of 4 or more on any subscale The scores are interpreted by comparing them to age
suggests caseness. There is no specific cutoff suggested, appropriate norms. The manual provides norms for
nor are norms available, for older adults. However, adult non-patients, psychiatric outpatients, inpatients,
some evidence has shown that when using the binary and adolescent non-patients, but not for older adults.
scoring method with a population of older adults, the However, Hale, Cochran, and Hedgepeth (1984) have
best cut-off score for each subscale is 3/4 provided age-relevant norms for community-dwelling
(Pappassotiropoulos, Heun, & Maier, 1997). older adults.
The GHQ was developed to be used with adults Although the BSI was developed for adolescents
and adolescents. Although it is one of the most and adults, some research has been done on its use with
commonly used measures to detect the presence of older adults. High internal consistency has been found
 Aging & Mental Health 9

in samples of cognitively impaired older adults Kogan et al., 2000; Stanley et al., 2001; Wetherell
( ¼ 0.82; Fisher, Segal, & Coolidge, 2003), older et al., 2003). Correlations with self-report measures of
medical outpatients ( ¼ 0.89; Petkus, Gum, King- depression were lower and showed some evidence of
Kallimanis, & Wetherell, 2009; Petkus et al., 2010), discriminant validity (r ¼ 0.12–0.51; Diefenbach et al.,
community-dwelling older adults ( ¼ 0.72–0.79; 2009; Hopko et al., 2003; Kogan et al., 2000; Senior
Fisher et al., 2003; Pektus et al., 2009), and older et al., 2007; Stanley et al., 2001; Wetherell et al., 2003).
adult psychiatric outpatients ( ¼ 0.72–0.90; Wetherell There is concern that some older adults have difficulty
et al., 2007, 2010). We found no study reporting test– completing and interpreting the content of the reversed
retest reliability data for the BSI in samples of older items of the PSWQ (Stanley et al., 2003; Wetherell
adults. There has been little effort to validate the BSI et al., 2003). In order to respond to this problem,
with older adults. One study found that the BSI did not Hopko et al. (2003) eliminated eight of the original
distinguish between medically ill older patients with items to create an abbreviated scale (PSWQ-A). This
and without an anxiety disorder (Wetherell et al., scale has shown good psychometric properties in
2007), whereas another found that it could discrimi- samples of older adults (Crittendon & Hopko, 2006;
nate between homebound older adults with and Hopko et al., 2003; Nuevo, Mackintosh, Gatz,
without anxiety disorders (Petkus et al., 2010). There Montorie, & Wetherell, 2007).
were no studies identified in our review in which the Two studies identified in our review used the
BSI was used to assess the effects of mental health PSWQ to evaluate the effects of CBT for treating
treatments for older adults. Somatic symptoms of the late-life anxiety. In a study examining the efficacy of
BSI are grouped in a somatization subscale and a CBT treatment relative to a minimal contact control
therefore provide some assurance scores on anxiety- group in older adults with GAD, an analysis of simple
related subscales; they are not inflated by the presence effects demonstrated significant improvements on the
of symptoms better explained by a medical condition. PSWQ as well as in other anxiety measures for the
Although the BSI is simple and covers a wide range of CBT group (Stanley et al., 2003). In the aforemen-
symptoms, the lack of supporting psychometric evi- tioned study (Wetherell et al., 2003), the main effects
dence for older adults, especially validity evidence, for time were significant for the PSWQ as well as most
severely limits its usefulness with older adults. of the other anxiety and depression measures. These
two studies are supportive of the treatment sensitivity
of the PSWQ, but additional studies assessing the
Penn State Worry Questionnaire effects of mental health treatments for late-life anxiety
The PSWQ (Meyer et al., 1990) is a 16-item self-report with the PSWQ are needed. Overall then, although the
questionnaire designed to evaluate pathological worry. lack of norms for older adults is problematic, initial
The items are rated on a five-point scale and summed evidence suggests that the PSWQ and its abbreviated
to provide a score ranging from 16 to 80, with higher form may be useful in assessing worry in older adults.
scores reflecting higher levels of worry. Eleven items
are worded in the direction of pathological worry,
whereas the remaining items are worded to indicate the Symptom Checklist 90-R
absence of worry. Although there are no specific cut- The SCL-90-R (Derogatis, 1994) is a 90-item self-
off scores, the mean score for individuals with gener- report questionnaire designed to evaluate a wide range
alized anxiety disorder is between 60 and 68. There are of psychological problems and symptoms of psycho-
no norms for older adults, but a cut-off score of 50 has pathology. It covers nine primary symptom dimensions
been suggested for use with samples of older medical (somatization, obsessive-compulsive, interpersonal
patients (Stanley et al., 2003). sensitivity, depression, anxiety, hostility, phobic anxi-
The PSWQ was developed and normed for younger ety, paranoid ideation, and psychotism) and has three
adults, but several studies have examined its psycho- global indices of distress which give an overall sense of
metric properties in samples of older adults. High psychological distress. The items are rated on a five-
internal consistency has been found in samples of older point scale. A score of 50 or lower on any scale is
adults diagnosed with generalized anxiety disorder considered to be in the normal range and a score of 65
( ¼ 0.81–0.89; Beck, Stanley, & Zebb, 1995; Stanley or above is considered to be a clinical case. The cut-off
et al., 2001; Wetherell et al., 2003), in home care scores have not been validated for use with older
residents ( ¼ 0.79; Diefenbach et al., 2009), and in adults. The manual provides norms for adult and
community-dwelling older adults ( ¼ 0.80–0.91; Beck adolescents non-patients, adult psychiatric outpatients,
et al., 1995; Hunt, Wisocki, & Yanko, 2003; Senior and adult psychiatric inpatients; no norms are pro-
et al., 2007). Moderate to high test–retest reliability has vided for older adults. Furthermore, we were unable to
been found in samples of older adults (r ¼ 0.54–0.78; find any studies examining its psychometric properties
Hopko et al., 2003; Stanley et al., 2001). The PSWQ with older adults and the reviewed articles that used
showed adequate concurrent validity by virtue of the SCL-90-R did not report any psychometric infor-
significant correlations with other self-report measures mation. There were no studies identified in our review
of anxiety (r ¼ 0.29–0.79: Andreescu et al., 2008; in which the SCL-90-R was used to assess the effects of
Diefenbach et al., 2009; Hopko et al., 2003; mental health treatments for older adults. Overall, the
10 Z. Therrien and J. Hunsley

lack of evidence for the psychometric soundness of the in older adults. The questionnaire uses an agree/
SCL-90-R with older adults means that it is currently disagree response choice format, with the number of
not an appropriate measure for use with this ‘agree’ responses added for the total score. The
population. maximum score is 20, with higher scores indicating
higher anxiety. According to the original study, the
optimal cut-off score to identify GAD in older adults
Worry Scale is 10/11, and 8/9 to identify other anxiety disorders
The WS (Wisocki et al., 1986) is a 35-item self-report (Pachana et al., 2007). Similar results have been found
questionnaire developed to measure worries in the in other studies of psychogeriatric patients (Byrne
areas of health (17 items), finances (five items), and et al., 2010; Pachana et al., 2007).
social conditions (13 items). The items are rated on a The GAI was developed and normed with samples
five-point scale ranging from 0 (never) to 4 (much of of community-dwelling older adults and older adults
the time) and summed to provide a score ranging from receiving psychiatric services. Excellent internal con-
0 to 140, with higher scores reflecting higher levels of sistency has been shown in the original study of
worry. No norms or cut-off scores are provided in the community-dwelling older adults ( ¼ 0.92; Pachana
original study. However, a mean total score of 35.4 et al., 2007) and older adults receiving psychiatric
for individual with GAD and of 11 for non-anxious services ( ¼ 0.93; Pachana et al., 2007). High internal
control have been reported in the literature (Stanley, consistency has also been found in other samples of
et al., 1996). Community samples of active older adults community-dwelling older adults ( ¼ 0.90–0.92;
have yielded total mean scores ranging from 10.4 to Andrew & Dulin, 2007; Byrne et al., 2010; Pachana
17.4 and samples of homebound older adults et al., 2007), in psychogeriatric samples ( ¼ 0.93;
have found mean scores ranging from 17.3 to 23.7 Pachana et al., 2007), and in an older adult sample
(Wisocki, 1994). receiving home care ( ¼ 0.93; Diefenbach et al., 2009).
The WS has been created specifically for use with Lower, but still good, internal consistency was found
older adults. The WS total score is associated with in older adults with mild cognitive impairment
excellent overall internal consistency in both GAD ( ¼ 0.76; Rozzini et al., 2009). Researchers have
( ¼ 0.81–0.93, Stanley et al., 1996, 2001) and non- described the data obtained from the instrument as
anxious samples ( ¼ 0.93–0.94; Hunt et al., 2003; having sound test–retest reliability (Pachana et al.,
Stanley et al., 1996) and good internal consistency for 2007), but not report precise reliability information.
its subscales ( ¼ 0.76–0.95; Hunt et al., 2003; Stanley One study conducted with older adults with mild
et al., 1996). Adequate test–retest reliabilities have been cognitive impairment showed good test–retest reliabil-
found in samples of older adults for the total scale ity (r ¼ 0.86; Rozzini et al., 2009). Moderate to strong
(r ¼ 0.69–0.70, Stanley et al., 1996, 2001) and the correlations between the GAI and other anxiety
subscales (r ¼ 0.58–0.80; Stanley et al., 1996, 2001). measures show evidence of concurrent validity, with
Concurrent validity for the WS has been shown by correlations ranging from 0.58 to 0.86 (Byrne et al.,
significant correlations between the scale and measures 2010; Diefenbach et al., 2009; Pachana et al., 2007).
of anxiety (r ¼ 0.54–0.63, Stanley et al., 2001; Wisocki, However, relatively high correlations (r ¼ 0.65–0.79)
1988; Wisocki et al., 1986). However, the scale also between the GAI and depression measures provide
showed high correlations with measures of depression limited evidence of discriminant validity (Andrew
(r ¼ 0.500.78 Wisocki, 1988; Wisocki et al., 1986). As et al., 2006; Diefenbach et al., 2009; Paukert, 2009).
noted previously, examining a CBT intervention for There were no studies identified in our review in which
GAD, Stanley et al. (2003) found that improvements the GAI was used to assess the effects of mental health
were noted for the CBT group on the WS and other treatments for older adults. Taken together, although
anxiety measures. much more research evidence is needed, initial evidence
The original scale has been revised and expanded suggests that the GAI is likely to be a useful tool for
to include 88 items in six dimensions (finances, health, detecting anxiety in older adults.
social conditions, personal concerns, family concerns,
and world issues). Its psychometric properties are
currently under investigation, with initial evidence
indicating that they are similar to the original scale Discussion
(Hunt et al., 2003; Watari & Brodbeck, 2000). Overall The aims of this review were to evaluate the mental
then, although more research is needed on discriminant health research literature on older adults in order to (1)
validity and treatment sensitivity, initial evidence identify the most commonly used anxiety measures and
suggests that the WS may be useful in assessing (2) determine how appropriate the measures are for
worry in older adults. clinical or research use with older adults by examining
their psychometric properties. Results of our literature
search indicate that, with more than 90 measures used
Geriatric Anxiety Inventory to evaluate anxiety in older adults, no clear consensus
The GAI (Pachana et al., 2007) is a 20-item self-report exists among researchers on which anxiety measure is
questionnaire designed to measure anxiety symptoms more appropriate to use when evaluating a
 Aging & Mental Health 11

geriatric population. Out of the most commonly used Test–retest reliability of the two measures has been
measures, only one (GMSE) was developed specifically shown to be adequate. Convergent validity for these
for older adults. Therefore, it seems that, with the lack measures has been suggested by moderate correlations
of anxiety measures created and validated for older with other self-report measures of anxiety. The PSWQ
adults, researchers and clinicians often need to rely on showed good discriminant validity with low correla-
anxiety measures created for younger adults when tions with self-report measures of depression but the
assessing older populations. However, because many BAI show only limited discriminant validity. Although
differences exist between younger and older adults, it is both measures were initially developed for younger
unlikely that a single measure can adequately assess populations, subsequent evidence suggests that they
anxiety across the entire adult life span. For example, might be good choices when selecting an anxiety
compared to younger adults, older adults report more measure for a geriatric population. The GMSE, a
somatic symptoms, which means that anxiety measures semi-structured interview created for older adults that
that were not specifically developed for older adults includes anxiety clusters, is the only measure that
may not provide an accurate assessment of anxiety provides norms and cut-off scores validated for older
(Fuentes & Cox, 1997; Wetherell & Gatz, 2005). As a adults. That, combined with its appropriate psycho-
result, it is extremely important for clinicians and metric properties, suggests that it is a good option for
researchers to have access to valid and reliable mea- assessing anxiety in older adults. However, the GMSE
sures with established age-appropriate norms. has not been used in research published in the past
Most of the assessment measures we reviewed lack decade, which may suggest that more recent measures
sufficient evidence for their psychometric soundness may be more appropriate to measure anxiety in older
when used with older adults. Several critical consider- adults.
ations limit their use with geriatric populations. First, Indeed there is a growing number of anxiety
although each reviewed measure showed adequate instruments designed for use with older adults. Both
internal consistency (  0.70), the existing data often the WS and the GAI were specifically developed for
come from a single published study and mostly from older adults and showed sufficient psychometric evi-
samples of older psychiatric outpatients. Both replica- dence to warrant their use in assessing anxiety in older
tion and extension of previous findings are necessary adults. In many instances, these instruments might be
in order to determine if the measures are good general preferred over the more commonly used measures that
screening measures for anxiety in older adults or were not developed for the geriatric population. Both
whether they may have relevance only for a specific measures demonstrated high internal consistency in
subgroup of older adults. Second, only four measures older psychiatric outpatients as well as in community-
(BAI, GMSE, PSWQ, and STAI) showed evidence of dwelling older adults. Test–retest reliability of the two
adequate test–retest reliability in older adult samples. measures has been shown to be good. Convergent
The lack of test–retest reliability needs to be taken into validity for these measures has been provided by
consideration if the measure is to be used numerous significant correlations with other self-report measures
times, such as when evaluating treatment effects. The of anxiety. But, because of the extent to which these
third limitation resides in the lack of evidence for both measures have been found to correlate with measures
discriminant and concurrent validities in most mea- of depression, further research focusing on discrimi-
sures. Three measures showed adequate concurrent nant validity is certainly required.
validity with older adults (GMSE, BAI, and PSWQ),
four showed limited concurrent validity (GADS,
HARS, HADS, and STAI), whereas no information
was found for the other three measures. Every Coexistence of somatic symptoms
measure, except the PSWQ, either showed limited or In this review, six of the most commonly used
low discriminant validity. Also, very few instruments measures (BAI, BSI, GHQ, GADS, HARS and SCL-
have adequate normative data for older adults, which 90-R) were weighted heavily with somatic symptoms,
severely limit their clinical value. Only the GMSE has which makes it difficult to distinguish between anxiety
been normed for older adults and provides cut-off symptoms and symptoms of other health problems (or
scores validated for this population. Finally, few even normal aging) among the geriatric population.
measures had evidence of treatment sensitivity, sug- This can be problematic as the experience of anxiety
gesting the urgent need for studies assessing the varies greatly in younger and older adults. Not only do
sensitivity of commonly used anxiety instruments to older adults experience more somatic symptoms when
the effects of mental health treatments for older adults. anxious but they are also likely to have coexisting
Taken together, three of the most commonly used physical conditions that may produce anxiety-like
measures showed sufficient psychometric evidence to symptoms. It is therefore critical for the measures to
warrant their use in assessing anxiety in older adults. be able to distinguish between anxiety-like symptoms
The BAI, a measure of general anxiety, and the PSWQ, caused by a medical condition from symptoms caused
a measure of worry, have both demonstrated high by an anxiety disorder. By using a measure that
internal consistency in older psychiatric outpatients includes many somatic symptoms, a high proportion of
as well as in community-dwelling older adults. non-anxious older adults experiencing symptoms of a
12 Z. Therrien and J. Hunsley

medical condition may fall within the range used to this might have resulted in an overly narrow review of
identify clinical anxiety in a younger population. This the published literature on anxiety measures used with
must be taken into consideration when using these older adults. However, our search strategy did ensure
measures with older adults, particularly with older that we considered the vast majority of journals that
adults in medical settings or with a medical condition. publish research on the health and/or mental health of
older adults.

Coexistence of depression
Another issue that must be taken into consideration Conclusions
when evaluating the presence of anxiety in older adults The present systematic review shows that the anxiety
is the frequent coexistence of anxiety and depression in measures most commonly used with older adults are
later life. Studies suggest that as much as 38–46% of mostly measures developed for a younger population.
older adults meeting criteria for a mood disorder also Although there is empirical support for the use of some
meet criteria for an anxiety disorder (Beekman et al., of these measures, the majority of measures lack
1998; Flint, 1994; Lenze et al., 2000). As these sufficient evidence of their psychometric soundness
comorbid conditions can increase the complexity of when used with older adults. The STAI was found to
anxiety assessment and diagnosis, it is important that be the most commonly used measure in the reviewed
the measures differentiate anxious and depressive articles. However, an examination of its psychometric
symptoms in older adults. In this review, none of the properties yielded mixed results, suggesting that it does
most commonly used measures showed adequate not yet show sufficient supporting psychometric evi-
evidence of discriminant validity with respect to dence and should, therefore, be used with caution when
mood disorders, which is likely to lead to a high assessing older adults. The HADS was also frequently
misclassification rate among older adults. Two mea- used, but the lack of psychometric evidence for this
sures (GADS and HADS) include both a depression instrument suggests that it is not be a good option
and an anxiety subscale but research indicates that, for when evaluating older adults until more research
both measures, the anxiety and depression subscales examines its validity and reliability. The GMSE was
are highly correlated and therefore might not distin- created for older adults and shows preliminary evi-
guish between the disorders (Davies et al., 1993; Flint dence of psychometric soundness. However, this mea-
& Rifat, 2002; Koloski et al., 2008). Consequently, sure has not been used in recent studies and, therefore,
researchers and clinicians should be careful when using might not be the most appropriate choice when
these measures until more solid evidence of discrimi- assessing anxiety in older adults. Both the PSWQ and
nate validity is obtained. Obviously, information the BAI have shown good psychometric properties,
beyond what is available from these instruments must suggesting that they may be useful tools to measure the
be considered when making any diagnostic presence of anxiety in older adults. However, consid-
formulations. ering the limited research on the psychometric prop-
erties of the measures, clinicians and researchers must
be cautious and carefully consider the strengths and
Limitations weaknesses of each measure before deciding which one
Findings from this review must be interpreted within to use for a specific purpose. Although the evidence
the limitations of systematic reviews in general. A key to date is somewhat limited, the measures specifically
issue in selecting articles for review was deciding which developed for older adults, such as the WS and the
studies to include and which to exclude. In particular, GAI, should be seriously considered by clinicians and
we required that studies have samples in which all researchers when assessing anxiety in a geriatric
participants were at least 65 years of age. We excluded population.
from consideration, therefore, a number of articles in A major shortcoming evident in the reviewed
which some research sample included participants in measures is the inclusion of somatic symptoms of
their late 50s or early 60s. Accordingly, this limited the anxiety that often overlap with the symptoms of
number of studies we examined in detail. Setting the normal aging, comorbid conditions, and medication
inclusion criteria to allow samples in which all partic- side effects. The use of measures heavily weighted for
ipants were at least 55 years of age would certainly somatic symptoms should be avoided in medical
have increased the number of studies available for our settings or in samples presenting with a medical
review, but at the cost of including too broad a range condition. Caution should also be used when assessing
of ages for our intended focus on older adults. The older adults who present possible depressive symp-
search was also limited to articles published in either toms, as many of the anxiety measures are highly
French or English as those were the languages under- correlated with measures of depression. Additionally,
stood by the authors. Furthermore, we searched for most measures do not present age appropriate norms
studies in the most important databases for psychology or clinically relevant cut-off scores, which greatly limit
(PsycINFO) and medicine (PubMed), but other their use with older adults. For the evaluation of
databases were not considered. It is conceivable that anxiety in older adults to be more evidence-based,
 Aging & Mental Health 13

there is a pressing need for the validation of measures Byrne, G., Pachana, N., Goncalves, D., Arnold, E., King, R.,
that were created for young adults and that are used & Khoo, S. (2010). Psychometric properties and health
with older adults. Research is needed on the develop- correlates of the Geriatric Anxiety Inventory in Australian
ment and validation of anxiety measures created community-residing older women. Aging and Mental
specifically for older adults that evaluate a wide Health, 14, 247–254.
Cairney, J., Corna, L.M., Veldhuizen, S., Herrmann, N., &
range of anxiety symptoms and disorders.
Streiner, D.L. (2008). Comorbid depression and anxiety in
later life: Patterns of association, subjective well-being, and
impairment. The American Journal of Geriatric Psychiatry,
References 16, 201–208.
Chambless, D.L., & Hollon, S.D. (1998). Defining empiri-
Alwahhabi, F. (2003). Anxiety symptoms and generalized cally supported therapies. Journal of Consulting and
anxiety disorder in the elderly: A review. Harvard Review Clinical Psychology, 66, 7–18.
of Psychiatry, 11, 180–193. Cheung, Y.B. (2002). A confirmatory factor analysis of the
American Psychological Association Presidential Task Force 12-item General Health Questionnaire among older
on Evidence Based Practice (2006). Evidence-based prac- people. International Journal of Geriatric Psychiatry, 17,
tice in psychology. American Psychologist, 61, 271–285. 739–744.
Ames, D., Flynn, E., Tuckwell, V., & Harrigan, S. (1994). Christensen, H., Jorm, A.F., Mackinnon, A.J., Korten, A.E.,
Diagnosis of psychiatric disorder in elderly general and Jacomb, P.A., Hendersen, A.S., & Rodgers, B. (1999). Age
geriatric hospital patients: AGECAT and DSM-III-R differences in depression and anxiety symptoms: A
compared. International Journal of Geriatric Psychiatry, structural equation modelling analysis of data from a
9, 627–633. general population sample. Psychological Medicine, 29,
Andreescu, C., Belnap, B.H., Rollman, B.L., Houck, P., 325–339.
Ciliberti, C., Mazumdor, S., . . . , Lenze, E.J. (2008). Clarke, D., & Clarkson, J. (2009). A preliminary investiga-
Generalized Anxiety Disorder Severity Scale validation in tion into motivational factors associated with older adults
older adults. The American Journal of Geriatric Psychiatry, problem gambling. International Journal of Mental Health
16, 813–818. and Gambling, 7, 12–28.
Andrew, D.H., & Dulin, P.L. (2007). The relationship Cohen, L.L., La Greca, A.M., Blount, R.L., Kazak, A.E.,
between self-reported health and mental health problems Holmbeck, G.N., & Lemanek, K. (2008). Introduction
among older adults in New Zealand: Experiential avoid- to special issue: Evidence-based assessment in
ance as a moderator. Aging and Mental Health, 11, pediatric psychology. Journal of Pediatric Psychology, 33,
576–603. 911–915.
Ayers, C.R., Sorrell, J.T., Thorp, S.R., & Wetherell, J.L. Copeland, J.R., Beekman, A.T., Dewey, M.E., Hooijer, C.,
(2007). Evidence-based psychological treatments for late- Jordan, A., Lawlor, B.A., . . . , Wilson, K.C. (1999).
life anxiety. Psychology and Aging, 22, 8–17. Depression in Europe: Geographical distribution among
Beck, A.T., Epstein, N., Brown, G., & Steer, R.A. (1988). An older people. British Journal of Psychiatry, 174, 312–321.
inventory for measuring clinical anxiety: Psychometric Copeland, J.R., Dewey, M.E., & Griffiths-Jones, H.M.
properties. Journal of Consulting and Clinical Psychology, (1986). A computerized psychiatric diagnostic system and
56, 893–897. case nomenclature for elderly subjects: GMS and
Beck, J.G., Stanley, M.A., & Zebb, B.J. (1995). Psychometric AGECAT. Psychological Medicine, 16, 89–99.
properties of the PSWQ in older adults. Journal of Clinical Copeland, J.R.M., Kelleher, M.J., Duckwort, G., & Smith,
Geropsychology, 1, 33–42. A. (1976). Reliability of psychiatric assessment in older
Beck, J.G., Stanley, M.A., & Zebb, B.J. (1996).
patients. International Journal of Aging and Human
Characteristics of generalized anxiety disorder in older
Development, 7, 313–322.
adults: A descriptive study. Behaviour Research and
Copeland, J.R., Kelleher, M.J., Kellett, J.M., Gourlay, A.J.,
Therapy, 34, 225–234.
Cowan, D.W., Barron, G., & de Gruchy, J. (1975). Cross-
Beck, J.G., Stanley, M.A., & Zebb, B.J. (1999). Effectiveness
national study of diagnosis of the mental disorders: A
of the Hamilton Anxiety Rating Scale with older gener-
comparison of the diagnoses of elderly psychiatric patients
alized anxiety disorder patients. Journal of Clinical
admitted to mental hospitals serving Queen’s county, New
Gerpsychology, 5, 281–290.
York, and former Borough of Camberwell, London.
Beck, A.T., & Steer, R.A. (1990). Manual for the Beck
British Journal of Psychiatry, 126, 11–20.
anxiety inventory. San Antonio, TX: Psychological
Copeland, J.R.M., Kelleher, M.J., Kellett, J.M., Gourlay,
Corporation.
A.J., Gurland, B.J., Fleiss, J.L., & Sharpe, L. (1976). A
Beekman, A.T., Bremmer, M.A., Deeg, D.J., van Balkom,
semi-structured clinical interview for the assessment of
A.J., Smit, J.H., DeBeurs, E., . . . , van Tilburg, W. (1998).
diagnosis and mental state in the elderly: The Geriatric
Anxiety disorders in later life: A report from the long-
Mental State Schedule I: Development and reliability.
itudinal aging study Amsterdam. International Journal of
Geriatric Psychiatry, 13, 717–726. Psychological Medicine, 6, 439–449.
Boey, K.W., & Chiu, H.F. (1998). Assessing psychological Copeland, J.R., Prince, M., Wilson, K.C., Dewey, M.E.,
well-being of the old-old: A comparative study of GDS-15 Payne, J., & Gurland, B. (2002). The geriatric mental state
and GHQ-12. Clinical Gerontologist, 19, 65–75. examination in the 21st century. International Journal of
Bryant, C., Jackson, H., & Ames, D. (2009). Depression and Geriatric Psychiatry, 17, 729–732.
anxiety in medically unwell older adults: Prevalence and Costa, E., Barreto, S., Uchoa, E., Firmo, J., Lima-Costa, M.,
short-term course. International Psychogeriatrics, 21, & Prince, M. (2006). Is the GDS-30 better that the GHQ-
754–763. 12 for screening depression in elderly people in the
14 Z. Therrien and J. Hunsley

community? The Bambui Health Aging Study (BHAS). Flint, A.J., & Rifat, S.L. (2002). Factor structure of the
International Psychogeriatrics, 18, 493–503. Hospital Anxiety and Depression Scale in older patients
Crittendon, J., & Hopko, D.R. (2006). Assessing worry in with major depression. International Journal of Geriatric
older and younger adults: Psychometric properties of an Psychiatry, 17, 117–123.
abbreviated Penn State Worry Questionnaire. Journal of Forman, S., Fagley, N.S., Steiner, D.D., & Schneider, K.
Anxiety Disorders, 20, 1036–1054. (2009). Teaching evidence-based interventions: Perceptions
Cully, J.A., Graham, D.P., Stanley, M.A., Ferguson, C.J., of influences on use in professional practice in school
Sharafkhaneh, A., Souchek, J., & Kunik, E. (2006). psychology. Training and Education in Professional
Quality of life in patients with chronic obstructive Psychology, 3, 226–232.
pulmonary disease and comorbid anxiety or depression. Fuentes, K., & Cox, B.J. (1997). Prevalence of anxiety
Psychosomatics, 47, 312–319. disorders in elderly adults: A critical analysis. Journal of
Dale, B., Saevareid, H., & Soderhamn, O. (2009). Testing and Behavior Therapy and Experimental Psychiatry, 28,
using Goldberg’s General Health Questionnaire: Mental 269–279.
health in relation to home nursing, home help, and family Goldberg, D. (1978). Manual of the General Health
care among older, care-dependent individuals. International Questionnaire. Windsor: NFER.
Journal of Mental Health Nursing, 18, 133–143. Goldberg, D., Bridges, K., Duncan-Jones, P., & Grayson, D.
Davies, K.N., Burn, W.K., Mckenzie, F.R., Brotherwell, (1988). Detecting anxiety and depression in general
J.A., & Wattis, J.P. (1993). Evaluation of the Hospital medical settings. British Medical Journal, 297, 897–899.
Anxiety and Depression Scale as a screening instrument in Goldberg, D.P., & Hillier, V.F. (1979). A scaled version of
geriatric medical inpatients. International Journal of the General Health Questionnaire. Psychological Medicine,
Geriatric Psychiatry, 8, 165–169. 9, 139–145.
Dawson, P., Kline, K., Wiancko, D.C., & Wells, D. (1986). Gurian, B.S., & Miner, J.H. (1991). Clinical presentation of
Preventing excess disability in patients with Alzheimer’s anxiety in the elderly. In C. Salzman & B. Lebowitz (Eds.),
disease. Geriatric Nursing, 7, 298–301. Anxiety in the elderly: Treatment and research (pp. 31–44).
de Beurs, E., Beekman, A.T., van Balkom, A.J., Deeg, D.J., New York, NY: Springer Publishing Co.
van Dyck, R., & van Tilburg, W. (1999). Consequences of Hale, D.W., Cochran, C.D., & Hedgepeth, B.E. (1984).
anxiety in older persons: Its effect on disability, well-being Norms for the elderly on the Brief Symptom Inventory.
and use of health services. Psychological Medicine, 29, Journal of Consulting and Clinical Psychology, 52, 321–322.
583–593. Haley, W.E. (1996). The medical context of psychotherapy
de Graaf, R., Bijl, R.V., Spijker, J., Beekman, A.T.F., & with the elderly. In S. Zarit & B. Knight (Eds.), A guide to
Vollebergh, W.A. (2003). Temporal sequencing of lifetime psychotherapy and aging: Effective clinical interventions in a
mood disorders in relation to comorbid anxiety and life-stage context (pp. 221–239). Washington, DC:
substance use disorders: Findings from the Netherlands American Psychological Association.
Mental Health Survey and Incidence Study. Social Hamilton, M. (1959). The assessment of anxiety states by
Psychiatry and Psychiatric Epidemiology, 38(1), 1–11. rating. British Journal of Medical Psychology, 32, 50–55.
Dennis, R.E., Boddington, S.J., & Funnell, N.J. (2007). Self- Haworth, J.E., Moniz-Cook, E., Clark, A.L., Wang, M., &
report measures of anxiety: Are they suitable for older Cleland, J.G. (2007). An evaluation of two self-report
adults? Aging and Mental Health, 11, 668–677. screening measures for mood in and out-patient chronic
Derogatis, L.R. (1994). Administration, scoring and proce- heart failure population. International Journal of Geriatric
dures manual. Minneapolis, MN: National Computer Psychiatry, 22, 1147–1153.
System. Henderson, A.S., Duncan-Jones, P., & Finlay-Jones, R.A.
Derogatis, L.R., & Spencer, P.M. (1982). The Brief Symptom (1983). The reliability of the Geriatric Mental State
Inventory (BSI): Administration, and procedures manual-I. Examination. Community survey version. Acta
Baltimore, MD: Clinical Psychometric Research. Psychiatrica Scandinavica, 67, 281–289.
Diefenbach, G.J., Stanley, M.A., Beck, J.G., Novy, D.M., Himmelfarb, S., & Murrell, S.A. (1984). The prevalence and
Averill, P.M., & Swann, A.C. (2001). Examination of the correlates of anxiety disorders in older adults. Journal of
Hamilton scales in assessment of anxious older adults: A Psychology, 116, 159–167.
replication and extension. Journal of Psychopathology and Holmbeck, G.N., & Devine, K.A. (2009). Editorial: An
Behavioral Assessment, 23, 117–124. author’s checklist for measure development and validation
Dienfenbach, G., Tolin, D., Meunier, S., & Gilliam, C. manuscripts. Journal of Pediatric Psychology, 34, 691–696.
(2009). Assessment of anxiety in older home care Hopko, D.R., Stanley, M.A., Reas, D.L., Wetherell, J.L.,
recipients. The Gerontologist, 49, 141–153. Beck, J.G., Novy, D.M., & Averill, P.M. (2003). Assessing
Edelstein, B.A., Woodhead, E.L., Segal, D.L., Heisel, M.J., worry in older adults: Confirmatory factor analysis of the
Bower, E.H., Lowery, A.J., & Stoner, S.A. (2008). Older Penn State Worry Questionnaire and psychometric proper-
adult psychological assessment: Current instrument status ties of an abbreviated model. Psychological Assessment, 15,
and related considerations. Clinical Gerontologist, 31(3), 173–183.
1–35. Huber, P., Mulligan, R., Mackinnon, A., Nebuloni-French,
Fisher, B.M., Segal, D.L., & Coolidge, F.L. (2003). T., & Michel, J.P. (1999). Detecting anxiety and depression
Assessment of coping in cognitively impaired older in hospitalised elderly patients using a Brief Inventory.
adults: A preliminary study. Clinical Gerontologist, 26, European Psychiatry, 14, 11–16.
3–12. Hunsley, J., & Mash, E.J. (2007). Evidence-based assessment.
Flint, A.J. (1994). Epidemiology and comorbidity of anxiety Annual Review of Clinical Psychology, 3, 29–51.
disorders in the elderly. The American Journal of Hunsley, J., & Mash, E.J. (2008). A guide to assessment that
Psychiatry, 151, 640–649. work. New York, NY: Oxford University Press.
 Aging & Mental Health 15

Hunt, S., Wisocki, P., & Yanko, J. (2003). Worry and use of Mottram, P., Wilson, K., & Copeland, J. (2000). Validation
coping strategies among older and younger adults. Journal of the Hamilton Depression Rating Scale and
of Anxiety Disorders, 17, 547–560. Montgomery and Asberg Rating Scales in terms of
Institute of Medicine (2001). Crossing the quality chasm: A AGECAT depression cases. International Journal of
new health system for the 21st century. Washington, DC: Geriatric Psychiatry, 15, 1113–1119.
National Academy Press. Mowry, B.J., & Burvill, P.W. (1990). Screening the elderly in
Johnson, G., Burvill, P.W., Anderson, C.S., Jamrozik, K.K., the community for psychiatric disorder. Australian and
Stewart-Wynne, E.G., & Chakera, T.M. (1995). Screening New Zealand Journal of Psychiatry, 24, 203–206.
instruments for depression and anxiety following stroke: Naughton, C., Feely, J., & Bennett, K. (2007). A clustered
Experience in the Perth community stroke study. Acta randomized trial of the effects of feedback using academic
Psychiatrica Scandinavica, 91, 252–257. detailing compared to postal bulletin on prescribing of
Johnston, M., Pollard, B., & Hennessey, P. (2000). Construct preventative cardiovascular therapy. Family Practice, 24,
validation of the Hospital Anxiety and Depression Scale 475–480.
with clinical populations. Journal of Psychosomatics Nuevo, R., Mackintosh, M., Gatz, M., Montorio, I., &
Research, 48, 579–584. Wetherell, J.L. (2007). A test of the measurement
Kabacoff, R.L., Segal, D.L., Hersen, M., & Van Hasselt, invariance of a brief version of the Penn State Worry
V.B. (1997). Psychometric properties and diagnostic utility Questionnaire between American and Spanish older
of the Beck Anxiety Inventory and the State-trait Anxiety adults. International Psychogeriatrics, 19, 89–101.
Inventory with older adult psychiatric outpatients. Journal Pachana, N. (2008). Ageing and psychological disorders. In
of Anxiety Disorder, 11, 33–47. E. Rieger (Ed.), Abnormal psychology (pp. 421–457).
Kenn, C., Wood, H., Kucyj, M., Wattis, J., & Cunane, J. Sydney: McGraw Hill.
(1987). Validation of the Hospital Anxiety and Pachana, N.A., Byrne, G.J., Siddle, H., Koloski, N., Harley,
Depression Rating Scale (HADS) in an elderly psychiatric E., & Arnold, E. (2007). Development and validation of
population. International Journal of Geriatric Psychiatry, the Geriatric Anxiety Inventory. International
2, 189–193. Psychogeriatrics, 19, 103–114.
Kessler, R., Berglund, P., Demler, O., Jon, R., Papassotiropoulos, A., Heun, R., & Maier, W. (1997). Age
Merikangas, K., & Walters, E. (2005). Lifetime prevalence and cognitive impairment influence the performance of the
age-of-onset distributions of DSM-IV disorders in the General Health Questionnaire. Comprehensive Psychiatry,
National Comorbidity Survey replication: Erratum. 38, 335–340.
Archives of General Psychiatry, 62, 768. Paukert, A. (2009). The roles of social support, self-efficacy,
Kogan, J.N., Edelstein, B.A., & McKee, D.R. (2000). and optimism in physical health’s impact on depressive
Assessment of anxiety in older adults: Current status. and anxious symptoms among older adults. Dissertation
Journal of Anxiety Disorders, 14, 109–132. Abstracts International: Section B: The Sciences and
Koloski, N.A., Smith, N., Pachana, N.A., & Dobson, A. Engineering, 69, 5788.
(2008). Performance of the Goldberg Anxiety and Petkus, A.J., Gum, A.M., King-Kallimanis, B., & Wetherell,
Depression Scale in older women. Age and Ageing, 37, J.L. (2009). Trauma history is associated with psycholo-
464–467. gical distress and somatic symptoms in homebound older
Kvall, K., Ulstein, I., Nordhus, I., & Engedal, K. (2005). The adults. The American Journal of Geriatric Psychiatry, 17,
Spielberger State-Trait Anxiety Inventory (STAI): The 810–818.
state scale in detecting mental disorders in geriatric Petkus, A.J., Gum, A.M., Small, B., Malcarne, V.L., Stein,
patients. International Journal of Geriatric Psychiatry, 20, M.B., & Wetherell, J.L. (2010). Evaluation of the factor
629–634. structure and psychometric properties of the Brief
Lenze, E.J., Mulsant, B.H., Shear, M.K., Schulberg, H.C., Symptom Inventory-18 with homebound older adults.
Dew, M.A., Begley, A.E., & Reynolds, C.F. III. (2000). International Journal of Geriatric Psychiatry, 25, 578–587.
Comorbid anxiety disorders in depressed elderly patients. Regier, D.A., Boyd, J.H., Burke, J.D., Rae, D.S., Myers,
The American Journal of Psychiatry, 157, 722–728. J.K., Kramer, M. , . . . , Locke, B.Z. (1988). One-month
Lenze, E.J., Rollman, B.L., Shear, M.K., Dew, M.A., prevalence of mental disorders in the United States.
Pollock, B.G., Ciliberti, C., . . . , Reynolds, C.F. III. Archives of General Psychiatry, 45, 977–986.
(2009). Escitalopram for older adults with generalized Rozzini, L., Chilovi, B.V., Peli, M., Conti, M., Rozzini, R.,
anxiety disorder: A randomized controlled trial. Journal of Trabucchi, M., & Padovani, A. (2009). Anxiety Symptoms
the American Medical Association, 301, 295–303. in mild cognitive impairment. International Journal of
Malakouti, S.K., Fatollahi, P., Mirabzadeh, A., & Zandi, T. Geriatric Psychiatry, 24, 300–305.
(2007). Reliability, validity and factor structure of the Scheick, J.I. (1991). Anxiety in the elderly: Treatment and
GHQ-28 used among elderly Iranians. International research. New York, NY: Spring.
Psychogeriatrics, 19, 623–634. Schuurmans, J., Comijs, H., Emmelkamp, P.M., Weijnen,
Mendlowicz, M.V., & Stein, M.B. (2000). Quality of life in I.J., van den Hout, M., & Van Dyck, R. (2009). Long-term
individuals with anxiety disorders. American Journal of effectiveness and prediction of treatment outcome in
Psychiatry, 31, 1602–1607. cognitive behavioral therapy and sertraline for late-life
Meyer, T., Miller, M., Metzger, R., & Borkovec, T.D. (1990). anxiety disorders. International Psychogeriatrics, 21,
Development and validation of the Penn State Worry 1148–1159.
Scale. Behaviour Research and Therapy, 28, 487–495. Seignourel, P.J., Kunik, M.E., Snow, L., Wilson, N., &
Morin, C.M., Landreville, P., Colecchi, C., McDonald, K., Stanley, M. (2008). Anxiety in dementia: A critical review.
Stone, J., & Ling, W. (1999). The Beck Anxiety Inventory: Clinical Psychology Review, 28, 1071–1082.
Psychometric properties with older adults. Journal of Senior, A., Kunik, M., Rhoades, H., Novy, D., Wilson, N.,
Clinical Geropsychology, 5, 19–29. & Stanley, M. (2007). Utility of telephone assessments in
16 Z. Therrien and J. Hunsley

an older adult population. Psychology and Aging, 22, State Examination. International Journal of Geriatric
392–397. Psychiatry, 6, 647–680.
Seva, A., Sarasola, A., Merino, J.A., & Magallon, R. (1991). Watari, K.F., & Brodbeck, C. (2000). Culture, health, and
Validity of the scaled version of the General Health financial appraisals: Comparison of worry in older
Questionnaire in a geriatric Spanish population. The Japanese Americans and European Americans. Journal of
European Journal of Psychiatry, 5, 32–36. Clinical Geropsychology, 6, 25–39.
Skopp, N.A., Novy, D., Kunik, M., Daza, P., Adams, J.H., Wetherell, J.L., & Areàn, P.A. (1997). Psychometric
Senior, A., & Stanley, M. (2006). Investigation of cognitive evaluation of the Beck Anxiety Inventory with
behavior therapy. The American Journal of Geriatric older medical patients. Psychological Assessment, 9,
Psychiatry, 14, 292. 136–144.
Snaith, R.P., & Zigmond, A.S. (1994). The Hospital Anxiety Wetherell, J.L., Ayers, C.R., Nuevo, R., Stein, M.B.,
and Depression Scale manual. Windsor: NFER-Nelson. Ramsdell, J., & Patterson, T.L. (2010). Medical
Spielberger, C.D., Gorsuch, R.L., Lushene, R., Vagg, P., & conditions and depressive, anxiety and somatic symp-
Jacobs, G.A. (1983). Manual for the State-Trait Anxiety toms in older adults with and without generalized
Inventory STAI (Form Y): Self-evaluation questionnaire. anxiety disorder. Aging and Mental Health, 14,
Palo Alto, CA: Consulting Psychologists Press. 764–768.
Spinhoven, P.H., Ormel, J., Sloekers, P.P., Kempen, G.I., Wetherell, J.L., Birchler, G.D., Ramsdell, J., & Unutzer, J.
Speckens, A.E., & Van Hemert, A.M. (1997). A validation (2007). Screening for generalized anxiety disorder in
study of the Hospital Anxiety and Depression Scale geriatric primary care patients. International Journal of
(HADS) in different groups of Dutch subjects. Geriatric Psychiatry, 22, 115–123.
Psychological Medicine, 27, 363–370. Wetherell, J.L., & Gatz, M. (2005). The Beck Anxiety
Stanley, M.A., Beck, J.G., Novy, D.M., Averill, P.M., Inventory in older adults with generalized anxiety disorder.
Swann, A.C., Diefenbach, G.J., & Hopko, D.R. (2003). Journal of Psychopathology and Behavioral Assessment, 27,
Cognitive-behavioral treatment of late-life generalized 17–24.
anxiety disorder. Journal of Consulting and Clinical Wetherell, J.L., Gatz, M., & Craske, M.G. (2003). Treatment
Psyhchology, 71, 309–319. of generalized anxiety disorder in older adults. Journal of
Stanley, M.A., Beck, J.G., & Zebb, B.J. (1996). Psychometric Consulting and Clinical Psychology, 71, 31–40.
properties of four anxiety measures in older adults. Wisocki, P.A. (1988). Worry as a phenomenon relevant to
Behaviour and Research Therapy, 34, 827–838. the elderly. Behavior Therapy, 19, 369–379.
Stanley, M.A., Novy, D.M., Bourland, S.L., Beck, J.G., & Wisocki, P.A. (1994). The experience of worry among the
Averill, P.M. (2001). Assessing older adults with general- elderly. In G.C.L. Davey & F. Tallis (Eds.), Worrying:
ized anxiety: A replication and extension. Behaviour Perspectives on theory, assessment and treatement
Research and Therapy, 39, 221–235. (pp. 246–261). New York, NY: Wiley.
Stanley, M.A., Wilson, N.L., Novy, D.M., Rhoades, H.M., Wisocki, P.A., Handen, B., & Morse, C.K. (1986). The
Wagener, P.D., Greisinger, A.J., . . . , Kunik, M.E. (2009). Worry Scale as a measure of anxiety among homebound
Cognitive-behavior therapy for generalized anxiety disorder and community active elderly. The Behavior Therapist, 9,
among older adults in primary care: A randomized trial. 91–95.
Journal of the American Medical Association, 301, Wolitsky-Taylor, K., Castriotta, N., Lenze, E.J., Stanley,
1460–1467. M.A., & Craske, M.G. (2010). Anxiety disorders in older
Thygesen, E., Saevareid, H.J., Lindstrom, T.C., Nygaard, adults: A comprehensive review. Depression and Anxiety,
H.A., & Engedal, K. (2009). Predicting needs for the 27, 190–211.
nursing home admission – does sense of coherence delay Yu, D.S., Lee, D.T., Woo, J., & Hui, E. (2007). Non-
nursing home admission in care dependant older people? A pharmacological interventions in older people with heart
longitudinal study. International Journal of Older People failure: Effects of exercise training and relaxation therapy.
Nursing, 4, 12–21. Gerontology, 53, 74–81.
Turina, C., Perdona, G., Bianchi, L., Cordioli, L., Burti, L., Zigmond, A.S., & Snaih, R.P. (1983). The Hospital
Micciolo, R., & Copeland, J.R. (1991). Interobserver Anxiety and Depression Scale. Acta Psychiatrica
reliability of the Italian version of the Geriatric Mental Scandinavica, 67, 361–370.