Anxiety disorders in older adults: Looking to DSM5 and beyond…

Anxiety in late life was for many years the ‘Cinderella’ of psychiatric disorders, often overshadowed by the focus on depression and dementia, and receiving little attention in research and clinical domains. As highlighted by an editorial published in this journal several years ago [1], the scientific study of anxiety in older age has a relatively brief history. Recently, however, there has been increased recognition of the prevalence and clinical consequences of anxiety disorders in older adults and lively debate about their nature and most appropriate diagnostic criteria [2, 3, 4]. The current evidence reveals that anxiety in older adults is more common than depression in community samples [5], often preceding depressive disorders [6]; moreover, co-morbid anxiety and depression has a poorer outcome than either condition alone [7]. Anxiety disorders are even more prevalent in clinical settings [8], and can have serious consequences for recovery from illness [9] and quality of life [10], as well as substantially increasing disability levels [11]. Therefore, anxiety disorders in older adults should be regarded as conditions of great public health importance.

Anxiety disorders tend to be under-recognised in older adults, and there are significant practical and methodological problems in obtaining accurate epidemiological data for this population [4]. The revision of the Diagnostic and Statistical Manual of Mental Disorders (DSM) [12] currently underway provides an impetus to review the criteria for anxiety disorders in older people and consider how they might be improved in DSM5. As part of this process, the Advisory Committee to the DSM5 Lifespan Disorders Workgroup convened a group of researchers to propose changes to the DSM that could promote better detection of anxiety disorders in older people. This commentary summarizes recommendations made by that group [2], and raises other issues that extend beyond the changes being considered for DSM5.

The workgroup started from the premise that anxiety is increasingly heterogeneous in later life, and may be qualitatively different from anxiety as experienced by younger adults. For example, older adults show decreased autonomic nervous system activity [13], which may either reduce the prevalence of panic disorder in older adults or make it harder to detect in this age group [14]. Furthermore, detection of anxiety by health care professionals may be made more difficult for a variety of reasons. Older people are more sensitive to the stigma associated with disclosure of psychiatric symptoms [15] and are less likely than younger adults to report symptoms of anxiety unless carefully prompted to do so by well-informed clinicians [16]. Moreover, impairment in work or social relationships (a key criterion for the diagnosis of all anxiety disorders) may not be readily apparent if the person is retired or socially isolated, and thus more able to avoid anxiety-provoking situations [17]. Although some older adults learn positive ways of coping with their symptoms, others may learn maladaptive ways to cope with their anxiety and reduce its impact on their functioning or distress, or have lived with their symptoms for so long that they are no longer recognised as a problem [18]. In sum, diagnosing anxiety disorders in older adults is difficult, in both clinical settings and in research, and this has major public health implications; for example, research on fear of falling in elderly has been hampered by the lack of diagnostic clarity, and, consequently, federally-funded treatment studies of fear of falling in older adults are rare. This problem must be addressed, beginning with DSM5.

Our committee suggested a number of ways in which the DSM5 might help facilitate the recognition of anxiety in older patients [2]. First, we addressed two issues central to the conceptualization and diagnosis of anxiety disorders, namely the assessment of avoidance and excessiveness. We recommended that DSM5 provide guidelines for the thorough assessment of avoidance in older adults to accompany the diagnostic criteria. These include giving age-relevant examples of avoidance, and emphasising the need to actively explore the functional consequences of anxiety, which may be less obvious given role changes in later life (e.g., retirement). For example, the current DSM criteria for posttraumatic stress disorder include the person not expecting to have a career, children, marriage, or a normal life span. Older adults are likely to have already experienced these events, so further clarification about that symptom is needed. Age-related stereotypes may increase the likelihood that avoidance of previously enjoyed activities is seen as an understandable part of “getting older”, rather than the result of anxiety. For example, clinicians may too quickly accept a patient’s decision to stay home rather than go to a family gathering as normative for old age, rather than probe the possibility that this behaviour reflects anxious avoidance. With respect to “excessiveness,” we suggest that rather than inquiring about the patient’s self-perception of the excessive nature of worry, it may be more useful to ask more objective questions such as how much time is spent worrying, whether the older person thinks that they worry more than other older adults, and whether other people think of him/her as a worrier. Examples of these prompts should be provided in the text accompanying the diagnostic criteria [2].

A third recommendation concerned the provision of guidelines to assess co-morbid anxiety and depression. Anxiety and depression are highly comorbid [6, 19] and the diagnosis of Mixed Anxiety/Depression may be particularly relevant for older adults, whose symptoms may not meet strict diagnostic criteria for either disorder [4] or may demonstrate a fluctuating symptom pattern [20]. Mixed anxiety and depression is a syndrome with poorer prognosis than anxiety or depression alone [7], but little is known about the best treatment approaches [16]. Thus, the proposal in DSM5 for Mixed Anxiety-/Depression diagnosis would be welcomed both for its potential clinical utility and its relevance for generating much-needed research in older adults.

Diagnosing anxiety in the context of medical illness is another significant challenge in older adults, who frequently experience physical illnesses and may be taking multiple medications; yet, detecting clinically significant anxiety in this context is important, given the deleterious consequences of mental disorders on the outcome of medical illnesses, and the potential benefits to the patient of addressing both physical and mental health concerns. Currently, the DSM excludes cases in which an anxiety disorder is “due to the direct physiological effects of a general medical condition”, a phrase that is understandable at the theoretical level, but almost impossible to interpret in clinical practice. The presence of a medical illness does not preclude the existence of co-morbid anxiety that can exacerbate functional decline and increase the use of health services [21]. The medical and anxiety symptoms can be hard to distinguish, however, especially in situations where medical conditions or medications mimic anxiety symptoms; examples include diabetes and thyroid disease, and the use of corticosteroids. Moreover, even if certain somatic symptoms are directly attributable to a medical condition, pathological behaviours in response to those symptoms may not be (e.g., panic attacks due to chronic obstructive pulmonary disease leading to agoraphobic avoidance of situations that physiologically do not cause panic, such as wearing certain clothing or spending time in smaller rooms within the home). Our committee was concerned that a lack of clarity regarding when to use the “due to a GMC” exclusion might exclude cases from diagnosis and treatment, and under-estimate the prevalence of the anxiety disorders in older adults in epidemiological studies. Therefore, we made suggestions that the text contain guidance for distinguishing symptoms of anxiety from physiological manifestations of illnesses or their treatment. Useful lines of enquiry may include whether the anxiety pre-dates or follows the onset of the medical condition, whether there are other life events that could account for increased anxiety, and whether recently started medications are known to have side-effects that mimic anxiety symptoms.

Two issues with regard to anxiety are likely to be of particular relevance to older adults: the comorbidity of anxiety with dementia, and the existence of specific fears of older people, most notably fear of falling. With regard to the former, the DSM currently gives no guidance for assessing anxiety in the presence of dementia in older adults, and research in this area is sparse, possibly because of the lack of valid criteria for diagnosing anxiety in people with dementia [22]. The course of anxiety in dementia appears to be complex, with a likely bidirectional relationship between anxiety and cognition [23], and some studies suggest that anxiety is more common in people with vascular dementia than those with an Alzheimer’s pathology [24]. Our committee recommended that the DSM include clinically relevant guidelines for assessing anxiety in dementia, for example, by suggesting careful questioning of the chronology of the cognitive and anxiety symptoms, as well as seeking corroboration from an informant, even at early stages of decline. Eliciting the individual’s level of understanding and insight into their cognitive impairment may also be informative, as it is likely that anxious worry decreases as insight declines, to be replaced with behavioural manifestations of anxiety, such as agitation. [25].

Fear of falling is said to be the most commonly reported fear of older people, with point prevalence estimates ranging from 29% – 54% [26,27]. It shares many features with phobic disorder: physiological arousal when confronted by the feared situation, and marked avoidance of such situations. In this syndrome, however, the issue of perceived excessiveness is of particular concern. In a study by Gagnon et al. [28], for example, among a sample of 48 people with fear of falling, most of whom manifested moderate avoidance, only one deemed their anxiety to be excessive. Thus, the remaining 47 failed to meet criteria for a phobic disorder. This is a good example of an age-specific fear that “falls through the cracks” of the current DSM classification system.

Some of the suggestions made by our group are likely to be accommodated within the DSM revision process. There are already indicators that the “excessiveness” criterion will be replaced by the more useful term “out of proportion to the actual danger posed” for diagnoses such as specific phobia and social anxiety, but will be retained in the diagnostic criteria for GAD, and this will benefit older adults. It is disappointing to note that the proposed revisions released so far [29], while retaining a number of points of guidance for identifying symptoms in children, do not yet incorporate analogous guidance with respect to older people. For example, Criterion C for Social Anxiety Disorder states that in children the fear or anxiety may be expressed in crying, tantrums, freezing, clinging, shrinking or refusal to speak. No examples are given for identifying symptoms in older adults, or for different presentations of anxiety in the context of cognitive decline and dementia. A further shortcoming is that although the impairment criterion has been harmonised across the diagnostic categories, it remains very general, with reference only to social or occupational functioning.

The currently proposed addition of a Mixed Anxiety-Depression diagnostic category [29] is to be welcomed. The proposed criteria for this disorder are the presence of three or four symptoms of depression including depressed mood or anhedonia, accompanied by “anxious distress.” The latter is defined by having two or more of the following symptoms: irrational worry, preoccupation with unpleasant worries, having trouble relaxing, motor tension, and fear that something awful may happen. As yet, it is unclear whether the diagnostic criteria proposed will validly capture a mixed diagnostic picture for older adults. One concern is that older adults do not always endorse depressed mood, and some authors suggest that depression without sadness is characteristic for older adults [30]. Although motor tension is listed as a symptom of anxious distress, other somatic concerns typically endorsed by older people, such as pain and sleep disturbance, do not appear on this list. Once again, the accompanying text could provide guidance on this.

Therefore, the most promising opportunity to improve the detection of anxiety disorders in older adults will be through the revision of the accompanying text in the DSM. Here, there is enormous scope to provide clinically relevant material that will assist clinicians. For example, in relation to impairment and avoidance, older adults would benefit from examples that show some sensitivity to their social worlds, where caring duties or club participation may be more pertinent than job or school participation. The text could also include age-specific examples of avoidance (e.g., avoiding purchasing needed items for fear of “going broke” or avoiding asking for help for fear of being a burden), and how to assess whether an anxiety is “out of proportion” or “excessive.” According to the DSM5 website, there are no current plans to revise the criteria for Anxiety Disorder Associated with a Known General Medical Condition, other than consideration being given to the deletion of Criterion C (which excludes the diagnosis if the symptoms are better accounted for by another mental disorder, such as Adjustment Disorder). Here again, explication of the “direct physiological effects” criterion will need clarification in the accompanying text.

In summary, it appears that the proposed revisions to the DSM5 criteria for the anxiety disorders will be an improvement, but they may be insufficient for enhancing the detection and delineation of clinically significant anxiety syndromes in older adults. We have proposed additional revisions, primarily to accompanying text, to enhance detection of late life anxiety. We conclude by looking beyond DSM5 and offer three recommendations for promoting a better understanding of anxiety in later life.

First, we need to maintain a clinical and research focus on the anxiety disorders across the lifespan, so that knowledge in this domain keeps pace with the developments in depression and dementia. To this end, Field Trials and other studies of anxiety disorders should include adequate representation of older participants recruited from a wide range of settings so as to capture the variability in presentations of anxiety across the spectrum of community, medical, and residential environments. This would ensure that greater attention be paid to issues that are unique to older adults (such as anxiety in the context of dementia and fear of falling). Second, an eventual move away from a largely categorical diagnostic system towards a more dimensional approach is likely to benefit older adults. This can be illustrated by reference to Generalized Anxiety Disorder (GAD) – possibly the most prevalent anxiety disorder among older people [31]. GAD is a controversial diagnosis and some authors point out the high comorbidity with depression, given that up to 40% of people who are initially diagnosed with GAD subsequently develop depression or mixed anxiety-depression [20]. A more dimensional approach may benefit older adults by reducing the reliance on meeting syndrome criteria, thus also helping to address the problem that older adults may experience psychiatric symptoms at a level that is clinically significant [32] while not meeting current DSM diagnostic criteria. [4].

Third, attitudinal change on the part of both clinicians and the public is necessary. Clinicians and researchers can challenge stereotypical assumptions about older people, including that avoidance of activities is normative. Moreover, anxiety and its manifestations need to be viewed within the broader canvas of multiple influences on a happy and healthy old age. The promotion of physical health to prevent diseases such as cardiac, neurological, or pulmonary conditions that may cause or exacerbate anxiety is already a prominent goal of public health policy. Mental health clinicians also have an important part to play in promoting successful rehabilitation from medical conditions, so that their potentially disabling consequences can be minimised. Thus, significant improvements in mental health will likely result from making improvements to our diagnostic criteria, promoting physical health, and encouraging more positive attitudes towards ageing by challenging negative stereotypes.