Undertreatment of late-life anxiety

Unfortunately, the vast majority of older people with anxiety disorders are not receiving appropriate treatment. A Dutch population-based study showed that less than 10% of people aged 55 years or older with anxiety disorders received psychological treatment or antidepressant drug therapy. ^{Reference De Beurs, Beekman, van Balkom, Deeg, van Dyck and van Tilburg3} A more recent population-based study in Canada showed that only 20.6% of older persons with anxiety disorders had consulted any health service for their anxiety problem in the past year. ^{Reference Grenier, Préville, Boyer, O'Connor, Béland and Potvin2} These low treatment rates are problematic as the consequences of late-life anxiety disorder are considerable. Late-life anxiety disorder is associated with increased healthcare consumption, reduced quality of life and high disability rates. When left untreated, it has a tendency to become chronic and substantially increases the risk for developing major depression. Furthermore, between 25 and 43% of older people with anxiety disorders use benzodiazepines chronically (e.g. Grenier et al ^{Reference Grenier, Préville, Boyer, O'Connor, Béland and Potvin2} ), the majority of whom without ever having received a trial of serotonergic antidepressants or CBT. Although guidelines leave room for the use of benzodiazepines, these agents are never considered first choice. Based on their risk-benefit ratio, their use becomes even less favourable with increasing age, and is reserved for short-term crisis management or a final step if CBT and serotonergic antidepressants have failed.

Several factors contribute to the low treatment rates of late-life anxiety disorder. Ageism may be an important explanation as it hinders both adequate detection and, if recognised, adequate treatment. Physicians, therapists, older patients as well as next of kin, often interpret anxiety symptoms and avoidance behaviour in later life as normal and more or less acceptable, when in reality the patient may have a psychiatric disorder. Furthermore, as a result of the ageing process itself, somatic morbidity increases and may partially overlap with symptoms of panic and anxiety. In addition, symptoms of arousal as well as anxiety cognitions tend to decline with increasing age, resulting in some authors characterising anxiety in later life as a less devastating condition compared with anxiety disorder earlier in life. Nonetheless, avoidance behaviour, an important negative consequence of anxiety, does not seem to decline with age.

Available evidence

The need to properly treat late-life anxiety is increasingly recognised, acknowledged by the emergence of several meta-analyses ^{Reference Pinquart and Duberstein4-Reference Gould, Coulson and Howard7} and an even greater number of review articles. In summary, these articles conclude that late-life anxiety disorder can be efficiently treated with both serotonergic antidepressants and CBT. These conclusions, however, are rather tentative as the same reviews and meta-analyses point to the lack of well-designed RCTs in older age groups. This scarcity has even led to the combining of randomised and non-randomised studies within some meta-analyses. ^{Reference Pinquart and Duberstein4,Reference Thorp, Ayers, Nuevo, Stoddard, Sorrell and Wetherell6} A closer look at the RCTs included in these meta-analyses (or published thereafter) reveals three other important limitations.

First, many studies focus on anxiety symptoms instead of anxiety disorders. The challenge of clinical practice, however, is to detect and treat those patients who need treatment. Patients who are likely to have a favourable course (i.e. those without a psychiatric disorder) should not be overtreated. All RCTs restricted to patients with anxiety disorders were confined to generalised anxiety disorder or included a heterogeneous sample of patients with generalised anxiety disorder, panic disorder or social phobia. To date, only one RCT has been conducted in panic disorder specifically, ^{Reference Hendriks, Keijsers, Kampman, Oude Voshaar, Verbraak and Broekman8} whereas no clinical trials are available that evaluate treatment of social phobia, hypochondria, obsessive-compulsive disorder or PTSD in later life.

Second, most studies are relatively small. The 16 RCTs evaluating any form of psychotherapy ^{Reference Pinquart and Duberstein4-Reference Gould, Coulson and Howard7} included on average 16 patients (range 5-30) in the active treatment condition. Recently, 70 older patients with generalised anxiety disorder received CBT within a larger RCT. ^{Reference Stanley, Wilson, Novy, Rhoades, Wagener and Greisinger9} Although pharmacotherapy trials (n = 15) in this field have included more patients, these trials remain small, with an average number of 43 patients in the active treatment condition. ^{Reference Pinquart and Duberstein4}

Third, one may argue about the definition of ‘later life’, as the average age of patients in the available studies is just below 70 years. ^{Reference Pinquart and Duberstein4} The relatively low average age is due to the fact that many studies include patients from the age of 55 years and over. This age cut-off contrasts with clinical practice where old age psychiatry departments generally provide care for patients aged 65 years and above or for patients with significant somatic comorbidity. None of the available trials has assessed somatic comorbidity systematically, let alone the impact of somatic comorbidity on treatment feasibility and treatment effects.

Towards treatment possibilities in later life

Despite these limitations, consensus seems to emerge that in later life, antidepressants may be more effective than CBT and that the cognitive aspects of CBT are less effective. Behavioural techniques, especially relaxation therapy, seem the most effective CBT components in later life. ^{Reference Thorp, Ayers, Nuevo, Stoddard, Sorrell and Wetherell6} Cognitive-behavioural therapy might thus be less effective in older persons compared with their younger counterparts. Up until now, direct comparisons between older and younger patients with anxiety disorders were not available. In this issue, Wetherell and colleagues have analysed the results of the Coordinated Anxiety Learning and Management (CALM) study by age. ^{Reference Wetherell, Petkus, Thorp, Stein, Chavira and Campbell-Sills10} The CALM study examines the effectiveness of collaborative care for generalised anxiety disorder, panic disorder, social phobia or PTSD in 1004 patients within a randomised controlled design. Wetherell et al's paper is the first direct comparison of treatment outcome between younger (age 18-59 years) and older patients (age 60-75 years) with anxiety disorders, including a relatively large number of older patients (n = 134/1004, 13.3%). The results confirm the hypothesis that late-life anxiety disorder is more treatment resistant compared with anxiety disorder in younger patients. Limited by low patient numbers, post-hoc analyses per anxiety disorder suggest that in later life the best results can be achieved in patients with social phobia or PTSD. The lower treatment effectiveness argues for the development of age-specific strategies in later life. Some recent developments are interesting. First, enhanced CBT may improve treatment outcome. ^{Reference Mohlman, Gorenstein, Kleber, de Jesus, Gorman and Papp11} Enhanced CBT consists of regular CBT with the addition of learning and memory aids designed to increase homework compliance, strengthen memory for therapeutic techniques, and to facilitate the use of these techniques. Second, third-wave CBT seems promising for older people, such as acceptance and commitment therapy and mindfulness-based stress reduction. ^{Reference Wetherell, Afari, Ayers, Stoddard, Ruberg and Sorrell12} Although third-wave CBT is quite heterogeneous and includes a diversity of techniques, an important common factor is the abandonment or cautious use of content-oriented cognitive interventions typically part of second-generation CBT. As cognitive interventions are assumed to be less effective in later life, third-wave CBT may offer promising alternatives. Third, if antidepressants are indeed the most efficacious strategy in later-life, clinical trials should also focus on augmentation strategies. Recently, a pilot study suggested that drug therapy can be efficiently augmented with personalised modular CBT. ^{Reference Wetherell, Stoddard, White, Kornblith, Nguyen and Andreescu13} Furthermore, as older people benefit most from relaxation exercise, ^{Reference Thorp, Ayers, Nuevo, Stoddard, Sorrell and Wetherell6} it is worthwhile to examine the effect when augmented with pharmacotherapy.

It is not difficult to conclude that significant work has to be done in this area. I hope the paper by Wetherell and colleagues, ^{Reference Wetherell, Petkus, Thorp, Stein, Chavira and Campbell-Sills10} that confirms the lower treatment efficacy for late-life anxiety disorder, will motivate the field to develop more age-specific treatment strategies.