Should antidepressant medication be used in the elderly?

Abstract

Late-life depression is a serious illness accompanied by medical morbidity, cognitive decline and risk of suicide. Antidepressant medications are a cornerstone of treatment for depressed elders. Although they are optimally provided in conjunction with psychotherapy, in many cases they are used alone. Recently, concern has developed over modern antidepressant medication, including concerns about their ultimate efficacy and particular risks that may be seen in older adult populations. Ultimately, antidepressant medications are effective for many individuals and continue to play an important role in treating depressed elders, although the potential risks must be weighed with the patient and their families. Current data do not support restriction of their use and untreated depression has serious negative health consequences. Patients need treatments with better efficacy and safety, including new pharmacological options and better access to and dissemination of nonpharmacological treatment.

Keywords:

• adverse events
• aging
• antidepressant
• depressive disorder
• geriatrics
• psychotropic medications
• side effects
• suicide
• treatment

Depression in older adults is common, disabling and accompanied by high levels of medical morbidity and disability [1]. Major depressive disorder occurs in 5% of community dwelling elders, while up to 16% of older adults have clinically relevant depressive symptoms. Beyond personal suffering, depression is associated with a wide range of negative outcomes, including cognitive decline and dementia, increased family caregiving burden, poor medical outcomes, risk of suicide and high mortality. Although the occurrence of depression is clearly shaped by environmental stresses and experiences, it is also a ‘brain disorder’ characterized by altered function of key neural circuits. Coexisting medical illness contributes to alterations in brain function and so predispose individuals to depression [2].

Given its negative repercussions, depression obviously requires treatment. Concordant with American Psychiatric Association guidelines, expert consensus guidelines for older adults propose that optimal treatment consists of antidepressant medication coupled with psychotherapy [3]. Unfortunately, depression is often unrecognized; even when diagnosed, depression in older adults is often untreated or undertreated [4]. In addition, despite evidence that learning-based psychotherapies (e.g., cognitive behavioral therapy, problem solving therapy, and interpersonal therapy) are effective treatments for depression, psychotherapy treatment for depression is declining in the USA and remains uncommon among depressed elders [5]. The reasons for these changes are complex and may include limited provider availability and poor reimbursement for psychotherapy by Medicare.

Thus, the majority of treated elders receive only antidepressant medications. However, safety concerns have arisen over the use of antidepressants in older adults. Others have questioned the risk/benefit ratio of antidepressants, even proposing restrictions on their use [6]. Such positions may reflect depression’s stigma or a lack of appreciation of the serious adverse consequences of untreated depression [7]. This article discusses a perspective on benefits and risks of these medications and then presents a pragmatic path going forward.

Potential benefits in older populations

We have good evidence supporting the efficacy of modern antidepressants in treating depression, including their benefit in older populations [1,8]. Admittedly, the data are less robust for older populations who are often excluded from clinical trials. In addition, interpreting these data can be challenging. There is publication bias wherein many clinical trials registered with the US FDA are unpublished. While reported effect size values for published trials are higher in journal publications than in FDA reviews [9], the problem may lie in clinical trial design decisions that artificially inflate placebo response rates [10]. This results in lower effect sizes and increased difficulty detecting a therapeutic signal.

Other work finds that successful antidepressant treatment provides real-life benefit. Depression management using both medications and psychotherapy significantly improves physical function, quality of life and lowers mortality rates [11,12]. Such benefits may be maintained for years following treatment. Meanwhile, although late-life depression has high rates of relapse, maintenance antidepressant treatment can reduce relapse rates [13]. Importantly, antidepressant medications also benefit other common psychiatric illnesses including anxiety disorders.

Suicide is a significant risk in older adult populations where it is almost twice as frequent as in the general population. In comparison with younger populations, suicidal ideation decreases with age, but older adults who develop suicidal thoughts are at a higher risk of actually completing suicide. Results of analyses conducted by the FDA found that while antidepressant use was associated with increased rates of suicidal thoughts and behavior in children, antidepressant use significantly decreases the risk of suicidal thoughts and behavior in geriatric patients aged over 65 years (odds ratio = 0.37; 95% CI, 0.18–0.76) [14]. Although one may argue that treatment emergent adverse events are not always captured in trial registry reports [15], findings supportive of the FDA’s conclusions showing decreased risk in older adults are observed even when analyzing patient-level data [16]. In older adults, suicidal ideation and behavior are strongly related to severity of depression, and reduction in suicidality mirrors improvement in depression severity [16,17]. Ultimately, depression is the most salient and modifiable risk factor for suicide and effective antidepressant treatment is the best way to reduce suicide rates [17].

Depression is also associated with cognitive impairment and increased risk for dementia [1]. Although other psychotropic medications such as benzodiazepines negatively affect cognition, modern antidepressant medications do not appear to carry the same risk. Indeed, aside from agents with anticholinergic activity, antidepressant use may result in improved cognitive performance [18]. However, this effect does not appear to modify the underlying risk of dementia.

Assessing potential risks in older populations

Depression in older populations is accompanied by age-related changes in drug metabolism, medical morbidity and polypharmacy. This complicates our ability to definitively identify causal relationships between antidepressant exposure and adverse outcomes. Although observational studies can identify medication adverse events, a prerequisite is that the adverse effect is distinct from the illness [19]. This is problematic in late-life depression that is associated with a range of medical illnesses across multiple organ systems [20]. Although many studies examining the relationship between depression and physical health do not account for antidepressant use, depression is associated with immune system and hypothalamic–pituitary–adrenal axis dysfunction [20], which in turn negatively affects end organs. Thus, it is not always possible to disentangle what adverse health outcomes are related to antidepressant use and what are influenced by the occurrence and persistence of depression. This limits our ability to identify factors necessary to determine causality, such as temporality and specificity [19]. Moreover, conclusions from observational studies comparing depressed individuals taking antidepressants with nondepressed populations not taking antidepressants may not accurately identify true risks associated with drug exposure.

That being said, antidepressants have specific risks in older adult populations that are not observed in younger adults. One particular concern is that antidepressants increase the risk of falls, osteoporosis and fractures. Falls are a known risk for the older tricyclic antidepressants as they increase the risk of orthostatic hypotension. More recently, based on numerous observational studies, the updated Beers Criteria highlighted the selective serotonin reuptake inhibitors along with most other psychotropic drugs as increasing these risks. However, these data do not meet the level of rigor needed to determine causality and may be confounded by observations that many risk factors for falls are also risk factors for depression, while depression is itself associated with increased falls risk [19]. Although falls are a potential risk, the current evidence does not support cessation of antidepressant use in older depressed adults. Perhaps a more useful approach would be to provide assessment and intervention for falls in depressed elders.

The lack of clarity in differentially assigning risk to either antidepressant use or to depression itself plagues other associations, such as reports associating antidepressant use with vascular disease. However, this logic does not affect our understanding of risks that bear a clear temporal relationship to antidepressant use, such as hyponatremia or antidepressant discontinuation syndrome.

In the end, all pharmacological treatments carry risks. The question is, does the benefit outweigh the risk? This question is not as clear as it may first appear. The definition of benefit may be variable, with some individuals achieving complete remission of depressive symptoms, while those with chronic, treatment-resistant depression may experience improved function and benefit even with incomplete resolution of symptoms. Likewise, future risks are, by definition, unknown and even acknowledged risks will not affect the majority of patients. In the end, before making sweeping changes to clinical practice, as proposed by others we need careful assessment of causality building on established criteria [19].

A pragmatic approach

From the literature, several points can be made. First, depression in older adults is accompanied by a host of negative outcomes, including suffering and disability, medical complications, cognitive decline and risk of suicide. Second, antidepressants are effective in treating depression in older populations, although they are not always as effective as we and our patients would like. Third, antidepressants have side effects and risks, some of which can be observed acutely while others may be longer term consequences.

For clinicians, the pragmatic answer is that antidepressants have a key role in treating depressed elders. We need to work to identify depression when it presents, discuss the potential risks and benefits for all available treatment options, and once a decision is made, carefully monitor the patient to assure symptom improvement and treatment tolerability. This is true for whatever treatment selected, whether it be antidepressant monotherapy, psychotherapy or brain stimulation therapy.

This is also a challenge to the scientific community. We need better treatments for depression. Many people develop side effects to current antidepressants or do not have a satisfactory response. Can we move beyond current drug mechanisms? Will a better understanding of novel agents such as ketamine lead to more robust treatments? Can we address past critiques of trial methodology to better assess the true risks of treatment? There is room to improve the quality of and accessibility of psychotherapy. We need better availability of evidence-based nonpharmacological treatments, such as efforts to validate internet-based psychotherapy. In the end, just as we need pragmatic approaches in how we currently treat depression, we need broad and pragmatic approaches to how we conduct research on depression interventions.

Financial & competing interests disclosure

This project was supported by National Institute of Mental Health (NIMH) grants R21 MH099218 and R01 MH102246. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.