Morbid Anxiety: Identification and Treatment

Anxiety disorders are an important public health issue by virtue of their prevalence, chronicity, comorbid conditions, and impact on quality of life. With a 12-month prevalence rate of 18% in U.S. communities, they are the most common psychiatric disorders (1). During their lifetime, one in four U.S. residents will have at least one anxiety disorder. Typically appearing early in life, anxiety disorders tend to have a chronic persistent or relapsing course over the lifespan; in this sense they cause even more enduring morbidity than substance use or mood disorders (2). Another consistent epidemiologic finding, to be discussed elsewhere in this issue (3), has been the twofold higher prevalence of anxiety disorders among women (4). Mental health comorbid conditions are the rule with anxiety disorders, especially with other anxiety, mood, and substance use disorders. Moreover, clinical anxiety can predispose to, complicate, and worsen outcomes in a variety of physical conditions, including cardiovascular (5) and respiratory diseases (6). Anxiety in later life worsens cognition and adaptation and is considered a putative risk factor for dementia (7). The burden of anxiety disorders on societies is dramatic; for example, in one earlier U.S. study, annual anxiety-related direct and indirect costs were in excess of $42.3 billion (8), and a 2010 European Union estimate was €74.4 billion (9).

In recent years, diagnostic and treatment options for anxiety disorders have advanced in precision and effectiveness. The pathogenesis of these conditions is still unfolding; however, because of major advances in our knowledge of fear neurocircuitry, neuroimaging, and neurogenetics, personalized care is on the horizon. In this clinical synthesis, emphasizing DSM-5 adult anxiety conditions (10), I highlight modern approaches to diagnosis, work-up, and evidence-based treatment.

Anxiety Phenomenology

Taken together, anxiety disorders are characterized by excessive fear, anxiety, and associated avoidance behaviors. Fear is defined as the response to an acute threat, whereas anxiety is conceptualized as anticipation of future threat. At a neural circuitry level, awareness of fear and anxiety states appears to be mediated via cortical circuits, whereas defensive responses to threats (associated behavioral and physiological responses) tend to be mediated via subcortical and brainstem structures and circuits (11, 12). Cardinal symptoms indicative of specific disorders include recurrent spontaneous panic attacks, excessive worrying, phobic avoidance, fear of negative social scrutiny, and separation fears. Typical anxiety and fear are usually brief, adaptive responses to a stressor, which resolve as the stressor abates. However, one can view typical anxiety and morbid anxiety on a spectrum of severity; for example, isolated panics are extremely common responses to stress (occurring in 20% of the general population) (13), in contrast to recurrent panics with anticipatory anxiety. Temperamental, cultural, and developmental factors can complicate the clinical judgment of “normal anxiety.” Morbid anxiety, by contrast, usually results in enduring distress and impairment in key areas of functioning.

The more dramatic clinical syndromes, such as panic disorder, tend to result in active help seeking and present with typical symptoms that are readily identifiable, whereas less dramatic disorders such as generalized anxiety disorder (GAD) present, not infrequently, with undiagnosed somatic complaints of fatigue, malaise, stomach discomfort, pain, shortness of breath, or palpitations. Social anxiety disorder, by its very nature, tends to present with complications, such as excessive alcohol use or depression, rather than with the patient expressing social difficulties. A majority of patients with anxiety are followed and treated in primary care settings. However, underdiagnosis and undertreatment continue to be persistent problems, whether patients are seen in primary care (14) or psychiatric settings (15).

DSM-5 Classification and Diagnostic Changes

Major changes to anxiety classification rolled out in separate sections of the DSM-5 include the recategorization of obsessive-compulsive disorder (OCD) spectrum disorders and of trauma and stress response disorders (10). Within the new anxiety disorders category, panic disorder and agoraphobia are identified as separate disorders that may co-occur (see Table 1). In the panic disorder section, there is a descriptive subsection outlining the panic attack specifier, which can be applied to any other anxiety or psychiatric disorder with associated panics (e.g., OCD, anorexia nervosa, posttraumatic stress disorder [PTSD]). Separation anxiety disorder has been added to the anxiety disorders category in recognition of the fact that this condition can also occur in adulthood (40% of cases occur after age 18) (16). Now in the social anxiety disorder section, one specifier/subcategory has been changed to “performance only.” More than two-thirds of all patients with social anxiety will have generalized social interactive fears or a mixture of social interactive and performance fears; thus, the “performance anxiety only” presentation is the exception (17).

Diagnostic and Differential Diagnostic Issues

To discriminate among the disorders, it is clinically useful to ask questions that probe thought content (e.g., “What are you afraid of?”; see Table 1, “Core Features/Fears”). Patients with panic disorder are afraid of their attacks; patients with social anxiety fear negative social evaluation; patients with agoraphobia fear situations from which escape might be difficult; patients with separation anxiety fear a disruption in attachment, and patients with GAD are anxiously preoccupied with day-to-day issues. In addition, it is routinely important to establish that the anxiety condition is not better explained by another psychiatric disorder (e.g., PTSD or OCD in the case of panic attacks; and PTSD, mood, or psychotic disorders in the case of GAD) or a medical condition (e.g., hyperthyroidism; substance overuse [e.g., caffeinism, stimulant abuse]; anemia; cardiac, respiratory, or neurological disorder).

A recent nosological research initiative, the National Institute of Mental Health (NIMH) Research Domain Criteria (RDoC) Project, is an important effort intended to move psychiatric diagnosis toward a dimensional and neurocircuitry-based framework, the longer term goal being to propel the field toward precision medicine (18). This overall strategy cuts across many different DSM-5 disorders to look at the neural underpinnings of common biobehavioral dimensions (19). Within the RDoC matrix of inquiry, symptom domains relevant to anxiety disorders include the negative valence domain, within which the constructs of acute threat (fear), potential threat (anxiety), and sustained threat (chronic stress) are clinically meaningful. A second domain of relevance is arousal/modulatory systems, and constructs of particular interest within this domain are arousal/hyperarousal and sleep-wake cycle disturbances (20). While RDoC framework research findings have yet to change clinical practice, the notion of considering anxiety disorders transdiagnostically or dimensionally has case formulation and treatment planning implications.

Approaches to Treatment

Treatment Planning

After a careful clinical assessment and diagnosis, case formulation for the patient with anxiety may be approached in a variety of ways. One simple scheme, readily adapted to the analysis of acute mood or anxiety cases, is the heuristic of a stress/diathesis model. One reviews illness risk factors and tailors the model to the patient’s history. The individualized explanatory model helps to make the patient’s symptoms and illness course understandable, providing a good segue into treatment discussions. Biological treatments, such as pharmacotherapies, can stabilize putative neurochemical problems, which are likely to be a component of predisposing factors (e.g., genetic diatheses, early trauma, and anxiety-prone temperament). Evidence-based psychotherapies, such as cognitive-behavioral therapy (CBT), target perpetuating emotional learning processes such as impaired fear extinction, anticipatory anxiety, and avoidance and safety behaviors. Thus, the major treatment modalities are conceptualized and presented as complementary, an approach that is synchronous with the clinical neuroscience literature regarding treatment mechanisms (21). Beyond review of psychoeducational material about typical and atypical anxiety (all that may be needed for some very mildly ill patients), the empirical literature strongly supports presenting combination CBT and pharmacotherapy as the treatment package of choice (especially for moderately to severely ill patients) (22, 23). Building a treatment alliance with the anxious patient involves being mindful of anxiety-related psychological issues that are likely to be present in varying degrees across the disorders. These issues including feelings of vulnerability; social isolation; a tendency toward cognitive, emotional, and behavioral avoidance; threat-biased cognitions; and sensation sensitivity. These factors can surface in the form of resistance to new treatment suggestions, limited adherence to medications or therapy tasks, or avoidance of treatment sessions. In the following review of evidence-based treatments, we concentrate our remarks on the more complex anxiety disorders (panic disorder, GAD, and social anxiety disorder). Separation anxiety treatments are covered in another article in this issue (24).

Evidence-Based Pharmacotherapies

In planning pharmacotherapy for anxiety, one adopts a target symptom approach. The goal of therapy is to treat to remission whenever possible, to promote optimal outcomes and prevent relapse. Regarding expectations about treatment response, presenting a spectrum model of treatment sensitivity is useful, with panic disorder and specific phobia being highly treatment sensitive and other complex anxiety disorders, such as GAD and social anxiety disorder, being only moderately treatment sensitive. In addition to achieving symptomatic improvement, pharmacotherapies promote improvement in overall quality of life, a fact not fully appreciated in earlier clinical trials (25).

Panic disorder.

Selective serotonin reuptake inhibitors (SSRIs) and serotonin-norepinephrine reuptake inhibitors (SNRIs) are generally considered first-line treatment options for anxiety spectrum disorders, and panic is no exception (26). As an aside, a new neuroscience-based nomenclature (NbN) is being adopted by many journals to characterize psychotropic agents in terms of their brain effects (in contrast to historical or marketing-driven nomenclature) (27). In NbN terms, SSRIs and SNRIs would be now be described as agents for anxiety that target the serotonin pharmacologic domain via reuptake inhibition mechanisms. Second-line options include older antidepressant medications such as the tricyclic antidepressants, of which imipramine is the best studied. Benzodiazepines (in NbN terminology, GABA-positive allosteric modulators) are good third-line treatments or adjunctive treatments to be offered to promote acute stabilization (28) (29). However, it is not unreasonable to prescribe benzodiazepines for the short-term and longer term care of patients who are intolerant of or nonresponsive to antidepressant agents (30, 31). Higher potency, shorter half-life benzodiazepines are preferable, with alprazolam and clonazepam being panicolytic agents approved by the Food and Drug Administration. Fourth-line agents include anticonvulsants such as gabapentin and pregabalin. Although second-generation antipsychotic agents have been studied in small to moderately sized trials and well tolerated, efficacy results have been mixed (32). In a recent systematic review of anti-anxiety treatment effects (23), pre- to posttreatment effect sizes (Cohen’s d) were as follows: 2.25 for SNRI therapy, 2.15 for benzodiazepine treatment, 2.05 for SSRI therapy, and 1.83 for imipramine (all large effect sizes). In contrast, the effect size for pill-placebo effects was 1.29. Thus, the different classes of anxiolytic/antipanic medications appear similarly effective, differing primarily on mechanism, tolerability/safety, and pharmacokinetic characteristics. SSRI/SNRI-resistant patients may benefit from augmentation treatment with agents in the other anxiolytic classes mentioned earlier (26).

GAD.

For treatment of an acute exacerbation of GAD, SSRIs are the first-choice medications. If the initial trial is not effective, some experts recommend a trial of another SSRI (33) before moving to an SNRI. The SNRIs duloxetine and venlafaxine have been extensively studied in GAD randomized controlled trials (RCTs) (34). Beyond these agents, other evidence-based medication options include buspirone, benzodiazepines, tricyclic antidepressants, pregabalin and gabapentin, and antihistamines (hydroxyzine). A meta-analysis of 21 placebo-controlled GAD trials from the world literature observed only modest treatment benefit from active medications: overall effect size (Cohen’s d)=0.39 (35). For adults with GAD, comparative effect sizes of the different agents were as follows: 0.5 for pregabalin, 0.45 for antihistamines, 0.42 for SNRIs, 0.38 for benzodiazepines, 0.36 for SSRIs, and 0.17 for azapirones (buspirone). A newer development in GAD pharmacotherapy is the observation that the second-generation antipsychotic agent quetiapine is an efficacious and tolerable monotherapy (36). In addition, in a recent active-comparator GAD RCT, the novel melatonin 1/2 receptor agonist, agomelatine, has demonstrated promising efficacy (37). SSRI/SNRI-resistant patients with GAD may benefit from augmentation therapy with an agent from the classes mentioned earlier. However, additional work is needed to improve the evidence base in this area (38).

Social anxiety disorder.

For the stabilization treatment of social anxiety disorder, SSRIs and the SNRI venlafaxine are the recommended initial pharmacotherapies (39, 40). An extended trial of up to 12 weeks may be necessary before calling a trial “failed.” There is less of an evidence base for social anxiety disorder compared with other anxiety disorders to guide long-term treatment and alternative treatment steps. However, SSRI-resistant patients may benefit from augmentation with a benzodiazepine, gabapentin, or buspirone. Recent empirical data support the benefits of clonazepam augmentation in SSRI-resistant social anxiety (41). Beyond this, monotherapy trials with a monoamine oxidase inhibitor (e.g., phenelzine), gabapentin, or a benzodiazepine are recommended. In a recent meta-analysis, the overall effects of pharmacotherapy were determined to be modest (Hedge’s g=0.39) (42). The comparative effect sizes of specific agents were as follows: phenelzine (g=1.14), paroxetine (g=0.49), venlafaxine extended-release (g=0.45), and moclobemide (g=0.23). Thus, there is a considerable need for more robust social anxiety disorder pharmacotherapies.

Personalized pharmacotherapy.

With the advent of commercially available pharmacogenomics screening platforms, there has been significant improvement (two- to threefold) in the accurate prediction of antidepressant treatment responses (43). Retrospective data in a mixed group of outpatients with depression and anxiety who were undergoing pharmacotherapy suggested that utilization of health care resources could be similarly predicted (44). Patients with prescriptions that were not recommended, or that were “red bin” (versus yellow or green bin) prescription choices, had higher service utilization. Thus far, clinical predictors/moderators of antidepressant response among patients with anxiety have been inconsistent (45). Regarding promising treatment response biomarkers, there is preliminary evidence for the validity of the serotonin transporter genotype variant 5-HTTLPR/rs25531, as well as some imaging measures of anterior cingulate cortex function (46). Furthermore, in a recent multisite, 10-week, open-label, sertraline trial for patients with social anxiety disorder (N=346), polymorphisms of the RGS2 gene (which encode for a G-protein-signaling regulator previously linked to social anxiety and 5-HT neurotransmission) predicted the likelihood of clinical remission (47). However, although this new technology is very promising, the translation of personalized care protocols to routine care requires additional research.

Evidence-Based Psychotherapies

Although various psychotherapies have been tested for anxiety disorders, CBT (especially individual CBT) appears to have the most robust efficacy signal across different disorders (48). A positive evidence base for other approaches, such as brief psychodynamic psychotherapy, mindfulness-based cognitive therapy (MBCT), and relaxation training, is also emerging. Initiatives via the Internet and other electronic media to improve the dissemination and acceptability of CBT are promising, and the effectiveness of these approaches can approach that of in-person CBT (49).

Panic disorder.

In a large, multisite (N=312), panic disorder trial, after acute treatment (12 weeks), both the CBT (panic control therapy involving interoceptive exposure) and imipramine treatment groups were superior to placebo, and the combination treatment group fared slightly better than those receiving monotherapies. However, combination treatment benefits improved by the end of maintenance treatment (6 months) (50). In a more recent two-site trial (N=201 patients), patients treated with panic-focused psychodynamic psychotherapy improved similarly to patients receiving CBT, supporting the robustness of dynamic approaches for panic (51). A recent Cochrane Database meta-analysis of panic therapy studies concluded that, although CBT was the most extensively studied modality and was often superior to control conditions, it was difficult to say that it was clearly better than alternative therapeutic approaches (52).

GAD.

A systematic review of GAD psychological treatments (N=41 studies involving 2,132 patients) identified CBT as the best tested and validated therapy with large effect sizes versus wait-list control (Hedge’s g=0.84) (53). In a recent 8-week GAD trial, the effectiveness of MBCT was observed to be similar to that of a CBT-based psychoeducational intervention, and both were found to be superior to usual care (54). In addition, applied relaxation has been a long-standing, evidence-based, therapeutic approach to the management of GAD symptoms (55).

Social anxiety disorder.

In a large meta-analysis of social anxiety therapies (N=101 trials, involving 13,164 patients), individual CBT emerged as the most robust psychotherapy compared with a wait-list control group (standard mean difference [SMD]=−1.19) (40). In contrast, psychodynamic therapy had an SMD of −0.62. Also, in this evaluation, individual CBT emerged as the only psychotherapy that separated from psychological placebo (SMD=−0.56) and was slightly better than several pharmacotherapies (SNRIs and SSRIs). In a large, multisite comparison social anxiety disorder trial of CBT (N=209), dynamic psychotherapy (N=207), and wait-list control (N=79) conditions, dynamic psychotherapy was similar to CBT in effectiveness after acute treatment (56), and both intervention groups had response rates of 70% at 24-month follow-up (57). Before this trial, there was little convincing evidence among patients with social anxiety supporting the effectiveness of psychodynamic therapy.

Is Combination Treatment Really Better?

Data from different studies and meta-analyses endorse the additional benefit of combination treatment (therapy plus medications) over monotherapy with either anxiolytic medicines or psychotherapies. One assessment of studies that added pharmacotherapy to CBT (a meta-analysis of 11 studies with 471 patients) observed a moderate degree of additional short-term benefit (58). Patients with panic disorder and those with GAD derived the most benefit from this strategy. In studies of the reverse approach (adding psychotherapy to antidepressant medications), results from 52 studies involving 3,623 patients (with panic disorder, OCD, and major depressive disorder) again supported the clinical superiority of combination treatment (22). Overall, these findings weigh against the view that anxiolytic pharmacotherapies impede effective psychotherapy; rather, they provide evidence of additive benefit.

Over the past decade, there has been much research interest in antianxiety treatment strategies that enhance exposure-related extinction learning. This translational effort was inspired initially by exciting preclinical findings that acute administration of d-cycloserine—an old drug used for tuberculosis treatment and a partial glycine site agonist at the NMDA/glutamate receptor complex—enhanced the extinction of conditioned fears (59). Early results of adapting this strategy to patients with specific phobias were very promising (60). This treatment approach is attractive in its strategic and minimal use of pharmacotherapy (low-dose d-cycloserine just before therapy sessions) to promote adaptive learning. Although there is a great deal of potential in this type of treatment paradigm for anxiety spectrum conditions and beyond (61), translation to clinical practice has been slow, owing to mixed results in larger scale clinical trials (62).

Complementary and Alternative Treatment Interventions

Although there have been encouraging antidepressant efficacy data for a number of complementary and alternative, integrative medications (CAIMs; e.g., St John’s Wort, omega-3 fatty acids, S-adenosyl-methionine, tryptophan), there are less convincing data supporting their anxiolytic spectrum of activity (63). There is one positive, moderate-sized, single-site, double-blind, placebo-controlled trial of kava-kava for the stabilization treatment of GAD (64). However, currently, this CAIM cannot be offered as a standard treatment option, and practitioners should be aware of the potential for rare but serious adverse events (e.g., hepatotoxicity). In one systematic review, for GAD and panic disorder, there was modest evidence that exercise was an effective strategy (63), whereas another meta-analysis of seven studies did not find sufficient evidence to offer exercise routinely as a treatment approach for anxiety (65). The latter conclusion was based on concerns regarding the validity of some of the study control conditions. Regarding the benefit of yoga for stress/anxiety symptoms, there is one positive trial in civilian PTSD, but there were no rigorous trials in DSM-5 anxiety conditions (66). Breathing techniques adapted from Sudarshan Kriya yoga are showing promise in the management of GAD and PTSD (67, 68). Other complementary therapy approaches, such as Swedish massage, may also be effective for GAD treatment (69). Thus, complementary treatment approaches may offer relief for individual patients, but additional positive data are needed before these interventions are systematically included in treatment plans.

Neuromodulation

With the success of stimulation therapies for depression, there has been much interest in the applicability of this class of therapies for morbid anxiety. Currently, sample size has been small for studies of these interventions for anxiety and stress disorders. Both effectiveness- and tolerability-wise, repetitive transcranial magnetic stimulation appears promising for treatment of panic disorder (70). However, confirmatory, larger scale trials are needed. In one small (N=8), open-label PTSD/major depressive disorder 8-week trial, external trigeminal nerve stimulation, as an adjunct to ongoing pharmacotherapy, was beneficial for mood and trauma symptoms (71). There was a case report for a patient with GAD that documented the anxiolytic benefit of transcranial direct current stimulation (72). Cranial electrical stimulation devices appear to have utility as an augmentation treatment for depression and could, with additional testing, prove to be of value for patients with anxiety (73). With a benign side effect profile relative to pharmacotherapies, this “third wave” of psychiatric therapeutics is likely to become an important option for patients resistant to the standard treatment modalities.

Summary and Conclusions

Patients with anxiety disorder are a prevalent yet underserved population in primary care and psychiatric care settings. This review, emphasizing the past 5 years of clinical research and opinion regarding anxiety diagnosis and treatment, has summarized the impact of DSM-5 for anxiety disorder classification and highlighted the major evidence-based clinical treatment options. DSM-5 anxiety disorders are now more similar to each other phenomenologically; anxiety conditions that were dissimilar (e.g., OCD and PTSD spectrum disorders) in DSM-IV were recategorized in DSM-5. The remarkable prevalence of adult separation anxiety disorder has led to its inclusion in the DSM-5 anxiety disorders category. Clinical experts and the clinical research literature of late have emphasized transdiagnostic approaches to assessment and treatment, partly inspired by the NIMH RDoC initiative, and partly in recognition of the fact that the major treatment options cut across diagnostic boundaries. While few new antianxiety treatments have emerged in the past 5 years, the treatment literature has matured to the point where many large systematic reviews and meta-analyses are available to inform current clinical practice. The paucity of new pharmacotherapies may be partly due to industry priorities shifting away from drug development for central nervous system disorders. In general, treatment effects for panic disorder are good, whereas interventions for other complex disorders (e.g., GAD and social anxiety disorder) are only modestly effective. Combination treatment with pharmacotherapy and psychotherapies (although not always practical) should be routinely offered across diagnoses, because the effects are likely to be additive. SSRIs and SNRIs are the first-choice pharmacotherapies across disorders, whereas benzodiazepines are recommended adjunctively as short-term interventions. Thus far, second-generation antipsychotic agents are not clearly beneficial for patients with anxiety, with the exception of those with GAD, for which quetiapine appears to be an effective monotherapy. Psychotherapy-wise, CBT—especially individual CBT—is a robust therapy option across disorders; however, patient access to qualified practitioners remains a challenge. The Internet and other electronic platforms for delivering CBT have partly addressed this problem. The evidence base for brief dynamic therapies has improved, especially for panic disorder and social anxiety disorder. Emerging treatment strategies such as complementary and alternative therapies, neuromodulation, cognitive enhancement, and personalized medicine are very promising. Of these options, aerobic exercise appears to be modestly effective for GAD and panic disorder, whereas effectiveness of other strategies awaits additional research.