The first known clinical case description of an anxiety disorder appeared in the medical corpus of the Ancient Greek physician Hippocrates. The description tells of Nicanor, a man who developed an extreme fear of a "flute girl" whom he encountered one night at a drinking party and who haunted him every night for many years. Five hundred years after this case description, the Ancient Roman Stoic philosophers Seneca the Younger and Cicero addressed the topic of anxiety at length, recognizing both its benefits and harms, depending on the severity and circumstances of the anxiety [1]. These texts reveal a sophisticated understanding of fear and anxiety among these ancient authors, even by modern medical standards. It was not until the 19th century that Charles Darwin noted essential similarities in the expression of fear and anxiety in mammals, reinforcing Seneca's notion that fear and anxiety are ultimately adaptive traits [2]. In its normal state, anxiety facilitates the management of potential future hazards [3,4,5]. In its extreme state, the individual regards it as excessive or distressing or it can cause impairment in the individual's daily life, thus constituting a disorder [6,7].

The analogy of a smoke detector demonstrates the adaptive and maladaptive aspects of anxiety [8,9]. Just as the function of a smoke detector is to signal potential fires so that one can take action to prevent harm, the function of anxiety is to signal any potential hazards so that preventive actions can be taken. In this analogy, an anxiety disorder is an extreme that renders the individual more sensitive to threat signals [10]. Although those with higher anxiety experience more false alarms (signals for a threat that does not occur), this is advantageous to the extent that it reduces the risk of a fatal miss. In other words, the costs associated with false alarms and misses are not equal: over-reacting to non-threats is generally less costly than failing to detect one danger. Nonetheless, living in a chronic state of high anxiety can take a long-term toll on an individual's health and quality of life, and in these cases, intervention is warranted.

In 1949, the Nobel Prize in Medicine was awarded to António Egas Moniz for his discovery of "a simple operation, always safe, [and] which may prove to be an effective surgical treatment in certain cases of mental disorder" [11]. Specifically, Moniz discovered the prefrontal leukotomy as a treatment for mental disorders, including anxiety disorders [12]. Since then, studies have found that damage to the ventromedial prefrontal cortex produces resistance against anxiety and depression [13; 14; 15; 16]. Despite the effective reduction of anxiety in these patients, it took many decades until research began to address the harms imposed by damage to the prefrontal cortex. For example, in addition to reducing anxiety, damage to the ventromedial prefrontal cortex also impairs self-regulation and decision-making and can induce sociopathic behaviors [17; 18; 19; 20; 21]. Similar patterns of anxiety reduction were also observed in one patient with focal bilateral lesions to the amygdalae who showed a similar pattern of impairment in her daily life as those with damage to the prefrontal cortex [22]. Although the prefrontal cortex and amygdala are critical structures in a neural network that is necessary for anxiety, these findings highlight the fact that damage to these structures comes with unintended consequences. These findings also highlight the more general point that, in treating anxiety disorders, it is also important to not abolish otherwise useful traits as it is to reduce the anxiety to a manageable level.

In 1949, the Nobel Prize in Medicine was awarded to António Egas Moniz for his discovery of "a simple operation, always safe, [and] which may prove to be an effective surgical treatment in certain cases of mental disorder" [11]. Specifically, Moniz discovered the prefrontal leukotomy as a treatment for mental disorders, including anxiety disorders [12]. Since then, studies have found that damage to the ventromedial prefrontal cortex produces resistance against anxiety and depression [13; 14; 15; 16]. Despite the effective reduction of anxiety in these patients, it took many decades until research began to address the harms imposed by damage to the prefrontal cortex. For example, in addition to reducing anxiety, damage to the ventromedial prefrontal cortex also impairs self-regulation and decision-making and can induce sociopathic behaviors [17; 18; 19; 20; 21]. Similar patterns of anxiety reduction were also observed in one patient with focal bilateral lesions to the amygdalae who showed a similar pattern of impairment in her daily life as those with damage to the prefrontal cortex [22]. Although the prefrontal cortex and amygdala are critical structures in a neural network that is necessary for anxiety, these findings highlight the fact that damage to these structures comes with unintended consequences. These findings also highlight the more general point that, in treating anxiety disorders, it is also important to not abolish otherwise useful traits as it is to reduce the anxiety to a manageable level.

Two main subtypes of panic disorder have been observed, diverging between individuals with respiratory and non-respiratory symptoms [23,24,25]. Determining the subtype may be informative for treatment purposes. Older adults with panic disorder experience fewer symptoms of panic compared with younger adults, and their panic attacks are also reported to be less intense and shorter in duration [26,27,28,29].

The standard procedure for anxiety disorder assessment is the structured diagnostic interview, which is administered by a trained professional. The structured interview consists of pre-determined questions that assess for relevant symptoms based on diagnostic criteria. For example, an interview for GAD would start by asking the individual questions about the presence of worry symptoms over the past six months. If the interviewee answers this question affirmatively, the interviewer would then ask the individual about the presence of secondary symptoms associated with the worry (e.g., sleep, irritability). If the individual responds affirmatively to the minimum number of secondary symptoms required for a diagnosis of GAD, the individual would then be queried about the presence of distress or impairment due to the worry. The key advantage of the structured interview is its standardized administration, procedure, and scoring, which minimize bias and error in assessment. Two commonly used structured interviews for the assessment of mental disorders are the Diagnostic Interview Schedule (DIS) and the Composite International Diagnostic Interview (CIDI) [56,57]. In addition, the Anxiety Disorders Interview Schedule (ADIS) is a structured diagnostic interview that was developed specifically for anxiety disorder assessment [58]. These interviews are regularly updated along with diagnostic criteria, as for example with new editions of the DSM. Structured interviews rely essentially on self-report; in addition to being administered by clinicians, they may also be conducted by trained lay persons and/or computer-assisted technology (as in epidemiologic surveys).

The Overall Anxiety Severity and Impairment Scale (OASIS) is a brief, transdiagnostic screening tool designed to assess for the severity of anxiety in the past week of the individual's life [69]. There are five items, each rated on a five-point scale (0 to 4), yielding a total possible score of 20. A raw score of 8 or greater indicates the presence of anxiety disorder based on validation against anxiety disorder diagnosis using the psychiatrist-administered SCID [70]. Raw scores of 10 and 12 indicate the presence of marked and severe anxiety, respectively, based on validation against the clinician-rated Clinical Global Impression-Severity (CGI-S) scale in a sample of individuals with any anxiety disorder ascertained using the Mini International Neuropsychiatric Interview (MINI) [71].

Compared with younger adults, older adults report fewer and less concrete anxiety symptoms across anxiety subtypes [40,41,42,43,45]. In addition to this, age-related neurocognitive impairment makes self-reporting a more difficult method of assessment [55]. For example, those with memory impairment can experience stressors that evoke negative effects without leaving memory traces [22,80]. Although informant report can be a way of effectively gathering information about observable (e.g., physical) symptoms, it is ineffective for identifying unobservable (i.e., subjective) symptoms [81].

Anxiety disorders are the most prevalent mental disorders in older adults [54,84]. The most prevalent subtypes are, in descending order, specific phobia, GAD, social anxiety disorder, and panic disorder. Table 5 displays the one-year prevalence of anxiety disorders, both overall and by subtype, in the NESARC and the Collaborative Psychiatric Epidemiology Surveys (CPES) of the United States. The prevalence of anxiety disorders is higher among women relative to men, and the prevalence of all anxiety subtypes decreases among persons 75 years of age or older. Previous studies have also reported ethnic differences in prevalence, such that Native and White Americans have the highest prevalence, and Hispanic and Asian Americans have the lowest prevalence of anxiety disorders [85]. Black Americans have a higher or lower prevalence of anxiety disorders depending on subtype; specific phobias and GAD are more prevalent, comparable to Native and White Americans, whereas panic disorder and social anxiety disorder are less prevalent, closer to levels observed in Hispanic and Asian Americans. The prevalence of anxiety disorders does not vary substantially by educational attainment or marital status.

Anxiety disorders are the most prevalent mental disorders in older adults [54,84]. The most prevalent subtypes are, in descending order, specific phobia, GAD, social anxiety disorder, and panic disorder. Table 5 displays the one-year prevalence of anxiety disorders, both overall and by subtype, in the NESARC and the Collaborative Psychiatric Epidemiology Surveys (CPES) of the United States. The prevalence of anxiety disorders is higher among women relative to men, and the prevalence of all anxiety subtypes decreases among persons 75 years of age or older. Previous studies have also reported ethnic differences in prevalence, such that Native and White Americans have the highest prevalence, and Hispanic and Asian Americans have the lowest prevalence of anxiety disorders [85]. Black Americans have a higher or lower prevalence of anxiety disorders depending on subtype; specific phobias and GAD are more prevalent, comparable to Native and White Americans, whereas panic disorder and social anxiety disorder are less prevalent, closer to levels observed in Hispanic and Asian Americans. The prevalence of anxiety disorders does not vary substantially by educational attainment or marital status.

The chronicity of a disease refers to its persistence. Persistence is defined here as the percentage of respondents who meet diagnostic criteria for an anxiety disorder at baseline and who then meet criteria again at follow-up. Data from the NESARC indicate that approximately 30% of older adults (55 years of age and older) have persistent cases of anxiety disorder, or chronicity, assessed over a three-year follow-up period. The most persistent subtypes were specific phobia (25%) and GAD (20%), followed by social anxiety disorder (16%) and panic disorder (10%) [31].

Importantly, the findings of a 2016 systematic review and meta-analysis of prospective, longitudinal studies suggest that a diagnosis of any anxiety disorder at baseline is not associated with increased risk of all-cause mortality at follow-up [5]. In fact, in a population study of Norwegians, high anxiety symptoms were associated with lower mortality among individuals with depression [103]. In a population study of a 1946 UK birth cohort, individuals who demonstrated lower levels of trait anxiety in adolescence were associated with higher risk of accident mortality at follow-up [104]. Thus, low anxiety (but not high anxiety) is associated with increased mortality risk, and some degree of anxiety is beneficial for survival. Some anxiety likely encourages individuals to engage in preventive health behaviors. For example, women who worry about the possibility of breast cancer are more likely to seek routine screenings, people who are more worry-prone are more likely to vaccinate than those who worry less, and smokers with higher worries about their health have been found to be more likely to quit [105,106,107].

The two strongest risk factors for anxiety disorders among older adults are female sex and younger age [84,108,109]. However, other risk factors have also been identified. Cigarette smoking is shown to be a major risk factor of anxiety disorder onset, while smoking cessation is associated with reduced anxiety, suggesting that smoking interventions would have a significant effect on anxiety disorder onset [110,111]. Another important risk factor of anxiety disorder onset in longitudinal studies is the occurrence of adverse life events, such as the ending of a relationship or the injury, illness, or death of a loved one [112,113,114].

The two strongest risk factors for anxiety disorders among older adults are female sex and younger age [84,108,109]. However, other risk factors have also been identified. Cigarette smoking is shown to be a major risk factor of anxiety disorder onset, while smoking cessation is associated with reduced anxiety, suggesting that smoking interventions would have a significant effect on anxiety disorder onset [110,111]. Another important risk factor of anxiety disorder onset in longitudinal studies is the occurrence of adverse life events, such as the ending of a relationship or the injury, illness, or death of a loved one [112,113,114].

The prevalence of anxiety disorder is substantially higher in medical versus community settings, and there is a particularly high prevalence of anxiety disorder in individuals with Parkinson disease and among caregivers of older adults [51,61,122,123,124,125]. Studies have demonstrated that, in part, the psychological distress (e.g., anxiety and depression) experienced by caregivers is linked to their patients' overall cognitive well-being, patient functional ability, and the reported caregiver burden [126,127,128,129].

Individual CBT was found to be effective for acute treatment compared with waitlist control groups. Pharmacologic interventions included anticonvulsants, benzodiazepines, monoamine oxidase inhibitors (MAOIs), noradrenergic and serotonergic antidepressants, selective serotonin reuptake inhibitors (SSRIs), and selective norepinephrine reuptake inhibitors (SNRIs). SSRIs and SNRIs were found to be the most effective class of pharmacological treatment compared with placebo control groups [133].

Exposure therapy is the treatment of choice for specific phobias [139,140]. This includes in vivo (real-life) and virtual reality exposure to phobic stimuli or situations. Virtual reality exposure therapy was introduced in the 1990s, and although it may have some treatment benefit, it has not been found to have strong efficacy [141]. A one-session exposure therapy treatment for specific phobias was pioneered more than 30 years ago with a suggested duration of two hours and was subsequently used to treat various specific phobia subtypes [142,143,144,145]. More recent studies suggest that one session does not always turn out to be adequate and that multiple sessions are generally more efficacious [140,146]. However, there may be some cases where the single-session approach is viable.

Pharmacotherapy is not a common treatment for specific phobias. However, studies have sought to supplement exposure therapy using pharmacologic approaches. One such intervention administers cortisol to augment exposure therapy due to its role in interfering with memory for fearful scenarios [147,148]. Although this treatment shows some efficacy, it does not seem to be particularly advantageous relative to exposure therapy alone. A second form of pharmacologic augmentation for exposure therapy, introduced more than 20 years ago, is the antibiotic D-cycloserine, which is thought to facilitate fear extinction due to its role as an N-methyl D-aspartate (NMDA) receptor agonist [149,150]. D-cycloserine has also been used to augment exposure therapy for social anxiety disorder, with studies suggesting that this antibiotic can produce a marginal benefit for treating specific phobias and social anxiety disorder when combined with exposure therapy [151]. However, while these studies mention that the antibiotic is of a low dosage, they do not mention that this marginal benefit needs to be traded off against the risk of accelerating antibiotic resistance, which is a pressing global public health challenge. Computational studies suggest that increasing administration of low doses of antibiotics (as these studies suggest doing in conjunction with exposure therapy) accelerates resistance [152,153].

Given that anxiety itself is an adaptive trait, anxiety disorders are better seen as poorly regulated defenses than as defects. As decades of lesion studies indicate, a lack of anxiety may also create non-trivial problems for individuals' lives. Low levels of anxiety are associated with higher mortality risk, and those who report greater worries about particular health problems are likely to seek medical care and take preventative or corrective action [39,103,104,105,106,107]. If some degree of anxiety is advantageous, then insufficient and excessive anxiety can both be considered maladaptive.