Introduction

Anxiety disorders (Table 1) are the most prevalent psychiatric disorders and are associated with a high burden of illness [8]. Women are twice as frequently affected as men. With a 12-month prevalence of 10.3%, specific phobias are the most common anxiety disorders [18], however, individuals suffering from isolated phobias rarely seek treatment. The second most common type is panic disorder with or without agoraphobia (PDA; 6.0%), followed by social anxiety disorder (SAD) (2.7%) and generalized anxiety disorder (GAD; 2.2%). There is no evidence that anxiety disorders have occurred more frequently in recent years or decades [21, 22]. These disorders often co-occur with other anxiety disorders, depression, somatoform, personality, and substance abuse disorders [19].

Table 1 Anxiety disorders: short description according to ICD-10 classification [32]

Current conceptualizations of anxiety disorders posit an interaction of specific genetic vulnerabilities which manifest in neurobiological alterations and environmental factors (including childhood adversity, stress, or trauma). An abundance of high-quality research addressed the neurobiological causes of anxiety disorders, including research in neuroimaging, genetics, neurochemistry (neurotransmitters such as serotonin, norepinephrine, dopamine or GABA, neuropeptides such as cholecystokinin, neurokinins, atrial natriuretic peptide, or oxytocin, the HPA axis, neurotrophic factors such as NGF and BDNF, immunology, CO2 hypersensitivity, neurophysiology and neurocognition) [5, 6]; however, at present, none of the putative biomarkers has proved sufficient and specific as a diagnostic tool for anxiety disorders.


A consensus panel of German experts started to develop the “S3 guideline” for anxiety disorders in the year 2008. The term S3 refers to the highest quality requirements as defined by systematic evidence search and a consensus statement [2]. The present paper introduces the first revision of the German Guidelines for Anxiety Disorders [7]. Since 2014, many new randomized controlled studies (RCTs) for the treatment of anxiety disorders have been published. Recommendations on newly introduced treatment strategies were also examined, including Internet interventions and virtual reality exposure therapy. The guideline revision was published in 2021 and is available online [4].

In Germany, costs of the treatment of anxiety disorders are reimbursed by the statutory health insurance providers. Drug treatment is fully reimbursed, and so are defined contingents of psychotherapy sessions. Psychotherapy is provided by psychotherapists, including certified physicians or psychologists.

Anxiety disorders are mostly treated on an outpatient basis. Indications for hospitalization include suicidality, chronic anxiety disorders unresponsive to standard outpatient treatments, or marked comorbidity, e.g., with major depression, personality disorders or substance abuse.

Methods

The guideline expert panel consisted of 35 specialists for psychiatry and psychosomatics/psychotherapy, general practitioners, psychologists, and other members from all relevant professional societies and patient self-help organizations (n = 29) involved in the treatment of anxiety disorders in Germany (for a list, see [4]).

Pre-existing international guidelines on the treatment of anxiety disorders were searched using the Guidelines International Network [15]. Since the first version, new widely recognized international guidelines have appeared or have been revised [1, 3, 20, 23, 26]. Recommendations in these guidelines were screened but the expert panel also performed its own research when discrepancies between existing guidelines were found, when certain subject areas were not adequately covered, or when new trials potentially altering the evidence level of a treatment approach were published since the publication of the reference guidelines. For the revision, randomized controlled studies (RCTs) on the treatment of anxiety disorders were searched which were published between 16/09/2013 and 20/06/2019. Literature search methodology followed the principles of the PRISMA Statement [24]. Inclusion criteria for RCTs were: original publication in a peer-reviewed journal; RCTs on the treatment of anxiety disorders in adults defined according to ICD or DSM (panic disorder/agoraphobia, generalized anxiety disorder, social phobia, or specific phobia); not exclusively subgroup analysis; use of a control group (i.e. for drug trials, a placebo or a validated reference drug; for psychotherapy trials, a waiting list, an active control or ‘psychological’ placebo, i.e., a supportive conversation with the patient, without use of specific psychotherapeutic techniques; treatment as usual (TAU) or another established form of psychotherapy as reference. Only commercially available drugs approved by the EMA (European Medicines Agency) for an anxiety indication were included in the recommendations.

In the original version, a total of 1299 records were screened by titles and abstracts. For the revised version, 638 newly found records identified through database and hand search were screened by title and abstract (see PRISMA statement; Fig. 1). Records were eliminated for the following reasons: only meeting abstract, letter, review, meta-analysis, pooled study, double publication, secondary analyses, drug or herbal preparation not licensed, no DSM/ICD anxiety disorder diagnosis, patients with mixed anxiety disorders, only comorbid patients, treatment of subgroups (e.g., only students; exception: studies with patients ≥ 65 years were analysed separately), children/adolescents < 18; treatment of non-responders/adjunctive treatment of non-responders, augmentation, or drug combination treatments, sample size of any of the arms in a study of < 10 at baseline. After reading full texts, additional articles were eliminated because of qualitative shortcomings. The quality of the RCTs was evaluated following the SIGN Statement criteria [30], including blinding of investigators, equal conditions for treatment and control groups, use of standard assessment instruments, intent-to-treat analysis, adequate statistical power and correctness of statistical evaluation. Methodological flaws led to the exclusion of trials or to downgrading of their evidence level. Common reasons for downgrading the evidence grade included small sample size (particularly in non-inferiority comparisons) and inappropriate statistical methods. If a primary efficacy measure was not stated, results were re-evaluated using a Bonferroni-Holm correction [16] for multiple testing. Eventually, 92 new records fulfilling the inclusion criteria were included in the analysis.

Fig. 1
figure 1

PRISMA statement [24]. Newly included studies since 16/09/2013

The guideline protocol also allowed including results from open studies, case series, and single case reports whenever there were no sufficient RCTs available for certain decisions. However, there was no actual case in which a decision about an evidence level had to be based on such studies. While the evidence categories were based exclusively on the efficacy of the various treatments studied, the recommendation grades also took risks into account, e.g., adverse effects of drugs (Table 2).

Table 2 Evidence levels and recommendations grades

According to the principles of the guideline, the level of evidence has been determined on the basis of available original studies. Only in the case of contradictory results, meta-analyses have been used.

All recommendations were agreed by voting of the group members. All participating societies had one vote, and all statements issued by the committee were only accepted if they received at least 75% of all votes cast. Moderation of the expert consensus meetings was done by the Association of Scientific Medical Societies (Arbeitsgemeinschaft der wissenschaftlichen medizinischen Fachgesellschaften, AWMF, www.awmf.org), a German overhead organization managing official guidelines for all fields of medicine.

Because of the large number of study reports included in the guideline (n = 606), references are not provided for every statement in this article; rather, the reader is referred to the full guideline text [4], in German only).

Results

Diagnosis

For the diagnosis of anxiety disorders diagnosed using ICD or DSM criteria should be used. For Germany, the International Classification of Diseases [32] in its German modification (ICD-10 GM) is the official diagnostic system. In primary care, anxiety disorders often go unrecognized [33]. Before diagnosing an anxiety disorder, other mental disorders, such as other anxiety disorders, major depression, personality and somatoform disorders, as well as physical illnesses such as coronary heart disease, pulmonary disease, and others have to be excluded.

Treatment

Treatment is indicated when a patient fulfills criteria for an anxiety disorder as defined by ICD or DSM, shows marked distress or suffers from the sequelae resulting from the disorder (e.g., suicidality, secondary depression or substance abuse).

Patients should be educated about their diagnosis, the possible backgrounds and the principles of action of the available treatment approaches.

Anxiety disorders can be treated with psychotherapy, drug treatment and other interventions (Table 3). It is mandatory by law for treating psychologists and physicians to inform patients about the risks and benefits of available treatments and possible alternatives which are more effective or better tolerated. The treatment plan should be chosen after careful consideration of individual factors, e.g., the patient’s preference, previous successful or unsuccessful treatment attempts, illness severity, comorbidities including substance abuse or suicide risk, availability of treatment methods, costs, waiting periods and others. All interventions should be performed on the basis of an empathic and reliable therapeutic relationship. It is suggested that treatment success should be monitored with standard ratings scales.

Table 3 German Guideline for the treatment of anxiety disorders: treatment recommendations for anxiety disorders in adults

Psychotherapy

Of all psychological interventions, cognitive behavioral therapy (CBT) has by far the best body of evidence. In the case of phobic disorders, confronting the patients with their feared situations in exposure sessions is a crucial ingredient of the therapy. Group CBT has also been studied in RCTs, but there is not sufficient evidence to conclude that it is as effective as individual treatment. For patients with SAD, however, it seems reasonable to conduct self-assurance training in groups; thus, psychotherapy for SAD should include both personal and group therapy sessions.

For specific phobias, only studies with behavioural therapy exist, which should be performed as exposure treatment.

In comparison to CBT, the evidence for the efficacy of psychodynamic therapy (PDT) is weaker. RCTs on PDT were markedly fewer in number, and lower in quality, than those on CBT, and some comparison studies indicated superiority of CBT over PDT. According to the guideline, patients with PDA, GAD, or SAD should be offered PDT only if CBT was shown to be ineffective or is unavailable, or if the adequately informed patient expresses a preference for PDT.

Systemic therapy has recently become eligible for reimbursement by the German health care system. However, the few available studies had serious methodological flaws and inconsistent efficacy results; therefore, this treatment modality only received the “0” recommendation, indicating that it only should only be offered if standard treatments have failed or are not available.

Most efficacy studies used treatment manuals guiding the intervention strategy. Therefore, to maintain quality standards, it is recommended that psychotherapy in daily routine practice should also be manualized.

The guideline committee did not provide recommendations regarding the necessary duration or number of psychotherapy sessions due to the lack of reliable data. Published randomized studies on psychological therapies had an average study duration of 12.4 weeks [9], therefore, little is known about the additional benefits of psychotherapies that have 25, 50 or more sessions. An analysis performed by the guideline committee revealed that there is not sufficient evidence that longer therapies are more effective than shorter ones (for details, see original guideline).

Specific phobias can be treated in only a few exposure sessions. Most studies for the treatment of specific phobias (including fear of spiders, heights, or flying) had only one session lasting between 1 and 3 h, demonstrating that such short interventions are effective.

Internet-based psychological interventions (IPI)

In the recent years, numerous studies have investigated internet-based psychological interventions (IPI), most of which were based on CBT approaches. These usually involve no personal contact or only minimal E-mail or telephone contact with the study staff.

In most RCTs, IPIs were more effective than a waitlist control. However, evidence showing that IPIs are as effective as individual CBT with face-to-face contact is insufficient. Moreover, the efficacy results of IPI studies are mostly based on non-blinded self-ratings, which may lead to an overestimation of effects sizes due to expectation effects [11]. Therefore, the committee decided that IPIs should not be used as monotherapy but can be used to bridge a waiting period until face-to-face psychotherapy is available, or as an add-on self-help measure accompanying standard psychotherapy or medication treatment.

Virtual and augmented reality exposure treatment

Virtual reality (VR) and augmented reality (AR) technologies have been introduced in the treatment of phobias. In AR exposure therapy, virtual elements are merged into the view of the physical world. Thus, the experience is more authentic, and costs are lower, because it is not necessary to program the complete virtual environment.

For PDA, there are not enough studies to support the use of VR. However, for SAD, VR exposure therapy can be used as an add-on self-help measure. For specific phobias (fear of spiders, heights, or flying), VR exposure therapy can be used when in vivo exposure is not available.

Pharmacotherapy

A large database of RCTs on the efficacy of medications for PDA, GAD, and SAD is available. Only for specific phobias, drug studies are scarce, and behavioral treatments should be preferred.

First-line drugs for anxiety disorders include the selective serotonin reuptake inhibitors (SSRI) and the serotonin-norepinephrine reuptake inhibitors (SNRI) (Table 3). Drug side effects are listed in Table 4. For panic disorder, the tricyclic antidepressant clomipramine may be a second-line option. The drug is as effective as the SSRIs and SNRIs but has more adverse effects. For GAD, the calcium modulator pregabalin was shown to be effective, but there are concerns about cases of overdoses and withdrawal syndromes associated with the drug. Therefore, pregabalin should not be used as the first option and should not be given to patients with a history of substance abuse.

Table 4 Adverse effects of anti-anxiety drugs

The guideline group advised against the use of benzodiazepines because of their abuse potential, despite their high effectiveness in treating anxiety. However, in exceptional cases (e.g., severe cardiac disease, suicidality, contraindications for standard medications, and other conditions), benzodiazepines can be used for a limited time period after their risks and benefits have been weighed carefully.

Other treatment approaches which were not considered as first- or second-line drugs due to lack of consistent evidence from RCTs, include the tricyclic opipramol (a drug which is only licensed in Germany and a few other countries), buspirone for GAD, and moclobemide for SAD.

Only in rare cases, emergency drug treatment is an option in treating acute panic attacks. Mostly, talking calmly with the patient and explaining that the attack is not due to a serious medical condition is sufficient. Only in severe cases, lorazepam 1.0–2.5 mg melting tablets may be given p.r.n.

Patients have to be informed about possible adverse effects, interactions, contraindications, and warnings, following the current summary of product characteristic.

Patients starting treatment with antidepressants (SSRIs, SNRIs, and TCAs) should be advised that these drugs generally take effect after a latency period of about 2 weeks (range 1–6 weeks). During the first weeks of treatment, some patients have a tendency to discontinue treatment with SSRIs and SNRIs due to initial jitteriness and nervousness. Compliance can be improved by informing patients about these potential adverse effects and by starting at half of the usually recommended dose. The drugs should be given in the morning or at midday to avoid insomnia which may occur during the first treatment weeks. In patients with hepatic impairment, a dosage adjustment or use of medications with primarily renal clearance (e.g., pregabalin) may be required.

The recommended medications have been studied in placebo-controlled relapse prevention studies mostly lasting from 6 to 12 month, as this is a requirement by the EMA for new drug application. According to these studies, it is advisable to continue drug treatment for 6 to 12 months after remission has occurred. In patients with a long history of recurrent and/or severe anxiety symptoms, longer treatments may be necessary.

To avoid discontinuation syndromes, the dose should be slowly tapered at treatment termination.

In patients who are unresponsive to medications, the addition of psychotherapy is generally recommended. If there is no response to the first drug after 4–6 weeks of treatment, a second standard drug should be given instead. In case of a partial response after 4–6 weeks, raising the dose can be considered first. Table 5 contains a stepwise plan for treatment options in case of drug inefficacy or intolerance. When medications are offered off-label to patients with treatment-unresponsive anxiety disorders which were effective in RCTs but were not licensed for the specific disorder (e.g. quetiapine or agomelatine), medicolegal issues have to be considered.

Table 5 Stepwise plan for drug treatment if the initial standard drug treatment was not effective or poorly tolerated (modified from [12])

Combining psychotherapy and medication

The guideline does not recommend to start with psychotherapy before considering pharmacotherapy or vice versa, as there is no evidence from clinical studies justifying such a stepwise approach. As more data favor the combination of both modalities than not, both can be started at the same time. If response to psychotherapy or pharmacotherapy is insufficient, treatment should be switched to the other modality. One meta-analysis found higher effect sizes for medications than for psychological therapies and no evidence that gains with psychotherapy can be weakened by concomitant drug treatment [9].

Treatment of anxiety disorders in older patients

Anxiety disorders are most common in the age between 30 and 50. The average age in clinical studies was 37 years for PDA, 41 for GAD and 35 for SAD [10]. Therefore, anxiety disorders are less common in patients over 65 years, with the exception of GAD, which may be common in the elderly (Table 6).

Table 6 New recommendations in version 2 of the German Guideline for anxiety disorders

Only few studies for the treatment of GAD have been performed with older patients. Some studies in GAD suggest that CBT is less effective than in adults aged 18–65. Studies showed that duloxetine, venlafaxine, pregabalin, and quetiapine are effective in patients over 65 years. In older patients, an increased sensitivity to drug side effects and interactions has to be considered, including anticholinergic effects, risk of orthostatic hypotension and electrocardiogram changes, risk of falling, and paradoxical reactions to benzodiazepines.

Pregnancy and breastfeeding

When pregnant women suffer from an anxiety disorder, the risk of an untreated anxiety disorder must be weighed against the risk of damage to the unborn child as a result of drug treatment. A large study suggested no substantial increase in the risk of cardiac malformations attributable to use of antidepressants during the first 3 months of pregnancy [17]. Antidepressants have been associated with increased risk of spontaneous abortions, stillbirths, early deliveries, respiratory distress, and endocrine and metabolic dysfunctions [27]. However, the current evidence suggests that the use of many antidepressants, especially the SSRIs, is favorable compared to exposing the mother to the risks of untreated depression or anxiety disorders [25, 28].

Likewise, a careful assessment of the risk/benefit balance has to be done when a mother is breastfeeding. In such cases, CBT should be considered as an alternative to medication treatment.

Other treatment options

Exercise (running 5 km three times a week) was shown to be effective in treating PDA [13, 31]. However, exercise was less effective than clomipramine and no more effective than a control condition, relaxation. Therefore, exercise can only be recommended as add-on treatment to standard treatments. For GAD, the only available study could not reliably demonstrate superiority of weightlifting or cycling of a waitlist control condition. According to a meta-analysis of RCTs with patients with anxiety and related disorders such as OCD or PTSD [29], exercise had a small but statistically significant effect compared to control conditions.

Although controlled studies on the usefulness of self-help groups are lacking, it was an expert consensus that patients should be encouraged to participate in such activities if appropriate. It may also be useful to integrate the family members of the affected patients into the treatment plan.

Various other treatments that have been studied in RCTs. However, the guideline committee did not find sufficient evidence to recommend the following treatments: Client-Centered Therapy, Interpersonal Therapy, Progressive Muscle Relaxation, Applied Relaxation, Eye Movement Desensitization and Reprocessing, Music/Dance/Art Therapy, Yoga, beta-blockers, phytotherapeutics, homeopathic formulations, and repetitive magnetic stimulation (for references, see [4]).

New recommendations in version 2 of the guideline are summarized in Table 6.

Discussion

Since the publication of the first version of the German guideline for the treatment of anxiety disorders for adults [7], no fundamental changes of the treatment recommendations have been formulated. Treatment with SSRIs/SNRIs and CBT is still the mainstay for anxiety disorders.

Among newer developments, the number of clinical studies examining Internet psychological interventions (IPIs) for anxiety disorders has surpassed the number of studies on psychotherapy with face-to-face contact in the recent years, perhaps because such trials are much easier to conduct and less expensive than studies with face-to-face psychotherapy. At present, these treatment programs have some advantages, because personal contacts can be avoided during the COVID-19 pandemic. Moreover, IPIs are less expensive, save therapist time, require less organizational efforts, save travel time, and can be used at any time of the day. However, a closer look at the quality of the studies is warranted [11]. Even if the IPI programs are very sophisticated and can be individualized for special anxiety manifestations of each participant, it is hard to believe that computer programs can adequately address the unique interpersonal, social, medical or occupational problems of the participating individuals at the same level as ‘real’ therapists.

In most IPI studies, participants could shortly contact study staff by E-mail or telephone. However, in many of the published trials, the ‘therapists’ at the other end were psychology students with incomplete psychotherapy education. In some studies, diagnoses were only made using online diagnostic tests, or by psychology students using structured interviews [11]. Only in 15% of the studies, diagnoses were made by psychiatrists or psychologists in a personal interview. In almost all studies, the efficacy results were based on unblinded self-ratings. On average, two thirds of the participants in IPI studies had an academic background and recruitment has mostly relied on advertisements in the internet or other media, which may affect the generalizability of the study results. Altogether, the effect sizes of IPIs obtained in these studies may have been overestimated.

Since the first version of the guideline, more studies have been published with the new technologies Virtual Reality (VR) and Augmented Reality (AR) which are used in the treatment of specific phobias such as the fear of spiders, heights, or flying. As there is no sufficient evidence yet that these methods are as effective as in vivo exposure therapy, the expert panel recommends to use only these technologies as adjunctive treatment to standard behavior therapy or in cases in which alternative treatments are not available. However, as research in this field is growing constantly [14] and the programs increasingly provide a realistic, true to life experience, especially with the introduction of AR, these methods may play an important role in the treatment of phobias in the future.

Since the 2014 version of the guideline, no new medications for anxiety disorders have emerged. Therefore, there was no change in the recommendations for psychopharmacological treatment. Although there are many unmet needs in the pharmacologic treatment, no putative novel anxiolytic agents will be available in the near future [34]. One reason why pharmaceutical companies are increasingly reluctant to develop new drugs is that the patents of all current available anti-anxiety drugs have expired and very inexpensive generics are produced in China or India because of the lower production costs, while, at the same time, the costs of developing new drugs have increased dramatically.

The applicability of the present guideline is not only restricted to the special situation in Germany. It may also be useful for developing evidence-based treatment plans for adults with anxiety disorders in other countries, as it is based on a thorough world-wide evaluation of RCTs. Today, most drugs are developed for the international market, and the major principles of psychotherapy are not substantially different in the world-wide perspective.