Anxiety is the “fever” of psychiatry. Fever is a completely nonspecific symptom, which happens in many illnesses, and frequently without any illness. Fever “disorders” are meaningless. Similarly, anxiety is a nonspecific symptom, which is quite common, sometimes related to other illnesses, and sometimes just part of one’s personality. Most anxiety “disorders” are not scientifically valid illnesses.
In other words, the key concept here is a diagnostic hierarchy, as explained here. If anxiety happens as part of other illnesses, then anxiety is not itself a separate illness. (Just as fever occurring with pneumonia is not a separate fever “disorder”).
Let’s examine the usual main set of anxiety diagnoses.
Obsessive-compulsive disorder (OCD) is the most legitimate anxiety disease. It can happen with medical illnesses, like streptococcal infection in childhood. It can be inherited genetically reasonably strongly. It can exist on its own, without any presence of other illnesses like manic-depressive illness or schizophrenia. On the other hand, it often occurs as part of mania or depression, and is not itself an independent illness in those settings. The problem of bipolar/OCD “comorbidity” is discussed further here. Schizophrenia can also cause OCD. In those settings, treating mood illness or schizophrenia will cure the OCD symptoms that are caused by those other diseases. When OCD happens by itself, serotonin reuptake inhibitor (SRI) treatment, sometimes high dose, is the standard treatment. Dopamine blocker (antipsychotic) augmentation of SRIs is also proven effective.
Panic disorder involves panic attacks, defined as sudden extreme anxiety that is unprovoked, happening “out of the blue.” These panic attacks often awaken someone from sleep (despite the absence of nightmares) or can occur in otherwise innocuous situations (while calmly watching television on a couch). They can be associated with “agoraphobia”, or fear of being crowds or in public spaces. In the vast majority of cases, panic attacks happen in the context of depression, sometimes mania, and sometimes psychosis. In those cases, there is no panic “disorder” that needs independent treatment with SRIs, but rather panic attack symptoms will improve once the underlying mood or psychotic illness is treated. It is often helpful to use short-term benzodiazepine medication (like lorazepam) for immediate relief of panic attack symptoms when they occur. In the small minority of cases where no mood illness is present, panic attacks can be managed with longer-term SRI treatment, but this is usually not necessary, since panic attacks are brief and intermittent. They do not continue indefinitely throughout life without some other underlying illness, and thus indefinite SRI treatment long-term is inappropriate.
Generalized anxiety disorder (GAD) was a term created in 1980 for DSM-III, without any prior history in centuries of psychiatric practice and research. It was created to replace the older term “neurotic depression”; this history is reviewed here. GAD was meant to reflect those patients who were seen as chronically “neurotic”, or anxious, all the time. In the last century, hundreds of well-validated scientific studies of personality have identified a personality trait of “neuroticism”, which occurs in all individuals. Those who are high on this anxiety personality trait will be chronically anxious to a relatively high degree. This is not an illness, but a personality trait. Again, anxiety that lasts a long time often is caused by another illness, like depression or bipolar illness or psychotic illnesses. In those cases, SRI treatment is not necessary, although it can be optional as a short to medium term treatment, along with benzodiazepines. The anxiety will not improve until the underlying mood or psychotic illness is treated. (SRIs should be avoided in these settings when bipolar illness is present, however, as explained here). When all those conditions are carefully ruled out, then long-term anxiety reflects being high on the personality trait of neuroticism. Since personality traits can only be modified at best, but never fully changed, long-term medication treatment will not have much effect in that case. Some use of low-dose SRIs, with appropriate explanation to patients that personality traits are being treated and not illnesses, may be modestly helpful.
Post-traumatic stress disorder (PTSD) is discussed in detail here. Again the concept of a diagnostic hierarchy is central, and this condition should not be diagnosed when apparent PTSD symptoms only happen in the context of depressive, manic, or psychotic episodes. However, people can have PTSD symptoms outside of those mood or psychotic states, in which case, like OCD, a legitimate anxiety condition can be present. Unlike OCD, which seems to be primarily genetic or sometimes infectious in etiology, PTSD requires, by definition, an environmental trauma. A central diagnostic problem these days, in our view, is that the phrase “trauma” has been extended so broadly as to include common daily occurrences, such as car accidents or the death of a pet. In its origins, PTSD related to war trauma or to childhood sexual abuse. We recommend limiting the diagnosis to those conditions of severe and uncommon trauma, rather than to the kinds of common life experiences that happen to the entire population. Further discussion of this diagnostic matter can be found here. In general, PTSD symptoms do not need to be treated pharmacologically, because there is no underlying disease to treat; modest symptomatic benefit has been reported with SRI, benzodiazepines, and dopamine blockers (antipsychotics). But such benefit is limited, and thus we recommend short to medium-treatment at the lowest doses possible, not routine long-term treatment, which has never been shown to be effective in randomized trials.