I quit reading after the author misrepresented the symptoms of Generalized Anxiety Disorder, presenting only the first two specific criteria under C, and not the full list. This is a deceitful way to play to the lay person’s sense that disorder criteria are constructed to be overly inclusive, to pathologize normative behaviors.
An article I read last year featured quotes from purported psychologists saying that oppositional defiant disorder is just a child saying no to his or her parent. The article also described one of the symptoms as “Actively refuses to comply with majority’s requests.” This is a complete bullshit statement. The symptom is “often actively refuses to comply with adults requests or rules.” The misrepresentation turned the symptom from an undesirable behavior into sort of a desirable, non-conformist rebel kind of behavior.
The real question is how well these symptoms work empirically. If they were pathologizing normative behavior, when the symptoms are used they should identify about 50% of the population.
If they are not identifying valid conditions, then people who meet the criteria at one point should be no more likely than other people to meet criteria again later in their life.
If they are not identifying valid conditions but are instead pathologizing normative behavior, the people who are identified as meeting criteria should have the same kind of outcomes as anyone else. Their outcomes should be normative ones, right?
What you see when the symptoms are empirically tested is that in fact, only small proportions of the population are identified with a given disorder, which may range from something like 0.5% for a disorder like selective mutism or 1% for schizophrenia, to 6 or 7% for ADHD or 9% for oppositional defiant disorder. An estimate for the prevalence of depression in a single year is about 6 to 7%. So it is hard to argue that the criteria identify normative groups of people.
When criteria are met for a particular disorder, those people are much more likely than the average person to meet criteria again when measured later in time. This is evidence for the reliability of the criteria, and it is further evidence that the disorder criteria are not haphazard collections of items that people in general are likely to meet at any given time.
And in terms of outcomes, people who meet criteria for a disorder are much more likely to experience poorer outcomes than people who do not. This includes higher rates of other psychopathology, poor functioning in social, occupational and other settings, poor health outcomes, and higher rates of morbidity and mortality. These outcomes are also to some extent disorder specific. For instance, depression places individuals at higher risk for suicidality than anxiety disorders. Individuals with anxiety disorders are, in general, at lower risk for antisocial or aggressive behavior. People with conduct disorder are at greater risk to have a criminal record, to engage in aggressive and antisocial behavior and to show poor functioning in multiple settings than others. People with ADHD are at higher risk for vehicle accidents than others.
Additionally, factor analyses tell us something about the nature of the construction of disorder symptoms. A factor analysis is a way of looking at a bunch of items, determining how much each one is associated with each of the other items in the data set, and identifying whether there are some groups of items that tend to be associated with one another in a different way from other groups of items.
Factor analyses have shown, across researchers and data sets, that the symptoms tend to align with one another in ways that are highly concordant with the diagnostic criteria. This helps to further suggest that the disorder constructs are not willy-nilly amalgams of items.