Recently one of the authors of this paper gave a talk about ADD in which he cited this study to support the validity of adult ADD. This paper is an analysis of the National Comorbidity Survey (NCS) epidemiological study. In the NCS project, researchers knocked on doors all across the United States to identify the frequency of various DSM-based psychiatric diagnoses. This is the study that is used for basic citation of frequency of many diagnoses. In this analysis, the NCS lead researchers collaborated with ADD experts to identify the frequency of ADD diagnosis in the young to middle age adult population of the US, based on a sample of 3197 subjects, aged 18-44.
Overall, they found that ADD was diagnosable in 3% of this general population sample. This was defined as persistence of childhood ADD. To clarify, 8.1% of the NCS adult sample was identified as having been diagnosable retrospectively with childhood ADD. Of this group, ADD persisted into adulthood in 36.3%. So the overall adult prevalence rate was 2.94%.
This study is the basis for many claims. One is the idea that ADD is common in adults, happening in about 3% of the general population. Further, it supports the claim that ADD persists in a substantial portion of children into adulthood, namely about one-third of persons.
Note that if we take these results at face value, one could turn the interpretation around and conclude that most children - namely 2/3 - with ADD will not have persistence into adulthood. These days, many clinicians practice as if the majority, or almost all, cases of ADD in children or adolescents will persist into adulthood. It is routine practice to continue amphetamines for childhood ADD into young adulthood. Once the growing child enters college, amphetamines are not stopped. They often are continued well past college, into the decade of the 20s. Yet, this study, so often cited by supporters of the adult ADD diagnosis, shows that the reverse should be the case: In 2/3 of children and adolescents, amphetamines should be stopped before they reach the age of 18, if such treatment is based on the claim that they will continue to meet the diagnostic definition of adult ADD.
The most important finding in this NCS analysis, though, is one that the authors don’t emphasize. The researchers looked at “comorbidities” of adult ADD, based on the careful identification of other DSM diagnoses, which was the main purpose of the NCS epidemiological study.
In this analysis of comorbidities, the NCS study found that persons diagnosable with adult ADD could also be diagnosed at the same time with the following diagnoses in the percentages provided in Table 3 of the paper (page 1447):
The absolute frequency of these diagnoses was 15.0% for MDD, 10.4% for bipolar disorder, and 7.6% for dysthymia.
We will put off the discussion of dysthymia for now, for later discussion in this article on the subject of mood temperaments. It is notable, though, that over one-half of all cases of apparent adult ADD involved the mood temperament of dysthymia. Since depressive symptoms include poor concentration, this “comorbidity” becomes a major logical problem. How can you say that people have a chronic cognitive condition of poor attention (the claim of adult ADD) when they have another chronic cognitive condition of poor attention (dysthymia)?
Turning to other conditions among the adult ADD subjects, there were high rates of other diagnoses, as well, especially anxiety disorders (49.9%, post-traumatic stress disorder, PTSD; 34.4% generalized anxiety disorder, GAD) and substance abuse (40.2% alcohol abuse, 31.4% other drug abuse).
Now one can certainly claim, as the authors do, that these are “associations” with adult ADD; namely, that if you have the adult ADD diagnosis, you’ll also run into other problems, such as abusing drugs or alcohol. This is not unreasonable; the same interpretation exists with mood illness, for instance, where many patients abuse substances as a result of having mood illness (often as self-medication).
In the list of comorbidities, many overlap with each other; for instance, one can have PTSD and also meet criteria for GAD and dysthymia. But two diagnoses are exclusive; you cannot make one if the other is present: MDD and bipolar disorder.
If we look at those two exclusive mood diagnoses, MDD was present in about 39% and bipolar disorder was present in about 45%. Since they are mutually exclusive, this means that a total of 84% of all subjects with adult ADD also met criteria for either MDD or bipolar disorder.
Now, here is the key question: Is this a mere association?
There are three possibilities:
The third possibility is biologically implausible. Nature doesn't tend to give two diseases every time to patients who are unlucky enough to get one disease. You don't get diabetes 84% of the time that you get diagnosed with cancer.
That leaves the other two causal possibilities: either the mood illnesses cause apparent “adult ADD”, or adult ADD causes the apparent mood illnesses.
The last claim is made by ADD experts, when pressed on this issue. Adult ADD is such a terrible experience, they claim, that patients become depressed a lot, hence the 39% prevalence of apparent MDD. They also become quite anxious about half the time, and have terrible life experiences, hence the 34-50% prevalence of GAD and PTSD.
Adult ADD explains almost everything - except…
There is a deep flaw in this rationale: ADD proponents cannot claim that adult ADD causes mania. They cannot explain away the 45% prevalence of bipolar illness, which means that those subjects experienced manic or hypomanic episodes. There is no biological or clinical rationale that can claim that adult ADD causes manic or hypomanic episodes.
This leaves only one other possibility. Even if interpreted mostly charitably, about one-half of apparent cases of adult ADD likely represent manic or hypomanic episodes, i.e., bipolar illness.
It is well known that depression and anxiety both are mental states that impair concentration and cognitive function. Thus, if we allow that at least some of those apparent cases of ADD were caused by anxiety or depressive illnesses, as opposed to the reverse, then we can make the following claim: Almost half of all cases apparent adult ADD are caused by bipolar illness, and another subgroup are caused by depressive and anxiety illnesses.
This analysis would lead to a conservative interpretation that the majority of cases of claimed adult ADD are caused by either bipolar or depressive or anxiety illnesses.
In other words, the 3% prevalence rate of adult ADD is not what it seems. Independent adult ADD, not attributable to another psychiatric cause, would be a minority of that 3% prevalence, perhaps about 1/3 or so, leading to about a 1% true adult ADD prevalence rate.
This 1% or so rate would be consistent with the concept of an extreme of a normal trait. As noted in the special article, attention is a normal cognitive trait, which, like most physical traits, is distributed on a statistical normal curve. This means that 95% of observations occur within 2 standard deviations in either direction from the 50th percentile. In other words 2.5% of the population will be at one extreme or the other. 1% would be observed at over 2 standard deviations from the mean. This observation would be expected with any normal trait; it doesn't reflect a disease process necessarily.
One could take another view. If we accept that anxiety and depressive states cause inattention and cognitive impairment, and we consider those conditions as primary to apparent ADD, then we could say that almost all cases of apparent adult ADD reflect either bipolar illness, depression, or anxiety conditions.
In this latter interpretation, adult ADD more or less disappears as a concept.
When the PL editor raised the bipolar “comorbidity” aspect of the NCS study with one of the NCS authors recently, the NCS author wasn't even willing to admit the 45% comorbidity rate. This reaction is an example of how human beings ignore data that conflict with their beliefs.