I agree with whomever that thought this new layout was "cumbersome" - it is surely so. It makes keeping track of points from previous posts more difficult than before, and it deleted my post while I was typing it!
And I apologize for my last post. Sometimes, I get in a rush and bored, so I don't proofread and leave some passages wanting for clarity.
I will attempt to clarify my first paragraph (if you want to call it that). A person with GAD or any other anxiety disorder is diagnosed so because of s/he is exhibiting a set of symptoms found in the medical criteria for that disorder. GAD doesn't have an extensive criteria and is little more than persistent, excessive worry given a particular situation or event. SAD is the same, except the anxiety stems from social situations (or thoughts pertaining to social situations), especially those involving strangers. The Diagnostic and Statistical Manual of Mental Disorders (4th Ed., Text Revision) does list as a requirement, however, that the *symptoms* (anxiety) are not cause by another mental disorder. So when I wrote the trend is going both ways (between high I.Q. and the lower anxiety group and high I.Q. and higher anxiety group, with both higher I.Q. being the trend to both extremes), I did so because of GAD is little more than persistent inappropriate and excessive worry, which means you have a group of otherwise normal people with excessive worry, and the two groups presumably don't differ in any important way other than their habitual (trait) anxiety levels.
SO you have a group of excessively worried (GAD, high I.Q. population), slightly worried (normal and GAD, low I.Q. population), and hardly worried (typical high I.Q. population) individuals, and the only factor which differentiates the groups is levels of habit (trait) anxiety. Application of a GAD label after this distinction has already been made is tautologous, and hardly relevant since, in this context, unless we assume that *official* diagnosis is (at least in part) responsible for the controverting data, further distinction seems to be misnomer, which distracts from accurate interpretation of the data.
Lo and behold - clarity! (A comforting friend.)
At this point, we are still left with the paradoxical issue of two distinct trends for the same trait - unless, that is, we admit there is some other variable affecting the data (like two separate groups :]).
In the link you provided (I did not read the attached study), it states that the study found that participants had "two or more" psychiatric disorders. I believe this might confound the data, making it much less reliable than a pure GAD-to-I.Q. study. Also, time is a considerable factor since these were childhood scores (I'm sure we can both point out the myriad of problems this fact alone might bring.) and the study is based on finding from the participants in adulthood. It is even possible that their low I.Q. caused them to make poor lifestyle decisions (like heavy drug use), which triggered their disorders, among a whole host of other possible explanations. This is besides the point, though - the point is that between the two studies, the first more clearly addresses the issue of current WMC (and possibly I.Q.) in persons with anxiety disorders and is probably more reliable for purposes of this discussion.
The study makes a distinction (Correct me if I'm wrong on this.) between "trait" and "state" anxiety. Trait anxiety being (of course) the more permanent trait and an integral part of the individual's personality; and state anxiety depending mostly on temporary conditions.
The study found that only under high state anxiety conditions did the trait high anxiety group perform as poorly (or well :]) as the normal population, low WM group, but under normal conditions, trait high anxiety individuals had higher visual WMC. I don't see this as a major sticking point since both groups perform more poorly under higher state anxiety conditions, and naturally, if it is true that higher WM capacity would imbue more protection against higher state stress, both lower trait anxiety in some individuals (normal, high I.Q.) and higher trait anxiety in others (SAD, high I.Q.) should be similar in how they perform under stressful and ideal conditions given their an increasingly (negative or positive) extreme position on the trait anxiety scale and its corresponding higher WMC benefits. In other words, low stress situation, both high I.Q. groups do well; high stress situation, both do more poorly with respect for stress severity and initial WMC capacity.
As far the physiological aspect goes - so far, evidence implicates an overactive limbic system, but I agree more information is needed to pinpoint an exact mechanism. In fact, I am not even so sure that I can even support the notion that it is the brain which *fully* determines the mind, and not the mind which partly determines the brain, especially since we don't fully understand the mind yet. But that is more a philosophical ramble, and I will admit, not unlike my previous musings on creativity, more of a passing thought than relevant. :)
Also, some individuals do well on normal I.Q. tests, some don't and do better on untimed, culture-fair tests. My guess is that hyperconcentration (probably based on salience-based systems) may be key here, which might indicate that, as to a probable lesser degree in the normal population, motivation is a big factor when it comes to real world application.