Attention deficit hyperactivity disorder (ADHD) may be the most common childhood DSM-IV diagnosis [1
], with 30% to 50% of cases persisting into adulthood [2
]. Research in pediatric ADHD has progressed to documenting differences in neuropsychological functioning at the level of the DSM-IV subtypes: Combined
(ADHD-CT, both inattentive and hyperactive-impulsive symptoms), Predominantly Inattentive
(ADHD-IA), and least common, Predominantly Hyperactive-Impulsive
(ADHD-HI). The psychiatric classification system for ADHD, with its various subtypes, has evolved substantially over the years. DSM-IV defined the disorder based on the best available research at the time, but there is debate as to whether the exiting subtypes provide an accurate description of the disorder. Some researchers [2
] hypothesize that the ADHD-IA subtype may constitute a different type of attention disorder compared to the ADHD-HI and ADHD-CT subtypes. Barkley [2
] is most associated with this view. He conceptualizes the ADHD-HI and ADHD-CT as, centrally, exhibiting deficits in behavioral inhibition that in turn lead to impairment in executive functioning and working memory (WM). According to this hypothesis the ADHD-IA subtype is a fundamentally distinct disorder with problems in attention arising from noninhibitory mechanisms. In contrast to this hypothesis, others note (e.g., [5
] that the ADHD-CT and ADHD-IA subtypes both share inattention as a common dimension and therefore the ADHD-CT and ADHD-IA should both be impaired on tests requiring attention, processing speed, vigilance and WM. Indeed, there is evidence that inattentive symptoms are most associated with neuropsychological impairment across the ADHD-CT and ADHD-IA subtypes [5
]. Thus, one could conclude that individuals with the ADHD-CT may be most impaired because they experience the combined symptoms of inattention plus inhibition.
The results for documenting neuropsychological differences according to DSM-IV nosology for ADHD subtypes have been ambiguous at best. Several pediatric studies found no or limited differences between ADHD-CT and ADHD-IA subtypes on a series of executive functioning measures [3
]. However, other researchers have detected subtype differences or demonstrated differential performance across subtypes when making comparisons to control groups [9
While the findings regarding subtype differences in children are mixed, even less is known about the presence of neuropsychological differences between subtypes among ADHD adults. Gansler et al. [11
] compared ADHD-HI to ADHD-IA adults and found that the types of executive functioning deficits exhibited differed across groups. The ADHD-HI group showed perseverative responding on the Wisconsin Card Sort Test, while the ADHD-IA group demonstrated difficulty in WM on the Auditory Consonant Trigrams (ACT) Test. Also, the ADHD-IA group emitted a higher rate of Continuous Performance Test (CPT) commission errors, a somewhat counterintuitive finding because one would tend to associate such errors with impulsivity rather than inattention.
Murphy and colleagues [12
] examined neuropsychological functioning in a large sample of ADHD adults. The authors found no subtype differences on tests measuring interference control, inattention, response inhibition, WM, and verbal-ideational fluency. They suggest that the disparity between their results and Gansler's [11
] may reflect 1) differences in the WM and response inhibition tasks used; or 2) problems related to the current DSM-IV diagnostic criteria in which adults who formerly may have been classified as ADHD-CT as children, are diagnosed with the ADHD-IA subtype as adults due to a reduction of the hyperactive/impulsive symptoms and thereby no longer meet criteria for the ADHD-CT subtype.
The inclusion of WM subtests in the battery of neuropsychological tests in previous studies in ADHD is quite logical given the growing body of literature suggesting WM impairments associated with the disorder [2
]. WM, or the ability to hold information in mind, manipulate it, and use it to guide behavior, is a key component of executive functioning. A recent meta-analysis of neuropsychological performance in adults with ADHD [19
] found effect sizes for tests measuring WM and verbal memory to be among the highest within 10 functional domains. Gallagher and Blader [20
] suggest that "stressful" WM tests may be more sensitive to detecting deficits in adults with ADHD. Specifically, they highlight studies using the Paced Auditory Serial Addition Task (PASAT) [21
]. This test requires active manipulation of stimuli in which an action is required on the presented stimuli. This is in contrast to short-term memory tasks where previously presented stimuli are simply reiterated. The PASAT already has been shown to discriminate between a sample of male ADHD-CT adults and controls on a behavioral and functional neuroanatomical level [23
]. A review of neuropsychological performance in adults with ADHD [17
] found the PASAT produced one of the largest effect sizes in comparisons between controls and ADHD adults. Yet, we do not know if the PASAT is useful for discriminating between subtypes. One reason the PASAT may be more sensitive to detecting subtype differences is its ability parametrically to assess processing speed capabilities. It does this by using varying interstimulus intervals (ISI) between stimulus presentations.
We expect that both ADHD subtypes will exhibit impaired processing speed given its relationship to inattention [5
], but that the ADHD-IA group will be more impaired by shorter ISIs reflective of greater processing speed difficulties given previous research [5
]. Similarly we expect both subtypes to produce the most common type of error on the PASAT [26
], failures to respond, or omission errors due to inattention. The number of consecutive correct answers, or Longest Sequence of Consecutive Correct Answers, is another measure of WM functioning on the PASAT that appears to be a sensitive measure of functioning in both ADHD [27
] and neurologically-impaired populations [28
]. We hypothesize that the ADHD-CT type will perform more poorly on this measure than the ADHD-IA type because this measure will be sensitive to errors reflective of response inhibition and inattention.
This study also tested for gender differences in WM performance. The ADHD literature tends not to find significant differences between adult males and females [31
] or boys and girls with ADHD [34
], but only a few studies [8
] have examined the effect of gender on executive functioning performance. There is, however, some suggestion of gender differences according to subtype on symptom comorbidity [37
], neuropsychological functioning [25
], and motor performance [38
Gender differences in WM paradigms are relatively sparse with most studies examining memory differences in relationship to gender, rather than explicitly studying WM performance. Recent brain imaging studies, however, have begun to explore gender differences in relationship to brain activation [39
] and found that gender can significantly affect brain activation. An fMRI study [41
] suggests that men and women may use different neural substrates and perhaps strategies to perform WM tasks. Speck et al. [41
] found that women performed more accurately but also more slowly in comparison to the male participants on WM paradigms. These data suggest that WM tasks with a speed component may be more influenced by gender. Furthermore, there were strong lateralization differences with male participants displaying symmetric activation or right brain dominance, versus female participants who were more likely to activate the left hemisphere during the WM tasks. The authors concluded that the differences in lateralization are due to either differences in the use of problem-solving strategies by gender or differences in the neuroanatomy used to solve the tasks. We decided to explore gender differences in this study given that there is also some evidence for lateralization differences in how individuals with ADHD perform WM tasks [24
] and thus, it is plausible that ADHD may interact with gender on executive functioning performance.
The objectives of this study are to test for differences in WM between adults with and without ADHD and between ADHD-CT and ADHD-IA adults. We hypothesize that the ADHD participants will exhibit WM deficits in contrast to healthy control (NC) participants, and that there will be subtype-specific WM errors. We hypothesize that gender will have a significant effect on WM tasks, particularly those that require a greater processing speed load. It is also expected that faster interstimulus interval (ISI) presentation rates on the PASAT will be more sensitive in detecting differences between the NC and ADHD groups due to demands in processing speed.