A Chinese study team performed a systematic search of peer-reviewed journal literature to identify randomized controlled trials (RCTs) examining the efficacy of cognitive training as a treatment for youths with ADHD.
Seventeen RCTs with a combined total of 1,075 participants met standards for inclusion in a series of meta-analyses. Seven RCTs used waitlist controls, seven used placebo training, two used treatment-as-usual, and one used active knowledge training. Participants were unmediated in four RCTs, with varying proportions of medicated participants in the remaining thirteen.
A meta-analysis of 15 RCTs, with a combined 789 participants, assessed changes in inattention symptoms following treatment, as rated by parents or clinicians. It found a small-to-medium effect size improvement in symptoms of inattention. There was no indication of publication bias, but between-study heterogeneity was very high.
But that gain vanished altogether when combining only the six RCTs that were blinded, meaning the symptom evaluators had no idea which participants had received cognitive treatment and which participants had not. There was zero difference between the treatment and control groups. Significantly, between-study heterogeneity also diminished markedly, becoming low to moderate.
A second meta-analysis, of 15 RCTs with a combined 723 participants, assessed changes in hyperactivity/impulsivity symptoms following treatment, as rated by parents or clinicians. It found no significant difference between participants who received cognitive training and controls. There was no sign of publication bias, and between-study heterogeneity was moderate-to-high.
The three remaining meta-analyses looked for improvements in executive functions, using the Behavior Rating Inventory of Executive Function (BRIEF).
A meta-analysis of 13 RCTs, with a combined 748 participants, found a small-to-medium effect size improvement in the global executive composite index of BRIEF, as evaluated by parents. There was no sign of publication bias, and between-study heterogeneity was moderate-to-high.
But that improvement again disappeared altogether when considering only the five RCTs that were blinded. Between-study heterogeneity also became insignificant.
A meta-analysis of 6 RCTs with 401 participants found no significant improvement in the behavioral regulation index of BRIEF. Heterogeneity was negligible.
Finally, a meta-analysis of 7 RCTs with 463 participants also found no significant improvement in the metacognition index of BRIEF. In this case, between-study heterogeneity was high.
While acknowledging that “when analyses were set in blinded measures, effect sizes were not statistically significant,†the author nevertheless concluded, “In summary, multiple cognitive training alleviates the presentation of inattention and improves general executive function behaviors in children with ADHD.†This suggests an underlying bias on the part of the study team in favor of treatment even when not supported by best (i.e., blinded) methodological practices.
A Chinese study team performed a systematic search of peer-reviewed journal literature to identify randomized controlled trials (RCTs) examining the efficacy of cognitive training as a treatment for youths with ADHD.
Seventeen RCTs with a combined total of 1,075 participants met standards for inclusion in a series of meta-analyses. Seven RCTs used waitlist controls, seven used placebo training, two used treatment-as-usual, and one used active knowledge training. Participants were unmediated in four RCTs, with varying proportions of medicated participants in the remaining thirteen.
A meta-analysis of 15 RCTs, with a combined 789 participants, assessed changes in inattention symptoms following treatment, as rated by parents or clinicians. It found a small-to-medium effect size improvement in symptoms of inattention. There was no indication of publication bias, but between-study heterogeneity was very high.
But that gain vanished altogether when combining only the six RCTs that were blinded, meaning the symptom evaluators had no idea which participants had received cognitive treatment and which participants had not. There was zero difference between the treatment and control groups. Significantly, between-study heterogeneity also diminished markedly, becoming low to moderate.
A second meta-analysis, of 15 RCTs with a combined 723 participants, assessed changes in hyperactivity/impulsivity symptoms following treatment, as rated by parents or clinicians. It found no significant difference between participants who received cognitive training and controls. There was no sign of publication bias, and between-study heterogeneity was moderate-to-high.
The three remaining meta-analyses looked for improvements in executive functions, using the Behavior Rating Inventory of Executive Function (BRIEF).
A meta-analysis of 13 RCTs, with a combined 748 participants, found a small-to-medium effect size improvement in the global executive composite index of BRIEF, as evaluated by parents. There was no sign of publication bias, and between-study heterogeneity was moderate-to-high.
But that improvement again disappeared altogether when considering only the five RCTs that were blinded. Between-study heterogeneity also became insignificant.
A meta-analysis of 6 RCTs with 401 participants found no significant improvement in the behavioral regulation index of BRIEF. Heterogeneity was negligible.
Finally, a meta-analysis of 7 RCTs with 463 participants also found no significant improvement in the metacognition index of BRIEF. In this case, between-study heterogeneity was high.
While acknowledging that “when analyses were set in blinded measures, effect sizes were not statistically significant,†the author nevertheless concluded, “In summary, multiple cognitive training alleviates the presentation of inattention and improves general executive function behaviors in children with ADHD.†This suggests an underlying bias on the part of the study team in favor of treatment even when not supported by best (i.e., blinded) methodological practices.
Sleep disorders are one of the most commonly self-reported comorbidities of adults with ADHD, affecting 50 to 70 percent of them. A team of British researchers set out to see whether this association could be further confirmed with objective sleep measures, using cognitive function tests and electroencephalography (EEG).
Measured as theta/beta ratio, EEG slowing is a widely used indicator in ADHD research. While it occurs normally in non-ADHD adults at the conclusion of a day, during the day it signals excessive sleepiness, whether from obstructive sleep apnea or neurodegenerative and neurodevelopmental disorders. Coffee reverses EEG slowing, as do ADHD stimulant medications.
Study participants were either on stable treatment with ADHD medication (stimulant or non-stimulant medication) or on no medication. Participants had to refrain from taking any stimulant medications for at least 48 hours prior to taking the tests. Persons with IQ below 80 or with recurrent depression or undergoing a depressive episode were excluded.
The team administered a cognitive function test, The Sustained Attention to Response Task (SART). Observers rated on-task sleepiness using videos from the cognitive testing sessions. They wired participants for EEG monitoring.
Observer-rated sleepiness was found to be moderately higher in the ADHD group than in controls. Although sleep quality was slightly lower in the sleepy group than in the ADHD group, and symptom severity slightly greater in the ADHD group than the sleepy group, neither difference was statistically significant, indicating extensive overlap.
Omission errors in the SART were strongly correlated with sleepiness level, and the strength of this correlation was independent of ADHD symptom severity. EEG slowing in all regions of the brain was more than 50 percent higher in the ADHD group than in the control group and was highest in the frontal cortex.
Treating the sleepy group as a third group, EEG slowing was highest for the ADHD group, followed closely by the sleepy group, and more distantly by the neurotypical group. The gaps between the ADHD and sleepy groups on the one hand, and the neurotypical group on the other, were both large and statistically significant, whereas the gap between the ADHD and sleepy groups was not. EEG slowing was both a significant predictor of ADHD and of ADHD symptom severity.
The authors concluded, “These findings indicate that the cognitive performance deficits routinely attributed to ADHD … are largely due to on-task sleepiness and not exclusively due to ADHD symptom severity. … we would like to propose a simple working hypothesis that daytime sleepiness plays a major role in cognitive functioning of adults with ADHD. … As adults with ADHD are more severely sleep deprived compared to neurotypical control subjects and are more vulnerable to sleep deprivation, in various neurocognitive tasks they should manifest larger sleepiness-related reductions in cognitive performance. … One clear testable prediction of the working hypothesis would be that carefully controlling for sleepiness, time of day, and/or individual circadian rhythms would result in a substantial reduction in the neurocognitive deficits in replications of classic ADHD studies.
Blog Post Submission for APSARD Blog
ADHD continues to be a significant and difficult challenge in the collegiate world. The symptoms of the disorder directly impact a person’s ability to manage the demands of college. Matriculating students are expected to rapidly obtain and deploy many self-management skills. Increased academic expectations demand a greater capacity for sustained attention. And the evolving social milieu can tax the emotion-regulation and social cognition of those with ADHD.
Having seen our patients struggle, the Association for Collegiate Psychiatry decided to submit a workshop for presentation at the 2019 APA meeting in San Francisco. While developing the presentation we discovered a wealth of recent ‘young adult’ follow-up data from longitudinal studies.1 Without exception, the studys’ findings reflected a significant decrease in functional outcomes across multiple domains of adult life. Further, we discovered that the new work coming from the TRAC observational study of college students has found troublesome rates of psychiatric comorbidity after the first year.2
This epidemiologic evidence supports devoting resources to the care of this cohort. But it appears that this has not penetrated the world of campus mental health treatment. At present, most post-secondary schools (to our knowledge, data is quite limited) lean toward policies that make it difficult for students with ADHD to be diagnosed or treated on campus. One obstacle is requiring evidence of a childhood diagnosis, which many children with high-IQ compensated ADHD may not have received. Another can be the demand for expensive and comprehensive neuropsychological testing even though the diagnostic value of that testing remains unclear.3 Some student health centers ask students to obtain prescriptions from the treaters they saw prior to coming to campus, even if those prescribers are out of state. Though these policies may be deployed in an effort to decrease diversion of stimulant medication, such hurdles may be difficult for the 18 year old ADHD student to navigate. The result is that many students with this predictably destructive condition go untreated.
The good news is this subject interests the collegiate community. Among other things, our APA workshop was selected to be the APA’s ‘Member’s Course of the Month’ for January 2020.4
Much work remains in developing and deploying diagnostic policies and treatment strategies that colleges and universities feel comfortable supporting. We mentioned the APSARD community during the workshop as a resource for professionals interested in ADHD. And we hope the wider ADHD research and treatment communities will join us in focusing our energy on this underserved and sometimes maligned group of students who need our help.
