Cognitive Training for Psychiatric Disorders

Psychiatric disorders are associated with impairments in neural system activity and connectivity across distributed networks that underlie cognition and social–emotional processes. Increasing evidence indicates that these neural system dysfunctions are not immutably fixed, but instead may be amenable to well-designed cognitive training interventions that target restoration of neural system operations (Browning et al, 2012; Klingberg et al, 2005; Subramaniam et al, 2012). An explicitly ‘systems neuroplasticity'-based approach to cognitive training is founded on the premise that during successful skill learning, disproportionately larger and better-coordinated neuronal populations represent the salient inputs and action outputs of the trained skill, resulting in an increased feed-forward signal strength from sensory regions as well as greater task-relevant feedback-inhibitory control from the prefrontal cortex to enhance representations of relevant stimuli, and to enable more efficient and accurate associative memory processes (Vinogradov et al, 2012).

Effective cognitive training must target the underlying neural impairments associated with a specific pathophysiology. For children with ADD, for example, Klingberg et al 2005 found that computerized visual working memory (WM) exercises drove improvements in non-trained visuospatial and executive tasks, indicating generalization of training. This group also found that WM training improved WM capacity, which was correlated with neural changes in D1 receptor density, indicating increased dopaminergic release during training (Klingberg, 2010; McNab et al, 2009). Browning et al (2012) investigated attentional bias modification (ABM) training in remitted patients with depression, and found that ABM reduced residual depressive symptoms and normalized the cortisol awakening response, suggesting that it may be a ‘cognitive vaccine' that reduces the neurobehavioral risk for future depression episodes.

Our group recently performed a double-blind randomized controlled trial of a set of computerized exercises that focused on early auditory and visual processing, WM and basic social cognition (vs computer games control condition) in individuals with schizophrenia. Our rationale was that schizophrenia is characterized by deficits in both early pre-attentive perceptual processes as well as higher-order attention and WM operations (Vinogradov et al, 2012). We found that the schizophrenia participants who received the targeted training showed behavioral improvements on (untrained) neuropsychological measures of verbal memory and on reality-monitoring tasks, thus indicating generalization of training effects. Further, after the intervention, neural activation patterns during reality monitoring, which were abnormal in these patients at baseline, began to resemble the patterns observed in healthy participants (Figures 1a and c), and predicted better social functioning 6 months later (Subramaniam et al, 2012).

Figure 1.

Whole-brain fMRI analysis of reality-monitoring activity reveals signal increase within: (a) the medial prefrontal cortex (mPFC) in 15 healthy comparison (HC) subjects, (b) the posterior cingulate cortex, rather than the mPFC, in patients with schizophrenia (SZ) prior to computerized cognitive training, and (c) the mPFC in only the group of SZ patients who completed 80 h of active computerized cognitive training (SZ-AT), which is similar to the neural activation patterns observed in the HC sample (see Subramaniam et al, 2012).

Together, these emerging data suggest that people with a range of neuropsychiatric illnesses can benefit from targeted cognitive training; that this type of training can ‘restore' aspects of behavior and neural system functioning; and that this training can be generalized to enduring improvements in real-world functioning (Browning et al, 2012; Klingberg et al, 2005; Subramaniam et al, 2012). Future studies must examine the specific intervention methods that promote maximal cognitive and neural system ‘restoration' in the neuropsychiatrically impaired brain—likely by combining targeted cognitive training approaches with cognitive enhancing medications and neuromodulation techniques such as transcranial direct current stimulation.

One of the authors, Dr Sophia Vinogradov, has financial disclosures to declare, as Dr Vinogradov is a paid consultant in Brain Plasticity, a company with a commercial interest in cognitive training software. Dr Vinogradov is also a consultant to Genentech, Amgen, and Hoffman-LaRoche. Dr Subramaniam declares no conflict of interest.