Investigating the effectiveness of Online Cognitive Trainings at improving mental health and wellbeing in the workplace in ‘at risk’ population

: (Alternative Format Thesis)

Abstract

Subclinical depression is highly prevalent in the general population and is associated with an increased risk of major depressive disorder, functional impairment, poor workplace performance, and increased mortality. Digital mental health interventions in the form of computerised cognitive training or digital interventions based on therapeutic techniques are proposed to be a low cost and highly accessible treatment option for this ‘at risk’ population. However, the effectiveness of these interventions is unclear, as is their mechanism of action. Executive functioning may play a key role in the onset and maintenance of depressive symptoms, though there is little research exploring executive function as mechanism of action for digital mental health interventions. In this thesis I aimed to: 1) evaluate the effectiveness of digital cognitive trainings and self-guided therapeutic interventions on symptoms of depression, wellbeing and executive functioning in subclinical populations; and 2) explore whether changes in executive functioning mediate intervention-related changes in depressive symptoms.

In Study 1 (chapter 2) I addressed both aims by conducting a randomised controlled trial investigating the effects of working memory training or cognitive restructuring training on depressive symptoms, executive functioning, and interpretative bias, as compared to a waitlist control condition. Cognitive restructuring training reduced depressive symptoms, b = -3.31, p = .043, 95% CI [-6.49, -0.05] and increased positive interpretative bias, b = 0.16, p = .001, 95% CI [0.06, 0.25], though there was no effect on executive functioning. Additionally, the working memory training group did not differ from controls on any outcome (all ps > .05). Additionally, the low adherence to the training protocol limited our ability to make inferences about the impact of training executive function on mental health or interpretative biases. In Study 2 (chapter 3) I explored solutions to this issue by investigating the effects of gamification on adherence to a food-based response inhibition training. We recruited 252 participants who were randomly allocated to receive either a standardised response inhibition training, a training with feedback elements or a training with social elements. We found no evidence that
gamification improved adherence to the training protocol, F (2, 248) = 0.17, p = .848, ηp2 = .001, or training motivation, F (2, 214) = 1.40, p = .250, ηp2 = .012.

Given the low adherence to online computerised cognitive trainings, and the popularity of mental health smartphone apps, in Study 3 (chapter 4) I investigated the effects of app-based interventions. I recruited 191 employed adults with subclinical symptoms of depression and anxiety and investigated the effectiveness of executive function training and self-guided CBT delivered via smartphone apps on mental health (depression and anxiety), workplace wellbeing, and executive function. Participants were measured at baseline, after 4 weeks, and after 12 weeks. App-based executive function training significantly reduced depressive, b = -3.11, p = .019, 95% CI [-5.67, -0.54]; and anxious symptoms, b = -2.66, p = .019, 95% CI [-4.85, -0.48], at 12 weeks. However, there was no effect at 4-weeks (immediately after completing the intervention), and no effect on workplace wellbeing or executive functioning (all ps > .05). There was no effect of app-based self-guided CBT on symptoms of depression and anxiety, or executive functioning at any time point (all ps > .05). There was an effect on workplace wellbeing at 4 weeks, b = 3.89, p = .042, 95% CI [0.20, 7.58], this was not sustained at 12 weeks. Additionally, there was no mediating effect of executive functioning on change in depressive symptoms. In Study 4 (chapter 5), to further investigate executive function as a potential mechanism of action, I used EEG methods to measure midline theta power, a neural marker of cognitive control, in a subsample of participants from Study 3 (n = 53). Consistent with the findings of Study 3, there was no effect of training group on midline theta, F (2,41) = 0.54, p = .589, ηp2 < .001. This study failed to provide support for the hypothesis that app-based interventions would work via improvements in cognitive control.

Overall, the evidence for the effectiveness of online digital cognitive trainings was mixed. App-based executive function training and cognitive restructuring were both effective at reducing depressive symptoms in ‘at risk’ working age samples, but working memory training, and an app employing CBT techniques, were not. App-based CBT techniques improved workplace wellbeing after 4 weeks of training, though these effects were not sustained at follow up (12-weeks), and there was no effect of app-based executive function training on workplace wellbeing at any time point. Adherence to online training tasks was low in Studies 1 and 2, though app-based interventions appeared to engage and retain participants. Finally, there was no evidence that change in executive functioning mediated intervention-related changes in depressive symptoms. The results of this PhD suggest app-based executive functioning training programmes may reduce depressive symptoms, and may be a low-cost, accessible intervention for people with subclinical symptoms of depression, though the effects for self-guided apps including CBT techniques are mixed.

Date of Award	19 Feb 2025
Original language	English
Awarding Institution	University of Bath
Supervisor	Katherine Button (Supervisor), Graeme Fairchild (Supervisor) & George Stothart (Supervisor)