Background
Late-life depression (LLD), defined as major depressive disorder in individuals over 60 years of age, is a severe psychiatric illness, and has a prevalence ranging between 1 and 16% [
20]. LLD has a poor long-term prognosis, and is associated with chronicity and high mortality [
32,
34,
39]. Research shows that relapse, recurrence, and chronicity are higher in LLD than they are in younger populations [
25], while in later life longer duration and chronicity are associated with further cognitive decline and somatic comorbidity such as cerebrovascular diseases and diabetes [
2,
15,
47]. Due to aging of the global population, the burden associated with LLD is expected to increase further, which underscores the necessity of appropriate, evidence-based, and cost-effective treatments that are tailored to older populations [
9].
Older adults with depression seem less responsive to standard treatment protocols than their younger counterparts. The Netherlands Study of Depression in Older Persons (NESDO) reported that only 18% of their participants diagnosed with major depressive disorder (MDD) had recovered after 2 years [
15], despite having received psychopharmacological and psychotherapeutic treatment from primary or specialized mental health-care services. In comparison, approximately 80% of a younger adult population recovered in a two-year period [
57]. Moreover, the efficacy of antidepressant medication is lower in older adults than it is in younger adults, which is possibly related to age-related cognitive impairments [
4,
49]. For example, a meta-analysis of 74 studies showed that the number needed to treat for antidepressants was six for adults, eight for persons over 55 years and 21 for people over 65 years [
59]. Furthermore, for older depressed patients access to therapy is limited [
16] and treatment often does not take into account underlying age-specific contributing factors. Arguably, add-on treatments aimed at distinct mechanisms of LLD may improve outcomes in this population.
Impairment in a broad range of cognitive functions [
51] is a common and disabling factor in LLD [
42]. Apart from manifesting in performance tasks, cognitive deficits have also been observed in the brain. On a structural level, temporary changes in frontal brain regions are related to depressive symptomatology [
1,
42] but particularly relevant are the changes discerned in the cognitive control network (CCN), which consists of the dorsolateral prefrontal cortex (DLPFC), the dorsal anterior cingulate cortex (dACC), and parietal regions [
3]. The CCN plays a crucial role in emotion regulation and abnormalities in the associated regions can result in weakened cognitive control over negative affect [
1,
3]. Cognitive control is implicated in a range of cognitive functions, such as working memory, attention span, and executive functions, loss of which can cause (increased) dysfunctional repetitive negative thinking, eliciting or aggravating depressive symptoms [
18]. This specific vulnerability pathway has been supported by multiple MDD studies in adults, underscoring the importance of dedicated interventions targeting this crucial function [
3,
8,
19]. Furthermore, in LLD the prevalence of impaired cognitive control functions is particularly high, which has also been associated with the reduced effectiveness of antidepressants [
4,
42,
49]. An added treatment that specifically targets impaired cognitive control in older adults coping with depression may then be the preferred approach.
In recent years, researchers have started evaluating computerized cognitive control training (CCT) programs. These interventions generally consist of repetitive performance of cognitive tasks, aiming to (re)activate specific processes that potentially help restore the cognitive function addressed. In general, CCT is associated with small to moderate improvements of cognitive skills, depressive symptoms, and global functioning (see [
35], and meta-analyses by Motter et al. [
43]; Legemaat et al. [
36]). These effects seem to complement the effects of treatment as usual (TAU). More specifically, CCT programs seem to be an effective and acceptable intervention for depressed older adults [
43]. Additionally, online CCT is a low-cost, highly scalable treatment option. However, the number of studies examined was small and the authors did stress the need to further examine whether the cognitive gains spread to other neurocognitive processes, symptoms, and overall functioning. Relatedly, a systematic review of different types of CCT for healthy older participants shows that after training specific executive functions, small to moderate gains in cognitive performance are observed [
45]. Interestingly, CCT appears especially effective when the applied tasks are continuously adapted to the participants’ performance level. Therefore, adding an adaptive form of CCT to current treatment of depressed older adults, might result in beneficial effects on cognitive functions, depressive symptoms and overall functioning. A randomized controlled trial (RCT) comparing such an adaptive CCT format to placebo add-on treatment in specialized health care would be the appropriate means to further test this.
The adaptive Paced Auditory Serial Addition Task (aPASAT) is a promising intervention aimed at enhancing cognitive control [
24,
56]. In this task, individuals are instructed to sum up the last two numbers they heard in an auditorily presented continuous stream of numbers. It is adaptive in that its presentation speed is continuously matched to the participant’s performance, inducing greater emotional reactivity (e.g., frustration, negative thoughts, some negative affect), which might account for the benefits derived from such tailored tasks [
11,
17]. Siegle et al. [
56] concluded that the aPASAT activates the cognitive control network by showing that the dorsolateral prefrontal cortex is engaged during its performance. They also tested the task in participants with MDD, in whom the cognitive training reduced depressive symptoms and rumination more so than was the case in a control group receiving treatment as usual (TAU) [
56]. Further studies continued to show beneficial effects of the aPASAT on depressive symptoms in various samples, where gains were possibly mediated by diminished rumination through improved cognitive functioning [
12,
53,
60]. Interestingly, Brunoni et al. [
12] observed, in a combined treatment group receiving CCT and transcranial direct-current stimulation, that sustained reduction in depressive symptoms was greatest in the older participants, even when controlling for changes in cognitive control throughout the training. This suggests that especially older depressed adults benefit from the aPASAT as an add-on to other treatment.
In summary, there is an urgent need for innovative, effective, and affordable treatments for LLD, which can be delivered alongside treatment as usual. The literature indicates that adaptive CCT approaches may be most effective in improving cognitive functioning and reducing rumination in older adults. The aPASAT is a relatively low-cost intervention that can be used as an add-on treatment to TAU, but large-scale clinical trials investigating its (cost-)effectiveness, working mechanisms, and feasibility in LLD are lacking. We aim to address these questions in the RCT described below.
Aims and hypotheses
Our primary objective is to investigate whether the addition of a CCT program to ongoing LLD treatment (i.e., pharmacotherapy and/or psychotherapy) will help reduce depressive symptoms more so than a placebo training. Outcomes will be assessed 1 month post-training (T1) and during follow-up assessments at three (T2), six (T3) and 12 months (T4). We expect symptom reductions in the participants having received CCT to be greater than in the participants having completed the placebo training at all timepoints. Our second aim is to see whether CCT also exerts effects on other clinical measures including rumination (both trait and state), cognitive emotion regulation, working memory functioning (near and far transfer), inhibition, and quality of life. We expect that participants in the intervention group will show lower rumination, more adaptive cognitive emotion regulation, more improved working memory functioning, inhibition, and quality of life than the participants in the placebo group. Our third aim is to identify working mechanisms of CCT. We will examine whether improved working memory and reduced rumination mediate the effects of CCT on depressive symptoms, hypothesizing that improvements in depressive symptoms will follow improvement of working memory and reduction of rumination in the intervention group. Fourthly, we aim to study individual differences that could account for differences in the effect of CCT on symptom change, where we expect that baseline cognitive control, depression severity, and age will be moderating factors. Our fifth aim is to determine the cost-effectiveness of our add-on CCT, we will conduct a health-economic evaluation and budget-impact analysis, anticipating that CCT will reduce mental healthcare consumption during the follow-up period more so than placebo training. Finally, for our sixth aim, we will be assessing the acceptability and feasibility of the intervention by conducting qualitative interviews with subsamples of participants having received CCT, therapists involved in mental health care for older persons and other stakeholders to inform the further implementation of CCT as an add-on intervention next to existing treatments for LLD.
Discussion
The proposed RCT will examine the (cost-)effectiveness of a cognitive control training (CCT) program as an add-on intervention to specialized depression treatment in adults over 60 years coping with late-life depression (LLD). We will be offering the aPASAT as the active intervention since it targets working memory, which has been found to be effective in reducing depressive symptoms, possibly by increasing cognitive control over repetitive negative thought intrusions. Its effectiveness will be compared to TAU coupled with a placebo intervention (sham training). Change in depressive symptoms, potential working mechanisms, and cost-effectiveness will be investigated post-training and three times during a one-year follow-up. Being a pragmatic trial, inclusion restrictions will be limited, to thus to enhance its ecological validity pertinent to depressed populations in real-world clinical settings.
In addition to change in depressive symptoms, we will study change in rumination and cognitive deficits (i.e., reduced working memory performance) as possible working mechanisms of CCT since both cognitive pathways have been shown to be addressed by the aPASAT [
8,
19], with rumination having been identified as a mediating factor between executive dysfunction and depression-related behavior in LLD [
63]. We will also be making a first exploration into the effect of CCT on cognitive coping and inhibition of irrelevant (emotional) stimuli to examine the role of these underlying factors of LLD. We will also be looking at age, baseline cognitive control and depression severity to see whether these individual factors affect the effectiveness of CCT differentially and to predict which patients might benefit most from the intervention for well-informed personalized prescription or referral. We also hope to establish whether the addition of CCT to TAU for LLD improves health-related quality of life and overall daily functioning, which often remains impaired even after symptom reduction [
14].
From a societal perspective, the cost-effectiveness of mental-health innovations is an important factor. We expect online CCT to reduce healthcare consumption in older adults, while, in and of itself, CCT is an inexpensive treatment. It is readily accessible from home and its standardized format allows easy monitoring and guidance, and systematic comparison of outcomes over time.
The results from the flanking implementation study will serve to inform a wider implementation of digital interventions targeting this growing population within and outside mental-health care, potentially lowering the need for more expensive face-to-face health services. Since the target demographic will become increasingly familiar with digital technologies, computerized interventions will become increasingly relevant, a trend which has been amplified by the COVID-9 pandemic and its ensuing lock-downs. In the implementation study, we will schedule qualitative interviews and focus group meetings to evaluate the appropriateness, acceptability and referral of online CCT. The goal of the implementation study is sustainable later scale up of the CCT intervention. Previous results on the acceptability of digital CCT formats in this population were promising [
6], but findings were often obtained in smaller pilot studies. The data we will be collecting will not only support an adequate implementation of the intervention under study but also offer insight into the acceptability and practical implications of digital interventions in general.
Despite the strengths of the trial, there are some limitations. First, there is no control group that receives only treatment as usual to monitor the natural course of depressive symptoms. Both groups follow a form of CCT – active or placebo. Therefore, the results of this trial will not be able to compare CCT treatment to TAU alone. Second, most patient will likely be treated within specialized mental healthcare centers, which limits the generalizability of the results.
In conclusion, new treatment strategies for late-life depression are urgently needed and we hope that, if proven effective, feasible and cost-effective, the intervention will be added to (inter)national guidelines as a recommended add-on to current depression treatment for older adults.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.