Mediating Effects of Neural Targets on Depression, Weight, and Anxiety Outcomes of an Integrated Collaborative Care Intervention: The ENGAGE-2 Mechanistic Pilot Randomized Clinical Trial

Study of a PST-trained voice-enabled artificial intelligence counselor for adults with emotional distress (SPEAC-2): Design and methods

BACKGROUND

Novel and scalable psychotherapies are urgently needed to address the depression and anxiety epidemic. Leveraging artificial intelligence (AI), a voice-based virtual coach named Lumen was developed to deliver problem solving treatment (PST). The first pilot trial showed promising changes in cognitive control measured by functional neuroimaging and improvements in depression and anxiety symptoms.

METHODS

To further validate Lumen in a 3-arm randomized clinical trial, 200 participants with mild-to-moderate depression and/or anxiety will be randomly assigned in a 2:1:1 ratio to receive Lumen-coached PST, human-coached PST as active treatment comparison, or a waitlist control condition where participants can receive Lumen after the trial period. Participants will be assessed at baseline and 18 weeks. The primary aim is to confirm neural target engagement by testing whether compared with waitlist controls, Lumen participants will show significantly greater improvements from baseline to 18 weeks in the a priori neural target for cognitive control, right dorsal lateral prefrontal cortex engaged by the go/nogo task (primary superiority hypothesis). A secondary hypothesis will test whether compared with human-coached PST participants, Lumen participants will show equivalent improvements (i.e., noninferiority) in the same neural target from baseline to 18 weeks. The second aim is to examine (1) treatment effects on depression and anxiety symptoms, psychosocial functioning, and quality of life outcomes, and (2) relationships of neural target engagement to these patient-reported outcomes.

CONCLUSIONS

This study offers potential to improve the reach and impact of psychotherapy, mitigating access, cost, and stigma barriers for people with depression and/or anxiety.

CLINICALTRIALS

gov #: NCT05603923.

Association Between User Interaction and Treatment Response of a Voice-Based Coach for Treating Depression and Anxiety: Secondary Analysis of a Pilot Randomized Controlled Trial

BACKGROUND

The quality of user interaction with therapeutic tools has been positively associated with treatment response; however, no studies have investigated these relationships for voice-based digital tools.

OBJECTIVE

This study evaluated the relationships between objective and subjective user interaction measures as well as treatment response on Lumen, a novel voice-based coach, delivering problem-solving treatment to patients with mild to moderate depression or anxiety or both.

METHODS

In a pilot trial, 42 adults with clinically significant depression (Patient Health Questionnaire-9 [PHQ-9]) or anxiety (7-item Generalized Anxiety Disorder Scale [GAD-7]) symptoms or both received Lumen, a voice-based coach delivering 8 problem-solving treatment sessions. Objective (number of conversational breakdowns, ie, instances where a participant's voice input could not be interpreted by Lumen) and subjective user interaction measures (task-related workload, user experience, and treatment alliance) were obtained for each session. Changes in PHQ-9 and GAD-7 scores at each ensuing session after session 1 measured the treatment response.

RESULTS

Participants were 38.9 (SD 12.9) years old, 28 (67%) were women, 8 (19%) were Black, 12 (29%) were Latino, 5 (12%) were Asian, and 28 (67%) had a high school or college education. Mean (SD) across sessions showed breakdowns (mean 6.5, SD 4.4 to mean 2.3, SD 1.8) decreasing over sessions, favorable task-related workload (mean 14.5, SD 5.6 to mean 17.6, SD 5.6) decreasing over sessions, neutral-to-positive user experience (mean 0.5, SD 1.4 to mean 1.1, SD 1.3), and high treatment alliance (mean 5.0, SD 1.4 to mean 5.3, SD 0.9). PHQ-9 (Ptrend=.001) and GAD-7 scores (Ptrend=.01) improved significantly over sessions. Treatment alliance correlated with improvements in PHQ-9 (Pearson r=-0.02 to -0.46) and GAD-7 (r=0.03 to -0.57) scores across sessions, whereas breakdowns and task-related workload did not. Mixed models showed that participants with higher individual mean treatment alliance had greater improvements in PHQ-9 (β=-1.13, 95% CI -2.16 to -0.10) and GAD-7 (β=-1.17, 95% CI -2.13 to -0.20) scores.

CONCLUSIONS

The participants had fewer conversational breakdowns and largely favorable user interactions with Lumen across sessions. Conversational breakdowns were not associated with subjective user interaction measures or treatment responses, highlighting how participants adapted and effectively used Lumen. Individuals experiencing higher treatment alliance had greater improvements in depression and anxiety. Understanding treatment alliance can provide insights on improving treatment response for this new delivery modality, which provides accessibility, flexibility, comfort with disclosure, and cost-related advantages compared to conventional psychotherapy.

TRIAL REGISTRATION

ClinicalTrials.gov NCT04524104; https://clinicaltrials.gov/study/NCT04524104.

Effects of a virtual voice-based coach delivering problem-solving treatment on emotional distress and brain function: a pilot RCT in depression and anxiety

Consumer-based voice assistants have the ability to deliver evidence-based treatment, but their therapeutic potential is largely unknown. In a pilot trial of a virtual voice-based coach, Lumen, delivering problem-solving treatment, adults with mild-to-moderate depression and/or anxiety were randomized to the Lumen intervention (n = 42) or waitlist control (n = 21). The main outcomes included changes in neural measures of emotional reactivity and cognitive control, and Hospital Anxiety and Depression Scale [HADS] symptom scores over 16 weeks. Participants were 37.8 years (SD = 12.4), 68% women, 25% Black, 24% Latino, and 11% Asian. Activation of the right dlPFC (neural region of interest in cognitive control) decreased in the intervention group but increased in the control group, with an effect size meeting the prespecified threshold for a meaningful effect (Cohen's d = 0.3). Between-group differences in the change in activation of the left dlPFC and bilateral amygdala were observed, but were of smaller magnitude (d = 0.2). Change in right dlPFC activation was also meaningfully associated (r ≥ 0.4) with changes in self-reported problem-solving ability and avoidance in the intervention. Lumen intervention also led to decreased HADS depression, anxiety, and overall psychological distress scores, with medium effect sizes (Cohen's d = 0.49, 0.51, and 0.55, respectively), compared with the waitlist control group. This pilot trial showed promising effects of a novel digital mental health intervention on cognitive control using neuroimaging and depression and anxiety symptoms, providing foundational evidence for a future confirmatory study.

Affective neural circuits and inflammatory markers linked to depression and anxiety symptoms in patients with comorbid obesity

Although we have effective treatments for depression and anxiety, we lack mechanistic understanding or evidence-based strategies to tailor these treatments in the context of major comorbidities such as obesity. The current feasibility study uses functional neuroimaging and biospecimen data to determine if changes in inflammatory markers, fecal short-chain fatty acids, and neural circuit-based targets can predict depression and anxiety outcomes among participants with comorbid obesity. Blood and stool samples and functional magnetic resonance imaging data were obtained at baseline and 2 months, during the parent ENGAGE-2 trial. From 30 participants with both biospecimen and fMRI data, this subsample study explored the relationship among changes in inflammatory markers and fecal short-chain fatty acids and changes in neural targets, and their joint relationship with depression and anxiety symptoms. Bivariate and partial correlation, canonical correlation, and partial least squares analyses were conducted, with adjustments for age, sex, and treatment group. Initial correlation analyses revealed three inflammatory markers (IL-1RA, IL-6, and TNF-α) and five neural targets (in Negative Affect, Positive Affect, and Default Mode Circuits) with significantly associated changes at 2 months. Partial least squares analyses then showed that changes in IL-1RA and TNF-α and changes in three neural targets (in Negative Affect and Positive Affect Circuits) at 2 months were associated with changes in depression and anxiety symptoms at 6 months. This study sheds light on the plausibility of incorporation of inflammatory and gastrointestinal biomarkers with neural targets as predictors of depression and comorbid anxiety outcomes among patients with obesity.

Associations between fecal short-chain fatty acids, plasma inflammatory cytokines, and dietary markers with depression and anxiety: Post hoc analysis of the ENGAGE-2 pilot trial

BACKGROUND

The microbiome-gut-brain-axis (MGBA) is emerging as an important mechanistic link between diet and mental health. The role of significant modifiers of the MGBA, including gut microbial metabolites and systemic inflammation, in individuals comorbid with obesity and mental disorders, is under-investigated.

OBJECTIVES

This exploratory analysis examined associations among microbial metabolites-fecal SCFAs, plasma inflammatory cytokines, and diet with depression and anxiety scores in adults comorbid with obesity and depression.

METHODS

Stool and blood were obtained from a subsample (n = 34) of participants enrolled in an integrated behavioral intervention for weight loss and depression. Pearson partial correlation and multivariate analyses determined associations among changes in fecal SCFAs (propionic, butyric, acetic, and isovaleric acids), plasma cytokines [C-reactive protein, interleukin 1 beta, interleukin 1 receptor antagonist (IL-1RA), interleukin 6, and TNF-α], and 35 dietary markers over 2 mo, and changes in SCL-20 (Depression Symptom Checklist 20-item) and GAD-7 (Generalized Anxiety Disorder 7-Item) scores over 6 mo.

RESULTS

Changes in the SCFAs and TNF-α at 2 mo were positively associated (standardized coefficients: 0.06-0.40; 0.03-0.34) with changes in depression and anxiety scores at 6 mo, whereas changes in IL-1RA at 2 mo were inversely associated (standardized coefficients: -0.24; -0.05). After 2 mo, changes in 12 dietary markers, including animal protein, were associated with changes in SCFAs, TNF-α, or IL-1RA at 2 mo (standardized coefficients: -0.27 to 0.20). Changes in 11 dietary markers, including animal protein, at 2 mo were associated with changes in depression or anxiety symptom scores at 6 mo (standardized coefficients: -0.24 to 0.20; -0.16 to 0.15).

CONCLUSIONS

Gut microbial metabolites and systemic inflammation may be biomarkers of importance within the MGBA, linking dietary markers, such as animal protein intake, to depression and anxiety for individuals with comorbid obesity. These findings are exploratory and warrant replication.

Association of Changes in Neural Targets and Dietary Outcomes among Patients with Comorbid Obesity and Depression: Post hoc Analysis of ENGAGE-2 Mechanistic Clinical Trial

BACKGROUND

Disruptions in brain circuits that regulate cognition and emotion can hinder dietary change and weight loss among individuals with obesity and depression.

OBJECTIVE

The study aimed to investigate whether changes in brain targets in the cognitive control, negative affect, and positive affect circuits after 2-mo problem-solving therapy (PST) predict changes in dietary outcomes at 2 and 6 mo.

METHODS

Adults with obesity and depression from an academic health system were randomly assigned to receive PST (7-step problem-solving and behavioral activation strategies) over 2 mo or usual care. Seventy participants (mean age = 45.9 ± 11.6 y; 75.7% women, 55.7% Black, 17.1% Hispanic, 20.0% White; mean BMI = 36.5 ± 5.3 kg/m2; mean Patient Health Questionnaire-9 depression score = 12.7 ± 2.8) completed functional MRI and 24-h food recalls. Ordinary least square regression analyses were performed.

RESULTS

Among intervention participants, increased left dorsal lateral prefrontal cortex (dLPFC) activity of the cognitive control circuit at 2 mo was associated with increased diet quality (β: 0.20; 95% CI: -0.02, 0.42) and decreased calories (β: -0.19; 95% CI: -0.33, -0.04), fat levels (β: -0.22; 95% CI: -0.39, -0.06), and high-sugar food intake (β: -0.18; 95% CI: -0.37, 0.01) at 6 mo. For the negative affect circuit, increased right dLPFC-amygdala connectivity at 2 mo was associated with increased diet quality (β: 0.32; 95% CI: -0.93, 1.57) and fruit and vegetable intake (β: 0.38; 95% CI: -0.75, 1.50) and decreased calories (β: -0.37; 95% CI: -1.29, 0.54), fat levels (β: -0.37; 95% CI: -1.50, 0.76), sodium concentrations (β: -0.36; 95% CI: -1.32, 0.60), and alcohol intake (β: -0.71; 95% CI: -2.10, 0.68) at 2 but not at 6 mo. The usual care group showed opposing associations. The 95% CIs of all between-group differences did not overlap the null, suggesting a significant treatment effect.

CONCLUSIONS

Among adults with obesity and depression who underwent PST compared with those under usual care, improved dLPFC-amygdala regulation of negative affective brain states predicted dietary improvements at 2 mo, whereas improvements in dLPFC-based cognitive control predicted dietary improvements at 6 mo. These findings warrant confirmatory studies. This trial was at clinicaltrials.gov as NCT03841682.

Associations between daily step count trajectories and clinical outcomes among adults with comorbid obesity and depression

Purpose

To examine the relationship between features of daily measured step count trajectories and clinical outcomes among people with comorbid obesity and depression in the ENGAGE-2 Trial.

Methods

This post hoc analysis used data from the ENGAGE-2 trial where adults (n=106) with comorbid obesity (BMI ≥30.0 or 27.0 if Asian) and depressive symptoms (Patient Health Questionnaire-9 score ≥10) were randomized (2:1) to receive the experimental intervention or usual care. Daily step count trajectories over the first 60 days (Fitbit Alta HR) were characterized using functional principal component analyses. 7-day and 30-day trajectories were also explored. Functional principal component scores that described features of step count trajectories were entered into linear mixed models to predict weight (kg), depression (Symptom Checklist-20), and anxiety (Generalized Anxiety Disorder Questionnaire-7) at 2-months (2M) and 6-months (6M).

Results

Features of 60-day step count trajectories were interpreted as overall sustained high, continuous decline, and disrupted decline. Overall sustained high step count was associated with low anxiety (2M, β=-0.78, p<.05; 6M, β=-0.80, p<.05) and low depressive symptoms (6M, β=-0.15, p<.05). Continuous decline in step count was associated with high weight (2M, β=0.58, p<.05). Disrupted decline was not associated with clinical outcomes at 2M or 6M. Features of 30-day step count trajectories were also associated with weight (2M, 6M), depression (6M), and anxiety (2M, 6M); Features of 7-day step count trajectories were not associated with weight, depression, or anxiety at 2M or 6M.

Conclusions

Features of step count trajectories identified using functional principal component analysis were associated with depression, anxiety, and weight outcomes among adults with comorbid obesity and depression. Functional principal component analysis may be a useful analytic method that leverages daily measured physical activity levels to allow for precise tailoring of future behavioral interventions.

Integrated collaborative care intervention for depression and obesity in primary care: translation from research to practice

The objective of this study was to present lessons learned about engagement, delivery modality and pandemic impact while delivering a collaborative care intervention with a socioeconomically, racially and ethnically diverse sample. Participants completed a post-intervention survey (n = 41) on experiences and preferred intervention delivery modality, coronavirus 2019 (COVID-19) Impact Survey (n = 50) and provided open-ended feedback about the intervention (n = 27). Intervention process data included attendance, modality, and withdrawals. Data were analyzed using descriptive statistics and inductive content analyses. Of 71 intervention participants, 6 (8%) withdrew before session 1. Completers adhered to intervention timeline better than withdrawals. Participants liked the in-person interaction, efficient coach support, accountability of in-person and Zoom vs. phone sessions and the flexibility and convenience of phone and Zoom vs. in-person sessions. A majority of participants reported experiencing pandemic impacts such as heightened emotional distress, decreased activity engagement, poorer eating behaviors and being unable to meet basic needs. Participants deviating from intervention timelines may be re-engaged by targeted outreach attempts. Videoconference has the potential for providing as-needed coaching. Future interventions may be optimized to account for and address areas impacted by the pandemic. Findings revealed specific strategies that can be implemented in future interventions to improve emotional and physical health among diverse populations.

Cross-trial prediction of depression remission using problem-solving therapy: A machine learning approach

BACKGROUND

Psychotherapy is a standard depression treatment; however, determining a patient's prognosis with therapy relies on clinical judgment that is subject to trial-and-error and provider variability.

PURPOSE

To develop machine learning (ML) algorithms to predict depression remission for patients undergoing 6 months of problem-solving therapy (PST).

METHOD

Using data from the treatment arm of 2 randomized trials, ML models were trained and validated on ENGAGE-2 (ClinicalTrials.gov, #NCT03841682) and tested on RAINBOW (ClinicalTrials.gov, #NCT02246413) for predictions at baseline and at 2-months. Primary outcome was depression remission using the Depression Symptom Checklist (SCL-20) score < 0.5 at 6 months. Predictor variables included baseline characteristics (sociodemographic, behavioral, clinical, psychosocial) and intervention engagement through 2-months.

RESULTS

Of the 26 candidate variables, 8 for baseline and 11 for 2-months were predictive of depression remission, and used to train the models. The best-performing model predicted remission with an accuracy significantly greater than chance in internal validation using the ENGAGE-2 cohort, at baseline [72.6% (SD = 3.6%), p < 0.0001] and at 2-months [72.3% (5.1%), p < 0.0001], and in external validation with the RAINBOW cohort at baseline [58.3% (0%), p < 0.0001] and at 2-months [62.3% (0%), p < 0.0001]. Model-agnostic explanations highlighted key predictors of depression remission at the cohort and patient levels, including female sex, lower self-reported sleep disturbance, lower sleep-related impairment, and lower negative problem orientation.

CONCLUSIONS

ML models using clinical and patient-reported data can predict depression remission for patients undergoing PST, affording opportunities for prospective identification of likely responders, and for developing personalized early treatment optimization along the patient care trajectory.