(© deagreez - stock.adobe.com)
In A Nutshell
- A smartphone app called HabitWorks helped people interpret ambiguous situations in less negative ways.
- Users also reported better day-to-day functioning and greater overall improvement than the symptom-tracking group.
- But on standard anxiety and depression questionnaires, the app did not outperform structured symptom tracking.
- The big takeaway: this looks promising as a brief mental health tool, but not yet as a proven standalone treatment.
A text from the boss that reads only “We need to talk.” Stomach drops. Mental resignation letter, drafted and ready. That knee-jerk leap to the worst possible explanation is what psychologists call interpretation bias, and it’s one of the engines that keeps anxiety and depression running in the background of millions of American lives. Researchers at McLean Hospital and Harvard Medical School have now built a smartphone app designed to break that mental habit, and a new trial shows it genuinely changes how people interpret the ambiguous moments of everyday life.
The catch? It didn’t reduce anxiety and depression symptoms any more than a structured symptom-tracking routine did.
The app, called HabitWorks, takes a deceptively simple approach. It presents users with vague everyday scenarios, like “Your boss wants to meet with you,” paired with a word that could frame that situation negatively (“criticize”), neutrally (“appointment”), or positively (“praise”). Users tap a button to say whether the word fits the situation, and the app gives them instant feedback that reinforces the more balanced, less catastrophic reading. Each session takes about five minutes and runs through 50 of these quick-fire pairings. The idea is that with enough repetition, the brain starts defaulting to less dire interpretations of life’s gray areas, like retraining the brain’s default response to ambiguous situations.
In a randomized controlled trial involving 340 adults across the United States who had at least mild symptoms of anxiety and depression, the app delivered strong results on some fronts while falling short on others, a pattern that has researchers rethinking which outcomes matter most when measuring whether a mental health tool is truly working.
How the HabitWorks Brain-Training App Study Worked
Recruited between March 2024 and February 2025, participants were a diverse group of American adults. The average age was about 33, roughly 57% were women, and the group included White, Black, Asian, multiracial, and Hispanic or Latine individuals. Everyone had to have at least mild symptoms of depression or anxiety to qualify, along with access to a smartphone. People with self-reported diagnoses of bipolar disorder or schizophrenia, or active symptoms of mania or psychosis, were excluded.
Participants were randomly split into two groups. One used the HabitWorks app, completing three of those five-minute brain-training exercises per week for four weeks. The other, the control group, simply filled out depression and anxiety symptom surveys three times a week for the same period. Both groups completed more detailed assessments at the start of the study, after four weeks, and again a month later.
What sets HabitWorks apart from earlier attempts at this kind of brain training is personalization. Before starting, users fill out a checklist about what kinds of situations tend to trigger their worry: social interactions, finances, perfectionism, physical symptoms, and so on. They also share demographic details like whether they have kids or a romantic partner. The app then pulls from a pool of 745 possible scenario-word pairings to build a custom set tailored to each person’s actual life. Previous versions of this type of training used the same generic scenarios for everyone, which users often found repetitive and irrelevant.
The Brain-Training App Changed Thinking Patterns But Not Symptom Scores
On the engagement front, HabitWorks exceeded expectations. The research team had set modest benchmarks going in, hoping that at least 25% of app users would still be active by week four. Instead, nearly 78% were still using the app in that final week, and about 44% achieved perfect adherence, meaning they completed all three exercises every single week. Users rated the app positively for both acceptability and usability, and reported feeling far more engaged with HabitWorks than the control group felt with their symptom-tracking surveys.
The app also did exactly what it was designed to do at a brain-training level. Compared to the symptom-tracking group, HabitWorks users showed clearly larger improvements in how they processed uncertain information. On a standardized test of interpretation bias, the app group improved their accuracy on non-threatening interpretations by about 14.6 percentage points more than the control group, and their accuracy on rejecting negative interpretations by about 33.4 percentage points more.
Beyond the thinking-pattern shifts, HabitWorks users also reported real-world improvements in how well they were functioning day to day. HabitWorks users reported significantly greater reductions in how much their mental health interfered with work, social life, home responsibilities, and close relationships compared to the control group. When asked to rate their overall improvement at the end of the four weeks, HabitWorks users also reported feeling meaningfully more improved than the tracking group did.
But when it came to the specific symptoms of depression and anxiety, the headline measures that most people would consider the whole point of a mental health tool, both groups improved by statistically similar amounts. Depression scores dropped significantly in the HabitWorks group and in the symptom-tracking group. Anxiety scores followed the same pattern. The app didn’t produce any additional benefit on these particular measures beyond what simply monitoring symptoms three times a week accomplished.
Why Tracking Your Symptoms Might Be a Treatment in Itself
This result surprised the research team, but it’s not without explanation. Symptom tracking, the act of regularly checking in with oneself about how one is feeling, is itself considered a therapeutic activity. It’s a core part of cognitive behavioral therapy and many other evidence-based treatments. The control group wasn’t doing nothing. They were engaging in structured self-monitoring three times a week, receiving weekly summaries of their symptom severity, and being contacted regularly by the research team. That level of attention and self-reflection may have been enough to move the needle on symptoms.
The researchers also pointed out that the questionnaires used for depression and anxiety, the PHQ-8 and GAD-7, capture specific clusters of symptoms over the past two weeks. Meanwhile, the measures where HabitWorks did outperform the control, like overall clinical improvement and daily functioning, are broader. The team suggested that HabitWorks may be doing something real and meaningful for people’s overall well-being that these narrower symptom scales aren’t built to detect.
The study also enrolled people from the general community with at least mild symptoms, not patients in a clinical setting seeking treatment for severe mental illness. Prior pilot studies of HabitWorks had tested the app in more acute settings, including among psychiatric inpatients, and the researchers acknowledged that the intervention might perform differently in populations with more severe symptoms or among people actively seeking help.
What Comes Next for Brain-Training Apps
After the four-week study period, participants in the symptom-tracking group were offered the chance to try HabitWorks, and about 64% chose to do so, a telling sign that even those in the control group were curious about or interested in the app.
HabitWorks appears to change how people process uncertain information, and people who use it report functioning better in their daily lives and feeling more globally improved. Those are outcomes that matter enormously to anyone living with the constant drag of anxiety and depression. But the lack of a specific edge over symptom tracking on standard depression and anxiety questionnaires means the app can’t yet claim to be a standalone treatment for those conditions based on this trial alone.
Still, for a five-minutes-a-day phone app that requires no therapist, no appointment, and no insurance, even modest real-world benefits in functioning and overall improvement could matter at a population level. An estimated one in five American adults experiences depression or anxiety symptoms at any given time, and fewer than a quarter receive adequate treatment. Tools that are engaging enough for people to actually keep using, and HabitWorks appears to be exactly that, could help fill an enormous gap, even if they turn out to work best as one piece of a larger puzzle rather than a cure-all.
Paper Notes
Limitations
The study enrolled a nonclinical community sample of adults with at least mild symptoms, which means results may not generalize to people with more severe clinical depression or anxiety, or to those actively seeking treatment. The symptom-tracking control condition was not a passive or inert control — it involved structured self-monitoring, weekly feedback, and regular contact, which may have therapeutic effects of its own, making it harder to detect differences between the two groups on symptom measures. The internal consistency of the measure used to assess one type of interpretation bias (benign accuracy scores on the Word-Sentence Association Paradigm) was poor at baseline, which could have affected the precision of those estimates. Additionally, the study relied entirely on self-report measures for clinical outcomes. Missing data ranged from about 10.5% to 20.7% across outcome measures due to participant dropout or survey noncompliance, though the researchers used statistical methods to account for this. Exploratory analyses conducted at one-month follow-up had reduced statistical power and potentially biased estimates because a large portion of the control group opted to use the HabitWorks app during that period, requiring their exclusion from those analyses.
Funding and Disclosures
This work was supported by the National Institute of Mental Health (R01MH12937) awarded to Courtney Beard, and by Harvard Medical School’s Livingston Fellowship and McLean Hospital’s Pope-Hintz Endowed Fellowship awarded to Alexandra L. Silverman. The researchers also acknowledged the HabitWorks Community Advisory Board and the Harvard Catalyst Coalition for Equity in Research for guidance on the project, and Frances Grace Hart for assistance with data collection.
Publication Details
Title: Randomized Controlled Trial of Smartphone-Based Interpretation Bias Intervention for Anxiety and Depression
Authors: Alexandra L. Silverman, Gabriela Kovarsky Rotta, Doah Shin, Irelee Ferguson, and Courtney Beard
Affiliations: Division of Depression and Anxiety Disorders, McLean Hospital; Department of Psychiatry, Harvard Medical School; Department of Psychology, University of Vermont
Journal: Accepted for publication at the Journal of Consulting and Clinical Psychology (Version 2, dated 12/7/25)
Preprint available at: https://osf.io/preprints/psyarxiv/9kxga_v2
ClinicalTrials.gov Identifier: NCT07025486







