A "growth mindset of mental health" intervention increases the likelihood of initiating self-help on a mental health app: a randomized controlled trial.

A “growth mindset of mental health” intervention increases the likelihood of initiating self-help on a mental health app: a randomized controlled trial

Mental health apps hold potential to improve access to treatment, but engagement remains low. One barrier may be the belief that mental health cannot be meaningfully improved. In a randomized controlled trial (N = 252), we examined whether a brief growth-mindset-of-mental-health intervention—portraying mental health as improvable, not fixed—can initiate mental health app engagement among people with mental health needs. Participants completed baseline well-being measures (stress, anxiety and depression, and positive mental well-being) and were randomly assigned to the growth-mindset-of-mental-health or active control condition. Over the subsequent three months, we tracked whether they re-engaged with the app overall (i.e., completed at least a second self-help feature), and with features categorized as low-, moderate-, and high-intensity. Analyses revealed that participants in the growth-mindset condition (vs. control) with poorer baseline mental well-being (i.e., higher stress, anxiety and depressive symptoms, and lower positive mental well-being) were significantly more likely to show repeat app engagement. Moreover, those with higher anxiety and depressive symptoms were more likely to show repeat engagement with moderate-intensity features, while those experiencing higher stress and lower positive mental well-being were more likely to do so with both low- and moderate-intensity features. Together, these findings demonstrate the benefits of the growth-mindset intervention for successful initiation of self-help on a mental health app. Trial Registration: The trial was registered on ClinicalTrials.gov (Identifier: NCT07071311) on July 25, 2025.

Introduction

Many people with mental health problems do not seek, access, or sustain professional treatment1. The treatment gap, defined as the discrepancy between individuals in need of care and those who receive treatment, remains alarmingly wide, with estimates ranging from 40% in high-income countries to as high as 85% in low- and middle-income countries2,3. Multiple factors contribute to this treatment gap, including stigma, financial constraints, and poor mental health literacy in high-income countries4, with low- and middle-income countries also facing shortages of mental health professionals and treatment facilities5.

In recent years, mental health applications (MHAs) have emerged as promising and scalable tools to bridge this gap6–8. Recent systematic reviews indicate a growing number of MHAs incorporating evidence-based strategies, such as cognitive behavioral therapy (CBT) and mindfulness, which have been shown to effectively reduce the severity of common mental health problems (e.g., stress, anxiety, and depression)9,10. These apps have demonstrated efficacy in randomized controlled trials (RCTs) and across various age groups, showing that they can equip individuals struggling with poor mental well-being with skills to manage their mental health effectively. Many of these MHAs are also either free or low-cost, widely available to anyone with a smartphone, and can be used privately and on-the-go11. Taken together, these evidence-based MHAs represent a scalable avenue with the potential to deliver reliable mental health support beyond traditional clinical settings.

Despite their potential, continuous engagement with MHA features remains a longstanding challenge12,13. Most users discontinue their use after minimal interaction14–18, which substantially limits the potential benefits of these features. This pattern is evident in a systematic review by Fleming et al. (2018) examining real-world uptake of digital mental health treatments, who found that up to 80% of users logged in only once after downloading the mobile app and did not return to complete another activity19. These findings were supported by an earlier review assessing the engagement rates of 12 publicly available MHAs, where the most common pattern of use was launching the app only once20.

These trends are evident even among widely used and empirically supported apps. For instance, in Australia’s most widely adopted CBT-based mental health app (MoodGYM), which has demonstrated efficacy across RCTs in reducing mental health symptoms, only 7.0% to 16.0% of users returned for further engagement21–25. Similarly, in apps that have promoted positive well-being, such as Happify Health, only 20.6% of users voluntarily returned to complete a second activity after their initial interaction26.

Efforts to improve engagement have often focused on technical features, such as enhancing user-friendliness, personalizing feedback, customizing self-help activities, and implementing features that boost social connection27,28. However, MHAs that have incorporated such features have not consistently demonstrated higher user engagement13,29. In addition to technical enhancements, participant characteristics such as demographic factors, baseline symptom severity, and familiarity with technology use have also been examined, but no single predictor has reliably predicted higher app engagement metrics across studies, and most proposed predictors show only small effects when examined in isolation15. In a comprehensive review of predictors of engagement with digital interventions, Linardon and Fuller-Tyszkiewicz (2024) found that engagement was best predicted by the aggregation of many small effects across multiple predictors. Collectively, these findings highlight that strategies centered on enhancing app features or relying on unmodifiable participant characteristics have had limited success in sustaining engagement with MHAs15,30. This is particularly concerning for people struggling with their mental health, as repeated engagement with MHA features has often been associated with greater mental health benefits (i.e., stronger dose–response relationships31–33), particularly among individuals with higher baseline symptom severity34.

Beyond considerations of app design and participant characteristics, psychological or attitudinal factors have been shown to significantly influence whether individuals pursue mental health support for themselves, yet these factors remain largely understudied in the context of MHA engagement35. One such factor is a person’s mindset, which can shape behavior in meaningful ways. Decades of research on fixed and growth mindsets demonstrates that one’s beliefs about the malleability of personal attributes can profoundly influence a wide array of behaviors36–39. A fixed mindset is the belief that a particular attribute (e.g., intelligence, personality, willpower, interest, opportunities) is relatively unchangeable. In contrast, a growth mindset reflects the belief that an attribute is malleable and, therefore, can be developed and improved40–42. For instance, middle school students with a growth mindset of intelligence—the belief that intelligence is malleable—showed significant academic gains over the following two years compared to those with a fixed mindset41,43,44. As another example, adults who viewed their shyness as a changeable trait were more likely to actively engage in social interactions, and adopt behaviors that mitigated the negative consequences of their shyness, relative to those who believed their shyness was relatively unchangeable45.

Critically, this perspective has been extended to beliefs about the growth versus fixed nature of attributes related to mental health. For example, people who endorsed a growth mindset of emotion—believing that emotions are malleable and can be regulated and improved—were more likely to adopt adaptive coping strategies, such as actively reappraising negative thoughts46. These strategies have been linked to positive outcomes, including greater subjective well-being and improved satisfaction in interpersonal relationships47. On the other hand, those with a fixed mindset of emotion relied more on avoidance-based strategies, such as suppression, substance use, distraction, or disengagement, all of which have been linked to greater psychological distress48–50. Additionally, Schroder et al. (2015) found that undergraduate students with a growth mindset about anxiety preferred seeking psychotherapy as their primary treatment, whereas those with a fixed mindset of anxiety favored medication-based treatments, attributing their mental health difficulties to uncontrollable genetic factors and feelings of helplessness48. Participants with higher levels of positive mental well-being also tended to endorse a stronger growth mindset about mental well-being at baseline (e.g., greater happiness, more satisfying interpersonal relationships)47, positive affect and overall flourishing51,52 were positively correlated with growth mindset of well-being measures (r = 0.20 to 0.52). In contrast, those who reported more severe mental health symptoms, including psychological distress53, depression and anxiety (r = −0.37 to −0.49)53,54, were more likely to hold fixed beliefs about mental well-being. That is, a fixed mindset may serve as a psychological barrier to engaging in proactive care.

Research has also shown that a growth mindset of well-being can be induced to produce positive outcomes. Howell et al. (2016) experimentally induced a growth mindset of well-being by presenting university students with information about (a) the effectiveness of well-being interventions, (b) the role of attitudes in supporting overall health, and (c) testimonials from older adults who had improved their health. Following these experimental tasks, students reported greater openness to adopting positive lifestyle changes, such as increasing their exercise levels, participating in recreational activities, and engaging with nature51. These findings suggest that interventions aimed at fostering a growth mindset about mental health could help increase individuals’ willingness and propensity to engage with self-help treatments to manage their own mental health.

Indeed, growth-mindset interventions have successfully promoted adaptive behaviors. For example, in the education domain, O’Keefe et al. (2023) developed a growth-mindset intervention that helped first-year undergraduate students understand that new academic interests can be cultivated, and are not fixed55. Among arts students who initially lacked an affinity for math and science, this intervention boosted subsequent interest and achievement in their required math and science courses, compared to an active control condition (see other examples, refs. 39,43,56,57).

In the present study, our primary goal was to test whether a growth-mindset-of-mental-health intervention (promoting the belief that mental health is improvable, not fixed) could boost the likelihood of repeat engagement with MHAs among employees with mental health needs, as compared to an active control group. To examine this, we conducted a two-armed RCT with working adults who had access to an evidence-based MHA, Intellect, through their organizations. Intellect is a digital mental health platform that serves as the official mental healthcare provider for various government agencies, insurance companies, healthcare organizations, hospitals, and corporations across Singapore and the Asia-Pacific region. The platform offers a broad range of self-guided mental health activities rooted in evidence-based practices, such as Cognitive Behavioral Therapy (CBT), Mindfulness, Acceptance and Commitment Therapy (ACT), and Positive Psychology, and some of these features have demonstrated efficacy in previous RCTs58–64.

After completing baseline measures of well-being, participants were randomly assigned to complete either the growth-mindset-of-mental-health intervention or the active control materials. We then assessed whether they showed repeat engagement (i.e., returning to the app to complete at least one additional self-help feature) over the subsequent 3-month period. This outcome was designed to assess successful initiation of engagement, rather than casual, exploratory use65. This outcome captures the likelihood of engagement beyond the initial use, and does not assess the frequency, duration, or depth of engagement with the app’s features. If the growth-mindset-of-mental-health intervention boosts the likelihood of repeat MHA engagement, it may suggest that viewing mental well-being as improvable could motivate people to take the important step of initiating engagement with mental health resources.

Our central hypothesis is that, among people with poorer mental well-being at baseline, the growth-mindset-of-mental-health intervention (vs. control) will increase the likelihood of repeat engagement with the app’s self-help features following the intervention. Employees with poorer well-being are not only more likely to view engagement with the MHA as personally relevant, but they are more likely to hold fixed beliefs about mental well-being66. As such, they may stand to benefit more from an intervention that instills a growth mindset of mental health, and consequently take more proactive steps to support their mental well-being.

Additionally, in secondary analyses we examined whether, among individuals with lower (but not higher) baseline mental well-being, the growth-mindset-of-mental-health intervention (vs. control) would increase the likelihood of repeat engagement with app features of varying intensities—low, moderate, and high—categorized by the increasing effort and time required to complete them. Low-intensity features included mood or stress tracking; moderate-intensity features included short exercises such as breathing techniques or thought-challenging activities; and high-intensity features included more effortful activities such as guided journaling or structured, psychoeducation programs.

Results

Analyses of excluded participants

Out of the 272 participants who enrolled in the study, 136 were randomly assigned to the growth-mindset condition, while 136 were allocated to the active control condition. Among the 136 participants in the intervention group, 11 were excluded due to (a) bot-detected fraudulent responses (3/136, 2.21%), (b) duplicate responses (7/136, 5.15%), as well as (c) hasty responses (< 4 min; 1/136, 0.74%). As for the 136 control group participants, 9 were excluded for similar reasons: (a) bot-detected fraudulent responses (3/136, 2.21%), (b) duplicate responses (1/136, 0.74%), (c) blanks, nonsensical or non-response answers (e.g., “NA,” “Yes,” “Don’t know”) for all writing activities (3/136, 2.21%), and (d) hasty responses (2/136, 1.47%). This exclusion rate did not differ between conditions (ps > 0.05). Moreover, independent sample t-tests and χ2 tests confirmed that excluded participants were not significantly different from our final sample on any demographic or baseline variables (ps > 0.05).

Descriptive analyses

Table 1 presents the descriptive statistics and intercorrelations of study variables for the full sample. Table 2 presents baseline descriptives by condition. No significant differences were found between conditions on the baseline variables (ps > 0.05). Baseline growth mindset of mental health also did not differ significantly between the growth-mindset condition (M = 4.31, SD = 0.841) and the active control condition (M = 4.29, SD = 0.831) on a 1–6 point scale, t(250) = −0.13, p = 0.897.Table 1Descriptive statistics and intercorrelations among baseline variables

Baseline variables M SD 1 2 3 4 5

1. Age 32.48 7.70

2. Gender 0.450 0.500 −0.031

3. Anxiety and Depression 7.96 2.53 −0.220*** −0.040

4. Perceived Stress 11.09 2.80 −0.095 −0.050 0.657***

5. Positive Mental Well-being 23.04 4.80 0.060 0.097 −0.572*** −0.695***

*p < 0.05; **p < 0.01; ***p < 0.001. Gender was coded as Male = 0, Female = 1.

Table 2Descriptive statistics for baseline variables by condition

Variables Growth-mindset condition (n = 125) Active control condition (n = 127) p-value

M (SD) / n (%) M (SD) / n (%)

Age 33.01 (8.13) 31.95 (7.24) 0.277

Gender Female 49 (39.2%) 63 (49.6%) 0.229

Male 74 (59.2%) 63 (49.6%)

Others 2 (1.6%) 1 (0.8%)

Anxiety and Depressive Symptoms 7.92 (2.43) 8.00 (2.65) 0.803

Perceived Stress 11.0 (2.71) 11.1 (2.89) 0.825

Positive Mental Well-being 22.9 (4.85) 23.2 (4.75) 0.592

Overall, 33.7% of the full sample showed repeat engagement with self-help features during the 3-month follow-up period (29.1% in the active control condition and 38.4% in the growth-mindset condition).

Primary analysis: overall repeat engagement of self-help features

Table 3 presents the main effects of condition, baseline well-being variables, and their interaction on repeat engagement with any of Intellect’s self-help features. As shown, no significant effects of condition or baseline well-being were observed when the baseline well-being variable was at its mean level. As hypothesized, however, the growth-mindset condition interacted with each of the three baseline well-being variables to predict a significantly higher likelihood of repeat engagement with the self-help features.Table 3Interaction of baseline well-being and condition on likelihood of repeat engagement of any self-help features

Variables Outcome: repeat engagement with self-help features (Yes/No)

β SE p-value LLCI ULCI

Moderator: Anxiety and Depression

Constant −0.798 0.634 0.208 −2.04 0.445

Condition 0.407 0.276 0.140 −0.133 0.947

Anxiety and Depression −0.062 0.077 0.420 −0.212 0.088

Condition × Anxiety and Depression 0.234 0.110 0.033 0.018 0.450

Moderator: Perceived Stress

Constant −0.784 0.623 0.209 −2.01 0.438

Condition 0.410 0.276 0.137 −0.131 0.951

Perceived Stress −0.053 0.069 0.447 −0.189 0.083

Condition × Perceived Stress 0.207 0.100 0.039 0.010 0.403

Moderator: Positive Mental Well-being

Constant −0.741 0.628 0.239 −1.97 0.491

Conditiona 0.381 0.279 0.173 −0.167 0.928

Positive Mental Well-being 0.036 0.042 0.390 −0.047 0.119

Condition × Positive Mental Well-being −0.149 0.059 0.013 −0.265 −0.032

LLCI lower limit of 95% confidence interval, ULCI upper limit of 95% confidence interval.

aReference category: Active Control; Age and dummy-coded gender (reference category: “Not Female”) were controlled for in all models.

Bold values indicate statistically significant effects at p < 0.05.

Figure 1 illustrates the conditional effects of treatment condition on likelihood of repeat engagement at +1 SD and −1 SD on each baseline well-being measure. These presented a consistent pattern: among participants with lower baseline well-being, those who completed the growth-mindset intervention (vs. control condition) were significantly more likely to return to complete at least another self-help feature. By contrast, there were no differences between conditions among participants with higher baseline well-being. Thus, the intervention was more effective for those experiencing mental health problems.

Specifically, the growth-mindset intervention (vs. active control) significantly increased the likelihood of repeat engagement with self-help features among participants with poorer baseline well-being (i.e., higher anxiety/depression, β = 1.00, SE = 0.390, p = 0.010, OR = 2.72; higher perceived stress, β = 0.995, SE = 0.394, p = 0.011, OR = 2.71; lower positive mental well-being, β = 1.10, SE = 0.397, p = 0.005, OR = 3.01). By contrast, there was no condition difference in the likelihood of repeat engagement among participants who were doing better at baseline (i.e., lower anxiety/depression, β = −0.184, SE = 0.393, p = 0.640, OR = 0.832; lower perceived stress, β = −0.161, SE = 0.393, p = 0.682, OR = 0.851; higher positive mental well-being, β = −0.323, SE = 0.401, p = 0.421, OR = 0.724).

Secondary analyses: repeat engagement with different types of self-help features

To determine the nature of participants’ repeat engagement, we tested the interaction of treatment condition and baseline well-being measures separately for low-, moderate-, and high-intensity self-help features.

Table 4 presents the results for the likelihood of repeat engagement with low-intensity self-help features. As shown, no significant effects of treatment condition or baseline well-being emerged at mean levels of the baseline well-being variable. However, treatment condition significantly interacted with two of the three baseline well-being variables: perceived stress and positive mental well-being.Table 4Interaction of baseline well-being and condition on repeat engagement with low-intensity features

Variables Outcome: repeat engagement with low-intensity features (Yes/No)

β SE p-value LLCI ULCI

Moderator: Anxiety and Depression

Constant −1.02 0.657 0.119 −2.31 0.264

Conditiona 0.242 0.284 0.394 −0.314 0.799

Anxiety and Depression −0.057 0.079 0.471 −0.211 0.097

Condition × Anxiety and Depression 0.180 0.112 0.108 −0.039 0.399

Moderator: Perceived Stress

Constant −1.06 0.648 0.103 −2.33 0.214

Conditiona 0.240 0.286 0.402 −0.320 0.800

Perceived Stress −0.069 0.072 0.338 −0.210 0.072

Condition × Perceived Stress 0.211 0.103 0.040 0.009 0.414

Moderator: Positive Mental Well-being

Constant −1.01 0.651 0.120 −2.29 0.263

Conditiona 0.213 0.289 0.461 −0.353 0.779

Positive Mental Well-being 0.037 0.043 0.398 −0.048 0.122

Condition × Positive Mental Well-being −0.131 0.060 0.030 −0.249 −0.013

LLCI lower limit of 95% confidence interval, ULCI upper limit of 95% confidence interval.

aReference category: Active Control; Age and dummy-coded gender (reference category: “Not Female”) were controlled for in all models.

Bold values indicate statistically significant effects at p < 0.05.

As shown in Fig. 2, the growth-mindset intervention (vs. control) led to an increased likelihood of repeat engagement with low-intensity features among participants with higher (+ 1 SD) baseline perceived stress (β = 0.839, SE = 0.405, p = 0.038, OR = 2.31) and among those with lower (−1 SD) baseline positive mental well-being (β = 0.850, SE = 0.404, p = 0.035, OR = 2.34). In contrast, there were no condition differences among participants with lower (−1 SD) baseline perceived stress (β = −0.345, SE = 0.408, p = 0.397, OR = 0.708) or higher (+ 1 SD) positive mental well-being (β = −0.407, SE = 0.414, p = 0.325, OR = 0.665).

Table 5 presents the results for moderate intensity self-help features. As shown, treatment condition interacted with all three baseline well-being measures to predict likelihood of repeat engagement with moderate-intensity self-help features.Table 5Interaction of baseline well-being and condition on repeat engagement with moderate-intensity features

Variables Outcome: repeat engagement with moderate-intensity features (Yes/No)

β SE p-value LLCI ULCI

Moderator: Anxiety and Depression

Constant −1.47 0.724 0.043 −2.89 −0.047

Conditiona 0.198 0.312 0.526 −0.412 0.810

Anxiety and Depression −0.049 0.085 0.565 −0.216 0.118

Condition × Anxiety and Depression 0.274 0.122 0.025 0.035 0.514

Moderator: Perceived Stress

Constant −1.44 0.713 0.043 −2.84 −0.047

Conditiona 0.204 0.312 0.513 −0.408 0.817

Perceived Stress −0.029 0.077 0.708 −0.179 0.121

Condition × Perceived Stress 0.228 0.112 0.042 0.009 0.446

Moderator: Positive Mental Well-being

Constant −1.32 0.710 0.062 −2.72 0.068

Conditiona 0.217 0.313 0.489 −0.397 0.830

Positive Well-being 0.038 0.047 0.417 −0.054 0.130

Condition × Positive Well-being −0.128 0.064 0.048 −0.254 −0.002

LLCI lower limit of 95% confidence interval, ULCI upper limit of 95% confidence interval.

aReference category: Active Control; Age and dummy-coded gender (reference category: “Not Female”) were controlled for in all models.

Bold values indicate statistically significant effects at p < 0.05.

As illustrated in Fig. 3, the growth-mindset intervention (vs. control) increased the likelihood of repeat engagement with moderate-intensity features for participants with higher (+ 1 SD) baseline levels of anxiety and depression (β = 0.895, SE = 0.423, p = 0.034, OR = 2.45), higher perceived stress (β = 0.849, SE = 0.424, p = 0.045, OR = 2.34), and lower (−1 SD) baseline positive mental well-being (β = 0.837, SE = 0.433, p = 0.05, OR = 2.31). By contrast, there were no condition differences for participants with lower (−1 SD) anxiety and depression (β = −0.494, SE = 0.456, p = 0.279, OR = 0.610), lower perceived stress (β = −0.425, SE = 0.459, p = 0.355, OR = 0.654), nor for participants with higher (+ 1 SD) positive mental well-being (β = −0.388, SE = 0.446, p = 0.385, OR = 0.679).

None of the baseline well-being measures interacted with treatment condition to predict the likelihood of repeat engagement with high-intensity self-help features (see Table 6).Table 6Interaction of condition and baseline well-being on repeat engagement with high-intensity features

Variables Outcome: repeat engagement with high-intensity features (Yes/No)

β SE p-value LLCI ULCI

Moderator: Anxiety and Depression

Constant −1.71 0.617 0.006 −2.91 −0.498

Conditiona 0.425 0.263 0.107 −0.091 0.942

Anxiety and Depression −0.042 0.072 0.562 −0.183 0.099

Condition × Anxiety and Depression 0.174 0.105 0.097 −0.031 0.379

Moderator: Perceived Stress

Constant −1.67 0.605 0.006 −2.85 −0.480

Conditiona 0.425 0.263 0.106 −0.090 0.940

Perceived Stress −0.016 0.065 0.810 −0.143 0.112

Condition × Perceived Stress 0.106 0.094 0.262 −0.079 0.291

Moderator: Positive Mental Well-being

Constant −1.65 0.606 0.006 −2.84 −0.463

Conditiona 0.409 0.264 0.121 −0.107 0.926

Positive Mental Well-being 0.012 0.040 0.763 −0.066 0.090

Condition × Positive Mental Well-being −0.080 0.056 0.153 −0.189 0.030

LLCI lower limit of 95% confidence interval, ULCI upper limit of 95% confidence.

aReference category: Active Control; Age and dummy-coded gender (reference category: “Not Female”) were controlled for in all models.

Bold values indicate statistically significant effects at p < 0.05.

Discussion

The present study investigated whether a brief, single-session growth-mindset-of-mental-health intervention could promote the successful initiation of engagement with self-help features on an evidence-based MHA, among employees with poorer baseline mental well-being. Consistent with our hypotheses, the intervention significantly increased the likelihood of repeat app engagement among participants who began the study with diminished well-being at baseline (i.e., higher levels of anxiety and depression, perceived stress, or lower positive mental well-being), relative to an active control condition. By contrast, and as predicted, the intervention had no significant effect on repeat self-help engagement among participants with relatively high well-being at baseline. Our secondary analyses revealed that these moderation effects were most evident for low- and moderate-intensity app features. For low-intensity features, the intervention condition showed increased repeat engagement among those with higher stress and lower positive well-being at baseline. For moderate-intensity features, the intervention condition showed increased repeat engagement among those with higher anxiety and depression, higher stress, and lower positive well-being. No effects of the intervention were observed for high-intensity features, regardless of participants’ mental health status at baseline.

Our findings show that an online, single-session growth-mindset intervention, when precisely designed to target a relevant psychological barrier, can meaningfully promote the successful initiation of MHA engagement among people who might otherwise be less likely to seek or persist with mental health support.

One possible explanation for this effect is that individuals with more clinically relevant symptoms may have a greater psychological need for change, which might increase their receptiveness to belief-shifting interventions. As shown in previous studies, individuals often perceive clinical symptoms to be more urgent, which can create a psychological state that heightens their motivation to seek relief67. When presented with the idea that mental health is malleable and can be improved through effort, they may be more likely to process the message deeply, recognize its personal relevance, and initiate seeking treatment. However, this proposed mechanism must be confirmed by future research.

Our study contributes to the growing body of research on wise interventions, which are brief, targeted interventions designed to modify specific psychological processes that may initiate recursive and lasting improvements in behavior68, as demonstrated by previous growth-mindset interventions55,56,69–71. Consistent with this framework, we targeted a specific belief that may undermine sustained engagement in mental health care, namely the belief that mental health is fixed and cannot be meaningfully improved. Based on previous research linking fixed mindsets to the avoidance of proactive behaviors for well-being48–50, we theorized that a fixed mindset of mental health would serve as an important psychological barrier to engaging with mental health resources, such as an MHA. Our intervention was designed to present the understanding that mental health is improvable and can be strengthened through intentional effort. After completing this treatment, participants in the growth-mindset condition showed a significantly higher likelihood of repeat engagement with self-help features on a digital mental health resource, relative to the control condition. We believe that these findings contribute to the literature on wise interventions by suggesting that a growth-mindset intervention can promote early treatment-seeking behaviors by initiating engagement on a MHA.

Our findings also speak to emerging directions in mindset science and wise interventions, particularly in response to recent calls to better understand “for whom” and under what conditions these interventions are most effective69,72. The long-term effects of mindset and other wise interventions hinge not only on shifting a targeted belief, but also on the affordances of the situation—that is, whether the surrounding environment supports or hinders the intervention’s intended effect—and on the needs of the person receiving the intervention. For instance, O’Keefe et al. (2023) developed a growth-mindset-of-interest intervention that encouraged incoming arts undergraduates to view academic interests as developable, not fixed. They found that the intervention was especially effective at boosting interest and achievement (i.e., grades) in math and science courses among students who did not initially have an interest in those subjects, but who nonetheless had opportunities to engage with those subjects55. Similarly, Paunesku et al. (2015) found that a growth-mindset-of-intelligence intervention had its strongest effects among students who were underperforming and at risk of dropping out57. These findings suggest that individual differences can play a powerful role in shaping the effectiveness of an intervention. In line with these studies and others, we found that employees who benefited most from our intervention were those experiencing greater mental health difficulties at baseline. The growth-mindset-of-mental-health intervention may have been particularly relevant to people with poorer baseline well-being. Importantly, they may have been supported by an environmental affordance: participants’ workplaces had made the Intellect app freely available to them, giving them an opportunity to act on the intervention’s message. Together, these personal and environmental conditions may have contributed to their increased likelihood of initiating self-help engagement on the MHA, an idea worth pursuing more deeply in future work.

The pattern of repeat engagement emerged only for the MHA’s low- and moderate-intensity features. Employees with higher perceived stress or lower positive well-being at baseline were significantly more likely to return to the app and engage with low-intensity features, such as mood tracking. Meanwhile, employees with poorer baseline mental well-being across all three measures—higher perceived stress, higher anxiety and depressive symptoms, and lower positive well-being—were significantly more likely to return to engage with moderate-intensity features, such as breathing techniques and thought-challenging exercises. Yet, we did not find a greater likelihood of repeat engagement with high-intensity features. One possible reason is that these features were more effortful and may not have matched participants’ immediate needs73. High-intensity features often demand a level of cognitive effort, time, and emotional investment that can feel overwhelming for individuals already experiencing significant psychological symptoms. For example, features like guided journaling or structured learning paths typically involve deep reflection of newly learned psychoeducational content, and skill application through writing, which may be particularly taxing for individuals struggling with their mental health. It is also possible that participants were less likely to engage with high-intensity features due to other concerns about confronting distressing emotions or perceiving a lack of short-term relief. In short, low- and moderate-intensity features may have felt more manageable and accessible than high-intensity features, and thus were more likely to be selected for repeat engagement.

In line with the Transtheoretical Model, which conceptualizes behavior change as a dynamic progression through five stages—precontemplation (not intending to change), contemplation (considering change), preparation (making small changes), action (actively engaging in new behavior), and maintenance (sustaining change over time)74,75—it is possible that the growth-mindset-of-mental-health intervention motivated participants to progress from precontemplation to preparation. This may help explain why the treatment showed higher initiation of engagement with lower- and moderate-intensity features, where the activities may have felt more feasible, but did not facilitate progression to more effortful high-intensity features that may correspond more closely to the action stage. Given that wise interventions, by targeting a specific fixed belief, can set in motion recursive processes that accumulate and amplify gains over time76–80, it is also possible that a three-month window was insufficient to observe further progression toward engagement with more intensive features. Alternatively, because participants were not drawn from a more severely distressed clinical population, there may have been less readiness or need to progress to more advanced stages of change. Future research could test these possibilities directly by assessing stages of change before and after intervention exposure, and by examining whether booster sessions would facilitate greater progression toward higher-intensity engagement over time. That said, in the present study, we did not evaluate whether engagement with these features led to subsequent improvements in mental health outcomes. Future research should also incorporate follow-up assessments to test whether repeat engagement with these evidence-based features leads to subsequent improvements in well-being.

From a practical standpoint, our findings carry important implications for mental health care delivery. While MHAs offer a promising avenue for expanding access to evidence-based mental health support, particularly for individuals who lack the time and resources to pursue in-person care, their potential continues to be significantly limited by persistently low engagement rates. Our findings suggest that existing efforts to promote engagement (such as technical enhancements16,27,28) may not be sufficient on their own to drive engagement. Instead, our research suggests that mindsets may play a critical role, at least in early-stage engagement. We argue that replacing the view that mental health is unchangeable with the belief that it is improvable may be an underutilized strategy for increasing repeat engagement with MHAs. It is possible that this mindset shift may serve as a motivational gateway that technical features alone cannot unlock. However, because we did not assess growth mindset change post-intervention, we cannot determine whether changes in mindset mediated the observed increase in repeat engagement with app features. Future research would benefit from longitudinal designs that monitor changes in mindsets about mental health over time and examine whether these changes predict subsequent patterns of app engagement. Nonetheless, our study is among the first to apply growth mindset theory to mobile app engagement, and also one of the very few to examine help-seeking using objective behavioral data.

The present study has several notable strengths. First, our primary outcome relied on objective behavioral engagement data rather than self-reported intentions or retrospective recall of engagement behaviors. This is important because prior research on a growth mindset related to well-being primarily examined self-reported openness to lifestyle change, rather than objective engagement with mental health support51. Numerous digital mental health interventions also rely on self-reported usage, which may be easily influenced by recall bias or social desirability81. By directly tracking participants’ app use, the present study provides a more ecologically valid assessment of the effects of a growth-mindset-of-mental-health intervention on initiating real-world help-seeking behavior.

Second, we examined moderation of intervention effects by baseline mental well-being, which enabled us to identify for whom the intervention was most effective. This focus on heterogeneity aligns with growing calls within the behavioral sciences to move beyond average treatment effects and to examine how interventions may be adapted to specific populations and contexts69,72.

Finally, the growth-mindset-of-mental-health intervention increased repeat self-help engagement within a predominantly Asian sample, a culture where mental health stigma remains relatively high82. Observing effects in this culturally constrained help-seeking environment suggests potential applicability beyond this specific setting, although the extent to which the intervention generalizes to cultures that are more open to mental health treatment remains an empirical question for future research.

Several limitations should also be noted. First, our app engagement data were collected over a 3-month period, which provides a meaningful snapshot of short-term behavioral use but does not capture longer-term engagement trajectories. Although it is notable that 38.4% of our participants in the growth-mindset condition showed repeat engagement, particularly in light of prior evidence that as few as 20% of digital mental health app users return after an initial login19, this finding does not indicate whether this early engagement would be sustained over time. As such, it remains uncertain whether the observed effects of the intervention would persist, intensify, or diminish with time. Given that wise mindset interventions have been theorized to initiate recursive processes that can amplify over time68,77,83, future studies should incorporate extended follow-up periods to assess long-term usage patterns.

Second, although our primary outcome was based on objective behavioral data, we were unable to analyze mental health outcomes longitudinally. It remains an open question whether increased and sustained engagement with the app translates into clinically meaningful improvements in mental health. Prior trials of the Intellect app have demonstrated such benefits59,61, so future research should examine whether behavioral engagement serves as a mediator between mindset interventions and improvements in psychological well-being.

Third, the study did not include a formal post-intervention manipulation check assessing growth mindset change. As a result, we cannot directly test whether the growth-mindset-of-mental-health intervention altered the targeted belief, nor whether any change in mindset mediated subsequent engagement behavior on the app. Although the intervention was designed to promote a growth mindset of mental health, alternative explanations cannot be ruled out, such as the possibility that the intervention increased participants’ attention to mental health content, which in turn may have led to re-engagement with Intellect’s self-help features. Future studies should include post-intervention assessments of growth mindset and conduct mediation analyses to clarify the psychological mechanisms through which belief-shifting wise interventions influence engagement with MHAs.

Fourth, our engagement outcome captured only whether participants returned to use the app beyond completing an initial self-help feature, and did not assess the frequency, duration, or depth of engagement with the app’s content. As a result, we cannot draw conclusions about the quality of engagement, such as the extent to which participants actively practiced or applied the skills provided by the app, or whether such engagement translated into improvements in mental health over time. Future studies can consider collecting more fine-grained app engagement metrics (i.e., time spent on each feature) alongside follow-up assessments of mental health outcomes to better understand how different forms of engagement relate to treatment benefits. Specifically, with these additional data, future researchers could test the full mechanism proposed in this study: whether exposure to the growth-mindset intervention leads to a stronger growth mindset of mental health; whether such changes increase the initiation of app engagement; whether early repeat engagement develops into sustained engagement; and ultimately, whether this process contributes to improvements in mental well-being over time.

Lastly, the generalizability of these findings should be considered in light of the study sample and context. The intervention itself (e.g., the reading and writing activities) may depend in part on participants’ educational attainment, and initiation of app engagement may also be influenced by digital literacy or familiarity with technology. The sample consisted primarily of employed adults with relatively mild symptoms on average and higher educational attainment. As such, it remains uncertain whether these findings would generalize to more severely distressed clinical populations, to individuals with lower educational attainment or digital literacy, or individuals without access to employer-supported digital mental health tools.

This study demonstrates that a brief, scalable growth-mindset-of-mental-health intervention can increase repeat engagement with a mental health app, particularly among those who may need support most. By targeting users’ beliefs about the changeability of mental health, this intervention offers a novel and promising strategy to address a persistent challenge in digital mental health: helping people successfully initiate treatment and support.

Method

Transparency and openness

We report how we determined our sample size, the statistical software used, the analyses conducted, the statistical findings, and all measures included in this study. We follow standards for reporting psychological research in scientific journals84. This information and the research materials are available at the study’s Open Science Framework repository (https://osf.io/nsw2t/overview).

This study received ethical approval from the Institutional Review Board of National University of Singapore (ID: NUS-IRB-2022-253). The trial was conducted in full compliance with ethical standards, including the Declaration of Helsinki. All participants provided informed consent prior to their involvement in the study. We registered the study protocol at ClinicalTrials.gov (Identifier: NCT07071311) on 07-25-2025, to provide a public record of the trial and promote transparency. Because registration occurred after the study began, this study was not preregistered.

Study design

We conducted a randomized controlled trial (RCT) to test the effectiveness of a brief intervention promoting a growth mindset of mental health, as compared to an active control condition that promoted the benefits of digital technology in the workplace. An active control condition was chosen as a conservative comparison, consistent with prior growth mindset RCTs, in which active controls are widely regarded as the gold standard in social-psychological intervention research39,55,56,85,86. The active control condition ensures that participants in both conditions engage in similarly-structured activities that vary only in their content, which helps control for expectancy effects and attention-related confounds. We tested whether the intervention would lead participants with lower baseline well-being—characterized by higher stress, anxiety and depression, or lower positive mental well-being—to show repeat engagement with self-help features on the Intellect app.

Procedure

Recruitment took place from March to September 2023. Employees from 83 organizations received an email newsletter coordinated by Intellect’s client relations teams in collaboration with the human resource managers from their respective organizations. This email provided a brief overview of our research study, along with instructions on how to download and register for an account on the Intellect app. The final section of the newsletter included a link to a Qualtrics survey, where employees could access the Participant Information Sheet (PIS) and provide their informed consent to participate. According to Intellect’s internal data, 9698 employees opened the newsletter. Of those, 2619 clicked on the survey link and reviewed the PIS, which outlined the study’s aims, eligibility criteria, data protection measures, participant rights, investigator contact details, and study incentives. The aims of our study were partially concealed, as employees were told that the purpose of the study was solely to examine the use of a mobile health application (“Intellect”) in the workplace. This partial concealment was necessary to avoid demand effects, where expectations of a desired behavior can lead to improvements in that behavior87. A total of 307 employees provided informed online consent to participate. After providing consent, employees completed pre-screening measures. Inclusion required participants to report (a) being able to read and understand English, (b) being 18 years or older, (c) being employed, and (d) not using other mental health mobile applications or services concurrently. Thirty-five participants were excluded for failing to meet any one of these eligibility criteria (see Fig. 4).

The remaining 272 participants proceeded to complete a demographic questionnaire and three baseline mental well-being assessments: (a) anxiety and depressive symptoms, (b) perceived stress, and (c) positive mental well-being. Additional measures unrelated to the state of participants’ mental health (e.g., self-stigma of seeking professional help, social support, marital status) were also collected and are detailed in Supplementary Material. Participants were then randomly assigned by Qualtrics to one of two conditions (i.e., growth-mindset intervention or active control); thus, participants and researchers were blind to treatment condition. After completing the treatment materials, participants were given a SGD $10 online gift card. Participants then either created a new Intellect account (N = 149, 59.1%) or signed into a dormant account they created at an earlier time (N = 103, 40.9%). The latter indicates that 40.9% of participants had previously downloaded Intellect before the study. Compensation was not contingent on downloading the app or using any app features during the 3-month follow-up period. Participants were also not instructed to complete any specific app features after the baseline session.

All responses to the treatment materials were reviewed for quality in four ways, which resulted in several exclusions (see Fig. 4). First, we excluded responses identified as likely bots by Qualtrics’ reCaptcha scores (i.e., scores under .50; see www.qualtrics.com/support/survey-platform/survey-module/survey-checker/fraud-detection). Second, we removed duplicate responses based on repeated email addresses. Third, survey completion times were reviewed. Based on our pilot testing, which showed an average treatment completion time of 10 min, and recommendations to exclude responses completed in less than 40% of the average completion time88, any responses submitted in under 4 min were excluded. Finally, participants who left all three writing activities in the treatment materials entirely blank or provided nonsensical or non-response answers (e.g., blank, “NA,” “Yes,” “Don’t know”) were excluded, as this was evidence of inattention. The final sample consisted of 252 participants. No participants were excluded based on their subsequent app use or repeat-engagement outcomes.

Treatment materials

This single-session growth-mindset intervention was developed using methods from the family of “Wise Interventions”68,83,89. It involved reading and reflective writing activities that portrayed mental health as improvable rather than fixed. To avoid eliciting demand, reactance, or stigmatization, the materials were not presented as an intervention or as addressing a deficit in employees. Instead, they were framed as part of an effort to gather insights to better support employees as a whole and improve their mental health. Participating employees were thus positioned as experts and co-creators of an intervention for other employees77,90.

Participants first read stories of three employees who shared their experiences overcoming mental health challenges in the workplace. Each described their struggles with workplace issues and their journey toward understanding that mental health is changeable, which includes taking proactive steps to improve it (e.g., deliberate self-care, or seeing a therapist). The stories represented diverse industries and professional roles, allowing participants to connect with a range of experiences. Since employees are likely to perceive their peers as credible and important sources of information, reading about how those peers proactively managed their mental health in response to workplace challenges may be particularly persuasive. This aspect of the material leveraged informational social influence91, as individuals often look to peers for guidance when forming beliefs and making decisions about their well-being. For example, one story illustrated how a director at a technology company suffering from anxiety attacks came to understand that his mental health was improvable and learned techniques to manage the anxiety after attending sessions with a clinician.

Next, participants wrote a brief essay reflecting on a time they successfully took action to improve their mental health while facing workplace challenges. They responded to the following prompt:

Take a moment and recall a time in your life when your mental health was poor but you took actions to improve your mental health. What did you do? In what ways did it improve your mental health and well-being?

They were informed that their essays would be used to support future employees in coping with mental health challenges at work. This activity leveraged the “saying-is-believing” effect, a technique whereby people advocate for a position to a relevant audience (i.e., other employees, in the present case), which strengthens their own endorsement of that belief92. By articulating that mental health is malleable and linking this idea to their own experiences, participants can come to endorse the belief more strongly for themselves (see Supplementary Material for more sample responses). As a representative example, one participant wrote:

People who believe that mental health can improve are more likely to take proactive steps toward improving it. Without the belief that change is possible, there’s little motivation to take action. This is why adopting the right mindset is essential.

Next, employees completed two additional brief essays, which again leveraged the saying-is-believing effect (see Supplementary Material for essay examples). In these essays, they offered advice to help other employees manage their mental health in response to the following prompts: “How might this understanding that mental health can improve help people with their own mental well-being, especially when work becomes stressful?” and “What can people do to improve their mental health when experiencing difficulties at work?” While writing, participants were also encouraged to draw on both the stories of other employees and their own experiences. By positioning participants as experts, helping to advise other employees at their company, the activity aimed to further strengthen their personal endorsement of a growth mindset68.

This control exercise was identical in approach and structure but instead promoted the use of technology in the workplace93. Unlike the intervention, the control exercise did not address mental health or its potential for development. Similar to the intervention, however, the control materials positioned employees as experts offering guidance to other employees, rather than as recipients of an intervention who were in need of help. These materials also emphasized a growth-oriented process, specifically focusing on the development of expertise in workplace technology as a crucial skill for overcoming challenges in the workplace. Therefore, these materials provided a rigorous comparison to test the effects of the intervention condition.

Participants first read stories about employees’ experiences managing their workload using workplace technology. Each described their challenges with learning and adapting to digital tools but ultimately came to see the value in technology use at work. This included viewing learning digital tools as manageable, and taking proactive steps, such as consulting with colleagues or making time to practice and improve. Different industries and roles were represented to resonate with participants’ own experiences at work.

Next, control participants wrote a brief essay about their personal experiences using digital tools to enhance their efficiency at work (see Supplementary Material for more sample responses). In response to the prompt “Take time to think about the digital technology you love most at work. What is the digital technology you use, and how does it help you?”, one representative response stated:

The growing adoption of online collaboration tools, such as Microsoft Teams and Microsoft Sharepoint, has greatly enhanced flexibility and efficiency. These tools allow us to communicate and collaborate on documents virtually at our convenience, enabling contributions to work without sacrificing personal time—This is an especially valuable benefit when working in a global organization. This aspect of digital technology is something I truly appreciate, as it has significantly supported me in my work.

After this reflection, participants in the control condition completed two additional essays, for which they were asked to offer advice to others on how to effectively use new technologies in the workplace. Their responses were guided by the following prompts: “Next, how do employees typically discover digital technologies at work?” and “How do employees apply those newly-discovered technologies at work?” Similar to the intervention materials, control participants were encouraged to draw from the employee stories they read and their own experiences with workplace technology.

Measures

Following past mindset intervention research, we collected participants’ age and gender for use as covariates because these variables have previously been associated with mindset-related outcomes56,94. (Additional demographic and baseline variables not used in present analyses include employment status, highest level of education attained, marital status, occupational position, length of time in their current employment role, prior app usage, and baseline growth mindset of mental health. Controlling for these additional factors did not change the significance or pattern of any of our findings).

Participants’ baseline (pre-intervention) anxiety and depressive symptoms were assessed with the widely used 4-item Patient Health Questionnaire (PHQ-4)95, which has demonstrated excellent internal consistency and convergent validity96. Two items assessed depressive symptoms over the past 2 weeks (e.g., “Over the past 2 weeks, how often have you been feeling down or hopeless?”), and the other two assessed anxiety levels (e.g., “Over the past 2 weeks, how often have you felt nervous, anxious or on edge?”). Each item was rated 0 = Not at all to 3 = Nearly every day. Higher PHQ-4 scores indicate more severe anxiety and depressive symptoms. The cutoffs for mild, moderate, and severe anxiety and depressive symptoms are 3, 6, and 9, respectively. The Cronbach’s alpha for PHQ-4 in our current sample was high (α = 0.799).

Participants’ baseline stress levels over the last month were assessed using the 4-item Perceived Stress Scale (PSS-4)97, which has good convergent validity with other aspects of mental well-being98. A sample item is “In the last month, how often have you felt that difficulties were piling up so high that you could not overcome them?” (1 = never to 5 = very often). Higher scores reflect more perceived stress. The Cronbach’s alpha in our sample was good (α = 0.775).

Participants’ baseline positive well-being was assessed using the 7-item Short Warwick-Edinburgh Mental Well-being Scale (SWEMWBS)99,100, which has demonstrated good internal validity and test-retest reliability101,102. Sample items include “I’ve been feeling optimistic about the future” and “I’ve been feeling relaxed” (1 = none of the time to 5 = all of the time). Scores are summed, ranging from 7 to 35, with higher scores indicating more positive well-being. The Cronbach’s alpha in the current sample was high (α = 0.876).

Engagement with app self-help features

Repeat app engagement was operationalized as the completion of at least one additional self-help feature beyond the first feature completed. This outcome captures whether participants returned to engage with another self-help feature on the app after completing an initial feature, but does not capture sustained or long-term engagement with the app. Nevertheless, completing another self-help feature reflects a more intentional and proactive effort to improve mental health. In contrast, using only a single feature is often interpreted as casual or exploratory use, rather than as evidence of successful initiation of self-help app engagement103–106. In this study, we assessed repeat engagement within 3 months of completing all assigned condition materials. Additionally, as secondary analyses, we assessed this with respect to distinct types of self-help features: low-, moderate-, and high-intensity. We also report app-use metrics for the overall sample in Table S2 of the Supplementary Material, including the types and average number of completed self-help features during the 3-month follow-up period.

Intellect’s full range of guided self-help features include “Mood and Stress Monitoring,” “Daily Toolkits,” “Rescue Sessions,” “Guided Journaling,” and “Learning Paths.” These features are categorized into three levels of intensity based on engagement time. Low-intensity features require relatively little effort and take less than 1 min, which include “Mood and Stress Monitoring.” Moderate-intensity features (“Daily Toolkits” and “Rescue Sessions”) require more effort and about 2 to 5 min to complete. Finally, high-intensity features (“Guided Journals” and “Learning Paths”) require the most effort and at least 10 min of focused engagement, including activities like writing and reflection. All three types of features on Intellect have been demonstrated to improve mental well-being in previous randomized controlled trials58–64.

“Mood and Stress Monitoring” allows users to log their current mood or stress levels. Users record their mood by sliding a scale and selecting from emojis labeled “terrible,” “not good,” “content,” “good,” and “happy.” A similar scale is used for stress, with options ranging from “very stressed” to “not stressed at all.” After logging their entries, users can view their previous moods and stress levels in a calendar format, allowing them to track emotional trends over time and gain insights into their emotional patterns. In line with recent studies107,108, mood-tracking features have been shown to help users identify patterns and potential causes of mood shifts, enabling them to take more informed actions to improve their well-being. We operationalized repeat engagement with low-intensity features by logging mood and stress levels at least twice (29.4% of the sample showed repeat engagement of low-intensity features).

“Rescue Sessions” are designed to alleviate in-the-moment anxiety through mindfulness and cognitive restructuring exercises. These sessions target common challenges, such as overthinking, high stress, low resilience, and burnout, and users may choose sessions that address their immediate concerns. Such challenges are often sustained by negative thought patterns109. Similar to CBT and mindfulness, “Rescue Sessions” guide users to either challenge the validity of their thoughts through Socratic questioning110,111 or to observe their thoughts from a distance and in a non-reactive manner112,113. Each session concludes with a recommendation for users to apply these techniques in their daily life.

Similarly, “Daily Toolkits” are brief mindfulness-centered features aimed at promoting a structured self-care routine throughout the day while addressing any negative thoughts that may arise. The morning segment consists of deep breathing exercises to focus on the present moment. Grounding techniques are introduced in the afternoon segment to help maintain mindfulness. Finally, the night segment promotes restful sleep through meditation and cognitive restructuring techniques114. We operationalized repeat engagement with moderate-intensity features by completing at least two “Rescue Sessions,” or “Daily Toolkits,” or at least one of each (23.4% of the sample showed repeat engagement of moderate-intensity features).

“Learning Paths” are self-guided, structured activities that provide in-depth psychoeducation and CBT-based self-help for various mental health problems. These features begin with an introduction to CBT principles, followed by guidance to enhance self-awareness of thoughts. Users then write about their own maladaptive thought–action cycles before receiving additional guidance to cognitively reframe these dysfunctional patterns. As an example, the “Feeling Depressed” learning path teaches users about common cognitive distortions, such as all-or-nothing thinking or overgeneralization115. Through structured exercises, users then reflect on their own distorted beliefs and write more balanced, evidence-based statements to reframe these negative thoughts.

“Guided Journals” prompt users to reflect on specific themes by recording their thoughts and then guiding them through cognitive restructuring techniques to reframe negative reflections116. These journals cover a broad range of themes, from challenges like ‘anxiety and depression’ and ‘poor self-esteem,’ to proactive improvements like ‘goal-setting’ and ‘sleep improvement,’ to maintaining mental well-being by highlighting positive factors (e.g., gratitude for positive events that occurred during the day). We operationalized repeat engagement with high-intensity features by completing at least two “Learning Paths” or “Guided Journals” or at least one of each (29.0% of the sample showed repeat engagement of high-intensity features).

Analytic approach

All analyses were conducted in SPSS V29.0. To test hypotheses, we used the PROCESS macro V.4.2. (Model 1) in SPSS117 to examine whether baseline well-being variables (anxiety and depression, perceived stress, or positive well-being) moderated the effect of treatment condition on the likelihood of repeat engagement of self-help features. Continuous independent and moderating variables were mean-centered. Following past mindset intervention research55, age (in years) and gender (1 = female, 0 = male) were included as covariates. Overall repeat engagement with self-help features was coded as a binary variable (0 = no repeat engagement, 1 = repeat engagement). Additionally, repeat engagement with each type of self-help feature (low-, moderate-, or high-intensity) was coded as binary (0 = no repeat engagement, 1 = repeat engagement). Since the outcome variables are binary, logistic regression was used for analyses.

For our primary analysis, we tested the interaction of condition and baseline well-being on overall repeat engagement. Because we had three distinct baseline well-being moderators, we conducted a separate model for each one (totaling three central models). Subsequently, to better understand which type of features drove our primary effects, we conducted those same three models separately for low-, moderate-, and high-intensity self-help features. Statistical significance was set at α = 0.05. Controlling for the age and gender covariates did not change the significance or patterns of our findings.

Supplementary information

References

1.Nochaiwong S Global prevalence of mental health issues among the general population during the coronavirus disease-2019 pandemic: a systematic review and meta-analysis Sci. Rep. 2021 11 10173 10.1038/s41598-021-89700-8 33986414 Nochaiwong, S. et al. Global prevalence of mental health issues among the general population during the coronavirus disease-2019 pandemic: a systematic review and meta-analysis. Sci. Rep.11, 10173 (2021).10.1038/s41598-021-89700-8 PMC8119461 33986414

2.Ndetei DM Mutiso V Osborn T Moving away from the scarcity fallacy: three strategies to reduce the mental health treatment gap in LMICs World Psychiatry 2023 22 163 36640407 10.1002/wps.21054 Ndetei, D. M., Mutiso, V. & Osborn, T. Moving away from the scarcity fallacy: three strategies to reduce the mental health treatment gap in LMICs. World Psychiatry22, 163 (2023).10.1002/wps.21054 PMC9840495 36640407

3.World Health Organization. mhGAP Intervention Guide for Mental, Neurological and Substance Use Disorders in Non-Specialized Health Settings (World Health Organization, 2016).27786430

4.Mojtabai R Barriers to mental health treatment: results from the National Comorbidity Survey Replication Psychol. Med. 2011 41 1751 1761 21134315 10.1017/S0033291710002291 PMC3128692 Mojtabai, R. et al. Barriers to mental health treatment: results from the National Comorbidity Survey Replication. Psychol. Med.41, 1751–1761 (2011).10.1017/S0033291710002291 PMC3128692 21134315

5.Bruckner TA The mental health workforce gap in low- and middle-income countries: a needs-based approach Bull. World Health Organ. 2011 89 184 194 21379414 10.2471/BLT.10.082784 PMC3044251 Bruckner, T. A. et al. The mental health workforce gap in low- and middle-income countries: a needs-based approach. Bull. World Health Organ.89, 184–194 (2011).10.2471/BLT.10.082784 PMC3044251 21379414

6.Goldberg SB Mobile phone-based interventions for mental health: a systematic meta-review of 14 meta-analyses of randomized controlled trials PLOS Digit. Health 2022 1 e0000002 35224559 10.1371/journal.pdig.0000002 Goldberg, S. B. et al. Mobile phone-based interventions for mental health: a systematic meta-review of 14 meta-analyses of randomized controlled trials. PLOS Digit. Health1, e0000002 (2022).10.1371/journal.pdig.0000002 PMC8881800 35224559

7.Marshall JM Dunstan DA Bartik W Clinical or gimmickal: the use and effectiveness of mobile mental health apps for treating anxiety and depression Aust. N. Z. J. Psychiatry 2020 54 20 28 31552747 10.1177/0004867419876700 Marshall, J. M., Dunstan, D. A. & Bartik, W. Clinical or gimmickal: the use and effectiveness of mobile mental health apps for treating anxiety and depression. Aust. N. Z. J. Psychiatry54, 20–28 (2020).10.1177/0004867419876700 31552747

8.Yin H Mobile mental health apps in China: systematic app store search J. Med. Internet Res. 2020 22 e14915 32716301 10.2196/14915 Yin, H. et al. Mobile mental health apps in China: systematic app store search. J. Med. Internet Res.22, e14915 (2020).10.2196/14915 PMC7418006 32716301

9.Linardon J The efficacy of app-supported smartphone interventions for mental health problems: a meta-analysis of randomized controlled trials World Psychiatry 2019 18 325 336 31496095 10.1002/wps.20673 Linardon, J. et al. The efficacy of app-supported smartphone interventions for mental health problems: a meta-analysis of randomized controlled trials. World Psychiatry18, 325–336 (2019).10.1002/wps.20673 PMC6732686 31496095

10.Linardon J Current evidence on the efficacy of mental health smartphone apps for symptoms of depression and anxiety: a meta-analysis of 176 randomized controlled trials World Psychiatry 2024 23 139 149 38214614 10.1002/wps.21183 Linardon, J. et al. Current evidence on the efficacy of mental health smartphone apps for symptoms of depression and anxiety: a meta-analysis of 176 randomized controlled trials. World Psychiatry23, 139–149 (2024).10.1002/wps.21183 PMC10785982 38214614

11.Anthes E Pocket psychiatry: mobile mental-health apps have exploded onto the market, but few have been thoroughly tested Nature 2016 532 20 24 10.1038/532020a 27078548 Anthes, E. Pocket psychiatry: mobile mental-health apps have exploded onto the market, but few have been thoroughly tested. Nature532, 20–24 (2016).

12.Boucher EM Raiker JS Engagement and retention in digital mental health interventions: a narrative review BMC Digit. Health 2024 2 52 10.1186/s44247-024-00105-9 Boucher, E. M. & Raiker, J. S. Engagement and retention in digital mental health interventions: a narrative review. BMC Digit. Health2, 52 (2024).

13.Valentine L A meta-analysis of persuasive design, engagement, and efficacy in 92 RCTs of mental health apps NPJ Digit. Med. 2025 8 229 10.1038/s41746-025-01567-5 40301581 Valentine, L. et al. A meta-analysis of persuasive design, engagement, and efficacy in 92 RCTs of mental health apps. NPJ Digit. Med.8, 229 (2025).10.1038/s41746-025-01567-5 PMC12041226 40301581

14.Baumel A Muench F Edan S Kane JM Objective user engagement with mental health apps: systematic search and panel-based usage analysis J. Med. Internet Res. 2019 21 e14567 31573916 10.2196/14567 Baumel, A., Muench, F., Edan, S. & Kane, J. M. Objective user engagement with mental health apps: systematic search and panel-based usage analysis. J. Med. Internet Res.21, e14567 (2019).10.2196/14567 PMC6785720 31573916

15.Linardon J Fuller-Tyszkiewicz M Exploration of the individual and combined effects of predictors of engagement, dropout, and change from digital interventions for recurrent binge eating Int. J. Eat. Disord. 2024 57 1202 1212 38410869 10.1002/eat.24175 Linardon, J. & Fuller-Tyszkiewicz, M. Exploration of the individual and combined effects of predictors of engagement, dropout, and change from digital interventions for recurrent binge eating. Int. J. Eat. Disord.57, 1202–1212 (2024).10.1002/eat.24175 38410869

16.Saleem M Understanding engagement strategies in digital interventions for mental health promotion: scoping review JMIR Ment. Health 2021 8 e30000 34931995 10.2196/30000 Saleem, M. et al. Understanding engagement strategies in digital interventions for mental health promotion: scoping review. JMIR Ment. Health8, e30000 (2021).10.2196/30000 PMC8726056 34931995

17.Szinay D Influences on the uptake of and engagement with health and well-being smartphone apps: systematic review J. Med. Internet Res. 2020 22 e17572 32348255 10.2196/17572 Szinay, D. et al. Influences on the uptake of and engagement with health and well-being smartphone apps: systematic review. J. Med. Internet Res.22, e17572 (2020).10.2196/17572 PMC7293059 32348255

18.Torous J Nicholas J Larsen ME Firth J Christensen H Clinical review of user engagement with mental health smartphone apps: evidence, theory and improvements Evid. Based Ment. Health 2018 21 116 119 29871870 10.1136/eb-2018-102891 Torous, J., Nicholas, J., Larsen, M. E., Firth, J. & Christensen, H. Clinical review of user engagement with mental health smartphone apps: evidence, theory and improvements. Evid. Based Ment. Health21, 116–119 (2018).10.1136/eb-2018-102891 PMC10270395 29871870

19.Fleming T Beyond the trial: systematic review of real-world uptake and engagement with digital self-help interventions for depression, low mood, or anxiety J. Med. Internet Res. 2018 20 e199 29875089 10.2196/jmir.9275 Fleming, T. et al. Beyond the trial: systematic review of real-world uptake and engagement with digital self-help interventions for depression, low mood, or anxiety. J. Med. Internet Res.20, e199 (2018).10.2196/jmir.9275 PMC6010835 29875089

20.Lattie EG Uptake and usage of IntelliCare: a publicly available suite of mental health and well-being apps Internet Inter. 2016 4 152 158 10.1016/j.invent.2016.06.003 PMC4936531 27398319 Lattie, E. G. et al. Uptake and usage of IntelliCare: a publicly available suite of mental health and well-being apps. Internet Inter.4, 152–158 (2016).10.1016/j.invent.2016.06.003 PMC4936531 27398319

21.Batterham PJ Neil AL Bennett K Griffiths KM Christensen H Predictors of adherence among community users of a cognitive behavior therapy website Patient Prefer. Adherence 2008 2 97 105 19920949 Batterham, P. J., Neil, A. L., Bennett, K., Griffiths, K. M. & Christensen, H. Predictors of adherence among community users of a cognitive behavior therapy website. Patient Prefer. Adherence2, 97–105 (2008).PMC2770409 19920949

22.Christensen H Griffiths KM Korten A Web-based cognitive behavior therapy: analysis of site usage and changes in depression and anxiety scores J. Med. Internet Res. 2002 4 e857 10.2196/jmir.4.1.e3 PMC1761927 11956035 Christensen, H., Griffiths, K. M. & Korten, A. Web-based cognitive behavior therapy: analysis of site usage and changes in depression and anxiety scores. J. Med. Internet Res.4, e857 (2002).10.2196/jmir.4.1.e3 PMC1761927 11956035

23.Christensen H A comparison of changes in anxiety and depression symptoms of spontaneous users and trial participants of a cognitive behavior therapy website J. Med. Internet Res. 2004 6 e124 10.2196/jmir.6.4.e46 PMC1550629 15631970 Christensen, H. et al. A comparison of changes in anxiety and depression symptoms of spontaneous users and trial participants of a cognitive behavior therapy website. J. Med. Internet Res.6, e124 (2004).10.2196/jmir.6.4.e46 PMC1550629 15631970

24.Christensen H Griffiths K Groves C Korten A Free range users and one hit wonders: community users of an Internet-based cognitive behaviour therapy program Aust. N. Z. J. Psychiatry 2006 40 59 62 16403040 10.1080/j.1440-1614.2006.01743.x Christensen, H., Griffiths, K., Groves, C. & Korten, A. Free range users and one hit wonders: community users of an Internet-based cognitive behaviour therapy program. Aust. N. Z. J. Psychiatry40, 59–62 (2006).10.1080/j.1440-1614.2006.01743.x 16403040

25.Neil A Predictors of adherence by adolescents to a cognitive behavior therapy website in school and community-based settings J. Med. Internet Res. 2009 11 e1050 10.2196/jmir.1050 PMC2762770 19275982 Neil, A. et al. Predictors of adherence by adolescents to a cognitive behavior therapy website in school and community-based settings. J. Med. Internet Res.11, e1050 (2009).10.2196/jmir.1050 PMC2762770 19275982

26.Carpenter J Seeing the “big” picture: big data methods for exploring relationships between usage, language, and outcome in internet intervention data J. Med. Internet Res. 2016 18 e241 27580524 10.2196/jmir.5725 Carpenter, J. et al. Seeing the “big” picture: big data methods for exploring relationships between usage, language, and outcome in internet intervention data. J. Med. Internet Res.18, e241 (2016).10.2196/jmir.5725 PMC5023946 27580524

27.Alqahtani F Orji R Insights from user reviews to improve mental health apps Health Inform. J. 2020 26 2042 2066 10.1177/1460458219896492 31920160 Alqahtani, F. & Orji, R. Insights from user reviews to improve mental health apps. Health Inform. J.26, 2042–2066 (2020).10.1177/1460458219896492 31920160

28.Borghouts J Barriers to and facilitators of user engagement with digital mental health interventions: systematic review J. Med. Internet Res. 2021 23 e24387 33759801 10.2196/24387 Borghouts, J. et al. Barriers to and facilitators of user engagement with digital mental health interventions: systematic review. J. Med. Internet Res.23, e24387 (2021).10.2196/24387 PMC8074985 33759801

29.Wu A Smartphone apps for depression and anxiety: a systematic review and meta-analysis of techniques to increase engagement NPJ Digit. Med. 2021 4 20 10.1038/s41746-021-00386-8 33574573 Wu, A. et al. Smartphone apps for depression and anxiety: a systematic review and meta-analysis of techniques to increase engagement. NPJ Digit. Med.4, 20 (2021).10.1038/s41746-021-00386-8 PMC7878769 33574573

30.Ng MM User engagement in mental health apps: a review of measurement, reporting, and validity Psychiatr. Serv. 2019 70 538 544 30914003 10.1176/appi.ps.201800519 Ng, M. M. et al. User engagement in mental health apps: a review of measurement, reporting, and validity. Psychiatr. Serv.70, 538–544 (2019).10.1176/appi.ps.201800519 PMC6839109 30914003

31.Daros AR Acceptability, engagement, outcomes, and dose–response associations of a mindfulness-based meditation app in individuals waiting for psychological services BMC Digit. Health 2025 3 19 10.1186/s44247-025-00155-7 Daros, A. R. et al. Acceptability, engagement, outcomes, and dose–response associations of a mindfulness-based meditation app in individuals waiting for psychological services. BMC Digit. Health3, 19 (2025).

32.Flett JA App-based mindfulness meditation for psychological distress and adjustment to college in incoming university students: a pragmatic, randomised, waitlist-controlled trial Psychol. Health 2020 35 1049 1074 32046499 10.1080/08870446.2019.1711089 Flett, J. A. et al. App-based mindfulness meditation for psychological distress and adjustment to college in incoming university students: a pragmatic, randomised, waitlist-controlled trial. Psychol. Health35, 1049–1074 (2020).10.1080/08870446.2019.1711089 32046499

33.Goldberg SB Testing the efficacy of a multicomponent, self-guided, smartphone-based meditation app: three-armed randomized controlled trial JMIR Ment. Health 2020 7 e23825 33245288 10.2196/23825 Goldberg, S. B. et al. Testing the efficacy of a multicomponent, self-guided, smartphone-based meditation app: three-armed randomized controlled trial. JMIR Ment. Health7, e23825 (2020).10.2196/23825 PMC7732708 33245288

34.Deady M Naturalistic evaluation of HeadGear: a smartphone app to reduce depressive symptoms in workers Behav. Ther. 2025 56 529 542 40287181 10.1016/j.beth.2024.01.001 Deady, M. et al. Naturalistic evaluation of HeadGear: a smartphone app to reduce depressive symptoms in workers. Behav. Ther.56, 529–542 (2025).10.1016/j.beth.2024.01.001 40287181

35.Bifftu BB Thomas SJ Win KT Users’ positive attitudes, perceived usefulness, and intentions to use digital mental health interventions: a systematic literature review and meta-analysis Comput. Biol. Med. 2025 190 110080 40158460 10.1016/j.compbiomed.2025.110080 Bifftu, B. B., Thomas, S. J. & Win, K. T. Users’ positive attitudes, perceived usefulness, and intentions to use digital mental health interventions: a systematic literature review and meta-analysis. Comput. Biol. Med.190, 110080 (2025).10.1016/j.compbiomed.2025.110080 40158460

36.Dweck CS Chiu CY Hong YY Implicit theories and their role in judgments and reactions: a word from two perspectives Psychol. Inq. 1995 6 267 285 10.1207/s15327965pli0604_1 Dweck, C. S., Chiu, C. Y. & Hong, Y. Y. Implicit theories and their role in judgments and reactions: a word from two perspectives. Psychol. Inq.6, 267–285 (1995).

37.O’Keefe, P. A. In The Complexity of Greatness: Beyond Talent or Practice 119–134 (Oxford Univ. Press, 2013).

38.Yeager DS Dweck CS Mindsets that promote resilience: when students believe that personal characteristics can be developed Educ. Psychol. 2012 47 302 314 10.1080/00461520.2012.722805 Yeager, D. S. & Dweck, C. S. Mindsets that promote resilience: when students believe that personal characteristics can be developed. Educ. Psychol.47, 302–314 (2012).

39.Yeager DS A national experiment reveals where a growth mindset improves achievement Nature 2019 573 364 369 10.1038/s41586-019-1466-y 31391586 Yeager, D. S. et al. A national experiment reveals where a growth mindset improves achievement. Nature573, 364–369 (2019).10.1038/s41586-019-1466-y PMC6786290 31391586

40.Crum AJ Salovey P Achor S Rethinking stress: the role of mindsets in determining the stress response J. Pers. Soc. Psychol. 2013 104 716 23437923 10.1037/a0031201 Crum, A. J., Salovey, P. & Achor, S. Rethinking stress: the role of mindsets in determining the stress response. J. Pers. Soc. Psychol.104, 716 (2013).10.1037/a0031201 23437923

41.Dweck, C. S. Mindset: The New Psychology of Success (Random House, 2006).

42.Dweck CS Yeager DS Mindsets: a view from two eras Perspect. Psychol. Sci. 2019 14 481 496 30707853 10.1177/1745691618804166 Dweck, C. S. & Yeager, D. S. Mindsets: a view from two eras. Perspect. Psychol. Sci.14, 481–496 (2019).10.1177/1745691618804166 PMC6594552 30707853

43.Blackwell LS Trzesniewski KH Dweck CS Implicit theories of intelligence predict achievement across an adolescent transition: a longitudinal study and an intervention Child Dev. 2007 78 246 263 17328703 10.1111/j.1467-8624.2007.00995.x Blackwell, L. S., Trzesniewski, K. H. & Dweck, C. S. Implicit theories of intelligence predict achievement across an adolescent transition: a longitudinal study and an intervention. Child Dev.78, 246–263 (2007).10.1111/j.1467-8624.2007.00995.x 17328703

44.Dweck, C. S. & Yeager, D. S. In Handbook of Wise Interventions: How Social Psychology Can Help People Change 9–35 (2020).

45.Beer JS Implicit self-theories of shyness J. Pers. Soc. Psychol. 2002 83 1009 12374434 10.1037/0022-3514.83.4.1009 Beer, J. S. Implicit self-theories of shyness. J. Pers. Soc. Psychol.83, 1009 (2002).12374434

46.Skymba HV Emotion mindsets and depressive symptoms in adolescence: the role of emotion regulation competence Emotion 2022 22 1255 33370145 10.1037/emo0000902 Skymba, H. V. et al. Emotion mindsets and depressive symptoms in adolescence: the role of emotion regulation competence. Emotion22, 1255 (2022).10.1037/emo0000902 33370145

47.Van Tongeren DR Burnette JL Do you believe happiness can change? An investigation of the relationship between happiness mindsets, well-being, and satisfaction J. Posit. Psychol. 2018 13 101 109 10.1080/17439760.2016.1257050 Van Tongeren, D. R. & Burnette, J. L. Do you believe happiness can change? An investigation of the relationship between happiness mindsets, well-being, and satisfaction. J. Posit. Psychol.13, 101–109 (2018).

48.Schroder HS The role of implicit theories in mental health symptoms, emotion regulation, and hypothetical treatment choices in college students Cogn. Ther. Res. 2015 39 120 139 10.1007/s10608-014-9652-6 PMC9037854 35474696 Schroder, H. S. et al. The role of implicit theories in mental health symptoms, emotion regulation, and hypothetical treatment choices in college students. Cogn. Ther. Res.39, 120–139 (2015).10.1007/s10608-014-9652-6 PMC9037854 35474696

49.Schroder HS Evaluating the domain specificity of mental health–related mind-sets Soc. Psychol. Personal. Sci. 2016 7 508 520 10.1177/1948550616644657 Schroder, H. S. et al. Evaluating the domain specificity of mental health–related mind-sets. Soc. Psychol. Personal. Sci.7, 508–520 (2016).

50.Schroder HS Mindsets in the clinic: applying mindset theory to clinical psychology Clin. Psychol. Rev. 2021 83 101957 33401130 10.1016/j.cpr.2020.101957 Schroder, H. S. Mindsets in the clinic: applying mindset theory to clinical psychology. Clin. Psychol. Rev.83, 101957 (2021).10.1016/j.cpr.2020.101957 33401130

51.Howell AJ Passmore HA Holder MD Implicit theories of well-being predict well-being and the endorsement of therapeutic lifestyle changes J. Happiness Stud. 2016 17 2347 2363 10.1007/s10902-015-9697-6 Howell, A. J., Passmore, H. A. & Holder, M. D. Implicit theories of well-being predict well-being and the endorsement of therapeutic lifestyle changes. J. Happiness Stud.17, 2347–2363 (2016).

52.Passmore HA Howell AJ Holder MD Positioning implicit theories of well-being within a positivity framework J. Happiness Stud. 2018 19 2445 2463 10.1007/s10902-017-9934-2 Passmore, H. A., Howell, A. J. & Holder, M. D. Positioning implicit theories of well-being within a positivity framework. J. Happiness Stud.19, 2445–2463 (2018).

53.Schroder HS The fixed mindset of anxiety predicts future distress: a longitudinal study Behav. Ther. 2019 50 710 717 31208681 10.1016/j.beth.2018.11.001 Schroder, H. S. et al. The fixed mindset of anxiety predicts future distress: a longitudinal study. Behav. Ther.50, 710–717 (2019).10.1016/j.beth.2018.11.001 31208681

54.Schroder HS Growth mindset of anxiety buffers the link between stressful life events and psychological distress and coping strategies Pers. Individ. Dif. 2017 110 23 26 10.1016/j.paid.2017.01.016 Schroder, H. S. et al. Growth mindset of anxiety buffers the link between stressful life events and psychological distress and coping strategies. Pers. Individ. Dif.110, 23–26 (2017).

55.O’Keefe PA Horberg EJ Dweck CS Walton GM A growth-theory-of-interest intervention increases interest in math and science coursework among liberal arts undergraduates J. Educ. Psychol. 2023 115 859 10.1037/edu0000798 O’Keefe, P. A., Horberg, E. J., Dweck, C. S. & Walton, G. M. A growth-theory-of-interest intervention increases interest in math and science coursework among liberal arts undergraduates. J. Educ. Psychol.115, 859 (2023).

56.O’Keefe PA Ramya SM Horberg EJ A growth-theory-of-interest intervention helps align science students with a new multidisciplinary curriculum Contemp. Educ. Psychol. 2025 81 102371 10.1016/j.cedpsych.2025.102371 O’Keefe, P. A., Ramya, S. M. & Horberg, E. J. A growth-theory-of-interest intervention helps align science students with a new multidisciplinary curriculum. Contemp. Educ. Psychol.81, 102371 (2025).

57.Paunesku D Mind-set interventions are a scalable treatment for academic underachievement Psychol. Sci. 2015 26 784 793 25862544 10.1177/0956797615571017 Paunesku, D. et al. Mind-set interventions are a scalable treatment for academic underachievement. Psychol. Sci.26, 784–793 (2015).10.1177/0956797615571017 25862544

58.Kosasih FR Efficacy of Intellect’s self-guided anxiety and worry mobile health programme: a randomized controlled trial with an active control and a 2-week follow-up PLOS Digit. Health 2023 2 e0000095 37224139 10.1371/journal.pdig.0000095 Kosasih, F. R. et al. Efficacy of Intellect’s self-guided anxiety and worry mobile health programme: a randomized controlled trial with an active control and a 2-week follow-up. PLOS Digit. Health2, e0000095 (2023).10.1371/journal.pdig.0000095 PMC10208477 37224139

59.Li MLY Min SLS Sündermann O Efficacy of the mHealth app Intellect in improving subclinical obsessive-compulsive disorder in university students: randomized controlled trial with a 4-week follow-up JMIR Mhealth Uhealth 2024 12 e63316 10.2196/63316 PMC11686029 39680884 Li, M. L. Y., Min, S. L. S. & Sündermann, O. Efficacy of the mHealth app Intellect in improving subclinical obsessive-compulsive disorder in university students: randomized controlled trial with a 4-week follow-up. JMIR Mhealth Uhealth12, e63316 (2024).10.2196/63316 PMC11686029 39680884

60.Ong WY Sündermann O Efficacy of the mental health app “intellect” to improve body image and self-compassion in young adults: a randomized controlled trial with a 4-week follow-up JMIR Mhealth Uhealth 2022 10 e41800 36409542 10.2196/41800 Ong, W. Y. & Sündermann, O. Efficacy of the mental health app “intellect” to improve body image and self-compassion in young adults: a randomized controlled trial with a 4-week follow-up. JMIR Mhealth Uhealth10, e41800 (2022).10.2196/41800 PMC9723978 36409542

61.Toh SHY Efficacy of the mental health app intellect to reduce stress: randomized controlled trial with a 1-month follow-up JMIR Form. Res. 2022 6 e40723 36515984 10.2196/40723 Toh, S. H. Y. et al. Efficacy of the mental health app intellect to reduce stress: randomized controlled trial with a 1-month follow-up. JMIR Form. Res.6, e40723 (2022).10.2196/40723 PMC9798264 36515984

62.Toh SHY Lee SC Sündermann O Mobile behavioral health coaching as a preventive intervention for occupational public health: retrospective longitudinal study JMIR Form. Res. 2023 7 e45678 37862086 10.2196/45678 Toh, S. H. Y., Lee, S. C. & Sündermann, O. Mobile behavioral health coaching as a preventive intervention for occupational public health: retrospective longitudinal study. JMIR Form. Res.7, e45678 (2023).10.2196/45678 PMC10625093 37862086

63.Yang Toh SH Preliminary effectiveness of an evidence-based mobile application to promote resilience among working adults in Singapore and Hong Kong: intensive longitudinal study Digit. Health 2023 9 20552076231178616 37274370 10.1177/20552076231178616 Yang Toh, S. H. et al. Preliminary effectiveness of an evidence-based mobile application to promote resilience among working adults in Singapore and Hong Kong: intensive longitudinal study. Digit. Health9, 20552076231178616 (2023).10.1177/20552076231178616 PMC10236254 37274370

64.Yokomitsu K Assessing the efficacy of the INTELLECT cognitive behavioral therapy mobile app for anxiety and depressive symptoms among at-risk Japanese employees: randomized controlled trial JMIR Mhealth Uhealth 2025 13 e60871 40554801 10.2196/60871 Yokomitsu, K. et al. Assessing the efficacy of the INTELLECT cognitive behavioral therapy mobile app for anxiety and depressive symptoms among at-risk Japanese employees: randomized controlled trial. JMIR Mhealth Uhealth13, e60871 (2025).10.2196/60871 PMC12212887 40554801

65.Liu C Uptake, adherence, and attrition in clinical trials of depression and anxiety apps: a systematic review and meta-analysis JAMA Psychiatry 2026 83 43 50 41259035 10.1001/jamapsychiatry.2025.3439 Liu, C. et al. Uptake, adherence, and attrition in clinical trials of depression and anxiety apps: a systematic review and meta-analysis. JAMA Psychiatry83, 43–50 (2026).10.1001/jamapsychiatry.2025.3439 PMC12631558 41259035

66.Schreiber C Schotanus-Dijkstra M Enhancing mental wellbeing by changing mindsets? Results from two randomized controlled trials BMC Psychol. 2024 12 77 10.1186/s40359-023-01470-2 38360675 Schreiber, C. & Schotanus-Dijkstra, M. Enhancing mental wellbeing by changing mindsets? Results from two randomized controlled trials. BMC Psychol.12, 77 (2024).10.1186/s40359-023-01470-2 PMC10870491 38360675

67.Doll CM Predictors of help-seeking behaviour in people with mental health problems: a 3-year prospective community study BMC Psychiatry 2021 21 1 11 10.1186/s12888-021-03435-4 PMC8414662 34479537 Doll, C. M. et al. Predictors of help-seeking behaviour in people with mental health problems: a 3-year prospective community study. BMC Psychiatry21, 1–11 (2021).10.1186/s12888-021-03435-4 PMC8414662 34479537

68.Walton GM Wilson TD Wise interventions: psychological remedies for social and personal problems Psychol. Rev. 2018 125 617 30299141 10.1037/rev0000115 Walton, G. M. & Wilson, T. D. Wise interventions: psychological remedies for social and personal problems. Psychol. Rev.125, 617 (2018).10.1037/rev0000115 30299141

69.Burnette JL A systematic review and meta-analysis of growth mindset interventions: for whom, how, and why might such interventions work? Psychol. Bull. 2023 149 174 36227318 10.1037/bul0000368 Burnette, J. L. et al. A systematic review and meta-analysis of growth mindset interventions: for whom, how, and why might such interventions work? Psychol. Bull.149, 174 (2023).10.1037/bul0000368 36227318

70.Schleider JL Weisz JR Reducing risk for anxiety and depression in adolescents: effects of a single-session intervention teaching that personality can change Behav. Res. Ther. 2016 87 170 181 27697671 10.1016/j.brat.2016.09.011 Schleider, J. L. & Weisz, J. R. Reducing risk for anxiety and depression in adolescents: effects of a single-session intervention teaching that personality can change. Behav. Res. Ther.87, 170–181 (2016).10.1016/j.brat.2016.09.011 PMC5127737 27697671

71.Schleider JL Mullarkey MC Chacko A Harnessing wise interventions to advance the potency and reach of youth mental health services Clin. Child Fam. Psychol. Rev. 2020 23 70 101 10.1007/s10567-019-00301-4 31440858 Schleider, J. L., Mullarkey, M. C. & Chacko, A. Harnessing wise interventions to advance the potency and reach of youth mental health services. Clin. Child Fam. Psychol. Rev.23, 70–101 (2020).10.1007/s10567-019-00301-4 31440858

72.Bryan CJ Tipton E Yeager DS Behavioural science is unlikely to change the world without a heterogeneity revolution Nat. Hum. Behav. 2021 5 980 989 10.1038/s41562-021-01143-3 34294901 Bryan, C. J., Tipton, E. & Yeager, D. S. Behavioural science is unlikely to change the world without a heterogeneity revolution. Nat. Hum. Behav.5, 980–989 (2021).10.1038/s41562-021-01143-3 PMC8928154 34294901

73.Geraghty AW Wood AM Hyland ME Attrition from self-directed interventions: investigating the relationship between psychological predictors, intervention content and dropout from a body dissatisfaction intervention Soc. Sci. Med. 2010 71 30 37 20400220 10.1016/j.socscimed.2010.03.007 Geraghty, A. W., Wood, A. M. & Hyland, M. E. Attrition from self-directed interventions: investigating the relationship between psychological predictors, intervention content and dropout from a body dissatisfaction intervention. Soc. Sci. Med.71, 30–37 (2010).10.1016/j.socscimed.2010.03.007 20400220

74.Klein M Mogles N Van Wissen A Intelligent mobile support for therapy adherence and behavior change J. Biomed. Inform. 2014 51 137 151 24858491 10.1016/j.jbi.2014.05.005 Klein, M., Mogles, N. & Van Wissen, A. Intelligent mobile support for therapy adherence and behavior change. J. Biomed. Inform.51, 137–151 (2014).10.1016/j.jbi.2014.05.005 24858491

75.Prochaska JO Velicer WF The transtheoretical model of health behavior change Am. J. Health Promot. 1997 12 38 48 10170434 10.4278/0890-1171-12.1.38 Prochaska, J. O. & Velicer, W. F. The transtheoretical model of health behavior change. Am. J. Health Promot.12, 38–48 (1997).10.4278/0890-1171-12.1.38 10170434

76.Kenthirarajah, D. & Walton, G. M. How brief social-psychological interventions can cause enduring effects. Emerg. Trends Soc. Behav. Sci.1, 1–15 (2015).

77.O’Keefe PA Lee HY Chen P Changing students’ beliefs about learning can unveil their potential Policy Insights Behav. Brain Sci. 2021 8 84 91 10.1177/2372732220984173 O’Keefe, P. A., Lee, H. Y. & Chen, P. Changing students’ beliefs about learning can unveil their potential. Policy Insights Behav. Brain Sci.8, 84–91 (2021).

78.Rosenthal, R. & Jacobson, L. Pygmalion in the Classroom (Holt, Rinehart & Winston, 1968).

79.Storms MD McCaul KD Attribution processes and emotional exacerbation of dysfunctional behavior N. Dir. Attrib. Res. 1976 1 143 164 Storms, M. D. & McCaul, K. D. Attribution processes and emotional exacerbation of dysfunctional behavior. N. Dir. Attrib. Res.1, 143–164 (1976).

80.Yeager DS Walton GM Social-psychological interventions in education: they’re not magic Rev. Educ. Res. 2011 81 267 301 10.3102/0034654311405999 Yeager, D. S. & Walton, G. M. Social-psychological interventions in education: they’re not magic. Rev. Educ. Res.81, 267–301 (2011).

81.Flett JA The peril of self-reported adherence in digital interventions: a brief example Internet Interv. 2019 18 100267 31890620 10.1016/j.invent.2019.100267 Flett, J. A. et al. The peril of self-reported adherence in digital interventions: a brief example. Internet Interv.18, 100267 (2019).10.1016/j.invent.2019.100267 PMC6926264 31890620

82.Subramaniam M Stigma towards people with mental disorders and its components—a perspective from multi-ethnic Singapore Epidemiol. Psychiatr. Sci. 2017 26 371 382 27018715 10.1017/S2045796016000159 PMC5647661 Subramaniam, M. et al. Stigma towards people with mental disorders and its components—a perspective from multi-ethnic Singapore. Epidemiol. Psychiatr. Sci.26, 371–382 (2017).10.1017/S2045796016000159 PMC5647661 27018715

83.Walton GM The new science of wise psychological interventions Curr. Dir. Psychol. Sci. 2014 23 73 82 10.1177/0963721413512856 Walton, G. M. The new science of wise psychological interventions. Curr. Dir. Psychol. Sci.23, 73–82 (2014).

84.Appelbaum M Journal article reporting standards for quantitative research in psychology: the APA Publications and Communications Board task force report Am. Psychol. 2018 73 3 25 29345484 10.1037/amp0000191 Appelbaum, M. et al. Journal article reporting standards for quantitative research in psychology: the APA Publications and Communications Board task force report. Am. Psychol.73, 3–25 (2018).10.1037/amp0000191 29345484

85.Eskreis-Winkler L Using wise interventions to motivate deliberate practice J. Pers. Soc. Psychol. 2016 111 728 27762575 10.1037/pspp0000074 Eskreis-Winkler, L. et al. Using wise interventions to motivate deliberate practice. J. Pers. Soc. Psychol.111, 728 (2016).10.1037/pspp0000074 PMC5091297 27762575

86.Walton GM Lifting the bar: a relationship-orienting intervention reduces recidivism among children reentering school from juvenile detention Psychol. Sci. 2021 32 1747 1767 34606384 10.1177/09567976211013801 Walton, G. M. et al. Lifting the bar: a relationship-orienting intervention reduces recidivism among children reentering school from juvenile detention. Psychol. Sci.32, 1747–1767 (2021).10.1177/09567976211013801 34606384

87.Good TL Sterzinger N Lavigne A Expectation effects: Pygmalion and the initial 20 years of research Educ. Res. Eval. 2018 24 99 123 10.1080/13803611.2018.1548817 Good, T. L., Sterzinger, N. & Lavigne, A. Expectation effects: Pygmalion and the initial 20 years of research. Educ. Res. Eval.24, 99–123 (2018).

88.Cobanoglu C Cavusoglu M Turktarhan GA Beginner’s guide and best practices for using crowdsourcing platforms for survey research: the case of Amazon Mechanical Turk (MTurk) J. Glob. Bus. Insights 2021 6 92 97 10.5038/2640-6489.6.1.1177 Cobanoglu, C., Cavusoglu, M. & Turktarhan, G. A. Beginner’s guide and best practices for using crowdsourcing platforms for survey research: the case of Amazon Mechanical Turk (MTurk). J. Glob. Bus. Insights6, 92–97 (2021).

89.Walton, G. M. & Crum, A. J. (eds.) Handbook of Wise Interventions (Guilford Press, 2020).

90.Walton GM Cohen GL A brief social-belonging intervention improves academic and health outcomes of minority students Science 2011 331 1447 1451 21415354 10.1126/science.1198364 Walton, G. M. & Cohen, G. L. A brief social-belonging intervention improves academic and health outcomes of minority students. Science331, 1447–1451 (2011).10.1126/science.1198364 21415354

91.Cialdini RB Goldstein NJ Social influence: compliance and conformity Annu. Rev. Psychol. 2004 55 591 621 14744228 10.1146/annurev.psych.55.090902.142015 Cialdini, R. B. & Goldstein, N. J. Social influence: compliance and conformity. Annu. Rev. Psychol.55, 591–621 (2004).10.1146/annurev.psych.55.090902.142015 14744228

92.Aronson E The power of self-persuasion Am. Psychol. 1999 54 875 10.1037/h0088188 Aronson, E. The power of self-persuasion. Am. Psychol.54, 875 (1999).

93.Okonofua JA A scalable empathic supervision intervention to mitigate recidivism from probation and parole Proc. Natl. Acad. Sci. USA 2021 118 e2018036118 33782121 10.1073/pnas.2018036118 Okonofua, J. A. et al. A scalable empathic supervision intervention to mitigate recidivism from probation and parole. Proc. Natl. Acad. Sci. USA. 118, e2018036118 (2021).10.1073/pnas.2018036118 PMC8040791 33782121

94.O’Keefe PA Horberg EJ Lee F Dweck CS Implicit theories of opportunity: when opportunity fails to knock, keep waiting, or start cultivating? J. Pers. Soc. Psychol. 2023 124 1146 36521117 10.1037/pspa0000330 O’Keefe, P. A., Horberg, E. J., Lee, F. & Dweck, C. S. Implicit theories of opportunity: when opportunity fails to knock, keep waiting, or start cultivating? J. Pers. Soc. Psychol.124, 1146 (2023).10.1037/pspa0000330 36521117

95.Kroenke K Spitzer RL Williams JB Löwe B An ultra-brief screening scale for anxiety and depression: the PHQ–4 Psychosomatics 2009 50 613 621 19996233 10.1176/appi.psy.50.6.613 Kroenke, K., Spitzer, R. L., Williams, J. B. & Löwe, B. An ultra-brief screening scale for anxiety and depression: the PHQ–4. Psychosomatics50, 613–621 (2009).10.1176/appi.psy.50.6.613 19996233

96.Staples LG Psychometric properties and clinical utility of brief measures of depression, anxiety, and general distress: the PHQ-2, GAD-2, and K-6 Gen. Hosp. Psychiatry 2019 56 13 18 30508772 10.1016/j.genhosppsych.2018.11.003 Staples, L. G. et al. Psychometric properties and clinical utility of brief measures of depression, anxiety, and general distress: the PHQ-2, GAD-2, and K-6. Gen. Hosp. Psychiatry56, 13–18 (2019).10.1016/j.genhosppsych.2018.11.003 30508772

97.Warttig SL New, normative, English-sample data for the short form perceived stress scale (PSS-4) J. Health Psychol. 2013 18 1617 1628 24155195 10.1177/1359105313508346 Warttig, S. L. et al. New, normative, English-sample data for the short form perceived stress scale (PSS-4). J. Health Psychol.18, 1617–1628 (2013).10.1177/1359105313508346 24155195

98.Lee EH Korean versions of the Perceived Stress Scale (PSS-14, 10 and 4): psychometric evaluation in patients with chronic disease Scand. J. Caring Sci. 2015 29 183 192 24660854 10.1111/scs.12131 Lee, E. H. et al. Korean versions of the Perceived Stress Scale (PSS-14, 10 and 4): psychometric evaluation in patients with chronic disease. Scand. J. Caring Sci.29, 183–192 (2015).10.1111/scs.12131 24660854

99.Ng Fat L Evaluating and establishing national norms for mental wellbeing using the short Warwick–Edinburgh Mental Well-being Scale (SWEMWBS): findings from the Health Survey for England Qual. Life Res. 2017 26 1129 1144 10.1007/s11136-016-1454-8 PMC5376387 27853963 Ng Fat, L. et al. Evaluating and establishing national norms for mental wellbeing using the short Warwick–Edinburgh Mental Well-being Scale (SWEMWBS): findings from the Health Survey for England. Qual. Life Res.26, 1129–1144 (2017).10.1007/s11136-016-1454-8 PMC5376387 27853963

100.Tennant R The Warwick-Edinburgh mental well-being scale (WEMWBS): development and UK validation Health Qual. Life Outcomes 2007 5 1 13 10.1186/1477-7525-5-63 PMC2222612 18042300 Tennant, R. et al. The Warwick-Edinburgh mental well-being scale (WEMWBS): development and UK validation. Health Qual. Life Outcomes5, 1–13 (2007).10.1186/1477-7525-5-63 PMC2222612 18042300

101.Kalok A The psychological impact of movement restriction during the COVID-19 outbreak on clinical undergraduates: a cross-sectional study Int. J. Environ. Res. Public Health 2020 17 8522 33212969 10.3390/ijerph17228522 Kalok, A. et al. The psychological impact of movement restriction during the COVID-19 outbreak on clinical undergraduates: a cross-sectional study. Int. J. Environ. Res. Public Health17, 8522 (2020).10.3390/ijerph17228522 PMC7698578 33212969

102.Mallett R Coyle C Kuang Y Gillanders DT Behind the masks: intolerance of uncertainty, perceived vulnerability to disease and psychological flexibility in relation to state anxiety and wellbeing during the COVID-19 pandemic J. Contextual Behav. Sci. 2021 22 52 62 34603949 10.1016/j.jcbs.2021.09.003 Mallett, R., Coyle, C., Kuang, Y. & Gillanders, D. T. Behind the masks: intolerance of uncertainty, perceived vulnerability to disease and psychological flexibility in relation to state anxiety and wellbeing during the COVID-19 pandemic. J. Contextual Behav. Sci.22, 52–62 (2021).10.1016/j.jcbs.2021.09.003 PMC8464037 34603949

103.Bernstein EE Patterns of skills review in smartphone cognitive behavioral therapy for depression: observational study of intervention content use JMIR Ment. Health 2025 12 e63497 39993308 10.2196/63497 Bernstein, E. E. et al. Patterns of skills review in smartphone cognitive behavioral therapy for depression: observational study of intervention content use. JMIR Ment. Health12, e63497 (2025).10.2196/63497 PMC11894344 39993308

104.Lattie EG Uptake and effectiveness of a self-guided mobile app platform for college student mental health Internet Inter. 2022 27 100493 10.1016/j.invent.2021.100493 PMC8749061 35036331 Lattie, E. G. et al. Uptake and effectiveness of a self-guided mobile app platform for college student mental health. Internet Inter.27, 100493 (2022).10.1016/j.invent.2021.100493 PMC8749061 35036331

105.Pham Q A library of analytic indicators to evaluate effective engagement with consumer mHealth apps for chronic conditions: scoping review JMIR Mhealth Uhealth 2019 7 e11941 30664463 10.2196/11941 Pham, Q. et al. A library of analytic indicators to evaluate effective engagement with consumer mHealth apps for chronic conditions: scoping review. JMIR Mhealth Uhealth7, e11941 (2019).10.2196/11941 PMC6356188 30664463

106.Wong HW Postsecondary student engagement with a mental health app and online platform (Thought Spot): qualitative study assessing factors related to user experience JMIR Ment. Health 2021 8 e23447 33797395 10.2196/23447 Wong, H. W. et al. Postsecondary student engagement with a mental health app and online platform (Thought Spot): qualitative study assessing factors related to user experience. JMIR Ment. Health8, e23447 (2021).10.2196/23447 PMC8052598 33797395

107.Caldeira C Mobile apps for mood tracking: an analysis of features and user reviews AMIA Annu. Symp. Proc. 2018 2017 495 29854114 Caldeira, C. et al. Mobile apps for mood tracking: an analysis of features and user reviews. AMIA Annu. Symp. Proc.2017, 495 (2018).PMC5977660 29854114

108.Schueller SM Understanding people’s use of and perspectives on mood-tracking apps: interview study JMIR Ment. Health 2021 8 e29368 34383678 10.2196/29368 Schueller, S. M. et al. Understanding people’s use of and perspectives on mood-tracking apps: interview study. JMIR Ment. Health8, e29368 (2021).10.2196/29368 PMC8387890 34383678

109.Watkins ER Constructive and unconstructive repetitive thought Psychol. Bull. 2008 134 163 18298268 10.1037/0033-2909.134.2.163 Watkins, E. R. Constructive and unconstructive repetitive thought. Psychol. Bull.134, 163 (2008).10.1037/0033-2909.134.2.163 PMC2672052 18298268

110.Overholser JC Beale E The art and science behind Socratic questioning and guided discovery: a research review Psychother. Res. 2023 33 946 956 36878221 10.1080/10503307.2023.2183154 Overholser, J. C. & Beale, E. The art and science behind Socratic questioning and guided discovery: a research review. Psychother. Res.33, 946–956 (2023).10.1080/10503307.2023.2183154 36878221

111.Vittorio LN Using Socratic questioning to promote cognitive change and achieve depressive symptom reduction: evidence of cognitive change as a mediator Behav. Res. Ther. 2022 150 104035 35016095 10.1016/j.brat.2022.104035 Vittorio, L. N. et al. Using Socratic questioning to promote cognitive change and achieve depressive symptom reduction: evidence of cognitive change as a mediator. Behav. Res. Ther.150, 104035 (2022).10.1016/j.brat.2022.104035 35016095

112.Raes F Mindfulness and reduced cognitive reactivity to sad mood: evidence from a correlational study and a non-randomized waiting list controlled study Behav. Res. Ther. 2009 47 623 627 19362705 10.1016/j.brat.2009.03.007 Raes, F. et al. Mindfulness and reduced cognitive reactivity to sad mood: evidence from a correlational study and a non-randomized waiting list controlled study. Behav. Res. Ther.47, 623–627 (2009).10.1016/j.brat.2009.03.007 19362705

113.Toh SH Kanter JW Lim CH Suendermann O Revisiting the monitoring and acceptance theory: findings from a daily diary, three-armed randomized controlled trial Mindfulness 2024 15 2536 2553 10.1007/s12671-024-02447-x Toh, S. H., Kanter, J. W., Lim, C. H. & Suendermann, O. Revisiting the monitoring and acceptance theory: findings from a daily diary, three-armed randomized controlled trial. Mindfulness15, 2536–2553 (2024).

114.Huberty JL Testing a mindfulness meditation mobile app for the treatment of sleep-related symptoms in adults with sleep disturbance: a randomized controlled trial PLoS ONE 2021 16 e0244717 33411779 10.1371/journal.pone.0244717 Huberty, J. L. et al. Testing a mindfulness meditation mobile app for the treatment of sleep-related symptoms in adults with sleep disturbance: a randomized controlled trial. PLoS ONE16, e0244717 (2021).10.1371/journal.pone.0244717 PMC7790277 33411779

115.Beck AT Thinking and depression: I. Idiosyncratic content and cognitive distortions Arch. Gen. Psychiatry 1963 9 324 333 14045261 10.1001/archpsyc.1963.01720160014002 Beck, A. T. Thinking and depression: I. Idiosyncratic content and cognitive distortions. Arch. Gen. Psychiatry9, 324–333 (1963).10.1001/archpsyc.1963.01720160014002 14045261

116.Roche K Pilot study on the impact of gratitude journaling or cognitive strategies on health care workers J. Pediatr. Health Care 2023 37 414 424 36882351 10.1016/j.pedhc.2023.02.002 Roche, K. et al. Pilot study on the impact of gratitude journaling or cognitive strategies on health care workers. J. Pediatr. Health Care37, 414–424 (2023).10.1016/j.pedhc.2023.02.002 PMC9985770 36882351

117.Hayes, A. F. PROCESS: A Versatile Computational Tool for Observed Variable Mediation, Moderation, and Conditional Process Modeling (2012).