This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Mental Health, is properly cited. The complete bibliographic information, a link to the original publication on http://mental.jmir.org/, as well as this copyright and license information must be included.
Caregivers play a pivotal role in maintaining an economically viable health care system, yet they are characterized by low levels of psychological well-being and consistently report unmet needs for psychological support. Mobile app–based (mobile health [mHealth]) interventions present a novel approach to both reducing stress and improving well-being.
This study aims to evaluate the effectiveness of a self-guided mobile app–based psychological intervention for people providing care to family or friends with a physical or mental disability.
In a randomized, single-blind, controlled trial, 183 caregivers recruited through the web were randomly allocated to either an intervention (n=73) or active control (n=110) condition. The intervention app contained treatment modules combining daily self-monitoring with third-wave (mindfulness-based) cognitive-behavioral therapies, whereas the active control app contained only self-monitoring features. Both programs were completed over a 5-week period. It was hypothesized that intervention app exposure would be associated with decreases in depression, anxiety, and stress, and increases in well-being, self-esteem, optimism, primary and secondary control, and social support. Outcomes were assessed at baseline, postintervention, and 3-4 months postintervention. App quality was also assessed.
In total, 25% (18/73) of the intervention participants were lost to follow-up at 3 months, and 30.9% (34/110) of the participants from the wait-list control group dropped out before the postintervention survey. The intervention group experienced reductions in stress (
This study demonstrates that mHealth psychological interventions are an effective treatment option for caregivers experiencing high levels of stress. Recommendations for improving mHealth interventions for caregivers include offering flexibility and customization in the treatment design.
Australian New Zealand Clinical Trial Registry ACTRN12616000996460; https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=371170
Caring for people living with physical or mental health difficulties can be a challenging role, one that is becoming increasingly common as trends in public policy move toward assisting people with disabilities to remain within their family environment for as long as possible [
Despite their challenging circumstances, caregivers have been reported to identify positive aspects associated with caregiving, including a sense of value in their role [
Compromised emotional well-being in caregivers (eg, mental disorders) may adversely impact care recipients. There is evidence, for example, that care recipients have poorer general health, mental health, and quality of life and exacerbated disability symptomatology when caregivers experience mental health problems [
Given the available evidence on the significance of caregiver burden, tailored interventions designed to reduce stress and promote well-being in carers are critically important. Among existing interventions, the primary psychological treatments are based on principles of cognitive behavioral therapy (CBT), which have been shown to reduce depression in caregivers [
Although these approaches show promise, caregivers face a number of barriers to accessing in-person treatment programs, including economic, geographic, and mobility factors; limited time to engage in interventions; and difficulties in finding and/or affording the cost of suitable alternative caregiver support to attend treatment [
Mobile app–based brief interventions offer a number of strengths over other digital delivery platforms. Their small size and portability allow an intervention to be readily accessed at times of greatest need [
This study is a randomized controlled trial of a mobile app–based, self-directed psychological intervention for people who are providing care to family or friends with a physical and/or mental disability. It was hypothesized that the intervention would produce a greater reduction in stress, depression, and anxiety as well as increased well-being, compared with control participants (hypothesis 1). To assess the broader impact, we also evaluated emotional well-being, self-esteem, optimism, primary and secondary control, and perceived social support (secondary outcomes; hypothesis 2). We hypothesized that these improvements in self-reports will be maintained for 3 months postintervention for primary outcomes (hypothesis 3) and secondary outcomes (hypotheses 4). Although the intervention was designed to provide a range of modules with different techniques that could each be useful for improving outcomes, we tested the possibility that the effect of intervention allocation was moderated by the number of treatment modules completed. In particular, we predicted that improvements in primary and secondary outcomes would be stronger for individuals allocated to the treatment condition who engage in more modules (hypothesis 5). We also explored the usefulness of this form of intervention through caregivers’ perceptions of the app’s engagement, functionality, aesthetics, information, and quality, expecting positive ratings across these metrics for the intervention (hypothesis 6).
The design of the trial was a 2 (condition:
Participants were recruited through a mix of traditional strategies and targeted social media advertising. Support was sought from caregiver organizations and services, who agreed to display study flyers (both in physical and digital forms) and allowed the research team to attend caregiver events and seminars for recruitment purposes. Social media advertising was conducted through Facebook, with separate advertisement campaigns targeting either Australians broadly or those with an interest in specific disability topics (eg,
To be eligible to join the study, participants were required to be (1) an Australian resident, (2) aged 18 years or older, (3) fluent in English, (4) helping to support a friend or relative with a physical or mental condition/disability, (5) able to access an Apple iOS mobile phone device (iPhone or iPad) with internet access for the duration of the study, and (6) not have participated in an electronic health (eHealth) intervention (any technology-based health intervention, including mobile apps) within the previous 6 months. Smartphone app literacy was also a de facto eligibility criterion but was assumed by the participant’s willingness to sign up for the study. A CONSORT flow diagram is provided below (
This Figure provides a CONSORT flow chart of participant numbers.
The required sample size was powered with the following assumptions: (1) a moderate group difference (SD 0.5) between the intervention and active control groups for the primary and secondary outcomes at postintervention; (2) power set at 0.80; (3) α set at .05 (2 tailed); (4) expected attrition rate of 20% for the intervention group [
StressLess is a 5-week, self-directed intervention, based on the principles of second- and third-wave CBTs [
This figure shows layout of the app.
In addition to the intervention modules, StressLess also supports users in self-monitoring their well-being through in-the-moment assessments (EMA). Participants were prompted to complete a self-monitoring assessment up to 4 times per day via the StressLess app notification function. Participants were prompted to complete either a
The active control involved the mobile app StressMonitor. This comprised the same self-monitoring EMA function and feedback bar chart as the StressLess intervention but did not contain any intervention modules. The inclusion of an active control condition enabled the statistical separation of effects because of novelty (or burden) of completing app-based self-monitoring of mood from treatment outcomes.
Items assessing participants’ caregiver roles were adapted from the Australian Bureau of Statistics’ (ABS) Survey of Disability, Aging, and Carers (SDAC) [
The primary outcomes were the participants’ stress levels, depression, anxiety, and subjective well-being. The first 3 variables were measured using the Depression Anxiety Stress Scale-21 [
The Personal Wellbeing Index (PWI) was used to assess the primary intervention outcome of participants’ subjective well-being [
Beyond the primary outcome measures listed earlier, the study also assessed additional secondary variables that were predicted to improve after completing the intervention. Affective mood was assessed using the Homeostatically Protected Mood Scale [
Self-esteem was assessed using the Rosenberg Self-Esteem Scale [
Optimism was assessed using the
Primary and secondary control were assessed using an abbreviated version of the Primary and Secondary Control Scale (PSCS) [
Social support was assessed using the Multidimensional Scale of Perceived Social Support [
The quality of the intervention app was assessed using the Mobile Application Rating Scale [
After providing informed consent via Qualtrics (by reading a plain language statement and then responding to a question about whether they consented) and meeting the study eligibility criteria, participants were invited to complete the baseline assessment as a web survey. Participants were then randomly allocated to either the active control or intervention arm using a 3:2 assignment in blocks of 5 created through Qualtrics (web-based survey provider of choice), with the expectation that attrition would be higher in the active control group because of lower incentive to remain in the study. Instructions were provided to participants detailing how to install the app (either StressLess or StressMonitor) on their mobile phone or iPad. The app is free and does not include any hidden costs. For both groups, the plain language statement provided via the baseline Qualtrics survey provided contact details for free helplines if the participants felt distressed at any stage because of the intervention. The StressLess and StressMonitor apps also contained these contact details in the app to remind participants that they could contact LifeLine (a free, Australian counseling service) if they felt distressed.
Following the download of the app, participants then completed 5 weeks with their assigned app, with weekly contact from the research team by either an email or phone call to maximize engagement. In more detail, a standard email was sent to all participants in each group in weeks 1, 2, 4, and 5 explaining an aspect of either the intervention or the active control program. For example, week 2 emails were titled
Participants then completed the postintervention assessment as a web survey and were reimbursed for their time with a $50 voucher. The postintervention survey was identical for participants from both groups, with the exception that the intervention group received the app quality measure. Furthermore, active control participants who completed the postintervention survey were provided with instructions on how to download the intervention app from the iOS app store. Finally, intervention participants were invited to complete a follow-up survey 4 months after completing the postintervention assessment.
Following the principles of intention-to-treat (ITT) analysis, individuals were retained in the group they were randomized to. Thus, even in cases where participants in the intervention group did not use the app at all (n=15), they were retained in the intervention group for the purposes of analysis. Missing data were handled using multiple imputation, with 50 imputations. By default, Mplus uses Monte Carlo Markov Chains with 100 iterations per imputation and chained equations to impute missing values for variables [
For the evaluation of efficacy, time was entered as a level 1 predictor (0=baseline and 1=postintervention). At level 2, group (0=control and 1=intervention) was included as a predictor of the dependent variable (DV) as well as a moderator of the level 1 relationship between time and DV scores. This latter effect (a cross-level interaction) was used to ascertain whether the rate of improvement in symptoms was greater for intervention participants than for those in the control group (hypotheses 1 and 2). Maintenance effects were tested similarly, although the time effect compared postintervention (coded 0) against the 4-month follow-up time point (coded 1; hypotheses 3 and 4). As the follow-up data were only collected for the intervention group, there was no level 2 predictor for group. Dose-response effects were tested with the intervention group only, by moderating the time effect by the number of modules completed (hypothesis 5). Each outcome variable was modeled separately. Descriptive statistics were reported for the evaluation of user ratings of the intervention (hypothesis 6). All effects were tested at
The final sample consisted of 183 caregivers;
Compared with national caregiver data available from the ABS [
Demographic characteristics of participants from the intervention and active control groups.
Variables | Intervention (n=73) | Active control (n=110) | Group differences | ||||
|
|
|
Chi-square (df) | ||||
Age (years), mean (SD) | 40.29 (6.51) | 39.21 (5.86) | 1.16 (179) |
|
.25 | ||
|
|
|
N/Aa | 1.52 (2) | .47 | ||
|
Male | 3 (4) | 5 (5) |
|
|
|
|
|
Female | 69 (95) | 104 (95.4) |
|
|
|
|
|
Other | 1 (1) | 0 (0) |
|
|
|
|
|
|
N/A | 8.90 (5) | .11 | |||
|
<15,000 (10,385) | 5 (7) | 5 (45) |
|
|
|
|
|
15,000-30,000 (10,385-20,771) | 19 (26) | 19 (17) |
|
|
|
|
|
31,000-60,000 (21,463-41,542) | 16 (22) | 16 (15) |
|
|
|
|
|
61,000-100,000 (42,235-69,237) | 34 (47) | 34 (31) |
|
|
|
|
|
101,000-150,000 (69,929-103,856) | 21 (29) | 21 (19) |
|
|
|
|
|
>150,000 (103,856) | 15 (21) | 15 (14) |
|
|
|
|
|
|
|
N/A |
|
|
||
|
Full-time paid | 11 (15) | 15 (14) |
|
0.07 (1) | .79 | |
|
Full-time study | 9 (12) | 10 (9) |
|
0.49 (1) | .48 | |
|
Full-time home | 31 (43) | 35 (32) |
|
2.16 (1) | .14 | |
|
Part-time paid | 21 (29) | 38 (34) |
|
0.67 (1) | .41 | |
|
Casual paid | 3 (4) | 10 (9) |
|
1.65 (1) | .20 | |
|
Part-time home | 19 (26) | 22 (20) |
|
0.92 (1) | .34 | |
|
Unemployed | 5 (7) | 8 (7) |
|
0.01 (1) | .91 | |
Number of care recipients, n (%) | 1.66 (0.82) | 1.49 (0.71) | N/A | 1.50 (182) | .13 | ||
|
|
|
N/A | 3.17 (4) | .53 | ||
|
Parent | 5 (7) | 6 (5) |
|
|
|
|
|
Spouse | 5 (7) | 8 (7) |
|
|
|
|
|
Child | 60 (82) | 85 (77) |
|
|
|
|
|
Friend | 2 (3) | 5 (5) |
|
|
|
|
|
Other | 1 (1) | 7 (6) |
|
|
|
|
|
|
|
N/A | 4.31 (3) | .23 | ||
|
<20 | 4 (6) | 12 (12) |
|
|
|
|
|
20-29 | 3 (4) | 11 (11) |
|
|
|
|
|
30-39 | 3 (4) | 5 (5) |
|
|
|
|
|
>40 | 59 (86) | 76 (73) |
|
|
|
|
|
|
|
N/A |
|
|
||
|
Sensory | 22 (42) | 30 (48) |
|
0.32 (1) | .57 | |
|
Intellectual | 36 (69) | 43 (68) |
|
0.01 (1) | .91 | |
|
Physical | 25 (48) | 25 (40) |
|
0.82 (1) | .37 | |
|
Psychosocial | 46 (88) | 58 (92) |
|
0.43 (1) | .51 | |
|
Head injury/stroke or acquired brain injury | 1 (2) | 3 (5) |
|
0.68 (1) | .41 | |
|
Other | 16 (31) | 25 (40) |
|
0.99 (1) | .32 |
aN/A: not applicable.
Descriptive statistics by group and time point for primary and secondary outcomes.
Variables | Active control, mean (SD) | Intervention, mean (SD) | |||||||||
Baseline | Postintervention | Baseline | Postintervention | Follow-up | |||||||
|
|
|
|
|
|
||||||
|
Stress | 18.82 (7.98) | 18.94 (9.03) | 17.03 (7.88) | 14.72 (7.49) | 12.79 (7.58) | |||||
|
Anxiety | 8.14 (6.76) | 8.61 (6.90) | 7.56 (7.60) | 6.11 (5.86) | 5.58 (5.81) | |||||
|
Depression | 10.95 (8.00) | 10.87 (8.58) | 11.33 (8.67) | 9.66 (7.71) | 7.14 (6.79) | |||||
|
Subjective well-being | 58.02 (15.18) | 54.72 (17.06) | 55.73 (16.15) | 57.98 (17.54) | 62.82 (17.61) | |||||
|
|
|
|
|
|
||||||
|
Mood affect | 52.92 (15.96) | 52.37 (19.35) | 55.48 (16.88) | 58.11 (15.56) | 64.27 (16.06) | |||||
|
Optimism | 5.74 (2.65) | 5.81 (2.89) | 6.33 (2.51) | 6.60 (2.25) | 7.41 (2.22) | |||||
|
Primary control | 40.41 (9.05) | 41.50 (10.38) | 40.51 (9.32) | 42.75 (7.09) | 42.82 (7.62) | |||||
|
Secondary control | 17.88 (6.96) | 15.39 (7.91) | 20.80 (6.93) | 19.60 (8.89) | 19.74 (7.53) | |||||
|
Self-esteem | 9.27 (2.33) | 10.06 (2.36) | 9.44 (2.77) | 10.15 (2.49) | 10.97 (2.32) | |||||
|
Support_Family | 16.56 (6.61) | 17.34 (6.79) | 16.39 (6.54) | 17.66 (6.21) | 19.87 (5.39) | |||||
|
Support_Friends | 18.62 (5.47) | 19.42 (5.94) | 17.93 (5.65) | 19.68 (5.43) | 20.08 (5.81) | |||||
|
Support_Others | 19.91 (5.89) | 21.24 (5.45) | 20.04 (5.72) | 20.91 (5.89) | 23.68 (4.29) |
Multilevel modeling indicated a significant time×group interaction by the postintervention time point for the primary outcomes of anxiety (
These significant interaction effects were followed up with simple effects testing to determine changes in outcomes for the control and intervention groups separately. Stress symptoms were significantly reduced in the intervention group (
Among the secondary outcomes, the group×time interaction was only significant for secondary control (
Among the primary outcomes, significant improvements were observed from postintervention to the 3-month follow-up for depression (
Among the secondary outcomes, significant improvements in symptoms were observed for emotional well-being (
In total, 58 of the 73 individuals allocated to the intervention arm viewed at least one module, although all 73 individuals were retained for analyses consistent with the principles of ITT. On average, participants in the intervention condition completed 2.55 out of the 5 modules (SD 1.05). Psychoeducation (56/58, 97%) and values modules (52/58, 90%) were the most commonly used modules, with less viewing of mindfulness (17/58, 29%), well-being (12/58, 21%), and behavioral activation modules (11/58, 19%).
The number of modules completed moderated the level of improvement in primary control from baseline to postintervention for the intervention group (
The overall quality of the app was rated highly, with a mean score of 3.94 out of a maximum score of 5 (SD 0.58). Participants rated their subjective quality of the app slightly lower (mean 3.19, SD 0.85). Within the subjective quality subscale, participants expressed that they would not choose to pay for the app (mean 2.22, SD 1.14), which was the only item to be rated with a mean score below 2.5. The app was rated particularly positively for its functionality (mean 4.19, SD 0.75), information (mean 3.96, SD 0.63), and aesthetics (mean 3.95, SD 0.63). Although all subscales were rated highly, the engagement subscale achieved the lowest mean score (mean 3.68, SD 0.65). Within the engagement subscale, the items assessing customization and interactivity were rated the lowest (mean 3.31, SD 0.85; and mean 3.47, SD 0.82, respectively).
The purpose of this study was to evaluate the efficacy of a mobile app–based, self-directed psychological intervention for individuals providing care to family or friends with a physical or mental condition. The sample predominantly consisted of mothers of children with a disability with high levels of care burden and stress. The intervention group experienced improvements in the primary outcomes of stress, depression, anxiety, and subjective well-being across the intervention period despite using only a small number of the treatment modules offered, with further improvements in mental health and outlook observed over the 3- to 4-month follow-up period. Participants rated the intervention app highly for its usability and quality, with the potential to improve the app design further through the addition of greater personalization and flexibility. Given the limited number of studies that have investigated the potential of mobile health (mHealth) tools for caregiver populations, the results of this study have important implications for future work in this field.
We found that caregivers initially presented with challenging caring contexts and elevated levels of distress. Importantly, the study sample differed in several ways from national survey data on caregivers in Australia (collected by the SDAC [
Intervention-related effects were observed despite the somewhat low usage across intervention modules. Although participants tended to not complete all modules provided by the app, the modules participants chose to complete appear to have been effective. This finding is consistent with the broader literature, which has found that the therapeutic techniques presented in each module are independently associated with improvements in mental health and well-being [
Although participants rated the quality of the StressLess intervention app highly, their feedback expressed a desire for greater personalization and flexibility in the app design. This suggests that caregivers may benefit from greater opportunities for customization and interactivity in the intervention app’s user experience. Evidence from the broader mHealth literature indicates that tailoring the user’s experience through personalized feedback, prompting, alerts, and reminders is more effective than providing static content to all users [
This study has several limitations. First, as noted earlier, the sample is not broadly representative of caregivers in national studies [
Overall, this study has important clinical implications for the design and effective treatment of mHealth interventions for caregivers experiencing stress. First, the results confirm prior studies showing that caregivers commonly report a need for support for their mental health and well-being, particularly in contexts with high levels of care burden [
Summary of the contents of the five modules of the StressLess intervention.
Summary of reliability estimates per group over time for outcome variables.
CONSORT-EHEALTH checklist (V. 1.6.1).
Australian Bureau of Statistics
cognitive behavioral therapy
Consolidated Standards of Reporting Trials
dependent variable
electronic health
ecological momentary assessment
intention-to-treat
mobile health
Primary and Secondary Control Scale
Personal Wellbeing Index
Survey of Disability, Aging, and Carers
The research was funded by the Deakin University-Australian Unity well-being partnership. As part of the funding agreement, a preliminary report for the study has been published [
MFT, BR, DH, ST, and KL conceptualized and designed the study, coordinated and supervised data collection, carried out the analyses, contributed to the interpretation of the data, and wrote the manuscript. LC, TC, SK, RC, and CO conceptualized and designed the study and critically reviewed and revised the manuscript for important intellectual content. BR developed the app used in the trial. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
The authors declare that they have no competing interests. Although BR made the app, there are no financial incentives to conflict with the aims of this manuscript, as the authors have made the app freely available to the public.