A commonly cited definition of mental health is the one postulated by the World Health Organization: “a state of well-being in which the individual realizes his or her own abilities, can cope with the normal stresses of life, can work productively and fruitfully, and is able to make a contribution to his or her community” [1]. Despite this definition — and numerous other readily used mental health definitions — incorporating positive facets of mental health, mental health care, and the research that underpins it is overwhelmingly focused on understanding and solving problems related to mental disorders [2]. The same applies for mental health intervention research, where historically, efforts have centered around treating or intervening in mental disorders, invariably focusing on “mental health conditions” or “mental illnesses” that significantly affect cognition, emotion, and behavior that may lead to dysfunction or disability [3,4]. Concurrently, research into mental health has focused on the related outcome of psychological distress [5], a state of emotional suffering that is typically characterized by symptoms of depression and anxiety [6].

Although a predominant emphasis on addressing psychological distress and mental disorder still persists, the COVID-19 pandemic has made the need to focus beyond mental disorders in clinical populations more salient. The far-reaching societal consequences of the pandemic increased the need to investigate how the general population can maintain positive and adaptive states of mental health and how these positive and adaptive states can be utilized to buffer against developing more complex problems [7,8]. Although the pandemic brought promotion of positive mental health, or alternatively, states of “mental well-being,” more front-of-mind, these outcomes have for decades already been advocated to be an important standalone mental health outcome and therapeutic avenue, in contrast to the “traditional” focus on merely reducing pathology [9,10]. Mental well-being specifically refers to a state where people generally feel good (ie, experience more positive than negative emotions, feel a sense of life satisfaction) and feel that they can function fully (eg, are able to self-actualize, self-realize, and have a sense of meaning) [11]. In other words, they perceive enjoyment and fulfilment with one’s life as a whole [12]. Decades of research by well-being pioneers such as Diener [13], Ryff [14], and Keyes [15], followed by researchers operating in the field of positive psychology [16], broached the important role of promoting mental well-being in the general population. A now large and established body of research clearly links the presence of mental well-being to a range of desirable physical health (eg, longevity and healthy aging, reduced hospital use) and mental health (eg, lower rates of suicide ideation, more healthy coping) outcomes [17-23].

Psychological distress and mental well-being do not merely occupy opposite ends of a single continuum, and further research is required to understand the complex relationship between the constructs, particularly in intervention research [24,25]. For example, high well-being indicators do not simply equate to being mentally healthy, as both well-being and the absence of distress or disorder are required to be mentally healthy [24,26]. Similarly, mental well-being is not simply the opposite of mental disorder, but rather is an outcome that can co-occur and be juxtaposed to the presence of mental disorder and psychological distress. A scoping review conducted in 2020 identified over 80 scientific publications providing supportive evidence that well-being and disorder/distress are considered to be negatively related but independent outcomes, with both aspects associated with good mental health [27]. First, empirical studies have shown the importance of maintaining and improving the mental state of well-being for the prevention of potential mental disorder [28-30]. Second, studies have shown that improving states of mental well-being in people with current mental disorders can impact rates of recovery [31,32].

While definitive research on the long-term impact of COVID-19 on mental health outcomes across the general population is heterogeneous and not clearly established, with more research needed to determine which population groups are most affected, academic research published since the outbreak of the pandemic generally points to a negative immediate consequence [33-39]. These adverse mental health consequences are often the result of societal restrictions and policies, rather than infection with (or fear thereof) the virus itself [40-42]. This is particularly the case for countries where COVID-19 caseloads have been comparatively low such as Australia. As of July 10, 2021, Australia had a total of 31,017 confirmed cases of COVID-19, resulting in 910 deaths (2.93% death rate) [43]. This death rate, 3.6 per 100,000, is very low compared with other countries such as the United States (184.9 per 100,000) and the United Kingdom (192.5 per 100,000) [44]. Despite this attenuated impact, various studies point to an immediate impact on outcomes of well-being and distress in Australia, particularly in at-risk population groups [45-48]. For instance, Batterham et al [49] found that participants in Australia who had financial distress, social impairment, and work impairment were most impacted. Similarly, Li et al [50] found that young people, who due to their age are already at higher risk of mental health problems, showed elevated mental health problems. University students, who generally display high rates of mental health problems and lower well-being [51], similarly demonstrated mental health problems immediately after the pandemic began [52].

As a result of the observed and anticipated consequences, there have been widespread calls to proactively intervene in mental health by targeting distress and the well-being of the general population [8]. The importance of addressing states of well-being and distress conjointly is advocated by proponents of dual-factor models [24,53], well-being therapy [54], the recovery movement [10], positive (clinical) psychology [2], and positive psychiatry [55]. These distinct streams in mental health care propose that, by taking both well-being and distress into account, the way we deliver mental health care across the spectrum of mental health, ranging from self-help options to treatment of recurrent lifetime disorders, can be improved. For example, mental health interventions incorporating a focus on well-being have the potential to prevent more severe mental disorder, can augment treatment, or can be used as early intervention for subclinical issues. In other words, such interventions can be used as a target to promote mental health in the general population using psychological interventions, addressing prodromal symptoms [26].

A range of behavioral and psychological interventions exist that have proven to positively impact mental well-being in nonclinical settings [56-58]. The most renowned of these psychological interventions stems from the work of positive psychology [59], but various distinct psychological interventions will have a significant impact, depending on a range of moderators including target population and delivery format [58]. One viable delivery format that has been shown to be able to have a sustainable impact in the general population is the use of internet-based solutions, which can be deployed at a larger scale and can be used irrespective of the presence of physical restrictions (eg, lockdowns) [60,61]. This is particularly the case for mental health interventions that target lower-intensity problems or issues related to well-being and mental health promotion, where support by mental health professionals is less warranted [62]. A rapid review by Rauschenberg et al [63] found good evidence for the short-term impact of online mental health interventions during COVID-19 for mental disorder, with evidence for mental health promotion interventions still being sparse.

A by-product of utilizing internet-based interventions with a stronger focus on well-being is that it can also help improve the current gap in service delivery for those experiencing symptoms of mental disorder, as it may help to address various challenges of the mental health system in clinical populations, including access issues, stigma, or “treatment resistance” to name a few [64,65]. For instance, providing complementary or integrated well-being intervention programs may reach clients and community members who do not respond to traditional treatment or do not access these due to associated stigma [66]. Adopting a strong emphasis on well-being also diminishes the reliance on using outcomes of distress and illness as the only indicators of treatment effectiveness. For instance, a poor response to psychological interventions is not uncommon, with nonresponder estimates of 30%-40% being documented [67]. Nonresponse in outcomes of distress is often seen as a lack of treatment impact, particularly in treatment models where the main focus is elimination of symptoms. As mental well-being is related to both prevention of mental disorder and recovery from illness, it plays a fundamental role in personal and functional recovery, which may be considered to be a proxy for future treatment impact [18,27].

In order to determine the merit of psychological interventions on improving mental health outcomes, scholars increasingly stress the importance of measuring both states of distress and well-being to best evaluate mental health intervention impact, a recommendation that comes with an often-overlooked nuance [68]. The impact of an intervention is often determined by comparing an average shift in scores of a cohort of participants using specific outcome measures [69]. For instance, an intervention targeting distress is thought to be beneficial if it can demonstrate an average significant and meaningful change in average scores [70] using, for example, a validated depression outcome measure such as the Depression Anxiety and Stress Scale (DASS-21) [71]. While average improvements are an important way to assess the potential impact of interventions on groups, it does not indicate whether interventions are efficacious or even suitable on an individual level [72].

Comparison of average changes furthermore obscure any possible differential impact of interventions on outcomes of mental well-being or distress within individuals [73]. Average improvements in mental well-being and mental disorder outcome measures at a group level do not necessarily mean that each individual demonstrates equal changes in both measures after receiving an intervention. These types of analyses leave intra- and even interindividual responses unclear and limited in their ability to detect how different individuals may respond in relation to different mental health outcomes. Thus, these generic approaches obscure important results that provide information as to what interventions work, on what dimensions they apply, and who specifically benefits from them. Subsequently, these studies provide limited guidance for practitioners and other stakeholders who wish to understand the nuances of how different interventions affect outcomes within (as opposed to between) individuals and then applying this knowledge to the way they provide mental health care [73].

Scientific studies investigating mental health interventions have not readily reported analyses of intraindividual change across both dimensions of mental health. A notable recent exception is a study by Trompetter et al [68], who investigated the impact of a self-help Acceptance and Commitment Therapy (ACT) intervention for people with clinical depression or anxiety. The authors utilized reliable change indices (RCIs) [74,75] to determine whether individuals demonstrated a meaningful change in their mental health outcomes. The researchers found that improvements in mental health or mental disorder did not necessarily co-occur. While 31% of participants improved across both dimensions, only 12% improved in aspects of positive mental health, and 57% improved in symptoms of distress. Additionally, results further differed for participants with depression and anxiety, showing how outcomes of well-being can interact differently with different outcomes of illness. Thus, analysis of the results using RCI not just enabled assessment of the overall effectiveness of the intervention for these 2 clinical populations but also gave better insight about for whom the intervention showed the most effect.

Additionally, approaches such as those taken by Trompetter et al [68] enable the assessment of change while taking baseline scores into consideration. Baseline scores have a clear impact on the summary of change that is potentially possible for each individual [76]. Positive reliable change cannot be expected if participants start at a baseline level that is not conducive to further improvement. Studies that focus on a reduction in pathology, both mental and physical, therefore typically use severity classifications to inform participant criteria for potential participant recruitment. For example, researchers might only include participants who meet the mild depression cutoff on measures such as the DASS-21 or Hamilton Anxiety and Depression Scales to ensure that participants actually show signs of current distress [77,78]. In contrast, studies that investigate the impact of interventions on mental well-being typically do not consider cutoffs as an important starting factor, neglecting the critical impact of baseline scores; this is an important relation to assess considering the precarious situation of mental health and mental health care and the calls for mental health promotion interventions for the general population during the COVID-19 pandemic [8,48,62].

This study aimed to advance the literature in a number of ways. First, it aimed to determine the impact of an internet-based psychological intervention on mental health outcomes during COVID-19 in Australia. It aimed to add to the currently lacking evidence [63] on the benefit of (internet-based) interventions to promote mental health and well-being, as opposed to targeting clinical symptoms in clinical populations. Second, it aimed to investigate whether the findings reported by Trompetter et al [68] can be replicated in a nonclinical, general population sample; the study aimed to determine differences in reliable change in outcomes of both mental well-being and psychological distress. Rather than using an intervention based on a clinical treatment approach (ACT in the case of the study by Trompetter et al [68]), our study aimed to show this differential impact of a universal, group-based, mental health intervention designed to promote mental health in the general nonclinical population. Finally, the study aimed to examine the effect of baseline differences in well-being and psychological distress on the effectiveness of the intervention.

This study was designed to establish whether a range of mental health outcomes would increase due to exposure to an internet-based psychological intervention and to test whether individuals would display improvements differentially across these outcomes using reliable change analysis. The study was an uncontrolled intervention study comparing data collected using an internet-based measurement tool at the beginning of the intervention compared to data by the end of the intervention. The study was approved by the Flinders University Human Research Ethics Committee (PN 2163).

Participants for this study were from 2 cohorts. The first consisted of adults (18 years or older) from the general population, while the second cohort consisted of adult university students. Participants from the general population signed up via an online website, which was promoted via local print media and radio, an email newsletter for the South Australian Health and Medical Research Institute (SAHMRI), and social media posts on Facebook and Twitter. The second cohort consisted of students who were recruited through one of the major universities in South Australia. No specific inclusion or exclusion criteria applied other than the requirement to be adult, understand the English language, and have access to an internet-enabled device with Zoom [79]. There was no face-to-face contact with any of the participants.

Recruitment was conducted over 6 months from March 2020 until July 2020. Recruitment procedures for the general population and the student population differed slightly. At the university, the study was advertised as a tailored, free, online program to improve mental health and well-being developed by SAHMRI and Flinders University. Promotional information for students was distributed via university emails, student associations, and social media including details about the program, facilitators, and the delivery format. Recruitment material highlighted the importance of well-being and mental health, and importantly, this period coincided with the ongoing impact of COVID-19. Recruitment material provided information about the development and individual components of the program. Interested participants provided their email address, name, and college and were sent an invitation to complete a measurement about their mental health and well-being 1 week before starting the first online session. Measurement was to be completed prior to commencing the intervention. Participants were not obliged to participate in the research study in order to attend the individual sessions. Within the general population, participants could enroll via a website with detailed information about the program or were recruited via partner organizations who were interested in promoting a well-being program to their staff.

Participants were sent an invitation to complete a measurement about their mental health and well-being before starting the initial online session. All participants were invited to attend 5 2-hour weekly sessions hosted online. While initially the study was planned to be conducted in person, COVID-19 restrictions in Australia required online delivery of the program via Zoom [79]. By the final session, participants were invited to complete a second measurement, which was used to determine the post-intervention score outcomes. This was emailed to participants 1 week before the final session, to be completed prior to attending the final session so that participants could track any changes observed since the original measure.

The “Be Well Plan” is a 5-week, group-based psychological program that helps participants create a personalized mental health and well-being plan by experimenting with a variety of resources and evidence-based activities to improve mental health and well-being. The program can be delivered in person in small- to medium-sized group settings (ranging between 10 and 40 participants) or alternatively online via platforms such as Zoom [79], the format utilized for this study.

Program facilitators for the training do not need to be trained mental health professionals, but rather are upskilled to be able to facilitate the training in an engaging and safe way, thereby improving the scalability of the program without further constraining existing health care resourcing. The trained trainers are required to participate in a minimum of 26 hours of face-to-face training. For this study, 6 trained trainers were involved in the delivery of the intervention with a variety of professional backgrounds including well-being research, counselling, workplace development training, and clinical psychology. Facilitators were either staff involved in the development in the intervention (n=4) or colleagues who have a professional training background and were involved in early testing rounds of the intervention (n=2).

The program was designed using a rigorous intervention development process called intervention mapping and among other techniques, relied on the use of co-design with end users and stakeholders [80]. The program aims to impact both outcomes of well-being and outcomes of distress by incorporating evidence-based activities aimed at promoting mental health. Included activities were derived from a large systematic review on distinct psychological interventions aimed at improving outcomes of well-being conducted by members of the research team [58]. A particular strength is that the program is tailored to participants’ unique mental health needs and interests of individual participants. A detailed description of the program and its development will be published in a separate manuscript. A general overview of the intervention can be found in Table 1.

Outcomes measured in this study included positive mental health and psychological distress, which are mental health outcomes that are most relevant to nonclinical populations.

The average participant age was 38.67 (SD 13.06) years. A total of 59 participants (59/89, 66%) were female, who had an average age of 40.05 (SD 14.07) years. Of the participants, 19 (19/89, 21%) were male, with an average age of 37.37 (SD 10.43) years. “Prefer not to say” was answered by 11 participants (11/89, 12%). Most participants were employed (69/89, 77%; average age: 40.71, SD 13.26 years) and not studying (57/89, 64%; average age: 43.25, SD 12.40 years). When comparing students (32/89, 36%; average age: 30.53, SD 9.98 years) to participants from the general population, it was noted that mental health baseline values were significantly worse for students for all outcomes: F_6,82=3.94, P=.002; Wilks Λ=.78; partial η²=.22.

Comparison of pre- and post-intervention scores showed a significant change in mental health variables across time: F_6,83=5.60, P<.001; Wilks Λ=.71; partial η²=.29. Table 2 displays the positive changes in all individual domains and the relevant test statistics, showing a significant positive change in all outcomes measured. Effect sizes for significant outcomes ranged between small and moderate, with the largest significant improvement noted in life satisfaction and the smallest improvements in anxiety.

Analysis of reliable change indicated that a total of 92% (82/89) of the participants demonstrated improvement in at least one of the domains of the outcomes measured. Of these 83 participants, 51% (42/82) showed both improvements in well-being and indicators of distress, whereas 29% (24/82) only showed improvement in well-being and 20% (17/82) only showed improvements in distress. Further, response patterns differed for the various distress categories.

Those who met the threshold for mild depressive symptoms and displayed a reliable improvement largely showed reliable change in both well-being and distress (25/35, 71%), with additional proportions demonstrating improvement in either well-being (5/35, 14%) or distress (5/35, 14%). Those participants who met the threshold for anxiety demonstrated reliable improvement in both anxiety and well-being (15/29, 52%) or well-being alone (13/29, 45%); only one participant (1/29, 3%) demonstrated reliable change in anxiety scores only. The majority of those who met the threshold for stress showed reliable change in well-being outcomes only (19/31, 61%), with 7 participants (7/31, 23%) showing a reliable change in both distress and well-being and 5 participants (5/31, 16%) showing reliable change in distress only.

Change in mental health outcomes significantly differed for those with “low” baseline values compared with those with “high” baseline scores before the intervention. As expected, significant changes were found for life satisfaction and distress across all categories, with those reporting lower baseline scores experiencing significantly lower change in outcome scores (see Table 3). Despite students showing significantly worse baseline mental health problems, no significant interaction effect was noted (F_6,82=1.70, P=.132; Wilks Λ=.89; partial η²=.11), showing that the change in the mental health outcome was not significantly different for students versus nonstudents.