Initial efforts to enhance the R2MR Mental Health Continuum Model

Abstract

Introduction:

The Road to Mental Readiness (R2MR) Mental Health Continuum Model (MHCM) is a mental health self-monitoring tool that allows individuals to self-map their mental health status on six domains using descriptors ranging in severity from healthy to reacting, injured, ill. The goals were to improve the validity of this tool by adding sub-domain descriptors from each domain (Study 1) and assessing whether the addition of these sub-domain descriptors to the domain descriptors resulted in more accurate self-mapping along the MHCM (Study 2).

Methods:

In Study 1, 159 Regular Force members completed a matching task in which they were presented with descriptors and asked to select the MHCM anchor that each belonged to. In Study 2, 274 Regular Force members completed several validated scales and mapped themselves on the MHCM using the descriptors selected in Study 1.

Results:

In Study 1, subsets of sub-domain descriptors were selected that optimized the accuracy mapping scores for all domains and most anchor categories. In Study 2, the addition of sub-domain descriptors enhanced the validity of the MHCM tool for all domains and was associated with the highest correlations with validated scales; most coefficients exceeded concurrent validity thresholds.

Discussion:

These findings provide initial support for the enhanced validity of a MHCM tool that should allow military members to better identify changes in their mental health status.

Lay Summary

Two studies were completed to provide an initial enhancement to the Road to Mental Readiness program’s Mental Health Continuum Model, a tool that helps military members to better understand, and detect changes in, their mental health and that more appropriately guides individuals to take the relevant action to address mental health concerns.

Introduction

Mental health training in the Canadian Armed Forces: the Road to Mental Readiness program

The mental health of Canadian Armed Forces (CAF) personnel is critical to their operational readiness.¹ Relative to the general Canadian population, military members are at higher risk for the development of certain psychological disorders, and the nature of their work increases their risk for exposure to traumatizing and morally injurious events.² The CAF has provided various programs and services focused on both the prevention and the treatment of mental health issues, including the Road to Mental Readiness (R2MR), an evidence-informed comprehensive mental health training and education program that has been effective in increasing mental health literacy and knowledge of mental skills and coping strategies and in improving attitudes toward care seeking and stigma reduction.^3–6 In line with evidence suggesting that repeated application and practice of skills in the training environment improves retention and effectiveness,^7,8 R2MR has developed tools to support the continued application of mental skills beyond the classroom, including a self-monitoring tool called the Mental Health Continuum Model (MHCM).

The MHCM was developed in collaboration with the U.S. Marine Corps in 2008,^9,10 as part of organization-wide efforts (by both U.S. and Canadian militaries) to create a common language and classification system for understanding stress that could be adopted by military members, leaders, family, and health professionals. It allows users to self-map their mental health status on six well-being domains (addictions, emotional well-being, performance and functioning, physical well-being, sleep, and social well-being) along a continuum that ranges in severity across four colour-coded anchors: healthy (green), reacting (yellow), injured (orange), and ill (red). The conceptualization of the MHCM was based on the understanding that, although epidemiological research has focused predominantly on dichotomous outcomes resulting from stress exposure (i.e., the presence or absence of disease), continuous assessments of stress-exposed members’ symptom burden revealed a broad distribution of distress, and such symptoms change over time.^11–14 Indeed, conceptualizing mental health as a continuum ranging from mental health to mental illness more accurately captures the nature of psychological functioning and well-being.^15–18

The MHCM tool provides a framework for individuals to monitor their mental health status, identify mental health concerns, and recognize when to take the appropriate action to address these concerns.¹⁹ Research with Royal Canadian Mounted Police cadets has demonstrated the effectiveness of regular mental health monitoring in improving mental health disorder symptom reporting.²⁰ Through this tool, CAF personnel are provided with mental health education that normalizes mental health fluctuations, creates an expectation of recovery and wellness, and highlights actions that individuals can take to maintain or regain their mental fitness. Given the stated preference of many CAF personnel to manage their own mental health conditions,²¹ the MHCM helps them to distinguish between early distress that can often be self-managed and more persistent symptoms that require professional assistance. Indeed, an online intervention found that depicting mental health and mental illness as a continuum, rather than as a dichotomy, was associated with reductions in stigma.²²

An initial study was conducted by D’Agata and colleagues to assess the validity of the current MHCM.²³ These researchers evaluated the extent to which self-mapping scores were correlated with validated scales in each domain and the extent to which CAF members could match the appropriate descriptors to their correct anchors. With respect to this matching task, they found that although matching accuracy was high for the healthy descriptors, it decreased for the descriptors for the reacting, injured, and ill anchors, suggesting potential issues with the face validity of the descriptors themselves. In terms of the self-mapping task (having participants select the descriptors that best described themselves), the authors found adequate agreement between MHCM self-mapping and the validated measures. Although the validated scale scores strongly predicted group membership for the healthy anchor, these rates decreased for the reacting, injured, and ill anchors, suggesting that participants might be underestimating their symptom severity on the current version of the MHCM tool. Altogether, these findings indicate that modifications to the descriptors might improve the validity of the current MHCM.

Current research

Our main goal was to improve the validity of the MHCM. Specifically, we were interested in evaluating whether the addition of sub-domain descriptors, in addition to revising the current domain descriptors, would improve the accuracy of self-mapping to the MHCM domains. The aim of the revisions was to improve the differentiation of the descriptors or response options by drawing on standardized scale content and construction. In Study 1, we tested the accuracy of the revised general domain descriptors and new sub-domain descriptors, identifying those with the highest accuracy rates. Then, in Study 2, we evaluated whether presenting these sub-domain descriptors in addition to the general domain descriptors (new vs. old) resulted in more accurate self-mapping along the MHCM relative to the general domain descriptors alone.

Methods

Study 1

Descriptor revisions

New general and sub-domain descriptors were created by a R2MR working group in 2020, based on a content analysis of relevant scales in each domain, and further revised by the authors.

Participants and procedure

Email invitations were sent to a random sample of 545 Regular Force CAF members in October 2023, and 159 participants completed the online survey (18% in French) using the Department of National Defence’s Qualtrics platform. No demographic information was collected from participants. Ethics approval for this project was obtained from the Department of National Defence’s Social Science Research Review Board. Participants were presented with four randomly selected descriptors from each domain one at a time, along with the domain title, and selected the anchor (healthy, reacting, injured, ill, or none) that each belonged to. The descriptor and domain order were randomized.

Results and discussion

Proportions associated with assigning each descriptor to each anchor were calculated, followed by mean accuracy scores for the four anchors and then mean accuracy scores for each descriptor overall. Our goal was to identify at least one descriptor per sub-domain, with accuracy rates above 8.3% representing greater-than-chance responding (because each anchor had 12 possible combinations of responses). Thus, we ensured that descriptors had greater than 10% accuracy on most anchors. However, this study was under-powered, so we also ensured that decisions were conceptually informed. For each domain, we calculated whether the results would be improved with the inclusion or exclusion of particular sub-domain descriptors by examining unique combinations of sub-domain descriptors with each of the domain descriptors. These findings are reported where relevant. The full results are presented in Supplementary Table 1.

Addictions

The domain descriptor demonstrated good accuracy (44.9% overall). We selected three sub-domain descriptors that demonstrated the highest overall accuracy (strong desire and causes life problems), that demonstrated better spread across anchors, particularly the injured and ill domains (difficulty to stop or go without), or both.

Emotional well-being

The domain descriptor performed well (51.1% overall). We wanted to ensure that the two central components of anxiety were represented in the anxiety sub-domain: 1) feelings of restlessness, nervousness, and being on edge and 2) inability to control and stop worrying.²⁴ One of the descriptors captured feelings of restlessness and nervousness but did not perform as well as others. We revised the item to also include being on edge. We added one descriptor that captured the frequency of worrying thoughts and ability to stop worrying thoughts. For the posttraumatic stress disorder (PTSD) sub-domain, we retained the highest performing descriptor that captured the ability to feel love for others (55.1% overall). However, we also wanted to capture other core components of PTSD, including intrusive memories, thoughts and behavioural avoidance, and problematic anger.^25,26 Three descriptors were added to this sub-domain to capture these components. For the depression sub-domain, we retained one of the higher performing descriptors (self-hatred). Although the hopelessness descriptor had low accuracy for the injured category, we retained it given its clinical relevance. We also added an item to capture anhedonia.

Performance and functioning

The general descriptor for this domain did not perform well (23% overall). We retained two sub-domain descriptors that captured talking negatively about work and spending more time at work than with family and friends, both of which performed well. We selected one resilience descriptor that captured difficulties in bouncing back from stressors and modified one that captured one’s ability to manage daily stress.

Physical well-being

The general descriptor performed adequately (30.8%). Regarding the physical symptoms of stress, we created a combined descriptor rather than selecting a subset of physical symptoms. The individual descriptors performed adequately (28.0%-49.8% overall). We also retained the activities-limited-because-of-physical-health descriptor that performed well (38.6% overall).

Sleep

The general domain descriptor performed well (40.7% overall). We selected the top performing descriptors for the sleep disturbance and sleep-related impairment sub-domains: trouble falling asleep, staying asleep, or waking too early and activities being disturbed by poor sleep. We combined the worry/sad at bedtime and stress-disturbing-sleep descriptors into a new overall worry, sadness, or stress-disturbs-sleep descriptor.

Social well-being

The domain descriptor for this domain performed well (40.4% overall). The top performing descriptor from each sub-domain was selected and included belonging to a community, having close supportive relationships, and having people to turn to for support.

Overall, our selected descriptors improved the accuracy of mapping scores for all the domains and for most of the anchor categories. All accuracy rates were lower than those previously reported except for the injured category for emotional well-being. However, there are methodological differences; for example, we used a multi-response method compared with a single-choice-response method (where chance accuracy is 8.3% vs. 20.0%, respectively). All our anchors had overall accuracy rates well above 8.3% (ranging from 29.0% for the ill category to 63.5% for the healthy category). In addition, our descriptors were presented in isolation, compared with presenting all four descriptors simultaneously. Another indicator of the descriptors’ validity is their utility in helping participants gain an accurate understanding of their well-being, which was the goal of Study 2.

Methods

Study 2

Participants and procedure

A total of 750 Regular Force members were invited to complete this online study in February 2024 (Study 2a). However, because of an error in data collection in which the social well-being domain scale was omitted from the initial survey, we emailed another 750 Regular Force members to complete a second survey (Study 2b) for the social well-being domain measures. For Study 2a, 183 participants completed the survey; however, 22 were removed for failing attention tests. Thus, Study 2a had a final sample of 161 participants (13% completed the study in French). Participants were mostly men (75%); many were between ages 35 and 44 years (39.8%) and had a post-secondary education (73.4%). Many were junior non-commissioned members (34.4%), were in the army (53.1%), and had on average 15.12 years of service (SD = 8.97). Most participants were not in a mental-health-related occupation (89.1%), had sought mental health treatment or services while serving in the CAF (62.5%), and had received R2MR training in the past (63.3%).

For Study 2b, 113 participants had completed some portion of the survey (15% in French). Their demographic profile was similar to that of the participants in Study 2a. Descriptive statistics for both samples are provided in Supplementary Table 2.

The experiment had several randomization settings within Qualtrics. For the sub-domain descriptors, participants chose the single response that best represented their current state (a multiple-choice format). See Figure 1 for a screenshot of the mapping task for the emotional well-being domain and the anger descriptor set for the PTSD sub-domain. The domain descriptors presented participants with the descriptors alongside the MHCM coloured banner; they ranged from healthy (green) to ill (red) and had a slider response scale (with scores ranging from 1 to 100) to indicate where along the continuum they fell (see Figure 2 for a screenshot).

Figure 1.

Sample sub-domain descriptor presentation for the mapping task

Figure 2.

Sample domain descriptor presentation for the mapping task

Measures

All measures were selected for having demonstrated strong psychometrics properties and for being validated in both English and French. Descriptive statistics for all scale scores are provided in Supplementary Table 3.

Addiction (alcohol consumption). The Alcohol Use Disorders Identification Test (AUDIT) is a 10-item scale created by the World Health Organization to assess alcohol consumption.^27,28 Total scores range from 0 to 40, with 1–7 indicating low-risk consumption, 8–14 indicating hazardous or harmful consumption, and 15 or higher indicating potential moderate to severe alcohol use disorder.

Addiction (gambling motives). The Gambling Motives Questionnaire-Financial (GMQ-F) contains 15 reasons for engaging in gambling behaviour.^29,30 Respondents indicate their frequency of each on a scale ranging from 1 (almost never/never) to 4 (almost always). Scale scores are calculated by taking the mean of the 15 items, with higher scores indicating greater engagement in gambling.

Emotional well-being (psychological distress). The Kessler Psychological Distress (K6) scale is a screening instrument for the identification of possible mental illness; it assesses the frequency of non-specific psychological distress in the past 30 days.^31,32 Participants indicate the extent to which they have experienced symptoms by selecting a response on a scale ranging from 0 (none of the time) to 4 (all of the time). Scores are calculated by taking the sum of the six items, with a cut-off of 13 or more indicating the likely presence of a serious mental illness.

Emotional well-being (suicidal ideation). The Suicidal Ideation Attributes Scale (SIDAS) is a five-item suicidal ideation screening tool on which participants indicate the extent to which they have experienced forms of ideation in the past month on a scale ranging from 1 to 10.^33,34 The total score is calculated by taking the sum of responses, and higher scores indicate greater ideation. A score greater than 1 indicates a risk of suicidal behaviour, and a score of 21 or higher indicates a high risk of suicidal behaviour.

Performance and functioning. The Utrecht Work Engagement Scale-9 (UWES-9) contains nine statements about feelings related to work,^35,36 with responses indicated on a scale ranging from 1 (almost) to 6 (always). Scores are calculated by taking the sum of items, with higher scores reflecting higher workplace engagement.

Physical well-being. Using the 15-item Physical Health Questionnaire (PHQ-15),³⁷ participants indicate the extent to which they experienced 15 physical symptoms during the past 7 days on a scale ranging from 1 (not at all) to 4 (constantly).

Sleep. The Sleep Condition Indicator (SCI) is a screening tool for insomnia disorder. Eight questions capture sleep habits during the past month,^38,39 and responses are indicated on a five-point scale. Total scores are calculated by summing all items, with higher scores reflecting better-quality sleep, and low mean scores (0–2) reflecting the threshold for insomnia disorder in the Diagnostic and Statistical Manual of Mental Disorders (5th ed.).²⁵

Social well-being. The Multidimensional Scale of Perceived Social Support (MSPSS) assesses an individual’s perceptions of social support from close others.^40,41 It has 13 items, and responses are indicated on a scale ranging from 1 (very strongly disagree) to 7 (very strongly agree). Scores are calculated by taking the mean of all items, with higher scores reflecting higher social support.

Results

Correlation analyses

We compared the self-mapping scores from each condition with validated scale scores within domains (see Supplementary Table 4). We standardized all descriptor mapping scores and then created aggregate scores by taking the average of respective items. Correlations of 0.75 or higher indicate adequate concurrent validity.⁴² For all domains, we also examined whether removing particular sub-domain descriptors improved findings. In addition, gender did not moderate these findings when examining correlation coefficients separately for men and women.

For addictions, the AUDIT was most strongly and significantly correlated with the old domain plus the sub-domain descriptors. The GMQ-F was not significantly correlated with the self-mapping conditions. For emotional well-being, both validated measures (K6 and SIDAS) were most highly and significantly correlated with the new domain plus sub-domain descriptors condition (exceeding 0.75 for the K6). For physical well-being, the PHQ-15 was most strongly and significantly correlated with the new domain plus sub-domain descriptor condition (exceeding 0.75). For performance, the UWES-9 was most strongly correlated with the old domain descriptor alone. However, by retaining the two descriptors significantly correlated with the UWES-9 (ability to focus energy and attention to achieve work objectives and talking negatively about one’s work), correlations for the two sub-descriptor conditions exceeded the previous highest values, with the new domain plus area descriptors most strongly correlated with the UWES-9. For sleep, the SCI was most strongly and significantly correlated with the new domain plus area descriptors (exceeding 0.75). For social well-being, the MSPSS was most strongly and significantly correlated with the old domain plus the sub-domain descriptor condition.

All coefficients for the selected descriptor conditions were higher than those reported by D’Agata and colleagues,²³ meaning that the descriptor scores were more highly correlated with validated measures in each of those domains than the previous descriptors, lending support for the improved validity of the revised MHCM tool (see Table 1). The retained descriptors for this revised MHCM tool are presented in Supplementary Table 4. Alternate presentations are also included in the supplemental file (Supplementary Figures 1–7).

Table 1.

Pearson correlations among standardized MHCM spectrum scores and validated scale scores by descriptor condition

Domain and scale	Condition
	Old domain	New domain	Old + sub-domain	New + sub-domain
Addictions
AUDIT	0.46*	0.40*	0.59*	0.34†
GMQF	0.14	0.21	0.20	0.26
Emotional Well-Being
K6	0.45*	0.63*	0.78*	0.90 *
SIDAS	0.48*	0.59*	0.51*	0.61 *
Performance & Functioning
UWES-9 (all sub-domain descriptors)	−0.56*	−0.51*	−0.46*	−0.40*
UWES-9 (sub-domain descriptors 1 and 2 only)	−0.56*	−0.51*	−0.63*	−0.69*
Physical Well-Being — PHQ-15	0.64*	0.58*	0.62*	0.86 *
Sleep — SCI	−0.74*	−0.80*	0.82*	0.89*
Social Well-Being — MSPSS	−0.67*	−0.22	−0.86‡	−0.37*

Note: Boldface indicates the experimental conditions with the best set of descriptors that were retained.

*p < 0.01.

^†p < 0.05.

^‡p < 0.001.

MHCM = Mental Health Continuum Model; AUDIT = Alcohol Use Disorders Identification Test; GMQF = Gambling Motives Questionnaire-Financial; K6 = Kessler Psychological Distress; SIDAS = Suicidal Ideation Attributes Scale; UWES-9 = Utrecht Work Engagement Scale; PHQ-15 = 15-item Physical Health Questionnaire; SCI = Sleep Condition Indicator; MSPSS = Multidimensional Scale of Perceived Social Support.

Mean scores on validated measures and MHCM anchors

We examined mean scale scores according to overall MHCM anchor scores and against scale cut-off criteria (see Supplementary Table 5). For all domains, means increased or decreased as expected for each anchor category. The mean AUDIT score for the injured group met the criteria for hazardous and harmful drinking, whereas those for the other groups did not. Regarding psychological distress, those who self-assessed as ill on average for the emotional well-being domain met the cut-off indicating a potential psychological disorder, whereas the other groups did not.

Predicted group membership of validated measures

We examined whether the scale scores significantly predicted overall MHCM mapping to the four anchors (see Table 2). Overall, the rates for correctly classifying participants’ overall domain anchors according to validated measures in those areas ranged between 36.7% and 61.5%, which were lower than those previously reported (although our analyses were more under-powered). Our findings align with previous research in that correct classification rates for the healthy and ill categories were highest and those for the reacting category were low. However, our classification rates for the injured category were higher on average than those previously reported. Also consistent with D’Agata and colleagues,²³ these findings improved when we collapsed the reacting and injured categories, suggesting that individuals struggle more with distinguishing between these states.

Table 2.

Correct classification of participants’ MCHM overall domain anchors based on validated scales

Domain and variable (scale)	Wilks’ Λ	N	χ2	df	Classification, %
					Total	Healthy	Reacting	Injured	Ill
Addictions: Alcohol (AUDIT)	0.82	108	24.81*	2	56.6	64.3	47.5	—	—
Emotional Well-Being
Distress (K6)	0.39	56	50.14*	3	57.1	92.3	50.0	37.5	66.7
Suicidal Ideation (SIDAS)	0.57	51	26.71*	3	45.1	91.7	8.7	61.5	66.7
Performance & Functioning: Work Engagement (UWES-9)	0.55	59	33.30*	3	54.2	85.7	52.9	37.5	100.0
Physical Well-Being: Physical Health (PHQ-15)	0.70	49	16.07*	3	36.7	62.5	47.1	5.6	66.7
Sleep: Sleep (SCI)	0.30	65	73.46*	3	61.5	77.8	40.0	69.6	87.5
Social Well-Being: Social Support (MSPSS)	0.39	51	45.12*	3	49.0	76.9	29.2	71.4	42.9

Note: Dashes indicate no participants mapped to that anchor.

^*p < 0.001

MCHM = Mental Health Continuum Model; AUDIT = Alcohol Use Disorders Identification Test; K6 = Kessler Psychological Distress; SIDAS = Suicidal Ideation Attributes Scale; UWES-9 = Utrecht Work Engagement Scale; PHQ-15 = 15-item Physical Health Questionnaire; SCI = Sleep Condition Indicator; MSPSS = Multidimensional Scale of Perceived Social Support.

Mental health screening scales with established cut-off scores

Finally, we examined the relationships between total MHCM anchor scores and the proportion of individuals who screened positive for a potential psychological disorder, suicidal ideation, hazardous and harmful alcohol use, and insomnia disorder. A total of 19.6% of those sampled (n = 56) screened positive on the K6 for a psychological disorder. Of those who screened positive, none self-assessed as healthy, 18.0% self-assessed as reacting, 54.5% as injured, and 27.3% as ill. Of those who screened negative, 28.9% self-assessed as healthy, 48.9% as reacting, 22.2% as injured, and none as ill. The single individual who screened positive for suicidal ideation self-assessed as ill on the MHCM. Of those who screened negative, 24.0% self-assessed as healthy, 46.0% as reacting, 26.0% as injured, and 4.0% as ill. Of those who screened positive for hazardous and harmful alcohol use (22.2% of those sampled; n = 54), 8.3% self-assessed as healthy, 83.3% as reacting, and 8.3% as injured. Of those who screened negative for hazardous and harmful alcohol use, 54.8% self-assessed as healthy, 42.9% as reacting, and 2.4% as injured. The single individual who screened positive for insomnia disorder self-assessed as ill on the sleep domain overall. In contrast, of those who screened negative, 14.0% self-assessed as healthy, 39.0% as reacting, 35.9% as injured, and 10.9% as ill.

General discussion

An initial study assessing the validity of the MHCM self-monitoring tool indicated that although the tool was performing adequately in some areas, further enhancements to the tool should be made. We proposed the addition of sub-domain descriptors in addition to revising the domain descriptors. The purpose of this two-study project was to identify the most valid subsets of sub-domain descriptors with the highest accuracy rates to include in Study 2 and then to evaluate the validity of this improved version of the MHCM.

Summary of findings

In Study 1, the selected descriptors improved the overall accuracy of mapping scores for all the domains. Although our accuracy rates were lower than those previously reported, our matching task was also methodologically more challenging. Overall, participants could categorize the descriptors with their appropriate anchors at well above chance level and with the pattern of findings mirroring those of D’Agata and colleagues.²³ Participants found the injured and ill descriptors more difficult to decipher and the healthy descriptors easiest to categorize.

In Study 2, we found that the addition of sub-domain descriptors enhanced the validity of the MHCM tool for all domains, such that the overall MHCM anchor scores were associated with higher correlations with validated domain measures, and four of the six domains had coefficients exceeding a concurrent validity threshold. We also found that mean scores on validated scales increased or decreased as expected across the MHCM anchors for all domains, lending further credibility to its validity, that those who self-assessed as injured in addictions met the criteria for hazardous and harmful drinking, and that those who self-assessed as ill for emotional well-being met the cut-off for a potential psychological disorder.

Our rates for correctly classifying participants’ overall domain anchors according to validated scales were lower than those previously reported, but our analyses were more under-powered. Nonetheless, our findings align with those of previous research in that classification rates for the healthy and ill categories were highest, and some individuals still struggled with distinguishing between the ill and injured states.

Finally, we found evidence for the potential validity of some domain descriptors in being able to discriminate between those who screened positive versus negative for potential disorders: 1) the emotional well-being descriptors for a likely psychological disorder and suicide ideation, 2) the addictions descriptors for those with and without problematic alcohol use, and 3) the sleep descriptors for potential insomnia disorder.

Strengths and limitations

Strengths of this research include that the descriptors generated and selected were based on validated scales in those respective areas and were conceptually relevant. Although the MHCM is not meant to serve as a diagnostic tool, we ensured that the domain most closely tied to mental health (emotional well-being) contained descriptors that cover the central components of anxiety, depression, and PTSD. In this way, if an individual is experiencing any of these symptoms, the MHCM tool should help individuals to identify them to some degree. In addition, our surveys were rigorous in their design, ensuring that randomization was applied whenever possible to limit the impact of response biases, order effects, self-consistency effects, and other confounding variables.

However, these strengths are qualified by some limitations. Although many analyses were well powered (e.g., correlations with validated scales, t-tests), some analyses were under-powered (i.e., discriminant functions for classifications, estimating descriptor parameters in the mapping task). This posed challenges in adequately comparing our classification rates with those previously reported and in our ability to project these findings beyond our samples. We did not find evidence to suggest that gender moderates these findings; however, a more robust assessment of fit based on demographic variables was not possible with this sample. It is worth noting, however, that our sample characteristics reflected what we would expect from a typical sample of CAF Regular Force members (i.e., more men and those younger and of lower rank, more army and fewer navy uniformed members, and more members with post-secondary education relative to those without post-secondary education).⁴³ Future work should focus on the potential utility of this version of the MHCM with a larger sample and based on relevant demographic variables.

Implications and future directions

The MHCM is a critical educational tool in the CAF, which highlights the continuous nature of mental health and the development of, and recovery from, mental illness. This tool promotes the idea that fluctuations between healthy and more distressed states is common, that movement in both directions is possible, and that ultimately there is always the possibility of a return to full health and functioning when temporary or more severe distress is experienced. The practical utility of this tool is to help increase members’ awareness and monitoring of their mental health status and to be able to identify changes in their health and performance. Such self-monitoring can help individuals to identify either existing or developing mental health concerns and then prompt individuals to take the appropriate action (e.g., seeking mental health treatment, seeking social support) to address these concerns.

These findings will have direct practical implications for the design of the MHCM tool in the R2MR mobile application, in addition to how the MHCM is taught in R2MR courseware. Learning about the various areas in each domain in which mental health symptoms and distress can present or manifest can help provide members with greater mental health literacy and identify signs of distress in themselves and others. Many forms of the MHCM model are currently being used by other organizations.^44–46 These findings could have implications for how others are currently conceptualizing, framing, and educating others on the MHCM.

One notable finding consistent with that reported by D’Agata and colleagues was that individuals had a more difficult time distinguishing between the injured and ill states and that collapsing across these anchors improved classification rates by validated scales.²³ Although a three-anchor continuum might possibly more parsimoniously categorize individuals, there is conceptual value in maintaining a distinction between injured and ill states when the goal is early detection of more persistent impairments in functioning and mental illness symptomatology and early care seeking. Recognizing when distress is treatment worthy (i.e., mental health literacy) is a predictor of care seeking behaviour, with delays in distress recognition linked with poorer treatment outcomes.^47–52

In addition, the stigma surrounding mental illness is still a prevalent concern in military populations.⁵³ In this way, the injured category provides an alternate and potentially less threatening choice than the more severe ill category, allowing individuals to acknowledge their distress and be encouraged to seek care while acknowledging biases (including self-stigma) that might prevent them from wanting to identify with the ill state. This would be an interesting avenue of research, in regard to assessing the relationships among stigma and MHCM mapping.

Conclusion

Through this research, we were able to enhance the MHCM tool by adding sub-domain descriptors that describe various states of functioning within each of the six domains of well-being. This enhanced MHCM should allow individuals to better map themselves along the continuum, identify detailed changes in their mental health status, and take the relevant action (e.g., engage in recovery strategies, seek social support, and seek mental treatment).

Biographies

Valerie M. Wood,, PhD, is a research and evaluation specialist for the Road to Mental Readiness program in the Canadian Forces Health Services Group, Department of National Defence. Her research has mainly focused on understanding individual differences in, and ways to support, the mental health, resilience, and well-being of military members and their families.

Suzanne Bailey, MSW, served in the Canadian Armed Forces as a social work officer and currently works with the Department of National Defence as the team lead for curriculum development for the Road to Mental Readiness (R2MR) program. As one of the founders of the R2MR program, she has been fortunate to lead and contribute to the development and delivery of mental health and mental performance training for Canadian Armed Forces personnel since 2008.

Kimberly Guest, MSW, has been practising clinical social work since 1995 after receiving her bachelor of social work from Carleton University in 1995 and her Master of Social Work from the University of Calgary in 1998. Kimberly is the Senior Curriculum Developer for the Road to Mental Readiness program, providing resilience and mental health curricula for the Canadian Armed Forces. Kimberly has always believed in the importance of education and early intervention in preventing long-term suffering and loss.

Funding Statement

No funding was received for this article.