Access to Mental Health Services: Precariously Employed Workers Experiencing Anxiety or Depression Encounter Barriers When Seeking Care

Abstract

Background: This study synthesizes findings on precariously employed workers’ self-reported feelings of severe or extreme anxiety and depression, along with their experiences accessing mental health services. Methods: This mixed-methods research included surveys (N = 259) and interviews (N = 40) with precariously employed workers in Ontario, Canada, conducted from November 2020 to July 2021. Inclusion criteria included: (i) not being directly employed, being self-employed, or a gig worker; (ii) not working full-time; (iii) not holding a permanent or open-ended contract; (iv) performing informal work; or (v) being recently unemployed. Results: The adjusted, statistically significant odds of reporting severe or extreme anxiety or depression were higher among workers with greater precarity (2.28), self-employed workers with no employees (3.61), gig or platform workers (3.08), workers earning less than 60% of the median income (2.75), and those unsure whether their hours would vary in the next three months (2.59). The odds were lower (0.22) for workers with some or little income variation in the previous three months. Interview participants described chronic stress, worry, anxiety, depression, and overall negative wellbeing linked to their precarious employment. Despite an increased need for mental health services, participants reported similar difficulties accessing them. Interpretation: To improve access to mental health services, sustainable intersectoral solutions with demonstrated potential are required, including increasing social and health expenditures, revising labor market legislation, and reorganizing the delivery of employer-dependent health services. Recommendations are made for solutions at various levels, including those that could be adopted by medical practitioners.

1. Introduction

Mental illness—defined as a disorder affecting mood, cognition, or behaviour that may hinder social and occupational functioning—is a major public health concern globally and in Canada [1,2,3]. The World Health Organization estimates that one in eight people worldwide lives with a mental disorder, with anxiety and depression among the most prevalent conditions [4]. National estimates suggest that one in three Canadians will experience mental illness during their lifetime [1,3], and approximately 15% seek related health services annually [2,5]. Based on 2012 data from the Canadian Community Health Survey—Mental Health, approximately one in fourteen Canadians experienced generalized anxiety disorder and about one in twenty-five experienced major depressive disorder in the previous year [1,6]. The economic burden of mental illness in Canada is substantial, estimated at over $50 billion annually in lost productivity, healthcare utilization, and social costs [7,8]. The COVID-19 pandemic further exacerbated population-level mental health, precipitating marked increases in anxiety, depression, and psychological distress across demographic groups in Canada and internationally [9,10,11], while simultaneously deepening pre-existing inequities in access to care. Consistent with these broader patterns, recent Canadian national polling data show that one in ten Canadians has faced persistently high levels of anxiety or depression following the pandemic [12]. These trends underscore the need to understand the structural and social determinants that shape mental health and its treatment.

Precarious employment (PE), characterized by job insecurity, low wages, and limited workplace protections, is increasingly prevalent in post-industrial economies [13,14,15,16] and has been strongly linked to poorer mental health outcomes [17,18,19,20,21,22,23,24,25]. PE encompasses a broad and heterogeneous range of work arrangements, including temporary contracts, on-demand and platform-based gig work, informal employment, involuntary part-time work, and self-employment without adequate social protection [14,26,27]. The expansion of digital labor platforms during the pandemic has accelerated the fragmentation of traditional employment relationships, giving rise to new forms of precarity that often fall outside existing regulatory frameworks [28,29]. In Canada, gig and non-standard work arrangements have grown considerably over the past decade [30], disproportionately affecting racialized workers, women, immigrants, and young adults [30,31]. The psychological mechanisms linking precarious work to poor mental health are well-documented and include chronic stress arising from income volatility, the unpredictability of work schedules, the erosion of workplace identity and social support, and the constant anticipation of job loss [18,23,32].

In addition to accelerating digital platform use, the COVID-19 pandemic produced other lasting shifts in professional practices across business and social psychological domains, particularly for precariously employed and frontline workers. For instance, evidence from Ontario shows that platform-based work has normalized algorithmic management, broadly defined as the re-allocation of decision-making from human supervisors to algorithms, or automated systems [33]. These new modes of work organization could undermine worker health and wellbeing or the adoption/enforcement of occupational health and safety measures, given they subject workers to ongoing performance monitoring, intense time pressures, and income volatility [34]. Notably, these patterns have persisted beyond the acute phase of the pandemic in gig and app-based sectors [35]. Similar dynamics were observed in care work, where irregular scheduling, multi job holding, limited paid sick leave, and non-permanent contracts exposed structural weaknesses in employment protections that continue to shape working conditions for personal support workers [36,37].

COVID-19 and the period following it made visible and institutionalized enduring stratifications in access to remote work, reinforcing occupational and income-based inequalities within labor markets and business practices. Canadian evidence demonstrates that work-from-home arrangements became normalized in higher wage, professional occupations, while low wage, public facing jobs—common among many precarious workers—remain largely in person, concentrating health and economic risks in sectors with weaker labor protections [38,39]. These divisions continue to shape occupational health exposure, flexibility, and bargaining power in post-pandemic labor markets.

From a social psychology perspective, these structural shifts have left enduring effects on worker wellbeing and occupational health and safety. Studies document sustained stress, burnout, post-traumatic stress symptoms, and weakened organizational trust among public health, care, hospitality, and homelessness service workers, alongside heightened intentions to leave front line occupations [40,41,42]. While emergency income support during the pandemic mitigated short term financial shocks, it did not resolve underlying gaps in labor protections or workplace support [43]. Post-pandemic policy responses, including new standards for platform work, reflect growing recognition of these issues, but also underscore that COVID-19’s most significant legacy lies in its revealing and reshaping of enduring inequalities in work organization, mental health, and employment security.

Workers in precarious jobs face unique barriers to accessing health services—particularly mental health care—due to insufficient benefits, unstable income, and a lack of employer support [18,22,44,45,46,47,48,49]. Canada’s fragmented mental health system [50]—in which many services are delivered privately and funded through employer-provided benefit plans—structurally disadvantages those in non-standard employment, who are least likely to hold such coverage [7,51]. This financing model creates a paradox whereby the populations with a high likelihood to experience mental illness due to occupational stress are simultaneously less likely to afford or access treatment [52,53]. Beyond financial barriers, precariously employed workers may also face temporal and logistical obstacles, including inflexible or unpredictable work schedules that conflict with appointment-based service delivery, fear of income loss associated with taking time off, and stigma related to disclosing mental health conditions to employers or within social networks [18,24,44]. Barriers to accessing health services can exacerbate health inequities and contribute to long-term societal costs [17,53,54,55,56]. Recent data show that both precarious work arrangements [30,57] and mental health needs are on the rise in Canada [3,12], while access to care remains inadequate [12,51,52,58,59,60]. For instance, Canadian national polling data indicate that more than half of individuals who are struggling with their mental health are not receiving the support they need [12], highlighting the urgency of addressing these intersecting challenges.

Despite a growing body of research on PE and mental health, significant gaps remain in the literature. Much of the existing evidence draws on European cohorts, relies on single-dimension measures of precarity, or focuses narrowly on employment insecurity without accounting for the multidimensional nature of PE [14,27]. Research examining the specific pathways through which PE shapes help-seeking behavior and access to mental health services in the Canadian context remains limited [22,24,47]. Furthermore, qualitative and mixed-methods approaches that can illuminate the lived experiences of precarious workers navigating mental health systems are underrepresented in the field [21]. This study directly addresses these gaps by adopting a multidimensional conceptualization of PE and integrating quantitative and qualitative data to capture both the prevalence and experiential dimensions of mental illness and service access among precarious workers in Ontario.

This paper presents findings from a mixed-methods study conducted in Ontario, Canada, between November 2020 and July 2021, as part of a broader investigation into the health effects of precarious and non-standard employment [18,45,61] carried out across six countries (Sweden, Belgium, Canada, Chile, Spain, and the United States) within a multi-year (2019–2025) research program undertaken by the Precarious Work Research (PWR) research consortium. The mixed-methods study described in this manuscript included 259 survey responses and 40 interviews with individuals meeting one or more criteria of employment precarity (e.g., self-employed, gig workers, informal or part-time workers, or recently unemployed). This manuscript focuses specifically on participants’ self-reported experiences of severe or extreme anxiety or depression, and their attempts to access mental health services. Our objectives are twofold: (1) to analyze associations between PE and mental illness, and (2) to examine how precariously employed individuals navigate access to care. After describing our methodology, we synthesize key findings and conclude with policy recommendations aimed at improving mental healthcare access for this deprived workforce.

2. Methods

2.1. Study Design and Sampling

This study employed a convergent mixed-methods design [62], integrating quantitative survey data with qualitative semi-structured interviews to examine the relationship between PE and self-reported mental health outcomes among workers in Ontario, Canada. We selected only one province, Ontario, as the study setting for both conceptual and practical reasons. In Canada, each province has its own labor market legislation, effectively creating distinct labor market environments [63]. Ontario, as the most populous and diverse province [64], with one of the country’s highest levels of educational attainment among its labor force [65], presents unique labor market characteristics that limit the applicability of broad generalizations across provinces. Additionally, because our research team is based in Ontario, we aimed to leverage existing partnerships with local community organizations to support and facilitate participant recruitment.

Survey inclusion criteria comprised workers aged 25–55 years who met at least one of the following characteristics at the time of survey completion or within the preceding three months: (i) not being directly employed by an employer, being self-employed (with or without employees), or engaged in gig/platform work; (ii) not working full-time; (iii) lacking a permanent or open-ended contract; (iv) engaged in informal employment (not paying taxes or making employer-based pension contributions); or (v) having experienced unemployment in relation to the COVID-19 pandemic. The age range was selected to capture workers in their primary working years while minimizing confounding from early-career transitions or pre-retirement dynamics.

Participants were recruited through a multi-pronged convenience sampling strategy. Survey recruitment involved advertisements posted on employment news media and social media channels, along with targeted outreach to organizations and individuals engaged in employment-related research and advocacy. The REDCap survey was provided as a clickable link in social media advertisements, whereas the recruitment materials shared with community organizations included the survey link in the form of a QR code. The survey was administered in English, November 2020 to June 2021, via REDCap online survey software version 4.0 [66]. A minimum sample size of five hundred was targeted to adequately power the study based on significance level of 0.05, and an expected event outcome of one in ten [67]. We used the 2012 Canadian prevalence data for generalized anxiety disorder and major depressive disorder [6] along with data documenting marked increases in anxiety, depression, and psychological distress across demographic groups in Canada and internationally during the COVID-19 pandemic [9,10,11] to estimate the expected event outcome. However, despite the survey remaining open for eight months to maximize participation and allow a larger number of respondents to complete it, ongoing recruitment efforts yielded limited success in enrolling eligible participants to meet the planned sample size.

To mitigate fraudulent or automated responses, only surveys containing valid postal codes, phone numbers, and email addresses were retained. Additionally, responses entered in close temporal proximity (within minutes) were manually reviewed for validity. Postal codes, phone numbers, and email addresses were collected to: (i) facilitate follow-up contact with eligible participants interested in interview participation, and (ii) ensure geographic representation across the province. Personal identifiers were separated from survey responses, with linking keys accessible exclusively to the project coordinator. No computer IP addresses were collected to further protect participant anonymity.

The total survey sample of eligible participants was 448, while the final sample for this analysis comprised 259 survey responses because only respondents who completed the second part of the survey, exploring links between employment, health, wellbeing, occupational health and safety, and household economic outcomes, including the question about anxiety/depression are included in this analysis. Given recruitment challenges during the COVID-19 pandemic, incomplete surveys were retained provided that key demographic and employment-related variables—including age and employment characteristics necessary for PE classification—were complete. This approach maximized analytical power while maintaining data integrity for primary study outcomes.

2.2. Survey Instrument and Measures

The survey consisted of two sections: (i) demographic and employment-related characteristics, and (ii) relationships between employment characteristics and health, wellbeing, occupational health and safety, and household economic outcomes. Across these sections, the instrument included a mix of multiple choice, categorical, frequency-based Likert scale, numeric entry, and one open-ended question at the end of the survey. Most questions were closed-ended to allow standardized coding across participants. Multiple choice questions used predefined response categories (e.g., employment arrangement, benefits received, household composition), while Likert type items captured frequency or degree (e.g., changes in hours, income variation, fear of requesting better working conditions). Several items required numeric responses, such as age, height, weight, and household size. Only one question—the final item inviting participants to describe how the COVID-19 outbreak affected their working lives—was fully open-ended. A small number of items included optional text boxes for clarification when respondents selected an “other” category. The instrument included mandatory screening questions related to age, recent employment, and job type to determine eligibility, but all other questions permitted skipping or selecting “not applicable,” consistent with the study protocol. No forced responses were used beyond what was necessary for inclusion/exclusion.

PE was operationalized using modified versions of the Employment Precariousness Scale (EPRES) [68] and the Multidimensional Precarity Index [19], which assess multiple dimensions of employment precarity, including temporariness, disempowerment, vulnerability, wages, and rights. A composite PE score was calculated for each participant and dichotomized at the median threshold (score > 15 = high PE; score ≤ 15 = low PE) to facilitate comparative analyses.

Given the study’s temporal overlap with the COVID-19 pandemic, we incorporated items assessing pandemic-related impacts on employment and health. Mental health was evaluated using a single item from the EQ-5D-5L quality of life instrument: “I am severely or extremely anxious or depressed.” [69]. While this serves as a screening-level measure rather than a diagnostic assessment, it provides a validated indicator of clinically significant psychological distress.

2.3. Statistical Analysis

Descriptive statistics were calculated to characterise the sample, and the prevalence of PE or other employment characteristics, and that of self-reported severe or extreme anxiety or depression. Logistic regression analyses, with each predictor analyzed separately, examined associations between PE status, other employment-related characteristics, and self-reported severe or extreme anxiety or depression, adjusting for relevant covariates. All quantitative analyses were performed using Stata version 15.1 [70].

2.4. Qualitative Interviews

Semi-structured interviews were conducted virtually between January and July 2021 by four trained interviewers using Zoom for Healthcare, a platform offering end-to-end encryption and compliance with HIPAA and PHIPA standards for health research [71]. Each interview lasted approximately 60–90 min. The interview guide explored participants’ employment characteristics and their perceived impacts on physical and mental health, wellbeing, and household economic stability.

Interview recordings were transcribed using Otter.ai web version, an AI-based transcription tool [72]. All transcripts were subsequently reviewed manually to ensure accuracy and contextual integrity. A theory-informed coding framework, developed collaboratively across countries participating in the broader international study, was applied. This framework underwent iterative refinement to reflect country-specific labor market conditions and healthcare contexts in Canada and the province of Ontario. The refinements made to the semi-structured interview guide were intended to incorporate Canadian and Ontario specific labor market attributes—such as employment legislation and labor protections—because these exposures were common to all participants, even though their impacts may have differed by immigration status, ethnicity, or other social identities. The interview guide used open-ended questions to allow respondents to draw on their own cultural, linguistic, and personal experiences, acknowledging the cultural heterogeneity within the participant group. As the study did not aim to compare experiences across specific cultural or national backgrounds, we did not develop multiple culturally tailored interview protocols. Instead, the semi-structured format enabled diverse cultural and experiential perspectives to emerge organically through participants’ narratives, which were integrated into the reflexive thematic analysis. This approach ensured consistency across interviews while remaining sensitive to participants’ varied backgrounds. Coding was performed using NVivo 12 [73] following multiple rounds of coder training and alignment exercises to enhance inter-coder reliability.

Data analysis employed an abductive thematic approach, integrating theory-driven codes with inductive insights emerging from the interviews [74]. Themes were synthesized in relation to the study’s key research questions, focusing on: (i) the links between employment characteristics and workers’ self-reported experiences of severe or extreme anxiety or depression, and (ii) workers’ experiences with mental illness and healthcare access. This synthesis provided nuanced, contextualized insights aligned with the mixed-methods research design [75].

2.5. Mixed-Methods Integration

The quantitative and qualitative components were integrated through a convergent parallel design [62], allowing findings from both components to be examined together and supporting a comprehensive understanding of both prevalence patterns and lived experiences [76]. Quantitative results informed the contextualization of qualitative findings, while interview data shed light on mechanisms underlying statistical associations.

2.6. Ethics

All procedures were approved by the St. Michael’s Research Ethics Board, MAP Centre for Urban Health Solutions, Li Ka Shing Knowledge Institute, Unity Health, Toronto, ON, Canada (REB 20-110). Informed consent was obtained from all participants prior to data collection.

3. Results

3.1. Surveys

Demographic characteristics of the survey sample are shown in

Table 1. Descriptive demographic characteristics of the survey sample and prevalence of severe or extreme anxiety/depression.

Demographic Characteristics	Convenience Survey Sample N (%)	Severely or Extremely Anxious or Depressed, n	% with Severe/Extreme Anxiety or Depression Within Category *	% of All Respondents with Severe/Extreme Anxiety or Depression by Category **
Age	259	51
25–35	91 (35.1%)	23	25.3%	45.1%
36–45	90 (34.7%)	17	18.9%	33.3%
46–55	78 (30.1%)	11	14.1%	21.6%
Gender	259	51
Male	45 (17.4%)	9	20%	17.6%
Female	200 (77.2%)	39	19.5%	76.5%
Gender variant/non-binary	10 (3.9%)	3	30%	5.9%
Prefer not to answer	4 (1.5%)	0	0%	0.0%
Immigration Status	259	51
Immigrant	33 (12.7%)	5	15.2%	9.8%
Non-immigrant	226 (87.3%)	46	20.4%	90.2%
Education Level	256	50
Primary or high school graduate	28 (10.9%)	4	14.3%	8.0%
Some college or university education	57 (22.3%)	16	28.1%	32.0%
College or university graduate	171 (66.8%)	30	17.5%	60.0%

Only respondents who completed the survey section assessing anxiety/depression were included in these analyses. Ns vary slightly across variables due to item-level missing demographic data. * “% within category” represents the proportion of participants within each demographic category who reported severe or extreme anxiety or depression (e.g., 23 of 91 participants aged 25–35). ** “% by category” represents the proportion of all participants reporting severe or extreme anxiety or depression who fall within each category (e.g., 23 of 51 participants). Percentages in this column sum to 100% within each demographic characteristic.

. Despite sustained efforts to recruit a balanced sample, the demographic distribution remains uneven. For example, the proportion of younger and non-immigrant workers is higher; female participants are overrepresented compared to males (77.2% vs. 17.4%), and a similarly disproportionate pattern is observed among participants with a college or university degree (66.8%). Within each demographic category, the proportion of workers indicating they are severely or extremely anxious or depressed was higher among younger workers, gender variant-non-binary, non-immigrants, and workers with some college or university education.

Employment characteristics, based on workers’ situation in the three months preceding survey completion, are displayed in

Table 2. Descriptive employment characteristics of the survey sample and prevalence of severe or extreme anxiety/depression.

Employment Characteristics	Convenience Survey Sample N (%)	Severely or Extremely Anxious or Depressed n	% With Severe/Extreme Anxiety or Depression Within Category *	% of All Respondents with Severe/Extreme Anxiety or Depression by Category **
Precariousness Employment Scale	240	48
Low PE (score of 15 or lower)	109 (45.4%)	14	12.8%	29.2%
High PE (score higher than 15)	131 (54.6%)	34	26.0%	70.8%
Employment Arrangement	229	44
Employed directly by the employer	164 (71.6%)	24	14.6%	54.5%
Employed through a temp agency	18 (7.9%)	4	22.2%	9.1%
Self-employed with no employees	24 (10.5%)	8	33.3%	18.2%
Self-employed with employees	1 (0.4%)	1	100%	2.3%
Gig/platform work	22 (9.6%)	7	31.8%	15.9%
Agreed Contract Length or Type	217	40
On-call or day-to-day basis	34 (15.7%)	9	26.5%	22.5%
Less than 6 months	50 (23%)	9	18.0%	22.5%
6 months to 1 year	25 (11.5%)	5	20.0%	12.5%
Longer than 1 year	22 (10.1%)	3	13.6%	7.5%
Permanent or open-ended	72 (33.2%)	13	18.1%	32.5%
End date or length of job unknown	14 (6.5%)	1	7.1%	2.5%
Work Hours	230	44
Part-time (<30 h per week)	90 (39.1%)	17	18.9%	38.6%
Hours vary from week to week (could be <30)	63 (27.4%)	16	25.4%	36.4%
Full time (≥30 h per week)	77 (33.5%)	11	14.3%	25.0%
Formal/Informal Employment	229	43
Formal	181 (79%)	32	17.7%	74.4%
Informal	48 (21%)	11	22.9%	25.6%
Income After Taxes	242	48
Less than 60% of median	159 (65.7%)	39	24.5%	81.3%
60% or more of median	83 (34.3%)	9	10.8%	18.8%
Income Variation in Past 3 Months	122	21
Very much	30 (24.6%)	10	33.3%	47.6%
Somewhat or a little	67 (54.9%)	5	15.0%	23.8%
Not at all	25 (20.5%)	6	24.0%	28.6%
Likelihood that Total Working Hours Will be Reduced in the Next 3 Months	225	47
Very likely	73 (32.4%)	16	21.9%	34.0%
Probably or maybe	53 (23.6%)	15	28.3%	31.9%
Not likely or Not at all likely	62 (27.6%)	8	12.9%	17.0%
Not applicable	37 (16.4%)	8	21.6%	17.0%

Only respondents who completed the survey section assessing anxiety/depression were included in these analyses. Ns vary slightly across variables due to item-level missing employment data. * “% within category” represents the proportion of participants within each employment category who reported severe or extreme anxiety or depression (e.g., 14 of 109 participants with a low PE score). ** “% by category” represents the proportion of all participants reporting severe or extreme anxiety or depression who fall within each category (e.g., 14 of 48 participants). Percentages in this column sum to 100% within each demographic characteristic.

. While workers were fairly evenly-distributed across the PE scale (high vs. low PE scores), other indicators of employment precariousness varied. Within each employment category, the percentage of workers who experienced severe or extreme anxiety or depression was higher among (i) workers with high PE scores; (ii) self-employed workers; (iii) on-call or on a day-to-day basis workers or those whose hours varied from week to week, as well as those unsure whether their hours will vary in the next three months; (iv) informal workers; and (vi) workers who earned less than 60% of the median income and whose income varied very much in the previous three months.

The unadjusted and adjusted logistic regression models estimating the relationship between severe or extreme depression or anxiety and employment characteristics are listed in

Table 3. Logistic regression models of predictors of severe or extreme anxiety or depression by employment precariousness score (high or low) and employment characteristics.

Model 1—Unadjusted
Explanatory Variables	Answer Categories	Severe or Extreme Anxiety or Depression
		Odds Ratio & Significance	95% CI Inf	95% CI Sup
Employment Precariousness Score N = 240 [Reference category = Low PE, score of 15 or lower]	High PE (score higher than 15)	2.38 *	1.22	4.84
Employment Arrangement N = 229 [Reference category = Employed directly by the employer]	Employed through a temp agency	1.67	0.51	5.49
	Self-employed with no employees	2.92 *	1.12	7.56
	Self-employed with employees	12,356,049.87	0	Inf
	Gig/platform work	2.72 *	1.01	7.37
Agreed Contract Length or Type N = 217 [Reference category = Permanent or open-ended]	On-call or day-to-day basis	1.63	0.62	4.31
	Less than 6 months	1	0.39	2.55
	6 months to 1 year	1.13	0.36	3.58
	Longer than 1 year	0.72	0.18	2.79
	End date or length of job unknown	0.35	0.04	2.01
Work Hours N = 230 [Reference category = Full time (≥30 h per week)]	Part-time (<30 h per week)	1.4	0.61	3.2
	Hours vary from week to week (could be <30)	2.04	0.87	4.8
Formal/Informal N = 229 [Reference category = formal employment]	Informal employment	1.38	0.64	3
Income after taxes N = 242 [Reference category = 60% or more of median]	Less than 60% of median	2.67 *	1.22	5.83
Income variation in past 3 months N = 122 [Reference category = Not at all]	Very much	1.58	0.48	5.21
	Somewhat or a little	0.26 *	0.07	0.93
Likelihood that total working hours will be reduced in the next 3 months N = 225 [Reference category = Not likely or Not at all likely]	Very likely	1.89	0.75	4.79
	Probably or maybe	2.66 *	1.03	6.91
	Not applicable	1.86	0.63	5.48
Model 2—Adjusted for Age Category, Gender, Education, Immigrant Status
Explanatory Variables	Answer Categories	Severe or Extreme Anxiety or Depression
		Odds Ratio	95% CI Inf	95% CI Sup
Employment Precariousness Score N = 240 [Reference category = Low PE (score of 15 or lower]	High PE (score higher than 15)	2.28 *	1.13	4.62
Employment Arrangement N = 229 [Reference category = Employed directly by the employer]	Employed through a temp agency	1.78	0.5	6.28
	Self-employed with no employees	3.61 *	1.32	9.88
	Self-employed with employees	17,308,659.43	0	Inf
	Gig/platform work	3.08 *	1.08	8.75
Agreed Contract Length or Type N = 217 [Reference category = Permanent or open-ended]	On-call or day-to-day basis	1.53	0.53	4.38
	Less than 6 months	0.8	0.29	2.21
	6 months to 1 year	0.97	0.29	3.27
	Longer than 1 year	0.64	0.16	2.58
	End date or length of job unknown	0.33	0.04	2.83
Work Hours N = 230 [Reference category = Full time (≥30 h per week)]	Part-time (<30 h per week)	1.33	0.56	3.14
	Hours vary from week to week (could be <30)	1.88	0.78	4.53
Formal/Informal N = 229 [Reference category = formal employment]	Informal employment	1.6	0.71	3.6
Income after taxes N = 242 [Reference category = 60% or more of median]	Less than 60% of median	2.75 *	1.22	6.17
Income variation in past 3 months N = 122 [Reference category = Not at all]	Very much	1.95	0.52	7.3
	Somewhat or a little	0.22 *	0.06	0.87
Likelihood that total working hours will be reduced in the next 3 months N = 225 [Reference category = Not likely or Not at all likely]	Very likely	1.8	0.7	4.62
	Probably or maybe	2.59 *	0.98	6.84
	Not applicable	1.82	0.58	5.72

* p < 0.05. Reference categories for each indicator are listed along with the indicator. Note: The odds ratio value of the “Self-employed with employees” category is problematic in both models due to the very small number of people in this group.

. The adjusted statistically significant odds of reporting severe or extreme anxiety or depression were higher among workers with high PE (2.28), self-employed with no employees (3.61) and gig/platform workers (3.08), workers earning less than 60% of the median income (2.75), and those unsure whether their hours will vary in the next three months (2.59). The statistically significant odds were lower (0.22) for workers with somewhat or little income variation in the previous three months.

3.2. Interviews

Table 4. Characteristics of the interviewees (N = 40).

Age Category	Low PE Score				High PE Score
	Female	Male	Gender Variant/Non-Binary/Not Stated	Total	Female	Male	Gender Variant/Non-Binary/Not Stated	Total
25–35	5	1	1	7	3	3	2	8
36–45	5	3	0	8	5	1	0	6
46–55	1	2	0	3	3	4	1	8
Total	11	6	1	18	11	8	3	22

shows the breakdown of participants by PE score, age, and gender. Most participants were college or university graduates.

As summarized below, participants described similar mental health concerns and struggles faced when seeking healthcare services, with few differences based on the level of employment precarity. These themes relating to workers’ mental health concerns and limited societal and workplace support are similar across the other five countries (Sweden, Belgium, Chile, Spain, and the United States) examined as part of the broader international study investigating the health effects of precarious and non-standard employment, and are reported in detail elsewhere [18,45,61].

Although the interview questions did not specifically ask about anxiety and depression, participants spoke at length about the chronic stress, worry, anxiety, depression and negative overall wellbeing that they experienced in relation to employment instability, uncertainty over the next paycheque, insufficient income to cover household bills and unpredictable or last-minute schedules. Workers felt they had no choice but to accept these conditions as they depended on the income. Participants also reported that the constant stress, volatile work schedules and low income put a strain on their relationships with family members and prevented them from forming new personal relationships, which further worsened their mental health and wellbeing. This quote from a woman with a high PE score encapsulates many of the experiences and feelings shared by participants.

“I lost a lot of weight because I wasn’t eating. But I ended up, you know, just kind of like running out of money … and that’s– that’s how I became homeless periodically. … I was still working, it just wasn’t enough to cover everything. So I remember becoming very suicidal at that point. … you can be working, you can be going to school, you can do everything right by your family, and things can still not work out.”

Despite expressing numerous mental health needs, participants spoke of the various struggles they faced when seeking health services. These included: (i) not having paid sick leave for health appointments; (ii) fearing asking for time off work for health appointments; (iii) high out-of-pocket medication costs; and (iv) mental health services (e.g., counselling and therapy) not being included in universal provincial health plans or employer benefits packages. This is because PE workers typically do not benefit from the same employer-based social and health benefits as their counterparts in full-time employment or with open-ended or permanent contracts. The following quote, from a male with a high PE score, illustrates the hopelessness felt by some interviewees:

“Not having benefits matters to workers and matters a great deal. This time around, I was not formally diagnosed with a bout of depression, but I was depressed. And part of the reason why I didn’t even go and get diagnosed is that I didn’t have a benefits package to pay for anything if that had been the diagnosis.”

3.3. Interpretation and Recommendations

Our study underscores the disproportionate mental health burden borne by precariously employed (PE) workers in Ontario, Canada. A significant proportion of respondents reported symptoms of severe or extreme anxiety and depression, which they attributed to job insecurity, unstable or insufficient income, and limited labor protections. Given that our study had low statistical power, and the demographic distribution of participants was uneven, the findings must be interpreted with caution and validated with larger samples of Canadian workers. These study’s findings corroborate recent longitudinal, cross-sectional, and population-based studies [20,77,78,79,80,81,82,83] along with qualitative studies [84,85,86,87] highlighting the psychosocial toll of PE. Despite evident clinical need, our respondents described numerous structural and financial barriers to accessing mental health care, including the absence of employer-provided benefits, lack of paid sick leave, and the incomplete public funding of psychological services in Ontario. These systemic gaps align with a growing body of literature identifying how PE not only exacerbates mental illness risk but also constrains access to care pathways [88,89].

Our findings reflect broader patterns in Canadian research demonstrating that workers in precarious arrangements often fall through the cracks of social and health safety nets, compounding their risk for poor health outcomes [47,90,91,92,93,94]. The dual burden of increased vulnerability and reduced care access demands urgent structural interventions [47,95,96,97]. Untreated mental illness among these workers could contribute to long-term societal costs—including productivity losses, caregiver burdens, and chronic disease burden on healthcare systems [98,99]. Additionally, PE could carry cross-generational impacts on youth mental health [100]. We advocate for several evidence-informed strategies to respond to the dual burden of heightened vulnerability to mental illness and constrained access to care experienced by workers in precarious employment:

• Enhancing public investment in universal, employment-independent mental health services, which has shown potential to reduce inequities in OECD nations [101].
• Reforming labor legislation to extend health and social protections to non-standard and gig economy workers [96], including mandating employer contributions to mental health coverage [102].
• Expanding access to workplace-based mental health interventions [103], and implementing evidence-based clinical protocols for the treatment of work-related psychological distress [104,105].
• Leveraging telehealth and integrated care models to remove logistical access barriers and broaden the reach of mental health professionals, particularly in under-resourced regions [106,107].

Although increasing public investment, expanding access to mental health services, and implementing legislative reforms are often difficult to achieve, particularly in the context of constrained budgets and competing policy demands, these approaches remain both necessary and feasible. Substantial economic and societal costs linked to untreated mental health [7,8,99] underscore the need for sustained structural interventions and investments. Importantly, there are numerous examples—within Canada and internationally—where evidence-informed public investment and regulatory action have successfully expanded access to mental health care, particularly for populations with limited workplace benefits or unstable employment. These precedents demonstrate that coordinated investment and policy change, while challenging, are achievable and can yield measurable improvements in mental health equity.

Canada has demonstrated that targeted public investment can expand access to mental health services for underserved individuals, including those without employer-provided benefits who cannot afford private therapy or who lack private insurance. The federally funded Wellness Together Canada platform (2020–2024) has provided free nationwide mental health support and has been widely used by individuals facing financial and insurance-related barriers to accessing care [108]. Similarly, Ontario’s Structured Psychotherapy Program—launched in 2017 and expanded province-wide as part of the Roadmap to Wellness initiative in 2020—offers no cost, cognitive behavioral therapy through primary care-linked and community-based settings, improving access to evidence-based psychotherapy for individuals without private insurance [109]. These initiatives, alongside federal–provincial bilateral mental health funding agreements [110], illustrate that public investment in mental health care is both feasible and already contributing to improved access in Canada. Comparable initiatives in other jurisdictions further demonstrate the feasibility of the strategies we propose. In Australia, the Better Access Initiative expanded publicly subsidized psychological services during and after the COVID-19 pandemic, resulting in substantial increases in service utilization and improved access for individuals facing financial barriers [111]. In the United Kingdom, the NHS Talking Therapies program (formerly IAPT) provides large scale, free psychological care and has consistently expanded access for low income, unemployed, and disabled individuals, contributing to reductions in inequities in service use [112].

Our findings should be interpreted in light of the broader, post-pandemic reorganization of work arrangement changes referred to in the introduction section. The normalization of algorithmic management in platform-based sectors, alongside persistent scheduling instability, limited paid leave, and weak employment protections in care and other front-line occupations, has reshaped the conditions under which PE workers experience and manage mental health challenges. These structural features—intensified monitoring, income volatility, and constrained autonomy—map directly onto the drivers of anxiety and depression reported by participants in our study, who consistently linked psychological distress to job insecurity, unpredictable earnings, and limited labor protections. In this context, COVID-19 did not simply produce a temporary shock but accelerated and entrenched organizational practices that continue to expose PE workers to heightened psychosocial risks. At the same time, the pandemic revealed and worsened lasting stratifications in access to flexibility and protection within labor markets. As remote work became normalized for higher wage professional workers, although the trend seems to be reversing now, PE workers remained concentrated in in-person roles with fewer protections and greater exposure to health and economic risks. Our findings extend this literature by showing that these occupational divides are mirrored in access to mental health care. Taken together, our results suggest that the mental health burden observed among PE workers cannot be understood solely as an individual or pandemic specific outcome. Rather, it reflects the interaction of post-pandemic labor market arrangements, employer dependent benefit systems, and longstanding gaps in publicly funded mental health care.

Further longitudinal research is required to examine the multidimensional effects of PE on mental health over time [23], with robust designs that enable stratification by race, gender, migration status, and education [113,114]. Intervention studies with tested evaluation indicators are also critical to identify and scale comprehensive multiprong initiatives [115] that measurably improve outcomes among PE workers [83,116,117].

Finally, we acknowledge several related limitations of our study: the small survey sample size and corresponding underpowered analysis, and a lack of disaggregated demographic analysis, due to small sample sizes per category and absence of race/ethnicity data. This impedes intersectional insights, which are crucial given the established differential impacts of PE across the social determinants of health [79,81,118,119,120,121,122,123,124,125,126]. Future research must prioritize inclusive sampling strategies to better capture the heterogeneity of experience within this population. Another study limitation is that, to maximize analytic sample sizes, participants with partial survey responses were retained for analyses in which relevant variables were complete; however, missing data may not have been random, potentially introducing bias and limiting comparability across analyses. While this approach is appropriate for studying precariously employed populations—where missingness itself may reflect structural vulnerability—it nonetheless warrants caution in interpreting effect estimates. Further, this study has several limitations related to survey sample selection. Participants were recruited through convenience and community-based channels, which may limit the representativeness of the sample and introduce self-selection bias. Individuals experiencing greater employment precarity or mental health concerns may have been more likely to participate, potentially inflating estimates of anxiety and depression. In addition, the cross-sectional survey design captures experiences at a single point in time and cannot establish causality. As a result, findings should be interpreted as indicative of patterns among precariously employed workers rather than as population-level prevalence estimates.

4. Conclusions

This mixed-methods study sheds light on the numerous access barriers encountered by precariously employed workers in Ontario, Canada, in seeking mental health support. Our findings reveal a clear association between employment precarity and heightened symptoms of anxiety and depression, driven by employment instability, low and unpredictable income, and limited workplace rights. Although Canada’s universal healthcare system guarantees access to essential services, the omission of comprehensive mental health coverage and employer-linked benefit gaps leaves precariously employed workers at a disadvantage. Many individuals delay or forego necessary care due to financial constraints or absence of paid leave, exacerbating mental health deterioration and widening health inequities.

These findings echo recent Canadian and international evidence showing that precarious employment both induces and perpetuates mental health inequities. To mitigate the personal and societal consequences of untreated mental illness—including lost productivity and rising chronic disease—targeted policy and health system reforms are imperative. These include increasing public investments in universally accessible mental health services, implementing labor protections that extend benefits to non-standard workers, and promoting integrated primary mental healthcare models. A multi-sectoral, evidence-informed response is urgently needed to reverse growing disparities, particularly as precarious employment continues to proliferate across global labor markets. Importantly, future research must integrate intersectional analyses and prioritize the voices of racialized and migrant workers to ensure equitable policy responses.

This study contributes to the literature by clarifying how employment precarity simultaneously increases mental-health risk and constrains access to care through structural features that remain insufficiently addressed within current health and labor systems. By integrating quantitative estimates of anxiety and depression with qualitative accounts of care-seeking, the findings highlight a compounded disadvantage in which workers with heightened mental-health needs face substantial barriers to support. Situating Ontario’s experience within a broader international context further demonstrates that these exclusions are not unique, but recur wherever access to mental-health care is closely tied to stable, full-time employment. Together, these insights advance understanding of how labor-market structures shape mental-health inequities and underscore the need for evidence-informed reforms that decouple access to care from employment status and address the psychosocial risks embedded in precarious work as such employment continues to expand globally.