1.0 Introduction

The coronavirus 2019 (i.e., COVID-19) pandemic has adversely affected population mental health worldwide [1], including the mental health of Canadian youth. Youth reported the largest reductions in mental health following the COVID-19 pandemic and restrictive measures, compared to any other age group in Canada, with 41.5% indicating they had excellent or very good perceived mental health in 2020, compared to 62.1% pre-pandemic in 2018 [2]. Further, 57% of adolescents report that their mental health has worsened since the onset of school closures and “lockdowns” [2]. School closures and lockdowns were likely associated with feelings of loneliness or social isolation, which can contribute to adolescents’ risk of poor mental health [3]. While, initially, early public health measures in Ontario reduced mental health out-patient visits in adolescents drastically in March/April of 2020 due to lockdowns and a shift to online services, usage of physician-based mental health services rose to above pre-COVID rates as of July 2020 and remained at this level until at least February 2021 [4]. Emergency department visits among adolescents for mental health-related services also increased in Canada following the pandemic by as much as 29.7% [5].

Data on the effects of the COVID-19 pandemic on income inequality are currently lacking, given the recency of the pandemic and the infrequency of income inequality measures. It is plausible that the pandemic widened the gap between the highest and lowest earners (i.e., income inequality) in Canada, with increases in unemployment rates disproportionately affecting the lowest earners, those less educated, immigrants, people of colour, and those with precarious employment [6]; that said, income inequality may have also been subsequently decreasing owing to the Canadian Emergency Response Benefit (CERB) and the Canadian Emergency Student Benefit (CESB), two programs employed during COVID-19 to mitigate some impacts of the COVID-19 disruption. While data are currently lacking, given the recency of the pandemic and the infrequency of income inequality measures, it appears as if government transfers (such as CERB and CESB) may have curtailed the rise of income inequality in the pandemic. Income inequality refers to the relative differences in incomes within a given group or area [7]. A wide gap in incomes is suggestive of a greater inequality in income, while a narrow gap suggests that there is not a wide distribution of incomes within a given group or area. In public health research, income inequality is most often measured using the Gini coefficient [8]. In Canada, the Gini coefficient has been increasing since the late 1980s and early 1990s [9, 10], especially in major urban centres [11]. For example, between 1990 and 2018, there was a 5.9% increase in the national Gini coefficient average in Canada [10].

2.0 Theory

The multitude of stressors associated with the COVID-19 pandemic may have led to diminished mental health among adolescents, and the effects are likely more pronounced in those from socioeconomically disadvantaged areas. Increases in income inequality associated with the pandemic, including related increases in income inequality associated with job loss and loss of wages in the pandemic [6], may represent a key factor contributing to greater vulnerability to adverse mental health in adolescents during this period. This supposition is in line with current theories on the association between income inequality and mental health. For example, income inequality is associated with increased divestment in human capital (e.g., cuts to social spending), which has been common during the COVID-19 pandemic, with divestments in human capital defined as cuts or losses in, for example, training programs or mental healthcare (e.g., counselling services). In turn, such divestments are associated with worsened mental health, given human capital sectors are oftentimes those that can improve health [12]. That said, income inequality was likely to subsequently reduce following the onset of the CERB, which provided monthly payments to individuals who were affected by job or wage loss associated with COVID-19. CERB is one example of investment in human capital, which could reduce the effects of income inequality. Moreover, income inequality is associated with lowered social cohesion, which was exacerbated during the COVID-19 pandemic [3]. Among adolescents, social relationships and connections are especially important for mental well-being, considering that higher engagement and solidarity with peers can mitigate the harmful impacts of stress. With the onset of lockdowns and work-from-home orders, adolescents grew increasingly isolated and experienced increased stress during the pandemic [3]. This isolation, in turn, may have been associated with poorer mental health.

While existing research has linked income inequality in schools to adolescent depression [13], it is unclear if the pandemic exacerbated the association between income inequality and adolescent mental health. This work is important considering the significant toll the COVID-19 pandemic had on youth mental health. Furthermore, and to inform equitable distribution of supports and resources for pandemic recovery is essential and the prevention of sustained impacts. Also, the COVID-19 pandemic provides an important learning opportunity on how the relationship between income inequality and adolescent mental health may have been impacted in such circumstances, which may help to better prepare for and inform policies and programs in the case of future events. The objective of this study is to examine if the COVID-19 pandemic exacerbated the association between income inequality and adolescent depression and anxiety.

3.0 Material & methods

4.0 Results

The study sample included 29,722 students across 43 Census divisions in British Columbia, Alberta, Ontario, and Quebec. The average age of the sample at baseline was 14.9 years [standard deviation (SD) = 1.5] and ranged between 12 and 19 years of age. The majority of the sample self-reported as white (76.3%), female (54.4%), and had weekly spending money of over $100 (20.9%). The mean Gini coefficient at the CD-level was 0.37 (SD = 0.03) and the median after-tax household income was $58,891.70 (SD = $8,843.50). The average proportion of visible minority households, single-parent households, and recent immigrant households were 9.9%, 15.4%, and 1.6%, respectively, across CDs. Details on the distribution of the sample are available in Table 1. Null models indicated an ICC for depressive scores of 0.02 (95% CI = 0.01, 0.04) within CDs and 0.27 (95% CI = 0.26, 0.28) for students over time within CDs. The ICCs for the anxiety scores were 0.03 (95% CI = 0.02, 0.05) within CDs and 0.32 (0.32, 0.34) for students over time within CDs.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Table describing the distribution of the COMPASS sample in 2018/19.

https://doi.org/10.1371/journal.pone.0293195.t001

In an unadjusted analysis, the results demonstrated that a one SD-unit increase in Gini coefficient was associated with a 0.08 score increase in depressive scores (95% CI = 0.04, 0.11). Students who completed the COMPASS survey after the onset of COVID reported 0.20-unit higher depressive scores (95% CI = 0.16, 0.24) compared to their responses pre-COVID. However, the association between income inequality and depressive scores did not change in magnitude by the onset of COVID-19. Findings were similar in the adjusted analyses, with students who completed their COMPASS questionnaire following the onset of COVID-19 reporting higher depressive scores (B = 0.14, 95% CI = 0.10, 0.18) compared to pre-COVID. However, the association between income inequality and depressive scores was eliminated, and no interaction observed between z-transformed Gini coefficient and COVID-19 onset when adjusting for relevant covariates. More details are available in Table 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Table describing the unadjusted and adjusted association between income inequality and depressive scores in COVID-19, COMPASS 2018–2021.

https://doi.org/10.1371/journal.pone.0293195.t002

For anxiety scores, the unadjusted model demonstrated an association between income inequality and anxiety scores (B = 0.09, 95% CI = 0.05, 0.13), COVID onset and anxiety scores (B = 0.19, 95% CI = 0.17, 0.21), and the interaction between z-transformed income inequality and COVID-19 onset and anxiety scores (B = 0.02, 95% CI = 0.001, 0.04). The adjusted analyses indicated that the association between income inequality on anxiety scores was changed following the onset of COVID-19 (B = 0.02, 95% CI = 0.0004, 0.03). More details are available in Table 3. A graphical representation of the interaction between z-transformed Gini coefficient and COVID-19 onset on anxiety scores is available in Fig 1. Fig 1 depicts the estimated probabilities of anxiety as associated with CD-level income inequality over time. This figure demonstrates how the gaps in anxiety scores between higher and lower income inequality groups widened after the onset of the pandemic. Further, the largest increases in anxiety scores were amongst those with the highest levels of income inequality at baseline and estimated probabilities of anxiety increased after COVID-19 regardless of income inequality level.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Line graph demonstrating the graphical representation of the interaction between z-transformed Gini coefficient (at values of -3, -1, 1, and 3) and COVID-19 onset (pre, versus peri-) on the predicted probability of anxiety scores, COMPASS 2018/19-2020/21.

https://doi.org/10.1371/journal.pone.0293195.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Table describing the unadjusted and adjusted association between income inequality and anxiety scores in COVID-19, COMPASS 2018–2021.

https://doi.org/10.1371/journal.pone.0293195.t003

5.0 Discussion

The results of this study indicated that the association between income inequality and anxiety was exacerbated by the onset of COVID-19, meaning that adolescents attending schools in higher income inequality areas had larger increases in anxiety symptoms from before to during the pandemic than those attending schools in lower income inequality areas. This study also contributes to the existing literature that reports an association between higher income inequality and increased depression and anxiety in adolescents [13, 24]. The results also indicated that income inequality was not associated with depressive score following the onset of COVID-19. Unsurprisingly, the results also indicated that following the onset of COVID-19, both anxiety and depressive scores increased in the sample, as our sample aged throughout the study period.

While income inequality and rates of adolescent adverse mental health are likely on the rise since COVID-19, the association between income inequality and mental health peri-pandemic had not yet been studied. This work is the first to quantify the joint effects of the COVID-19 pandemic and income inequality on adolescent mental health. While earlier work has examined the association between income inequality and adolescent mental health, most studies were cross-sectional in nature; as such, it is unclear if income inequality preceded mental health problems in these studies and the quality of existing evidence is quite low. This work shows that income inequality at baseline is associated with increases in depressive and anxiety symptoms over time.

While no association between income inequality and depression was observed in this study, the direction of effect is still what we might expect resulting from the COVID-19 pandemic, with higher income inequality being associated with greater increases in depressive scores. It is possible that these findings are attributable to the complex mechanisms that underlie mental health. For example, researchers have demonstrated that feelings of anxiety oftentimes present before those of depressive symptoms and these conditions are highly comorbid [25–27] and anxiety is more likely to emerge in pre- and mid-adolescence, whereas depression is more commonly onset during adolescence or in early adulthood [26]. Moreover, it is possible that the lack of statistical significance is owing a lag period, by which income inequality may take several years before the impacts incur. This, in tandem with the previous point, may explain the null association between income inequality and depression during the COVID-19 pandemic. Future longitudinal research is needed to examine potential delayed and/or sustained impacts of income inequality into the ongoing and recovery phases of the pandemic.

Mental health in adolescents has likely worsened since the onset of the COVID-19 pandemic, perhaps owing to worry regarding the virus, but also to isolation and social segregation resulting from public health measures (e.g., school closures, food insecurity, worries about the future). That said, results from a multi-national study indicated that countries that enacted stringent public health orders sooner actually had lower rates of depression in comparison to those that did not implement such measures with haste [28]. While some studies have demonstrated that mental health in children did not, in fact, worsen as a result of the pandemic [29], albeit this study was conducted amongst elementary-school-aged children (ages 9 to 12) who are at lower risk for mental health conditions given that most mental health conditions are onset in adolescence and young adulthood.

The current study used a large, school-based dataset to investigate the research question and validated measures of depression and anxiety (i.e., the CES-D and GAD-7, respectively) and included secondary school students across four Canadian provinces. One survey wave (2019/20) had a reduced response rate owing to the onset of the pandemic and shift to online learning and COMPASS distribution, so study weights were developed and applied to improve the generalizability to the typical COMPASS sample. Moreover, this study is the first to investigate the association between income inequality and adolescent mental health during COVID-19 and lends to a growing body of evidence that suggests that income inequality is harmful for adolescent mental health.

Existing research on mental health during COVID-19 is limited in that many report on mental health by asking individuals to compare how they are currently feeling peri-pandemic compared to pre-pandemic [30–33]. Measures of pre-pandemic mental health may be biased by individuals current emotional state when responding peri-pandemic. The current study addresses this limitation by using measures of mental health taken before the pandemic, and once again during the pandemic.

6.0 Conclusions

The current study demonstrated that income inequality is more highly associated with increased anxiety symptoms during COVID-19 than prior to COVID-19. These results are directly applicable to both school-level and upstream policies for mental health and demonstrate the importance income inequality. This study also contributes to the literature documenting the potential harms associated with COVID-19 and income inequality.