A muti-informant national survey on the impact of COVID-19 on mental health symptoms of parent–child dyads in Canada

The COVID-19 pandemic negatively impacted the mental health of children, youth, and their families which must be addressed and prevented in future public health crises. Our objective was to measure how self-reported mental health symptoms of children/youth and their parents evolved during COVID-19 and to identify associated factors for children/youth and their parents including sources accessed for information on mental health. We conducted a nationally representative, multi-informant cross-sectional survey administered online to collect data from April to May 2022 across 10 Canadian provinces among dyads of children (11–14 years) or youth (15–18 years) and a parent (> 18 years). Self-report questions on mental health were based on The Partnership for Maternal, Newborn & Child Health and the World Health Organization of the United Nations H6+ Technical Working Group on Adolescent Health and Well-Being consensus framework and the Coronavirus Health and Impact Survey. McNemar’s test and the test of homogeneity of stratum effects were used to assess differences between children-parent and youth-parent dyads, and interaction by stratification factors, respectively. Among 933 dyads (N = 1866), 349 (37.4%) parents were aged 35–44 years and 485 (52.0%) parents were women; 227 (47.0%) children and 204 (45.3%) youth were girls; 174 (18.6%) dyads had resided in Canada < 10 years. Anxiety and irritability were reported most frequently among child (44, 9.1%; 37, 7.7%) and parent (82, 17.0%; 67, 13.9%) dyads, as well as among youth (44, 9.8%; 35, 7.8%) and parent (68, 15.1%; 49, 10.9%) dyads; children and youth were significantly less likely to report worsened anxiety (p < 0.001, p = 0.006, respectively) or inattention (p < 0.001, p = 0.028, respectively) compared to parents. Dyads who reported financial or housing instability or identified as living with a disability more frequently reported worsened mental health. Children (96, 57.1%), youth (113, 62.5%), and their parents (253, 62.5%; 239, 62.6%, respectively) most frequently accessed the internet for mental health information. This cross-national survey contextualizes pandemic-related changes to self-reported mental health symptoms of children, youth, and families.


Methods
Study design and population. We collected data from an anonymous, voluntary, 10-min cross-sectional survey via Leger, a Canadian-based market research and polling firm (https:// leger 360. com), between April 20, 2022, and May 25, 2022, which was a time period when most Canadian provinces had eased social restrictions 34 . We used Leger's dynamic Leger Opinion (LEO) panel, an online pool of over 400,000 individuals recruited and validated through multiple methods who consented to be contacted for research purposes; at any given time this pool reflects a representative sample of Canadian residents with internet access. Respondents received LEO reward points after completing the questionnaire, which could be redeemed for gift cards and merchandise. Assuming children and youth aged 11-18 represent ~ 11% of the Canadian population (~ 4 million) 35 we recruited 1600 respondents (800 dyads (i.e., a group of two members)) to conduct subgroup analyses with a ± 3.5% margin of error at a 95% confidence level, and required that at least 15% of the sample included dyads who had lived in Canada for less than 10 years (5% less than 5 years) in order to understand their unique experiences. The (total) 85-item (English and French) electronic survey was administered to LEO panelists who identified as parents or legal guardians (> 18 years of age; hereafter referred to as parents) with at least one child (11-14 years of age) or youth (15-18 years of age) living in the same household; the oldest child or youth was selected if more than one was eligible. Age ranges for children and youth were selected to align with Statistics Canada standards and to adhere to institutional ethical requirements (e.g., age-tailored questions) 36 . Parents first completed their portion (45-items) independently followed by their child or youth (40-items), who completed their own portion independently. We followed the Checklist for Reporting Results of Internet E-Surveys (CHERRIES) guidelines (Supplemental Table 1) 37 . Survey development. We created a preliminary list of mental health questions based on findings presented in published articles identified in our scoping review 38 and systematic review 39 on strategies, approaches, and interventions to improve youth wellbeing during the COVID-19 pandemic and mapped them onto The Partnership for Maternal, Newborn & Child Health and the World Health Organization of the United Nations H6 + Technical Working Group on Adolescent Health and Well-Being consensus framework for defining, programming, and measuring adolescent wellbeing that is part of a broader program of work that includes a multistakeholder Call to Action to prioritize adolescent well-being. This framework includes five domains: (1) Good health and optimum nutrition; (2) Connectedness, positive values, and contribution to society; (3) Safety and a supportive environment; (4) Learning, competence, education, skills, and employability; and (5) Agency and resilience (Supplemental Table 2) 40 . Operational definitions for self-reported mental health symptoms are provided in Supplemental Table 3. Demographic, knowledge acquisition, and health literacy questions were based on the Coronavirus Health and Impact Survey (CRISIS) 41 . We developed a combination of continuous, categorial, Likert-type, and open-ended response options; Likert-type questions included a scale ranging from 1 (i.e., "a little") to 5 (i.e., "a lot"). Questions were iteratively refined by the core survey development team (JPL, SJM, RBM, DH, SH, PT) 42 and six citizen partners (three youth: MS, MH, SS, and three parents: KR, MS, AN). The order of the response options was randomized and attention checks (innocuous questions with a single correct answer) were randomly inserted throughout the questionnaire. One question was presented per screen and respondents were not given the opportunity to change their answers once they moved to the next screen; all questions included a "don't know" or "prefer not to answer" option that were excluded from analyses. www.nature.com/scientificreports/ questions, the Wilcoxon signed-rank test was used as it accounts for the paired nature of the data and the correlation present between parent and child 43 . Responses to nominal questions were compared between parent and child using McNemar's test for paired data 44 . In subgroup analyses, responses were stratified by prespecified (e.g., gender, geographic location, ethnicity) demographic characteristics. For the nominal questions, potential interaction by the stratification factors was assessed using a test of homogeneity of stratum effects, an extension of McNemar's test 45 . For clarity of presentation, "a little" or "a lot" better or worse are presented in aggregate as "better" or "worse." Two-sided p-values < 0.05 were considered statistically significant. All analyses were conducted using R version 4.2.1 46 .

Statistical analysis.
Patient and public involvement. We abided by the Canadian Institutes of Health Research (CIHR)guiding core principles of inclusiveness, mutual respect, support, and co-building 47 and adhered to the GRIPP-2 reporting guidelines for patient and public involvement 48 . Youth and parent involvement in the current project began in 2021; they participated in group discussion alongside other stakeholders (e.g., researchers, clinicians, decision makers). The research questions, protocol, and this paper were jointly developed with youth (MS, SS, MH) and parent (AN, MS, KR) partners on this team. All youth and family partners are compensated for their time.
Ethical considerations. All participants provided electronic informed consent on their own behalf; as the parent had significant knowledge of their child/youth, prior to submitting their own consent, the parent attested that they understood the information regarding their child/youth's participation and that their child/youth had the capacity to consent on their own behalf.  Table 4). Compared to their parents, children less frequently reported worsened mood (p < 0.001), anxiety (p < 0.001), irritability (p < 0.001), and inattention (p < 0.001) during the COVID-19 pandemic, while youth less frequently reported worsened mood (p < 0.001), anxiety (p = 0.006), and inattention (p = 0.028) during the COVID-19 pandemic.

Discussion
In this nationally representative, multi-informant survey we identified that over one-third of children, youth, and their parents self-reported that their mental health symptoms had worsened during the COVID-19 pandemic. We also found that children and youth were significantly less likely to self-report worsened mental health symptoms compared to their parents. Children and youth, compared to their parents, were significantly less confident in their ability to verify and understand information on mental health that was most frequently accessed via the internet. Children in our sample reported worsened symptoms of anxiety, inattention, and irritability while youth in our sample reported worsened symptoms of anxiety, inattention, and mood. These findings highlight developmental differences with regards to symptom presentation associated with the COVID-19 pandemic. Interestingly, children and youth presented with similar exacerbation in symptoms, including worsened symptoms of anxiety and inattention. In younger children, mood challenges often present as increased irritability, whereas low and depressed mood are symptoms that tend to emerge during adolescence and early to mid-adulthood 49,50 . Further, psychological adaptation matters in the context of our study, because we surveyed participants two years after the very early phase of the pandemic (May 2022 compared to May 2020), when short-lived changes in mental health symptoms had potentially diminished. Although mental health symptoms typically return to baseline levels following common life experiences 51,52 , the COVID-19 pandemic may have had different effects. Specifically, COVID-19 was unique in scale and social and economic consequences, and it is therefore likely that a   www.nature.com/scientificreports/ portion of the population will experience ongoing mental health difficulties. It will be important to identify and support children and youth who are most vulnerable [53][54][55] . Parents reported significantly worse mental health symptoms compared to their children and youth. Loss of routine occurred rapidly for many families during the COVID-19 pandemic related to changes in employment arrangements, school closures, and loss of access to activities outside of the home such as sports and extracurriculars 56 . Several studies have identified that children and youth in families that maintained a structured routine demonstrated lower rates of externalizing problems, over and above the effect of food insecurity, socioeconomic status, dual-parent status, maternal depression, and stress 57,58 ; these findings are consistent with prior work suggesting that lack of predictability is linked strongly to youth psychopathology [59][60][61][62] . However, maintaining routine and structure for parents was challenging as school closures were unpredictable and impacted many aspects of daily life and may have come at a cost to their own mental health 63 . The consequences of these difficulties were likely longstanding, related perhaps to the ways in which contextual risk permeates the structures and processes of family systems 64 . Our findings suggest that adequate mental health resources for parents are required as one component in the cascading process involving parental mental health and family processes for families to fully reap potential benefits associated with maintaining routine for children and youth during a pandemic.
Our results suggest that the mental health consequences associated with the COVID-19 pandemic are significant to children, youth, and their parents, including as a cost within the healthcare system 65 . During a health crisis that requires social lockdown, many parents spend increased time at home and thus have the ability to intentionally monitor not only their own mental health but also their child or youth's mental health with enhanced sensitivity to observe symptoms of worsened mental health 58 . Within this context, it is critical that parents have the knowledge and support required to inquire about or identify mental health difficulties within their families, especially among children and youth. A routine check-in for many children and youth is the pediatrician or family physician's office; additional at-home resources should be made available to families. The delivery of mental health resources urgently requires expansion to increase scalability while prioritizing equitable access (e.g., multiple languages and appointment time periods, affordable care) across diverse populations 66 .
Sociologists have applied an illness behavior perspective to the study of family burden, identifying that impacts on mental health for one family member have far-reaching effects on other family members as individuals and as members of a social system 67,68 . Concerns related to COVID-19 are not universally detrimental as, for example, individuals may experience a reduction in work-life or school-life conflict 69 or in stress from potential high-risk COVID-19 exposures at schools or in the workplace 70  www.nature.com/scientificreports/ employed parents to spend more time with their children or youth in place of work-life conflicts 72 , receipt of emergency financial benefits that alleviated financial constraint due to job loss 73 , and that emergency financial benefits may have exceeded wages from low-paying employments 74 may further protect against detrimental impacts. Additional dyadic quantitative and qualitative studies are required to better understand and uncover possible mechanisms by which social-economic characteristics are associated with mental health during periods of health crisis. Researchers, policymakers and public health practitioners have a unique opportunity to address disparities in mental health knowledge acquisition among children and youth 75,76 . Most parents in our study were aware of the mental health threat posed by the COVID-19 pandemic 77,78 ; however, we found an overall lack of confidence in mental health knowledge acquisition among children and youth including limited vetting of the knowledge they do have. For optimal uptake and sustained memory of knowledge, mental health literacy campaigns should be made in partnership with youth 79-81 and tailored to their preferences 82 . Our findings highlight the importance of healthcare professionals as conduits for accurate and valid information [83][84][85][86] . School or community healthcare education may be an accessible method for healthcare professionals to promote mental health to students 87 as children and youth are less likely to exist within these pre-established relationships 80 . Child and youth friendly practices can be improved by including empathetic and friendly staff 87 , offering flexible hours 79 , using appropriate communication skills 82 such as providing complex information in engaging, easy-to-understand language 88 , and conveying health information that supports their learning, so children and youth can be empowered to decide their own health views 89,90 . Mental health support delivered virtually may be a low-risk approach for increased access to care, especially for marginalized youth (e.g., racialized youth) and youth living in areas with limited access to in-person services 91,92 . This means 450 parents of youth were compared to their youth. Respondents who did not provide an answer were excluded from analyses. *, p < 0.05; **, p < 0.01; ***, p < 0.001; ns, not statistically significant. www.nature.com/scientificreports/ Improving mental health outcomes for children and youth in the recovery from the COVID-19 pandemic should be a research priority as we prepare for future pandemics. The rapid proliferation of child and youth mental health research during the COVID-19 pandemic impacted the quality of the execution of these studies that have been criticized for opting for timeliness at the expense of methodological quality 93,94 . Nonprobability or convenience samples were used in studies estimating prevalence of mental health disorders during the pandemic, which increases the likelihood of reporting bias 93 . We employed a nationally representative, multi-informant survey to explore changes in mental health symptoms from before to after the pandemic. Population-based cohort studies with longitudinal follow-up to monitor changes over time should be prioritized in future research.
Limitations. We used a cross-sectional survey design that allowed data collection from a large and representative sample of the Canadian population. However, our results have limited longitudinal applicability; we queried respondents to report retrospectively on perceived changes in mental health symptoms during the COVID-19 pandemic and cannot generalize our findings to changes in mental health symptoms that occurred in the post-pandemic period. We also relied on a volunteer panel (Leger's LEO panel) to recruit participants for compensation that may have introduced recruitment bias. Our survey was deployed online in English and French languages-Canada's two official languages-that excluded individuals without internet access or those who read and write exclusively in other languages (~ 9% and ~ 2% of the Canadian population, respectively) 95 .

Conclusions
Data from this cross-national survey highlights the importance of social factors in understanding pandemicrelated changes to mental health symptoms of children, youth, and their families. Understanding patterns of mental health and factors associated with changes are essential to ensure that services match the needs of the population served. Ongoing surveillance of mental health among children, youth, and their families, as well as system-level planning are important to facilitate effective mental health campaigns and efficient use and development of mental health resources in periods of health crisis.

Data availability
The data are not publicly available due to them containing semi-identifiable information that could compromise research participant privacy. Additional summary tables of count data are available from the corresponding author upon reasonable request. Ms. S. collected the data and edited subsequent drafts. Dr. T. conceptualized the study, wrote the original draft, and edited subsequent drafts. Dr. S. conceptualized the study, provided supervision, and edited subsequent drafts. Dr. F. conceptualized the study, provided resources and supervision, wrote the original draft, and edited subsequent drafts. J.P.L. and S.J.M. made substantial contributions to the conception and design of the work and drafted the work, approve the submitted version, and agree both to be personally accountable for each author's own contributions and to ensure that questions related to the accuracy or integrity of any part of the work, even ones in which the author was not personally involved, are appropriate investigated, resolved, and the resolution documented in the literature. All authors consent for publication.