The COVID-19 pandemic and transition phase has created unique stressors, particularly for families with children. Very quickly, families lost most of the externally generated daily structure afforded by school attendance, in-person work and recreation,Reference Hood, Zabatiero, Silva, Zubrick and Straker1 and had to adjust family life to frequently modified public health requirements. Before the COVID-19 pandemic, around a quarter of families had received childcare support from grandparents, extended family members or befriended parents.2 This support was no longer available for many families because of pandemic restrictions. To meet these increased demands, parents needed to engage family management practices to minimise household disruption and reduce dysfunctional family dynamics,Reference Barton, Brody, Yu, Kogan, Chen and Ehrlich3–Reference Nunes C, Roten, El Ghaziri, Favez and Darwiche6 especially during transitions between lockdowns and re-openings. Previous literature has described two family management practices associated with improved parent and child well-being:Reference Petren and Puhlman4 structured parenting and shared parenting.
Structured parenting
Structured parenting is a family management approach to provide children and adolescents with predictable and reliable caregiving routines, reasonable limits and scaffolding that correspond to a child's developmental needs.Reference Liu, Kryski, Smith, Joanisse and Hayden7 For example, younger children (ages 2–5 years) may require higher levels of parental monitoring and support in daily instrumental activities, such as personal hygiene, choosing and preparing healthy foods, dressing appropriately for the weather and limiting screen time, in addition to social and affective care. Most older children and adolescents may no longer need high levels of support in daily instrumental activities, so parental caregiving may shift to social, affective and academic support.Reference Liu, Kryski, Smith, Joanisse and Hayden7 In contrast to authoritarian or overprotective parenting,Reference Liu, Kryski, Smith, Joanisse and Hayden7 structured parenting is a flexible and consistent approach to help children and adolescents clarify their roles, increase personal responsibility, accomplish required tasks and develop behaviour regulation.Reference Petren and Puhlman4 Relevant to the demands of life during the COVID-19 pandemic, literature published before the pandemic indicates that families with absent or lower levels of structured parenting were more likely to be overwhelmed,Reference Davis, Myers, Logsdon and Bauer8 inefficient in meeting the demands of family life,Reference Petren and Puhlman4 less adaptable to new challengesReference Davis, Myers, Logsdon and Bauer8 and report higher levels of stress.Reference Miller, Song, Sturza, Lumeng, Rosenblum and Kaciroti9 Higher levels of structured parenting have also been associated with lower levels of parent and child mental health symptoms in the published literature.Reference Barton, Brody, Yu, Kogan, Chen and Ehrlich3,Reference Zajicek-Farber, Mayer and Daughtery5,Reference Liu, Kryski, Smith, Joanisse and Hayden7–Reference Churchill and Stoneman10
Shared parenting
Shared parenting refers to the satisfactory division of household and caregiving labour, including joint decision-making and shared parenting time.Reference Kline Pruett and DiFonzo11 Beyond a second parent simply ‘helping’ a primary parent, shared parenting involves both partners taking equal responsibility for household and parenting tasks,Reference Deutsch12 requiring support, coordination and communication between parents.Reference Nunes C, Roten, El Ghaziri, Favez and Darwiche6 Parents with higher levels of shared parenting have higher relationship satisfaction,Reference Tombeau Cost, Jonas, Unternaehrer, Dudin, Szatmari and Gaudreau13 better family functioning,Reference Petren and Puhlman4 lower parental stress,Reference Yalcintas and Pike14 better child–parent relationships,Reference Riina and McHale15,Reference Martin, Sturge-Apple, Davies, Romero and Buckholz16 lower levels of maternal internalising symptomsReference Yalcintas and Pike14,Reference Liu17 and better child well-being.Reference Newland18,Reference Sturge-Apple, Davies and Cummings19
Parent mental health
Given the chronic and variable stresses of COVID-19 pandemic waves, it is unsurprising that parents and children reported deteriorated mental health. A systematic review and meta-analysis on the mental health of mothers with children younger than 5 years reported that 26.9% of mothers met criteria for clinically significant depression symptoms.Reference Racine, Eirich, Cooke, Zhu, Pador and Dunnewold20 Focusing on the mental health of caregivers more broadly, another systematic review and meta-analysis reported that 27.4% of caregivers of children and adolescents up to 18 years of age reported clinically significant depression symptoms.Reference Panda, Gupta, Chowdhury, Kumar, Meena and Madaan21 Importantly, there were no differences in the rate of clinically significant symptoms in depression symptoms between male and female caregivers.Reference Panda, Gupta, Chowdhury, Kumar, Meena and Madaan21
Child and adolescent mental health
Likewise, several systematic reviews and meta-analyses have indicated that children and adolescents have experienced deteriorations in mental health related to the COVID-19 pandemic. Two systematic reviews on prevalence of clinically elevated depression in children was estimated at 25.2%Reference Racine, McArthur, Cooke, Eirich, Zhu and Madigan22 and 29.0%,Reference Ma, Mazidi, Li, Li, Chen and Kirwan23 with both meta-analysis reporting higher symptoms among adolescents compared with younger children. In a meta-analysis that specifically compared child and adolescent depression symptoms before and during the COVID-19 pandemic, symptoms were significantly higher early during compared with before the pandemic, but returned to baseline pre-pandemic levels by late 2020.Reference Robinson, Sutin, Daly and Jones24 In sum, the combination of unique and chronic stressors of the pandemic have had a deleterious impact on parent and child mental health, with potential longer-term consequences for parents and their children.Reference Ayano, Betts, Maravilla and Alati25–Reference Vafaeenejad, Elyasi, Moosazadeh and Shahhosseini27 As governments, healthcare organisations and individuals move beyond the acute phase of the pandemic, understanding individual and contextual factors that may contribute to improved parent and child mental health will be key to recovery.
Theoretical model
Bronfenbrenner's social–ecological model, which is based on concentric circles of growing influence on the child, emphasizes the importance of context in the development of an individual.Reference Tudge and Rosa28 At the centre of the model is the individual and their personal characteristics. The individual primarily inhabits the microsystem, including the immediate surroundings, which have the greatest impact on the individual. The microsystem encompasses proximal contexts such as family composition, and more distal ones, including school, paid work and neighbourhood quality. The larger context, conceptualised as the mesosystem, represents relationships, such as relationships with and between family members and interactions among the social–ecological levels. The exosystem includes interactions in community contexts. The macrosystem includes the influences of social, religious and cultural values. Finally, the chronosystem reflects the change over time. Within the context of the COVID-19 pandemic, school closures and lockdowns decreased access to community contexts. As such, different levels within the social–ecological model may have exerted more influence on parent and child well-being. For example, individual factors such as pre-existing mental health diagnoses for the child may have become more salient. Within the macrosystem, home confinement may have resulted in relationships among family members and interactions among the social–ecological levels (balancing parent paid work, online schooling and family demands) having a stronger influence on parent and child well-being. Also, the chronosystem reflecting changes over time in the lockdowns, openings and transition phases (Supplementary Fig. 1 available at https://doi.org/10.1192/bjo.2023.529), as well as changes in child development over 2.25 years, may have presented challenges to parent and child well-being. Our hypotheses are situated within the social–ecological model.
Given that both structured parenting and shared parenting have been associated with better parent and child mental health before the COVID-19 pandemic, we address the evidence gap on the role of structured and shared parenting on parent and child mental health during the pandemic and transition phase – a time of unpredictable stress. Previous literature has examined either structured parenting or shared parenting, without considering both constructs in the same model. Additionally, previous studies are also generally within a specific age group and do not span the breadth of child development. Finally, the examination of structured and shared parenting has often included a limited number of mental health outcomes or general well-being. We aim to identify family management methods that may help to rectify the sustained deterioration in parentReference Racine, Eirich, Cooke, Zhu, Pador and Dunnewold20,Reference Panda, Gupta, Chowdhury, Kumar, Meena and Madaan21 and childReference Racine, McArthur, Cooke, Eirich, Zhu and Madigan22–Reference Vafaeenejad, Elyasi, Moosazadeh and Shahhosseini27 mental health. Because different developmental periods involve different parenting demands,Reference O'Connor29 we hypothesised that structured and shared parenting may have been important to parent mental health for parents with younger children, but less so for parents with adolescents. Our first objective was to investigate associations between structured parenting and shared parenting with parent depression symptoms longitudinally among parents living in two-parent households with children aged 2–18 years. We hypothesised that both structured and shared parenting would be beneficial for parents’ mental health, with larger effect sizes for parents with younger children than parents with older children and adolescents. Our second objective was to investigate associations between structured and shared parenting, and child depression, hyperactivity/inattention and irritability symptoms longitudinally among children in this same group. We hypothesised that both structured and shared parenting would be beneficial for children's mental health, with larger effect sizes for younger children than older children and adolescents.
Method
Participants
This longitudinal cohort study was part of the Ontario COVID-19 and Kids Mental Health Study,Reference Korczak, Cost, LaForge-MacKenzie, Anagnostou, Birken and Charach30 embedded within two clinically referred mental health and neurodevelopmental cohorts and two community cohorts, described fully in our protocol.Reference Korczak, Cost, LaForge-MacKenzie, Anagnostou, Birken and Charach30 By inclusion of both clinical (e.g. children and adolescents with pre-COVID-19 mental health and/or neurodevelopmental diagnoses) and community samples, the impact of the COVID-19 pandemic could be examined across multiple relevant populations.
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. All procedures involving human participants were approved by all institutional research ethics boards (SickKids Research Ethics Board approval number 1000070222; Queen's University Health Sciences Research Ethics Board approval number 6005107; Western University Health Sciences Research Ethics Board approval number 115934; McMaster University Research Ethics Board approval number 10948; Holland Bloorview Rehabilitation Hospital Research Ethics Board approval number 0086) and all participants provided written informed consent/assent.
Parents of children aged 2–18 years (N = 1027) completed online questionnaires via REDCap version 13.5.4 for Windows (Vanderbilt University, Nashville, TN, USA; see https://projectredcap.org/) .Reference Harris, Taylor, Minor, Elliott, Fernandez and O'Neal31 To be included in these analyses, participants had to report living in a two-parent household. All data collection was completed from 15 April 2020 to 13 July 2022. For the parent depression outcome, response rates were 67.0% (Fig. 1). For child mental health outcomes, response rate varied by outcome: 85.6% for depression/emotion problems, 71.0% for hyperactivity/inattention and 83.5% for irritability/conduct problems (Fig. 1). To reduce participant burden, particularly for parents with older children whose mental health measures included more questions, not all questions were asked at every data collection point. Analysis was completed up to 31 August 2022.
Sample size flow chart.
Measures
Demographics
One parent from each family reported on household income, parent age, parent ethnicity/ancestry, parent role, number of people in the household, dwelling type (house, condominium, apartment), full-time employment of both parents, child age, child sex assigned at birth, child ethnicity/ancestry and child pre-COVID-19 psychiatric diagnoses, using items adapted from the Coronavirus Health Impact Survey (CRISIS) questionnaire (an instrument designed through international collaboration on mental health during the COVID-19 pandemic)Reference Merikangas, Milham, Stringaris, Bromet, Colcombe and Zipunnikov32 and the CRISIS – Adapted for Autism and Neurodevelopmental Conditions (CRISIS-AFAR).Reference Lai, Tint, Lunsky, Jachyra, Lin and Ameis33 We did not collect information on the exact date of child pre-COVID-19 psychiatric diagnoses, only that they had been made before the onset of pandemic restrictions in Canada.
Structured parenting
Structured parenting was assessed with one item drawn from the Parenting ScaleReference Lorber, Xu, Slep, Bulling and O'Leary34 (‘I am the kind of parent that … sets limits on what my child is allowed to do … lets my child do whatever they want’) and one item on routines in parenting constructed for the purposes of the current study (‘Throughout the day, I provide my child with … a clear and orderly routine … unstructured free time’). Items were measured on a seven-point Likert scale and were averaged. Higher scores indicate higher levels of structured parenting. Parents of children aged 2–5 years provided responses from April 2020 to November 2021 (Cronbach's α = 0.76, 95% CI 0.71–0.79). Parents of children aged 6–18 years provided responses from April to July 2020 (Cronbach's α = 0.70, 95% CI 0.66–0.77).
Shared parenting
Shared parenting was assessed with two items drawn from the Childbearing Attitudes QuestionnaireReference Tombeau Cost, Jonas, Unternaehrer, Dudin, Szatmari and Gaudreau13 (‘I am bothered by my partner's lack of involvement in the daily care of the child’; ‘Currently, is the sharing of household tasks a source of tension or conflict between you and your partner?’). Items are measured on a seven-point Likert scale and were averaged. Higher scores indicate higher levels of shared parenting. Parents of children aged 2–5 years provided responses from June 2020 to November 2021 (Cronbach's α = 0.76, 95% CI 0.72–0.79). Parents of children aged 6–18 years provided responses from April to July 2020 (Cronbach's α = 0.73, 95% CI 0.68–0.78). Structured and shared parenting were very weakly correlated (r = 0.04).
Parent depression symptoms
Parental depression symptoms were assessed with the Patient Health Questionnaire-8 (PHQ-8), comprising eight items measured with a four-point Likert scale.Reference Kroenke, Strine, Spitzer, Williams, Berry and Mokdad35 Responses were summed and higher scores indicate higher levels of depression symptoms. Parents provided repeated responses (median: 4; range: 1–11) on the PHQ-8 from April 2020 to May 2022.
Child depression symptoms
For children ages 2–5 years, parents completed the Strengths and Difficulties Questionnaire (SDQ) Emotional Subscale – Parent Report, comprising five items measured with a three-point Likert scale.Reference Croft, Stride, Maughan and Rowe36,Reference Goodman37 For children ages 6–18 years, parents completed the Revised Children's Anxiety and Depression Scale (RCADS-P), a ten-item depression subscale measured with a four-point Likert scale.Reference Chorpita, Moffitt and Gray38 Responses were summed for the SDQ Emotional Subscale and T-scores were created for RCADS-P. Higher scores indicate higher levels of symptoms. Parents responded repeatedly (median: 4; range: 1–14) from May 2020 to July 2022.
Child hyperactivity/inattention symptoms
In children aged 2–5 years, child hyperactivity was measured with the SDQ Hyperactivity Subscale – Parent Report, a five-item subscale measured with a three-point Likert scale.Reference Croft, Stride, Maughan and Rowe36,Reference Goodman37 In children aged 6–18 years, hyperactivity and inattention was assessed with the Strengths and Weaknesses of Attention-Deficit/Hyperactivity Disorder Symptoms and Normal Behavior Scale – Parent Report (SWAN), comprising 18 items measured with a seven-point Likert scale.Reference Swanson39 Sum scores were reversed, with higher scores indicating higher levels of symptoms. Parents responded repeatedly (median: 5; range: 1–14) from July 2020 to July 2022.
Child irritability symptoms
For children aged 2–5 years, irritability symptoms were assessed with the SDQ Conduct Problems Subscale – Parent Report, comprising five items measured with a three-point Likert scale.Reference Croft, Stride, Maughan and Rowe36,Reference Goodman37 For children aged 6–18 years, irritability was measured with The Irritability and Dysregulation of Emotion Scale, Brief – Parent Report (TIDES), comprising six items measured with a seven-point Likert scale.Reference Dissanayake, Dupuis, Arnold, Burton, Crosbie and Schachar40 Scores were summed, with higher scores indicating higher levels of symptoms. Parents responded repeatedly (median: 4; range: 1–14) from May 2020 to July 2022.
Data analysis
Data were analysed in R version 4.1.2 for Windows (R Foundation for Statistical Computing, Vienna, Austria; see https://www.R-project.org/) and RStudio version 1.2.1335 for Windows (RStudio Team, PBC, Boston, MA, USA; see http://www.rstudio.com/). As described above, data were collected at different times, using different developmentally appropriate measures. To account for differences in both data collection and in parenting behaviours across child development, we stratified the sample by child age group: 2–5 years old, 6–9 years old, 10–12 years old and 13–18 years old. These age groups were chosen a priori based on children's developmental stages requiring more or less explicit limit-setting and more or less instrumental caregiving. We used different pre-processing and data analysis for the age groups 2–5 years and 6–18 years, because of different frequency of exposure assessments. Data were analysed with mixed-effects linear models. Mixed-effects models allow for modelling of longitudinal data, and can handle uneven spacing of repeated measurements and variable numbers of repeated measures per participant/family. Reported results are the estimated fixed effects for the exposure of interest (e.g. structured parenting, shared parenting), after adjusting for the longitudinal nature of the data through fixed effects for changes in date, and random effects to account for correlations within family and participant.
Parents and children aged 2–5 years
We used the mice package for Windows (van Buuren & Groothuis-Oudshoom; see https://cran.r-project.org/web/packages/mice/index.html) for hierarchical imputation (n = 15) of missing item-level covariate and predictor data across multiple assessments (<50% missing data on all items required for analysis).Reference Lewin, Brondeel, Benmarhnia, Thomas and Chaix41–Reference Madley-Dowd, Hughes, Tilling and Heron43 As both the PHQ-8 and the SDQ rely on sum scores, participants had to complete all items on the outcome measure to prevent downward bias.
We analysed associations between repeated assessments of structured and shared parenting (data collection from April 2020 to November 2021) with repeated assessments of depression symptoms (data collection from April 2020 to July 2022), using linear mixed models (lme4 package for Windows (Bates et al; see https://cran.r-project.org/web/packages/lme4/index.html)).Reference Cnaan, Laird and Slasor44 First, we ran unadjusted models. To account for confounds, we analysed data in adjusted models. In the adjusted parent depression model, we included household income and parent ethnicity/ancestry as potential confounds. In adjusted child mental health models, child ethnicity/ancestry, child sex assigned at birth and PHQ-8 scores were included.
Parents and children aged 6–18 years
We used single-level imputation (n = 15; mice) for missing item-level covariate and predictor data (single assessments), if participants had <50% missing data on all items required for analysis.Reference Lewin, Brondeel, Benmarhnia, Thomas and Chaix41–Reference Madley-Dowd, Hughes, Tilling and Heron43 As the PHQ-8, SWAN and TIDES scales rely on sum scores, no outcomes were imputed to prevent downward bias.
We analysed associations between a single assessment of structured and shared parenting at the start of the COVID-19 pandemic (data collected once from April to July 2020) with repeated measures of parent depression symptoms (data collected from April 2020 through July 2022), using linear mixed modelsReference Cnaan, Laird and Slasor44 with the lme4 package for each age group separately. To account for confounds, we analysed data in adjusted models. In adjusted parent depression models, we included household income, parent role, parent ethnicity/ancestry and child pre-COVID-19 pandemic mental health diagnosis as confounds. These variables have been previously associated with parent mental health and with structured and shared parenting. In adjusted child mental health models, we included household income, child sex assigned at birth, child ethnicity/ancestry, child pre-COVID-19 pandemic mental health diagnosis and PHQ-8 at baseline as confounds. SWAN questions were not asked at baseline. To account for baseline levels of inattention/hyperactivity in the adjusted model, two items measuring hyperactivity and inattention from the CRISIS questionnaireReference Merikangas, Milham, Stringaris, Bromet, Colcombe and Zipunnikov32 were included.
Results
Participant characteristics
Participant characteristics are shown in Table 1. Because of low variance on specific variables (parent role, household income, previous mental health diagnosis), planned covariates were not included in some adjusted models in children aged 2–5 years.Reference Peduzzi, Concato, Kemper, Holford and Feinstein45 In contrast, most parents with children aged 6–18 years reported a pre-COVID-19 mental health diagnosis for their child. Therefore, we conducted sensitivity analyses to examine whether results depended on pre-COVID-19 child mental health diagnosis in this age group (Supplementary Material).
Participant characteristics
PHQ-8, Patient Health Questionnaire-8; SDQ, Strengths and Difficulties Questionnaire; RCADS-P, Revised Children's Anxiety and Depression Scale; TIDES, The Irritability and Dysregulation of Emotion Scale, Brief; n, number of participants.
a. For parents with children aged 2–5 years, results reflect an average across all measurements; for parents with children aged 6–18 years, results reflect an average of the baseline (only) measure.
b. Range: 1–7.
c. Range: 0–10.
d. Range: 35.16–126.08.
e. Range: 2–10.
f. Range: 0–8.
g. Range: −18 to 18.
Parent depression symptoms
Structured parenting was not associated with parent depression symptoms across the pandemic in any of the child age groups (Table 2). However, higher levels of shared parenting were associated with lower parent depression symptoms (Table 2, Fig. 2(a)).
Association between structured and shared parenting and mental health
Bold values were statistically significant. To account for testing three different outcomes (emotional problems, hyperactivity, conduct problem) for children aged 2–5 years, applying the Bonferroni correction (0.05/3), results in a critical value for statistical significance of 0.017; and to account for three different outcomes (depression symptoms, hyperactivity/inattention symptoms, irritability symptoms) for children aged 6–18 years, applying the Bonferroni correction (0.05/3), results in a critical value for statistical significance of 0.017. SDQ, Strengths and Difficulties Questionnaire; RCADS-P, Revised Children's Anxiety and Depression Scale; SWAN, Strengths and Weaknesses of Attention-Deficit/Hyperactivity Disorder Symptoms and Normal Behavior Scale – Parent Report; TIDES, The Irritability and Dysregulation of Emotion Scale, Brief.
a. The model with parents with children aged 2–5 years was adjusted for income and parent ethnicity; models with parents with children aged 6–18 years were adjusted for income, parent ethnicity, responding parent role and previous mental health diagnosis for child.
b. The model with children aged 2–5 years was adjusted for child gender assigned at birth, child ethnicity and parent depression; models with children aged 6–18 years were adjusted for income, child gender assigned at birth, child ethnicity, previous mental health diagnosis for child and parent depression
(a) Shared parenting and parent symptoms of depression. (b) Shared parenting and child symptoms of depression/emotional problems. (c) Structured parenting and child symptoms of irritability/conduct problems. (d) Shared parenting and child symptoms of irritability/conduct problems. All analyses stratified by child developmental groups: 2–5 years, 6–9 years, 10–12 years and 13–18 years. Statistically significant results are indicated by an asterisk (*) in the key.
Child depression symptoms
Structured parenting was not associated with child depression/emotion problems (Table 2). However, in children aged 2–5 years, higher levels of shared parenting were associated with lower levels of child emotional problems, but not depression, in other developmental groups (Table 2, Fig. 2(b)).
Child hyperactivity and inattention symptoms
Neither structured nor shared parenting were significantly associated with child hyperactivity or inattention/hyperactivity symptoms (Table 2).
Child irritability symptoms
Among children aged 2–5 years, higher levels of structured parenting and higher levels of shared parenting across the study were significantly associated with lower child conduct problems (Table 2, Figs. 2(b), 2(c)). We did not detect significant associations between structured parenting and child irritability in other developmental groups (Table 2, Fig. 2(c)). However, among children aged 13–18 years, higher levels of shared parenting at the start of the COVID-19 pandemic were associated with lower irritability symptoms across the study (Table 2, Fig. 2(d)).
Discussion
The objectives of this study were to examine the association of structured and shared parenting with parent and child mental health. Our findings suggest that higher levels of shared parenting are associated with fewer parent depression symptoms in parents with children in all age groups (2–5, 6–9, 10–12 and 13–18 years of age), but structured parenting was not associated with parent symptoms of depression in parents with children in any age group. Among children aged 2–5 years, higher levels of shared parenting were associated with fewer child emotion and conduct problems, but were not associated with hyperactivity symptoms. Among children aged 2–5 years, higher levels of structured parenting were associated with fewer child conduct problems, but were not associated with emotion problems or hyperactivity symptoms. Among children aged 6–9 years, neither shared parenting nor structured parenting were associated with depression, irritability or inattention/hyperactivity symptoms. Among children aged 10–12 years, neither shared parenting nor structured parenting were associated with depression, irritability or inattention/hyperactivity symptoms. Among adolescents aged 13–18 years, shared parenting was associated with less irritability, but was not associated with depression or inattention/hyperactivity symptoms. Among adolescents (aged 13–18 years), structured parenting was not associated with depression, irritability or inattention/hyperactivity symptoms.
Structured parenting
During the COVID-19 pandemic, externally supported structure was often absent.Reference Hood, Zabatiero, Silva, Zubrick and Straker1 This loss placed the burden of creating and maintaining structure on parents, such as keeping up with school-related tasks, sufficient physical activities and cognitive stimulation, and play. Pandemic-related lockdowns have been associated with reduced structured parenting.Reference Hood, Zabatiero, Silva, Zubrick and Straker1,Reference McGoron, Wargo Aikins, Trentacosta, Gomez and Beeghly46,Reference Glynn, Davis, Luby, Baram and Sandman47 Our findings on structured parenting and parent depression contrast studies conducted before the COVID-19 pandemic, which found that higher levels of structured parenting were associated with better well-being in mothersReference Churchill and Stoneman10 and decreased depression in parents.Reference Davis, Myers, Logsdon and Bauer8 In contrast to a cross-sectional, COVID-19-pandemic-related study showing an association between structured parenting and parent depression,Reference McGoron, Wargo Aikins, Trentacosta, Gomez and Beeghly46 our longitudinal results did not show any such association. Moreover, despite cross-sectional, COVID-19-pandemic-related associations between structured parenting and child mental health,Reference Glynn, Davis, Luby, Baram and Sandman47 we did not find much support for this association across domains and age groups. Structured parenting was only associated with decreased conduct problems in children aged 2–5 years, which is in agreement with previous literature.Reference Glynn, Davis, Luby, Baram and Sandman47
Structured parenting was not associated with parent mental well-being and, with the exception of conduct problems in children aged 2–5 years, was not associated with child mental health in any other age group or domain. This might be because parents prioritised other factors during this chaotic context rather than structuring the family environment. Given the number of concurrent changes to family life and the absence of externally supported structure, the consistent and effective implementation of structured parenting may have been more than parents were able or willing to take on, and thus may not have affected mental health. Other factors may have contributed more to parent and child mental health, such as lockdown policy stringency,Reference Aknin, Andretti, Goldszmidt, Helliwell, Petherick and De Neve48 COVID-19-related stress,Reference Rizeq, Korczak, Cost, Anagnostou, Charach and Monga49 material deprivationReference Rizeq, Korczak, Cost, Anagnostou, Charach and Monga49 or pre-existing child mental health diagnosis (see Supplementary Table S2). A core component of structured parenting is consistency, which would have been particularly challenging considering the frequent and short-notice changes to public health measures (Supplementary Fig. 1). Our findings support future investigations on the use and limits of structured parenting under the chaotic and inflexible life conditions that some families experience outside of the context of the COVID-19 pandemic.
Shared parenting
Shared parenting was associated with better parent mental health across all age groups and better child mental health (ages 2–5 years: emotion and conduct problems; ages 13–18 years: irritability). Sharing of household and childcare labour is a highly discussed topic, in research as well as in the general public.Reference Lewis50 Studies from before the COVID-19 pandemic indicate that women spend two to ten times more time on unpaid household and childcare labour than men, across all levels of income and all locations studied to date.Reference Ferrant, Pesando and Nowacka51 Even when fathers worked less and made less money than their female partners, they engaged in less housework on average.Reference Besen-Cassino and Cassino52 Before the COVID-19 pandemic, paternal involvement in childcare was associated with their own gender-role views, and not those of their female partner.Reference Bulanda53 However, during the pandemic, a different pattern emerged. For instance, a programme in Germany, intended to buffer economic and labour market problems, allowed employees to retain their jobs and salary but work reduced hours during lockdown phases. Fathers who participated in this programme engaged in more housework and childcare, particularly fathers with low or medium educational level.Reference Naujoks, Kreyenfeld and Dummert54 In the USA, parents who reported having no help in household and childcare labour were more likely to reduce paid work hours or completely drop out of the paid workforce. Thus, shared parenting may support workforce participation, with economic benefit for the family.Reference Zamarro and Prados55
Changes to the macrosystem in the COVID-19 pandemic context required parents to balance the demands of paid work (or the concerns of lost work), instrumental and affective care of children, negotiating/managing online school and other family demands in a time of great uncertainty. During the COVID-19 pandemic, both parents and children spent more time in the home, resulting in increased household labour and childcare responsibilities for parents. With a few exceptions (e.g. Sevilla and SmithReference Sevilla and Smith56), multiple reports indicate more equitable sharing of household and childcare labour by parents in diverse cultures during the COVID-19 pandemic.Reference Naujoks, Kreyenfeld and Dummert54,Reference Zamarro and Prados55,Reference Shafer, Scheibling and Milkie57–Reference Dib, Rougeaux, Vazquez-Vazquez, Wells and Fewtrell59 When the division of household labour and childcare was perceived to be more equitable, mothers reported fewer relationship problemsReference Waddell, Overall, Chang and Hammond60 and better coping skills.Reference Wei, Gao, Fewtrell, Wells and Yu61
Perhaps counterintuitively, we found that shared parenting had the largest effect on the mental health of parents with adolescent children compared with parents with younger children. This may reflect the different demands of parenting adolescents in particular.Reference Liu17 Likewise, we also found higher levels of shared parenting were associated with decreased symptoms of irritability in adolescents. Positive family dynamics (i.e. shared parenting) may have buffered the effects of social isolation for adolescents during the pandemic.Reference Cost, Crosbie, Anagnostou, Birken, Charach and Monga62 Adolescents who reported their parents had poor marital quality had higher rates of mental health problems as adults.Reference Hair, Anderson Moore, Hadley, Kaye, Day and Orthner63 Parenting conflicts, particularly those related to co-parenting, have also been associated longitudinally with worse parent–adolescent relationship quality,Reference Martin, Sturge-Apple, Davies, Romero and Buckholz16,Reference Hair, Anderson Moore, Hadley, Kaye, Day and Orthner63 and family conflict is the leading cause of mental distress in adolescents.Reference Wilkinson, Kelvin, Roberts, Dubicka and Goodyer64
We found that higher levels of shared parenting were associated with fewer emotional and conduct problems in children aged 2–5 years. We did not find any studies specifically examining the role of shared parenting in child mental health during the pandemic. However, a 2010 meta-analysis reported small effects between both parent cooperation and conflict and child externalising symptoms, with larger effect sizes within clinical populations.Reference Teubert and Pinquart65 In our sensitivity analyses, effect sizes for shared parenting were larger for parent and child depression symptoms when the child had a pre-existing mental health diagnosis compared with those who did not (Supplementary Tables 7 and 8). Similarly, there were small effects between both parent cooperation and conflict and child internalising symptoms, with larger effect sizes among younger children.Reference Teubert and Pinquart65 Although household labour and childcare have become more equitably shared between parents during the COVID-19 pandemic,Reference Naujoks, Kreyenfeld and Dummert54,Reference Zamarro and Prados55,Reference Shafer, Scheibling and Milkie57–Reference Dib, Rougeaux, Vazquez-Vazquez, Wells and Fewtrell59 it is unclear whether this adjustment will precipitate more durable changes in the division of household and childcare labour. This might benefit a broad spectrum of outcomes,Reference Petren and Puhlman4,Reference Tombeau Cost, Jonas, Unternaehrer, Dudin, Szatmari and Gaudreau13–Reference Martin, Sturge-Apple, Davies, Romero and Buckholz16,Reference Zamarro and Prados55 including the current findings on parent and child mental health.
Shared parenting is amenable to intervention. Two recent systematic reviews and meta-analysis identified positive effects of co-parenting interventions on parent well-beingReference Nunes C, Roten, El Ghaziri, Favez and Darwiche6,Reference Xiao and Loke66 for at-risk families and families with no known risks.Reference Nunes C, Roten, El Ghaziri, Favez and Darwiche6 Further, co-parenting interventions can be effective in non-traditional or multigenerational families.Reference Xiao and Loke66 Co-parenting interventions fostering shared parenting, even in times of crisis,Reference Cobham, McDermott, Haslam and Sanders67 might be beneficial for parents, children and labour participation.Reference Petren and Puhlman4,Reference Tombeau Cost, Jonas, Unternaehrer, Dudin, Szatmari and Gaudreau13–Reference Martin, Sturge-Apple, Davies, Romero and Buckholz16,Reference Zamarro and Prados55
Limitations and strengths
An important limitation in this study is our sample restriction to include only two-parent households. Future work should consider diverse family structures (e.g. single parents, multigenerational households) and diverse parents (e.g. LGBTQI+). This study was also limited by convenience sampling, which may not be representative of the general population in socioeconomic dimensions, and source bias, as all measures in this study were reported by parents. Understanding how different families and parents experiencing different contextual stressors may enact structured and shared parenting practices, and how these may buffer their own mental health and their child's mental health, will be important for clinical translation of these findings.
Despite these limitations, our study has several strengths, including longitudinal design in a large sample, enriched participation of parents whose children have a pre-existing mental health or neurodevelopmental disorder, and a powerful statistical approach making use of all available data. Our study is the first to our knowledge that examines the role of both structured and shared parenting in the same model. Further, we examine both parent and child mental health across the full span of child development, consider multiple mental health outcomes in children, and test how these family management systems affect children and adolescents with and without pre-existing mental health diagnoses in the same study.
In conclusion, shared parenting contributes meaningfully to parent and child mental health in two-parent households, even under times of stress. Shared parenting may be a target for intervention among parents and children experiencing mental health difficulties. Support of shared parenting responsibilities (parental leave policies guaranteeing paternal leave, flexible work-from-home policies, public health engagement, clinical interventions) may be a promising strategy to ameliorate parent and child mental health during the COVID-19 pandemic recovery and beyond.
Supplementary material
Supplementary material is available online at https://doi.org/10.1192/bjo.2023.529
Data availability
The data that support these findings are not publicly available because the data contain information that could compromise the privacy of the research participants.
Acknowledgements
The authors would like to gratefully acknowledge Charlie Keown-Stoneman, Xuedi Li, Anett Schumacher, Avalon Henry and Ronda Lo for their help with data management; Peter Szatmari for helpful comments on an earlier draft; and the families, children and youth who participated in this research.
Author contributions
K.T.C., P.M., C.S.B. and A.C. were responsible for study concept and design. K.T.C., P.M., E.U., D.J.K., J.C., E.A., S.M., E.K., R.S., J.M., P.A., C.L.B., S.G., R.N., C.S.B. and A.C. were responsible for acquisition, analysis or interpretation of data. K.T.C., P.M., E.U. and A.C. drafted the manuscript. D.J.K., J.C., E.A., S.M., E.K., R.S., J.M., P.A., C.L.B., S.G., R.N. and C.S.B. critically revised the manuscript for important intellectual content. K.T.C. and P.M. were responsible for statistical analysis. K.T.C, D.J.K., J.C., E.A., S.M., E.K., R.S., J.M., P.A., C.L.B., S.G., R.N., C.S.B. and A.C. obtained funding. P.M. and C.L.B. provided administrative, technical or material support. K.T.C., A.C. and C.S.B. supervised the study.
Funding
Funding for this study has been provided by the Canadian Institutes of Health Research (EG2-179443, WI2-179954, MS3-173092 to D.J.K.); the Ontario Ministry of Health (#700 to D.J.K.); Centre for Brain and Mental Health, SickKids, Leong Centre for Healthy Children, SickKids; and the Miner's Lamp Innovation Fund in Prevention and Early Detection of Mental Illness at the University of Toronto. Additional support for the Province of Ontario Neurodevelopmental Network cohort comes from the Ontario Brain Institute. Spit for Science was funded by the Canadian Institutes of Health Research (PJT-159462 to J.C., R.S.). The funders had no role in the design and conduct of the study, or collection, management, analysis and interpretation of the data.
Declaration of interest
K.T.C. reported receiving grants from The Hospital for Sick Children, University of Toronto, and the Canadian Institutes of Health Research; they have also acted as a statistical consultant for Unity Health and University of Toronto and conference/meeting attendance support from Merit Network. E.U. reported receiving grants from University of Basel Research Fund (3MS1064). D.J.K. reported receiving grants from Canadian Institutes of Health Research, Hospital for Sick Children, Gary Hurvitz Centre for Brain and Mental Health, the University of Toronto, and conference/meeting attendance support from the Canadian Paediatric Society and the Canadian Academy of Child and Adolescent Psychiatry. J.C. reported receiving grants from Canadian Institutes of Health Research, Ontario Brain Institute, and the Hospital for Sick Children Foundation. E.A. reported receiving consultation fees from Roche and Quadrant, research funding from Roche, in-kind support from AMO Pharma, editorial honoraria from Wiley and book royalties from APPI and Springer; she also holds a patent for the device Tully (formerly Anxiety Meter) and in-kind support for all Province of Ontario Neurodevelopmental Network data from the Ontario Brain Institute during the conduct of the study. S.M. reported receiving grants outside of the submitted work from the Cundill Centre for Youth Depression at the Centre for Addiction and Mental Health, and the Canadian Institutes of Health Research, royalties from Springer Publishing, and research support as the holder of the TD Bank Financial Group Chair in Child & Adolescent Psychiatry. E.K. reported receiving grants from Ontario Brain Institute, Canadian Institutes of Health Research and the Masonic Foundation of Ontario outside the submitted work. J.M. reported receiving grants from Canadian Institutes of Health Research, Ontario SPOR Support Unit, St. Michael’s Hospital Foundation, and the Lawson Centre for Child Nutrition, University of Toronto. R.S. receives grant funding from CIHR and Ontario Brain Institute, honoraria from Otsuka Pharmaceutical for scientific consultation, and equity in ehave. C.L.B. reported receiving grants from the Canadian Institutes of Health Research. S.G. reported receiving grants from Canadian Institutes of Health Research, McMaster University, Hamilton Health Sciences and HHS Foundation, the Masonic Foundation of Ontario, Empowered Kids Ontario, Azrieli Foundation, and the Ontario Brain Institute. R.N. has received grants from Ontario Brain Institute during the conduct of the study and has also received funding from Brain Canada, the Canadian Institutes of Health Research, and Hoffman-Laroche Ltd outside of the submitted work. C.S.B. reported receiving grants from Canadian Institutes of Health Research, Heart & Stroke Foundation of Canada, Physician Services Inc, The Edwin S.H. Leong Centre for Healthy Children, University of Toronto and Hospital for Sick Children, The Center for for Addiction and Mental Health, Walmart Canada Regional Community on addressing food insecurity in children admitted to hospital. A.C. reported receiving grants from SickKids Hospital and Leong Centre for Healthy Children. All other authors have no conflicts of interest to declare.
Open access
eLetters
No eLetters have been published for this article.