Families in the COVID-19 pandemic: parental stress, parent mental health and the occurrence of adverse childhood experiences—results of a representative survey in Germany

The SARS-CoV-2 coronavirus was declared a “public health emergency of international concern” by the World Health Organisation on January 30, 2020, and upgraded to a pandemic on March 11, 2020. In Germany, the first peak of infections occurred in spring 2020 and comprehensive, nationwide restrictions were implemented to slow down infection rates.

The restrictions are especially challenging for families, with home schooling, social distancing measures and lockdown situations having a profound and complex impact in the family context [1‐3]. These parenting challenges are compounded by the demands of working from home, economic difficulties and layoffs, and social restrictions on parents [1, 2, 4]. It has been suggested that the profound changes to everyday family life caused by the pandemic may fuel parental stress and intrafamilial tension, which may in turn lead to an increase in adverse childhood experiences (ACEs), including domestic violence, child abuse, and neglect [5, 6].

While evidence on the psychosocial sequelae of the COVID-19 pandemic for general population samples [7] and for children and adolescents [8] is emerging, the situation of families and specifically parental burden has less frequently been analysed so far. To date, few studies have yet focused on pandemic-related stress, parental stress and the occurrence of ACEs during the pandemic: an online survey of 420 caregivers in the United States conducted in April 2020 showed a moderate level of general caregiver stress, mild generalized anxiety, and average depression [9]. The role of parental stress was also analysed in a study conducted in Italy with 824 parents of children aged 1–14 years in early April 2020, when the nationwide lockdown period was extended for the first time [10]. Poorer parental coping with the lockdown measures was related to higher individual parent stress, parenting stress and child behavioural problems. Although these findings underline the broad effects of the pandemic within the family context, the online sample was not representative.

The Canadian Perspectives Survey Series examined family stress and perceived risk for domestic violence due to the COVID-19 pandemic in 4627 adults [11]. The results suggest that financial insecurities due to changes in working conditions were the main cause of an increase in family stress and in the perceived risk for domestic violence. Reduced social contacts were also related to higher concerns about domestic violence. However, the actual occurrence of domestic violence was not assessed. An online survey with 258 parents living in Singapore found that parental stress mediated the association between the perceived impact of COVID-19 and harsh punishment of children [12]. However, the sample size was relatively low and the sample was recruited via Facebook and community organizations, limiting the representativeness of the study and restricting its generalizability beyond Singapore. Two recent studies add evidence on the assumed relation between pandemic-related stress and the occurrence of or risk for child abuse. Brown et al. [13] analysed 183 parents of underaged children, mainly mothers, and showed that pandemic-related stress, anxiety and depression are related to parents’ perceived stress levels. Furthermore, parents who were receiving financial assistance already pre-COVID and parents with higher anxiety and depression showed increased child abuse potential. The actual occurrence of psychological maltreatment and physical abuse was assessed in a recent study by Lawson et al. [14]. The authors analysed risk factors for child abuse among 342 parents of 4–10-year-old children in an online survey. The loss of job due to the pandemic, higher parental depression and previous emotional maltreatment were related to emotional maltreatment within the week prior to data collection. For the occurrence of physical abuse, loss of job emerged as relevant risk factor as well, however, the effect was weakened if parents showed higher positive cognitive reframing. While this study provided important data on risk and resilience factors for the occurrence of child abuse, the sample was recruited via social media channels, limiting representativeness. Furthermore, the role of parental stress and other pandemic-related stressors for the occurrence and increase of ACE during the pandemic has not been analysed so far.

By surveying a representative sample of parents with underage children, this study aims to (1) generate representative data on pandemic-related stress, parental stress, parental subjective and mental health and the occurrence of ACEs; (2) describe risk factors for an increase in ACEs and (3) provide qualitative data on parents’ negative and positive experiences during the pandemic.

First, we analyzed data on pandemic-related stress and the other outcome measures. The level of pandemic-related stress varied across domains (see Fig. 1). Parents felt most stressed (ratings 4 or 5 on the Pandemic Stress Scale) by social distancing from family and friends (56.1%), closure of schools (55.9%), closure of childcare (52.1%), concerns about the health of others (47.0%), and restrictions on outside activities (46.2%). The total sum score of pandemic-related stress was M = 31.97 (SD = 10.96; range 1–70).

Parental stress at the time of the subjectively highest burden was significantly higher than pre-COVID-19 levels (M = 36.93, SD = 10.45, range 18–71 vs. pre-COVID-19 M = 34.72, SD = 10.63, range = 18–70; t(1023) = 12.474, p < 0.001); the effect size was small (d = 0.21). However, parental stress at the time of the subjectively highest burden was not higher than in two reference samples (US sample [20]: n = 116, M = 37.1, SD = 8.1; German sample [30]: n = 121, M = 37.18, SD = 7.70) and was significantly lower than in clinical reference samples [31] (parents in treatment for their child’s behaviour problems: n = 51, M = 43.2, SD = 9.1; parents in inpatient psychiatric treatment: n = 83, M = 41.9, SD = 9.4).

The mean score for parents’ general stress at the time of the subjectively highest burden was M = 5.28 (SD = 4.13, range 0–20) and thus in the low range [16].

Parents rated their overall health status at the time of the subjectively highest burden as significantly worse than pre-COVID-19 levels (M = 6.80, SD = 2.21 vs. pre-COVID-19 M = 7.34, SD = 2.04; t(1023) = 10.33, p < 0.001). The effect size was small (d = 0.31).

Regarding parent mental health, our sample reported significantly higher symptoms of anxiety and depression than the German normative data [21] (see Fig. 2); the effect sizes were small (depression d = 0.21; anxiety d = 0.12, total score d = 0.18). 12.3% of the sample scored above the 95th percentile of the PHQ-4 [21] for symptom levels for depression; 9.7% for anxiety; 7.4% for the total score.

As women usually report higher levels of subjective distress than men, we checked for gender differences. Mothers scored higher than fathers on general stress (d = 0.32, p < 0.001), anxiety (d = 0.28, p < 0.001), and depression (d = 0.28, p < 0.001) and showed poorer subjective health (d = 0.15, p = 0.018). No sex differences emerged for pandemic-related stress (d = 0.09, p = 0.160) and parental stress (d = 0.08, p = 0.196).

Pandemic-related stress and parental outcomes were significantly correlated, with moderate effect sizes (see Table 3).

To gain a more detailed picture of the associations between pandemic-related stress and parental outcomes, we conducted an item-wise analysis of the Pandemic Stress Scale (see Supplementary Material 3). Results showed a homogenous pattern: all outcomes were significantly related to the specific restrictions, with two exceptions (correlations subjective health with child care closures r = 0.002, and with job closures r = − 0.102). Coefficients of significant correlations were in the small to medium range (r = − 0.077 to r = 0.393).

This study is among the first to systematically analyse the effects of the corona pandemic on parental health and stress and to examine whether it was indeed related to an increase in the occurrence of ACEs. Overall, the negative effects on parental outcomes were small to moderate. While parental stress increased during the pandemic, it was still within the average range; subjective health decreased moderately; rates of anxiety and depressive symptoms increased, with small effects. Pandemic-related stress varied across domains; parents felt most stressed by social distancing measures and by the closure of schools and childcare facilities. Higher pandemic-related stress was related to poorer parental outcomes and an increase in ACEs. Furthermore, subgroups with an increase in ACEs were characterized by specifically high pandemic-related stress and poor parental outcomes.

Severe ACEs were reported by 6.5% of the sample, which is in line with current prevalence rates [32]. For the specific subtypes, covering less severe forms of ACEs, a lifetime prevalence of up to 32% was reported, also in line with the literature [32]. In the subgroup of parents who reported lifetime ACEs, between 16.4 and 55.4% reported an increase since the beginning of the pandemic. In these families, poorer outcomes were observed on all measures (pandemic-related stress, parental stress, general stress, subjective health, anxiety and depression) than in families reporting no change or a decrease. This adds to findings on the relation between parental anxiety and depression and risk for child abuse [13]. Especially, and consistently across categories, parental stress was higher in families where ACEs increased. Noteworthy, parents in this subgroup also reported higher pre-pandemic levels of parental stress than the families with a decrease or no change in ACEs (WDV d = 0.50, VEA d = 0.45). Including the score for pre-pandemic parental stress as covariate in the analyses did not change the results. Other studies conducted during the COVID-19 pandemic have also highlighted the relevance of parental stress for child-related outcomes, such as child behavior problems [10] and parent–child closeness [12]. Parental stress thus seems to be a central variable reflecting the impact of the pandemic on families—both on the parents themselves and on their children.

Parental stress was not the only factor related to an increase in the two most prevalent ACEs: witnessing domestic violence and verbal emotional abuse. Families with younger children were especially at risk, which may reflect the impact of childcare and school closures [33]. Contrary to our expectations, single parenting, parent sex, belonging to the COVID-19 risk group, risk of alcohol abuse, and low SES were not related to an increase in ACEs. Although low case numbers limit the ability to draw conclusions and may explain contradictions with other recent findings [13, 34], our data suggest that negative effects on parent or child level may be attributable less to general sociodemographic or socioeconomic factors and more to pandemic-specific socioeconomic factors, such as job losses and financial difficulties. This finding is in line with Beland et al. [11], who reported an association between financial worries, domestic violence and family stress for a population-based sample, and with Lawson et al. [14], who identified pandemic-related job loss as a risk factor for the occurrence of emotional maltreatment and physical abuse. In line with the literature [35], a parent’s own history of violence was related to an increase in ACEs, although the effects emerging were mixed. Taken together, our results suggest that beyond younger age and parent’s own history of violence, job losses and financial difficulties additionally contributed to an increase in ACEs during the pandemic. The results of our study underline that the effects of the pandemic are heterogeneous and the stress imposed by the pandemic might especially exacerbate in known risk groups, which corresponds to other studies (see also [12, 14, 34]). However, longitudinal data are needed to confirm these findings.

Parental stress and general stress were not much higher than usual in our sample, in line with other recent findings [9, 36]. However, in contrast to other studies [13], levels of anxiety and depression were also comparatively low, with only 7–12% of respondents scoring above the 95^th percentile of the PHQ-4. A recent meta-analysis of 17 studies conducted mostly in Asia found a much more pronounced increase in rates of anxiety and depression, of up to 18.7–50.9% [7]. These differences may be explained by the phase of the pandemic in which the data were collected [5]. Most studies published thus far were conducted during earlier, more acute stages of the pandemic; our study was conducted in August 2020, when the first peak of the pandemic was over in Germany, the lockdown measures had been relaxed, and many regions were on school holidays. This situation may also have biased parents’ ratings with respect to the subjectively hardest time, as recent data suggest a decline in anxiety and depression over the first 20 weeks after implementation of lockdown measures [37]. Furthermore, up to August 2020, the situation in Germany was more stable than e.g., in other European countries [38], which may also explain cross-national differences.

The qualitative data on the aspects that caused parents most stress validated and added to the quantitative data from the Pandemic Stress Scale. Participants reported being particularly stressed by social distancing, restrictions and childcare at home. About one-third of participants could not identify any positive aspects of the pandemic; however, parents perceived more time for themselves and for the family and an improvement in family cohesion as positive aspects of the pandemic [5, 39].

In sum, parental stress emerged as important target point for interventions. The qualitative data also identified promising areas for resource-oriented interventions to mitigate the detrimental effects of the pandemic [39]. Prospective studies on the mediating role of parental stress and interactions with potential risk and resilience factors for parent and child outcomes are needed to provide insights into the pathways between these variables, as first evidence suggests [14].

One strength of this study is the mixed mode approach to data collection, reducing the bias associated with collecting data online only. Unlike other parent samples, in which mothers tend to be heavily overrepresented, our sample included equal proportions of mothers and fathers. Mothers reported higher general stress, anxiety and depression, in line with both the literature in general [21] and data collected during the pandemic [34, 37]. Notably, mothers and fathers did not differ in the level of pandemic-related stress and parental stress reported. While this finding is in line with data reported in a validation study of the Parental Stress Scale [20], pandemic-specific comparison data are as yet scarce. While we aimed to yield a representative sample in terms of parental education, we observed an imbalance compared to population-based reference data for parental education and SES index [15, 29]. The share of low SES in our sample was smaller than expected [15]. This might be attributable to the requirement of German language skills for participation and the low share of migrant families within the sample [40]. Therefore, generalization of the results to families with migrant background, low education, low SES or other socioeconomic risks is limited.

Concerning the assessment of ACEs, the parent-report on lifetime occurrence of ACEs provided plausible data, as with 6.5% for severe experiences up to 32% for the less severe forms, the data correspond to prevalence estimations in the general population [32]. Regarding the change in the occurrence of ACEs during the pandemic, we can only draw conclusions for the parents affirming lifetime occurrence. Furthermore, we do not have information about the absolute frequency in occurrence. It should be noted that the wording of the items used to assess the ACE subtypes mainly reflected low severity levels on established classification systems [23, 28]. The items included only few sample situations for the specific subtypes and the broad range of adverse experiences within the subtypes was not covered. Consequently, the data cannot be generalized to more severe cases of these subtypes. Overall, 11.6–37.7% of the sample reported a decrease in ACEs and 27.1–46.2% reported no change, suggesting that a significant number of families may be less affected or even relieved during the pandemic. Further analyses on parental resources and predictors of positive outcomes during the pandemic are warranted.

For reasons of parsimony, we did not assess pre-COVID-19 data on all measures, which prohibited analyses of change in general stress and mental health. Furthermore, pre-COVID-19 data were assessed retrospectively and we only have parent’s self-reported data and child-related outcome measures were not included in the study. Given the cross-sectional design of the survey, the results are solely correlational in nature and causal inferences cannot be drawn. Prospective studies are needed to replicate the predictive role of the risk factors for an increase in witnessing domestic violence and verbal emotional abuse, as well as the interplay between the predictors.

The results of this study confirm that the pandemic has had a profound impact on families. As school closures and social distancing from family and friends were among the most burdening restrictions for parents, great caution is necessary in applying these measures. Furthermore, additional efforts need to be put into helping families deal with the restrictions and “the new normal” [5]. Structural interventions targeting parents’ working conditions, such as offering parental leave for one parent, might provide relief during phases of lockdown. Targeted low-threshold online interventions aiming to activate parents’ intra- and inter-personal resources are warranted. Support for dealing with specific lockdown measures, such as not only help with home schooling, but also addressing parents’ own needs and their role as parents, may be a promising approach to specifically address parental stress and mental health. Child welfare services need to be available and should actively explore the presence of ACEs, including child abuse and neglect to provide targeted interventions [2]. Our findings from a population-based sample indicate that rates of child abuse and neglect increased in about one-third during the pandemic. Further research on high risk groups, in socioeconomic and psychosocial terms, is urgently needed.