The purpose of this study was to investigate the interactive effects of concurrent adversity and prior maternal distress on later maternal distress and parenting quality. Specifically, we examined the interactive effect of COVID-19 related adversity and maternal distress assessed prior to the pandemic upon maternal distress and parenting assessed during the pandemic. A composite of maternal distress consisting of psychological distress, parenting stress and trait anxiety was used in this study, while parenting quality was examined using an observational measure of maternal sensitivity. As hypothesized, COVID-19 stress was directly associated with maternal distress assessed during the pandemic, but this was qualified by an interaction effect. Amongst mothers with higher levels of pre-pandemic maternal distress, COVID-19 related stress predicted greater maternal distress assessed during the pandemic, but this pattern was not observed amongst mothers with lower levels of pre-pandemic distress. This effect was robust, after accounting for maternal education, and the interaction remained significant when objective COVID-19 related stressors were considered. Importantly these results extend, but do not contradict, past research indicating that there was less of an increase in mental health symptoms amongst those who had higher levels of pre-pandemic distress [11]. As with other accounts, our examination of difference scores suggested less of an increase (or even, as in our case, a decrease, also see [8]) in distress amongst those with higher pre-pandemic scores. However, by additionally examining the moderating effect of COVID-19 perceived stress we were able to observe that this decline in distress was not the case for mothers who had high levels of pre-pandemic distress and felt high levels of COVID-19 stress. In fact, greater amounts of COVID-19 stress were positively associated with higher change scores only amongst mothers with higher levels of pre-pandemic distress. To our knowledge, no other group has examined such an interaction, which may be important in determining the complex way in which past and concurrent perceptions influence mental health.

In contrast, with regards to observed parenting, COVID-19 stress was not associated with maternal sensitivity assessed during the pandemic and there were no interactive effects with pre-pandemic maternal distress, even when pre-pandemic maternal sensitivity and maternal education were considered. Maternal sensitivity assessed during the pandemic was also not associated with maternal distress assessed during the pandemic. Hence, while results support our hypothesis concerning maternal distress, the hypothesis that parenting quality would be differentially affected by the interaction between pre-pandemic maternal distress and COVID-19 stress was not supported.

The current research’s study design helps to rule out a potential confound, present in much research examining the impact of prior mental health factors on the relation between stress and trauma on later functioning. That is, exposure to COVID-19 governmental restrictions was “assigned” to the entire study group – not just those with pre-existing risks – minimizing the likelihood that relations between pre-existing risk and post-adversity outcomes are due to an increased likelihood of risk begetting risk, as opposed to risk leading to greater vulnerability in the face of adversity. Moreover, the current work also considered relations between self-reported perceptions of COVID-19 distress and variation in objective stressors during the same time period and still uncovered a significant interaction between pre-pandemic maternal distress and COVID-19 stress (Additional File 6). Furthermore, to better explain the nature of this association we additionally conducted a mediational analysis examining whether risk begot increased perceptions of risk, as might be expected given research indicating that not only prior mood, but also change in experience during the pandemic, predicts subsequent mood [37]. In contrast to the significant findings concerning the moderating effect of pre-pandemic maternal distress on relations between COVID-19 stress and maternal distress assessed during the pandemic, post-hoc analyses found little support for mediational models. That is, there was no significant evidence that pre-existing distress led to the perception of more pandemic related distress and in turn greater post-pandemic onset distress. Rather, as indicated above, pre-existing distress may interact with subsequent concerns about routine activities (caring for others, housework, finances, hygiene, shopping for provisions, socializing, etc.) to impact subsequent distress. If replicated in a larger study, this would suggest that prevention programs should not only target well-being but also provide practical strategies for dealing with life’s daily hassles as they arise, and/or acknowledge that the skills needed to cope with major stressors and daily life may not always be interchangeable.

Recent research has reported high levels of psychological distress among the general population [7, 38] and parents [6, 39] during the COVID-19 pandemic, and that parents’ levels of depression and anxiety remained relatively consistent before and after the height of the pandemic [12]. The present study extends findings by revealing that while present mental health issues were moderately associated with COVID-19 related stress, within-individual increases in distress were more likely to occur among the mothers of pre-schoolers who had both higher levels of pre-existing distress and greater amounts of perceived COVID-19 stress. This demonstrates that pre-existing levels of risk moderates the extent to which adversity influences subsequent distress, such that those with lower levels of risk are buffered against adversity and are less likely to experience a decline in mental health.

The findings correspond with a previous study that found individuals with greater trait negative affect and detachment, or a history of stressful situations, were more likely to have mental health difficulties during the pandemic [38]. Past local retrospective research has also identified parental age, educational attainment, and parenting values as factors that influenced the extent to which COVID-19 influenced mental health [6]. The current study demonstrates that in addition to the factors above, prior psychological distress, anxiety and parenting stress also predisposes individuals to experience greater mental health distress in response to stress from the pandemic. Our findings align with the notion of individual trajectories in resilience (e.g., resilient, recovering, with chronic symptoms [40, 41]) following potentially traumatic events, with a substantial number of individuals being identified as experiencing chronic symptoms. However, much more than the objective severity of a given stressor (here, the pandemic and associated safety measures), biological and psychological influences seem to play a more important role in one’s individual trajectory towards resilience [42]. Recent evidence showed that individuals who are identified as developing chronic symptoms in response to a stressor have generally been experiencing poorer well-being and higher levels of anxiety [43]. In this context, an implication of the present findings is the importance of promoting and maintaining good mental health, as it may buffer against future difficulties coping with new stressful events as they unfold and enhance resilience at the individual level.

Maintenance of good mental health may be especially important for individuals working in professions that involve exposure to greater risk or high stress environments such as nursing, teaching, or the military. Good prior mental health may enable them to cope with stressful events despite occupational challenges and maintain resilience against developing symptoms of psychological distress. At the same time, however, this work also suggests that preparing individuals for the daily concerns that may arise in such challenging situations can be important, especially for those with higher levels of pre-existing distress. In the context of COVID-19 pandemic, special attention should perhaps be paid to healthcare workers, who experience high levels of psychological distress during and after virus outbreaks [44]. Similarly, this suggests that individuals with good prior mental health may be able to better cope with stressful life events such as divorce, illness, or death of loved ones. As such, social services that work with families who experience stressful events may want to pay attention to existing stressors or prior family functioning. Overall, from a policy perspective, the findings imply that better support should be provided to individuals who have pre-existing mental health difficulties or who are more likely to experience stressors, as they are more vulnerable to experiencing reductions in well-being in response to stressful events, even if their symptoms are not within the clinical range. These results may also suggest the need for routine prevention and support to be offered for those entering high stress fields.

In contrast to findings concerning maternal distress, there was no effect of COVID-19 stress on maternal sensitivity. This is unexpected because parenting quality is expected to change in response to environmental factors. For example, greater support for the primary caregiver led to the greatest improvements in child functioning [45], presumably due to changes in parenting quality. Maternal sensitivity has also been found to vary according to levels of socioeconomic status, parenting stress and internalizing symptoms [18]. However, past local research investigating the effects of COVID-19 found that parents with higher educational levels experienced overall higher levels of emotional well-being, compared to the less educated counterparts [6], as well as less notable decreases in income or cessation of outdoor time [46]. Hence, although our inclusion of relatively highly educated mothers may have decreased variation in the experience of COVID-19 related distress and so made it easier to examine the effects of relatively uniform adversity as a function of past well-being, it may also have decreased the average level of COVID-19 related distress experienced, and so limited our ability to detect any effects on parenting. Indeed, maternal sensitivity in this study was not associated with COVID-19 stress or maternal distress at both timepoints.

Accordingly, a limitation of this study is its small, predominantly Chinese, socioeconomically homogeneous sample, which could have obscured the association between COVID-19 stress and maternal sensitivity. Though we are not aware of any published quantitative accounts, amongst Singapore’s psychiatric community there is a sense that within clinical groups lower socioeconomic status related to greater declines in family functioning during the Circuit Breaker, perhaps in part due to being confined to comparatively smaller spaces with less opportunity to go outdoors (Helen Chen, personal communication, April 15, 2022). In addition, there is some reason to consider that reliance on social support as a buffer against stress may vary by ethnicity [47, 48], further limiting the generalizabilty of our findings, especially since COVID-19 restrictions intentionally limited social gatherings.

Second, it is possible that although all participants were exposed to the COVID-19 pandemic and its associated restrictions, individuals had unique experiences in their exposure to specific COVID-19 stressors that could have influenced levels of perceived stress from COVID-19. Still, analyses incorporating additional data about the COVID-19 experiences continued to reveal a significant interaction between pre-pandemic distress and perceptions of COVID-19 stress, though post-hoc tests were no longer significant (see Additional File 6). If possible, future research should account for objective sources of stress that correspond to areas in which participants rated perceived stress, to reduce the likelihood that higher perceived stress stems from increased exposure to a greater number of domain relevant stressors. The current research group will be following up with larger Singaporean cohorts in order to answer some of these questions, as well as to follow up on how COVID-19 experiences may influence future child development.

Finally, our questionnaire considered levels of COVID-19 stress at three different time periods. Mothers were asked to retrospectively report on COVID-19 related stress for the first two time periods and to consider recent and concurrent COVID-19 related stress for the third. Thus, our questionnaire may have been subject to memory bias [49]. Indeed, emotional state and perceptions concerning the passage of time during the COVID-19 pandemic may be linked [50]. We therefore conducted post-hoc analyses examining the potential association between the amount of time elapsed between the Circuit Breaker and questionnaire administration and COVID-19 stress at each time period (i.e., February to March 2020, April to May 2020, June 2020 onwards), as well as with overall COVID-19 stress and maternal distress assessed during the pandemic. We did not uncover significant associations. This, coupled with the fact that the various restriction phases in Singapore were clearly titled (e.g., Circuit Breaker, Phase One) defined specific group size limits [51], and were clearly communicated to the public via television and social media channels, suggests that memory biases may not have played a large role in explaining post-pandemic onset distress (see Additional File 8). Still, questionnaires prospectively querying stress at each phase may have been able to deepen our understanding of how interactions between past- and present- mental health and perceived stress unfold.