Introduction
Anxiety disorders are among the most prevalent clusters of mental disorders in children and adolescents (Bandelow & Michaelis,
2015; Kessler et al.,
2005; Remes et al.,
2016). Individuals with anxiety disorders suffer from excessive worry and anxiety, which impairs their daily functioning, including their social life or academic performance (Quilty et al.,
2003). The disorder often takes a chronic course, meaning that, without successful intervention, it tends to prevail (Beesdo et al.,
2009; Keller et al.,
1992). In order to develop successful interventions, it is important to gain insight into the mechanisms that play a significant role in how anxiety disorders develop.
Anxiety runs in families (Beidel & Turner,
1997; Eley et al.,
2015; Hudson et al.,
2011). Children with parents who had or have an anxiety disorder have a two- to threefold risk for developing an anxiety disorder, compared to children of parents without anxiety (Lawrence et al.,
2019; Telman et al.,
2018). To reduce this increased risk of anxiety in the offspring it is crucial to understand how the anxiety transmission unfolds in the family. Many studies have assessed the impact of both genetic and/or environmental influences in the familial aggregation of anxiety disorders (Eley et al.,
2015; Gregory & Eley,
2007; Hettema et al.,
2001). Genetic transmission explains approximately one-third of the variance in child anxiety (Hettema et al.,
2001). This leaves the majority of variance unexplained and attributed to environmental factors, alone and in interaction with genetic factors (Gregory & Eley,
2007). This is in line with a children-of-twins design study, where the relative influence of genetic and environmental factors was investigated and showed that environmental factors predominantly accounted for the parent–child transmission of anxiety (Eley et al.,
2015). This calls for research that elucidates the mechanisms involved in this transmission.
Children can acquire fears via others, including parents (also known as social fear learning, Rachman,
1977; Olsson et al.,
2007) in two ways. Firstly, children can acquire fear of a novel stimulus via modeling: observing others being fearful towards that novel stimulus (also known as vicarious fear learning). Within the family context, children can for example learn to fear a novel animal as a result of being exposed to parents’ anxious responses to that animal (Murray et al.,
2008). This vicarious fear transmission starts as early as in infancy, as children start seeking out information about novel stimuli from parents between 10 and 14 months of age (so called social referencing, Feinman,
1982; Nimphy et al.,
2023b). Secondly, children can learn to be fearful of a novel stimulus when they receive verbal information from others about the threatening/anxiety-provoking properties of this stimulus (also known as verbal (threat) information learning or instructional learning, Olsson et al.,
2007; Muris & Field,
2010). During early childhood, when children learn to speak, verbal information about novel stimuli from parents becomes especially salient (Berman,
2004). Rachman suggests that verbal information from parents and peers during childhood is the origin of most fears in daily life (Rachman,
1977). Findings of cross-sectional and longitudinal studies that assessed origins and potential mechanisms underlying childhood anxiety suggest a role of parental verbal threat information (Fliek et al.,
2017,
2019; Ollendick & King,
1991). Moreover, in a review, Muris and Field (
2010) argued that there is “clear support for the notion that the verbal provision of threat information may have fear-enhancing effects in children”. Therefore, in this meta-analysis, we focus on this verbal information-learning pathway by summarizing the empirical evidence on child acquisition of fear and anxiety via parental verbal threat information.
Besides the line of research investigating whether parental verbal threat information is related to child anxiety (symptoms), biased cognition, or general fearfulness, two distinct lines of research have studied child fear acquisition of
specific novel stimuli via parental verbal threat information. The first line of studies focuses on typically developing children and employs experimental designs, where parents are instructed/trained to express specific verbal information towards novel stimuli (i.e., Aktar et al.,
2022; Bell et al.
2015; Remmerswaal et al.,
2013). The second line of studies relies on naturalistic observations of anxious and non-anxious parents with their children, that investigate the relationship between parental verbal threat information about a novel stimulus and child fear responses to the stimulus in daily life (i.e., Nimphy et al.,
2023a; Radanović et al.,
2021; Remmerswaal & Muris,
2011; Setiawan et al.,
2018; Uy et al.,
2022). While the first line may enable us to draw causal inferences, the second line aims to capture anxiety transmission in daily life. Our meta-analysis will examine both complementary lines.
Importantly, parent-to-child transmission of fear serves an evolutionary adaptive purpose, namely helping children in recognizing and avoiding dangerous situations, to enhance their chances of survival (Feinman,
1985). However, parents with an anxiety disorder, who experience excessive fear and have a tendency to overestimate threat (American Psychiatric Association,
2013), may inadvertently express anxiety—even in the absence of a threat. Parents with higher trait anxiety make more negative statements about a novel stimulus to their children than parents with lower levels of trait anxiety (Muris et al.,
2010). Over time, children of anxious parents may develop heightened attention to threat signals or interpret the signals in a more negative manner (Aktar,
2022; Creswell et al.,
2010). Consequently, the influence of fear expressions on a child’s acquisition of novel stimuli might be more pronounced in children of anxious parents than in those with non-anxious parents.
Besides the role of parental anxiety, previous studies also investigated child characteristics such as temperament, general fearfulness, or anxiety symptoms as a potential moderator in the parent-to-child transmission of fear, possibly strengthening the effect that parent verbal anxiety expressions have on their children’s fear acquisition (Muris & Field,
2010; Percy et al.,
2016). For example, child behavioral inhibition (BI) is an important risk factor for developing social anxiety (see Clauss & Blackford,
2012). Moreover, BI was proposed to be a marker of enhanced vulnerability to environmental stressors, including parental anxiety expressions (Belsky & Pluess,
2009; Ingram & Luxton,
2005; Nigg,
2006). Nevertheless, findings regarding the moderating role of child anxiety dispositions in parent-to-child fear transmission are mixed, allowing no firm conclusion about a potential moderating role (Muris & Field,
2010).
The impact of the parental verbal threat information on child fear acquisition of novel stimuli might also depend on the child’s developmental stage, with children being more affected by parental anxiety expressions in earlier stages. As children develop increasingly advanced cognitive and emotional abilities, they gradually become more emotionally independent from their parents as they age (Morris et al.,
2007). In line with this idea, one study that investigated the relationship between parental verbal threat information on children’s fear of Covid-19, suggests that younger children might be more sensitive to parental verbal threat information (Uy et al.,
2022). They argue that older children may have greater emotion regulation capacity, which might dampen the impact of parental verbal threat information, compared to younger children. Younger children might also depend more heavily on their parents as sources of information than adolescents. However, this empirical finding still has to be replicated.
Currently, knowledge on the parent–child transmission of fear through parental verbal threat information and the moderating roles of child temperament and parental anxiety is based on narrative and systematic reviews (Emerson et al.,
2019; Muris & Field,
2010; Percy et al.,
2016). These reviews have concluded that parent–child transmission of fear via verbal threat information is a significant factor contributing to child acquisition of fear and anxiety. More specifically, the reviews argue that fear acquisition as a result of verbal threat information can manifest in children’s fearful and anxious cognitions (Muris & Field,
2010; Emerson et al.,
2019), heart rate (Muris & Field,
2010) and avoidant behavior to novel stimuli (Muris & Field,
2010; Percy et al.,
2016). Taken together, the findings summarized in these reviews also suggest the effect of verbal threat information on children’s cognitions, implicit associations, and behavior is noticeable for up to 6 months (Muris & Field,
2010).
This
meta-analysis aims to combine the available evidence from empirical studies to calculate the effect size of the relationship between verbal threat information and child fear and avoidance of a novel stimulus. In line with previous studies (Muris & Field,
2010), we included studies that assessed child fear or anxiety with behavioral (i.e. avoidance), physiological (i.e. elevated heart rate), or cognitive (i.e. fear belief) measures. We expected that verbal threat information from parents is positively correlated with childrens’ fear or avoidance towards a novel stimulus. Furthermore, we explored whether the relationship between parental verbal threat information and child fear of novel stimuli is stronger for children of parents with higher anxiety levels/an anxiety disorder, children with higher levels of anxiety dispositions, and younger children. By gaining more specific insights into the verbal threat information pathway, we aim to improve our theoretical understanding on fear learning mechanisms in childhood and possible practical applications in prevention efforts.
Discussion
This systematic review and meta-analysis systematically assessed the role of parental verbal threat information in the parent–child transmission of fears. The meta-analytic findings show that parental verbal threat information about novel stimuli can increase child fear—even after a single exposure to these stimuli (Hedges’ g = 1.26). In line with our systematic review, the meta-analytic findings did not reveal a moderating role of parental and child anxiety levels, or child age in this parent–child transmission of fears to novel stimuli. Below, we discuss each of these findings in turn.
Child Fear
This systematic review and meta-analysis revealed that parental verbal expressions about novel stimuli are linked to and can increase child fear reactions to these stimuli. These findings align with social fear-learning models (Olsson et al.,
2007; Rachman,
1977), and corroborate parental verbal threat information as a causal social fear-learning pathway. The average effect size of the impact of parental verbal threat information on child fear was large (Hedges’
g = 1.01 in all studies, Hedges’
g = 1.26 in experimental studies only, and Hedges’
g = 0.70 in correlational studies only). A recent meta-analysis that systematically assessed the effect (size) of the modeling of parental
nonverbal fear expressions (also known as vicarious learning) in infancy (Nimphy et al.,
2023b) found small to medium effect sizes (Hedges’
g = 0.39). Hence, the impact of parental
verbal fear expressions about novel stimuli appears to be larger on children’s fear of these stimuli, compared to the impact of parental
nonverbal expressions. While it could be possible that verbal expressions of anxiety are more direct and impactful on children’s reaction than nonverbal expressions, it is important to mention that multiple studies that are included in the current meta-analysis manipulated
both parental verbal threat information and non-verbal expressions of anxiety. The combined impact of nonverbal and verbal expressions of fear might explain the stronger effect size for our meta-analytic findings on fear learning via parental verbal threat information compared to fear learning via modeling.
Furthermore, in the current meta-analysis, studies predominantly assessed child fear through self-report questionnaires. Exposure to parent verbal threat information might only/to a larger degree impact children’s subjective fear levels, rather than the physiological or behavioral fear components. Since fear indices are often unrelated (Bradley & Lang,
2000), if children report more fear of a novel stimulus, it does not necessarily mean that children would also behave more fearful of the stimulus. Studies that only assess one fear index may not be able to capture the entirety of children’s fear reactions to the novel stimulus. Hence it is important to stress that our conclusions concern self-reported fears, rather than robustly holding across multiple fear indices, i.e., physiological or behavioral indices.
In our meta-analysis, we found a larger effect size on the link between parental verbal threat information and child fear in the experimental studies than in the correlational studies. While experimental studies investigated fear transmission in a lab by manipulating parental verbal information, the correlational studies assessed the relationship between naturally occurring communication of parental threat information and child fear of novel stimuli in daily life. The larger effect size in experimental studies might be explained by the increased control in the lab setting and reduction of the influence of confounding variables. Taken together, findings of our systematic review and meta-analysis revealed that parental verbal expressions about novel stimuli are linked to and can increase child self-reported fear of these stimuli.
Child and Parental Anxiety Dispositions
Based on susceptibility models, we expected children with higher anxiety levels/dispositions to be more susceptible to environmental stressors such as parental verbal threat information (Belsky & Pluess,
2009; Ingram & Luxton,
2005; Nigg,
2006). Against expectation, our meta-analysis did not reveal a moderating effect of child anxiety levels or disposition on child
fear (Hedges’
g = −0.03). Our systematic review revealed that the majority of the studies (4 out of 5) found no significant effect on child fear, and one in the opposite direction (decrease in avoidance). It could be that child anxiety dispositions, such as temperament make children more susceptible to parental verbal threat information (or nonverbal fearful expressions) in early life, rather than in childhood (see Nimphy et al.,
2023b). Moreover, instead of making children more susceptible to parental threat information, child anxiety dispositions in childhood might increase fearful responses to novel stimuli independent of parental information. Lastly, the anticipated moderating effects might not have been detected due to the strength and intensity of the experimental manipulation in most studies. In real life, threat-related information might be less explicit and more ambiguous, compared to the experimental manipulations. For instance, it is possible that children’s anxiety disposition plays a stronger role in fear acquisition if children are exposed to more ambiguous and less explicit verbal information. More ecologically valid designs are needed to investigate the role of anxiety dispositions in parent–child transmission of anxiety (for example, see Muris et al.,
2010). Since our interpretation is based on only five studies, more research is needed to investigate whether child anxiety dispositions are a risk factor for heightened fear acquisition after exposure to parental threat information. Nevertheless, until now, our findings suggest that the link between parent verbal threat information and child fear acquisition is not stronger for children with anxiety dispositions.
We explored whether children of parents higher in anxiety disposition are more susceptible to parental verbal threat information. However, our meta-analysis did not reveal a moderating effect of parent anxiety levels on child fear (Hedges’ g = 0.04). Our systematic review revealed that the majority of the studies (3 out of 4) found no significant effect on child fear. The only study that did find an effect was on a behavioral index of child fear, thus it remains possible that the predominant focus on subjective indices made this less visible/apparent. However, only a limited number of studies investigated the moderating role of parental anxiety in the link between parental verbal threat information on child observed fearful/avoidant behavior. Further research is needed to establish whether children of anxious parents might show increased fearful and avoidant behavior after parental verbal threat information compared to children of less anxious parents.
Another explanation for the finding that children of parents higher in anxiety disposition were more susceptible to parental verbal threat information might be that parental anxiety disorders rather than parent’s anxiety levels make children more susceptible to parental verbal threat information. For example, it could be that the repeated exposure to verbal threat information in families with anxious parents creates an anxiogenic environment and contributes to the familial aggregation of anxiety (also named chronic exposure, Perlman et al.,
2022). Additionally, anxious parents may be more inclined to endorse or facilitate their children’s anxious or avoidant reactions to novel stimuli and may opt to remove their children from situations where they could get exposed to these stimuli (Fisak & Grills-Tacquechel,
2007). These anxious parents may less frequently use adaptive strategies, such as providing a comforting object, reacting supportively, or demonstrating other problem-solving approaches, for regulating their children’s emotions (Stifter & Augustine,
2019). Consequently, these parental behaviors could potentially over time diminish children’s sense of self-efficacy for self-regulation and elevate their fears (Stifter & Augustine,
2019) and contribute to the heightened fear learning. Another possibility is, that rather than making children more susceptible to parental threat information, parent anxiety dispositions might act more as a risk factor for increased fear and anxiety responses independent of parental verbal information. Given the limited number of studies assessing the moderating role of parental anxiety levels, more research is needed to investigate its role in the parent–child fear transmission in community and clinical samples. Until now, the findings do not support a moderating role of parental anxiety levels in fear acquisition after a single exposure to parental verbal threat information.
Child Age
We examined if the impact of parental verbal information on children’s fear reactions to novel stimuli may differ across children’s age. We expected that younger children, who may have lower emotion regulation capacity to deal with parental verbal threat information, compared to older children, are more sensitive to this information and show increased fear learning. However, our meta-analysis did not reveal a moderating effect of child age on child fear (Hedges’ g = 0.05). Prospective studies, which investigate the parent–child transmission of fears over time, might help illuminate whether fear learning via verbal threat information differs across age.
Clinical Implications
By investigating social fear learning mechanisms and how they might differ between healthy and at-risk families, we may gain more insight into which specific pathways and factors to focus on in treatment or prevention strategies. In our meta-analysis, we found a large effect of parents’ verbal threat information about novel stimuli on child fear reactions towards these stimuli, independent of child or parental anxiety levels. While fear acquisition via this pathway can be seen as an adaptive response to potentially threatening and novel stimuli, it could be that in at-risk families, the exposure to parental verbal threat information in day-to-day life occurs more frequently or intensely, which could strengthen the impact of this fear learning pathway. To prevent child anxiety development via this route, prevention strategies could incorporate psychoeducation on parent-to-offspring social fear transmission. Given the large effects found in the verbal threat information pathway, prevention efforts could potentially target the (repetitive) verbal communication of the parent.
As parental verbal threat information can lead to fear acquisition towards novel stimuli in children, listening to parents’ positive or confident information may reduce or prevent fear acquisition. A recent systematic review assessed if children’s positive modeling (of parents, experimenters, and peers) in experimental studies can reduce or prevent fear acquisition to novel stimuli (Krause & Askew,
2022). Although their conclusions rely mostly on modeling rather than verbal information/instructional learning, from a limited number of studies, it might still be a promising pathway to reduce or prevent children’s fear acquisition to novel stimuli. Ultimately, gaining insight into children’s fear acquisition in developmentally sensitive designs and investigating potential strategies to reduce or prevent parent-to-child fear transmission is crucial to inform treatment and prevention efforts.
Limitations and Future Directions
This is the first meta-analysis on the effect of parental verbal threat information about novel stimuli on child fear of these stimuli. While this meta-analysis provides a less biased summary of existing studies on the parent–child transmission of fears via verbal threat information, this study still embodies the shortcomings of the individual empirical studies.
First, the studies included in our meta-analysis mainly consist of WEIRD (Western, educated, industrialized, rich, and democratic) samples, specifically predominantly Caucasian families with moderate to high SES (socio-economic status). Considering cultural factors when investigating children’s perception and reaction to parental emotional expressions is crucial (review by Nielsen et al.,
2017; Raval & Walker,
2019). To enhance the generalizability of our findings, future research investigating this fear-learning pathway should include more diverse samples, and/or compare this fear-learning pathway across multiple cultural environments.
Second, caution is warranted for the generalizing of our findings to real life parent–child fear transmission. The majority of studies, which are included in this meta-analysis, utilized an experimental design and tested the verbal learning effects in lab-based artificial contexts. While conducting experimental studies on this parent–child fear transmission pathway allows for stronger conclusions on causality (Kazdin,
2021), it may limit the generalizability of the findings to experiences in daily life. Children’s experience with the novel stimuli presented in the lab might not generalize well to their experience outside of the lab. Furthermore, in experimental studies, parents are trained to display specific verbal and nonverbal expressions of anxiety, which might also not represent how parent show fear in daily life. While children can be exposed to one parent’s reaction in the lab, in real life they might get exposed to conflicting emotional reactions from two parents/individuals, successively or simultaneously. These conflicting reactions may alter the child’s response to the novel stimuli. Hence, future research should assess this fear-learning pathway in multiple contexts, as well as investigate naturalistic observations in families with children or parents with an anxiety disorder.
It is also important to note that the majority of studies included in the systematic review and meta-analysis assessed fear reaction to non-social stimuli, such as animals. Thus, more research is needed to assess children’s fear acquisition via parental verbal threat information to social stimuli. Moreover, in multiple studies from our meta-analysis, children were not actually exposed to the novel stimulus. Rather, some studies asked children how they feel about or would react to the stimulus in a hypothetical encounter, or in anticipation of being exposed to the stimulus. Future research could try to disentangle the different effects of parental verbal threat information on children’s fear reaction in anticipation or as a reaction in an actual encounter with the novel stimulus, utilizing multiple fear indices, such as cognitive, behavioral, and physiological indices measured at multiple time points.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.