Anxiety is an adaptive mechanism for human survival but it can also be pathological and overwhelming (Bennett et al., 2015). Anxiety disorders are prevalent among children and adolescents, affecting 15%–20% of children and adolescents (Beesdo et al., 2009), and evidence suggests these rates are increasing (Lebrun-Harris et al., 2022). Even though anxiety is a significant public health concern, many young people with anxiety go untreated, causing academic, social, and family impairments (Chiu et al., 2016). In these cases, anxiety problems can persist into adulthood, increasing the risk of depression, substance dependence, and truncated educational attainment (Woodward & Fergusson, 2001). As such, the need for effective identification and inclusion of their caregivers in research and intervention is crucial.

Anxiety in families causes substantial mental health problems for parents and carers, and many of them struggle with their mental illness issues. Anxiety transmission and reinforcement are linked to cognitive, affective, and behavioral aspects (e.g., Creswell et al., 2006; Micco & Ehrenreich, 2008). Research has been mostly centered on behavioral factors through the identification of common rearing styles, practices, behaviors, and strategies presented by parents to children with anxiety disorders. Overinvolvement (e.g., intrusiveness, overprotection, excessive commands, control and regulation of the child’s activities and routines), negativity (e.g., rejection, opposition, criticism, negative affect) and modeling of anxiety or anxious rearing (e.g., McLeod et al., 2007) are linked to children’s anxiety. Parental behaviors influence children’s anxiety through direct and indirect pathways, reducing mastery and autonomy and reinforcing a child’s fear and catastrophic and irresolvable view of the world (Bögels & Brechman-Toussaint, 2006).

Affective factors such as a parent’s own emotions and emotional reactions have also been well documented. Parents’ emotional responses impact children’s perceptions, leading to negative emotions and stress coping. Suppose parents react with rejection or devalue the child’s feelings. In that case, this might communicate that their emotions are not valid or appropriate, which can increase the child’s negative emotions and abilities to cope with stress (Siegler et al., 2011).

Parental Modeling and Influence on Children’s Anxiety

Although significant advances in the comprehension of how parental behavioral and affective factors influence children’s anxiety have occurred during the last two decades, parental cognitive factors have received less attention. According to social learning theory (Bandura, 1977), observation, modeling, and communication of information (e.g., verbalizations of attributions, thoughts, expectations, perceptions, appraisals, inferences, or beliefs about the child, themselves, and the environment) are crucial learning mechanisms during childhood (see Fisak & Grills-Taquechel (2007), for a review). Preliminary research evidenced that parents’ cognitive styles, including certain beliefs and negative expectations about their offspring, might play a role in children’s cognitive and behavioral vulnerability to anxiety (e.g., Creswell et al., 2011). Moreover, the information-processing model of parental behavior posits that parents’ beliefs about children’s behavior guide their responses in parent-child interaction and the strategies they use to promote competent behavior, influencing children’s maladaptive responses, such as social withdrawal (Rubin & Mills, 1990). Parental expectations and cognitions are presumed to influence parenting practices that promote threats and reinforce children’s anxious cognitions, namely overinvolvement or modeling of anxious and avoidant responses (e.g., Askew & Field, 2008; Casillas et al., 2021). For instance, the research found that adults with a low sense of control over difficult caregiving situations were more likely to use intrusive, overcontrolling parenting strategies when interacting with their own (Guzell & Vernon-Feagans, 2004) or an unfamiliar child (Bugental et al., 1989). Also, Rubin & Mills (1990) explored how fathers and mothers would react to hypothetical scenarios in which their pre-school-age children behaved in an aggressive and withdrawn manner. Aggressive behaviors likely elicit parents’ anger, disappointment, and embarrassment, leading to high and moderate power strategies. Withdrawal behaviors led to feelings of worry and puzzlement and low-power strategies such as information-seeking and planning strategies. Mothers reported using a greater variety of strategies than fathers, and more substantial socialization pressures were evident for girls in the social domain. Further, fewer strategies were referred to when parents attributed the cause of withdrawal to the child’s stable traits.

Parental Emotional Regulation and Anxiety

An essential element of children’s emotional development is learning to regulate emotions, reflecting their emotional socialization processes. Children see how their parents think, display emotions, interact with the environment, and repeat what they see their parents do to deal with novel or complex situations (Morris et al., 2007). In this context, parents’ anxiety might play a role in their cognitive style, emotional reactions, and strategies to cope with adversity that might reflect parental anxiety. Parents of anxious-disordered children are likely to be anxious themselves (e.g., Last et al., 1991). There is evidence supporting that anxious mothers interpret ambiguous situations related to the child in a threatening manner (Lester et al., 2009), have little confidence in their coping ability, and respond to stressful situations with avoidance (Craske et al., 1989), react to children’s emotions with a less supportive style (Hurrell et al., 2015), expect their children to be more anxious and avoidant (Cobham et al., 1999), and catastrophize more (e.g., Becker & Ginsburg, 2011) than less anxious parents. Studies have also suggested that the association between parent’s and child’s anxiety is mediated by parental beliefs about the harmfulness of the anxiety (e.g., Francis & Chorpita, 2009) and by parental expectations about their child’s estimate of threat in ambiguous situations (e.g., Creswell et al., 2006). More recently, research has gathered evidence that anxious adults have certain beliefs about themselves as parents. A study showed that anxious parents had a greater sense of responsibility for their child’s actions and well-being than parents who scored within the normal range for anxiety. Parental anxiety was also associated with more intrusive and less warm behaviors in parent-child interactions (Apetroaia et al., 2015). When behavioral aspects are considered, parental anxiety seems to be linked with more withdrawn and less productive engagement in dyadic interactions (e.g., Schrock & Woodruff-Borden, 2010), more critical (Whaley et al., 1999), reduced tolerance towards children’s negative emotions (Creswell et al., 2013), which did not occur with non-anxious parents. In conclusion, it has not been clarified yet if parental maladjusted cognitions and behaviors are more due to parental anxiety or child anxiety (e.g., Gallagher & Cartwright-Hatton, 2009). Nevertheless, these findings highlight the importance of improving our knowledge about the role of parental cognitions and beliefs in both parental and children’s anxiety.

Despite the noteworthy improvements in research dedicated to parental factors associated with child anxiety in the past two decades (see Emerson et al., 2019, for a review), some gaps remain to be filled. First, to date, the links between parental cognitions and parenting behavior in the context of child anxiety disorders remain largely unmapped (Bögels & Toussaint-Brechman, 2006). Parental cognitions, emotions, and behaviors have rarely been integrated, especially when parents are confronted with the child’s distress in their everyday lives. Second, the studies assume that parents of anxious children share standard features. No study has characterized or differentiated these parents according to their cognitive, affective, and behavioral characteristics. Third, fathers are usually excluded from this research area or assumed to share the same characteristics as mothers. Fourth, despite literature consistently reinforcing the bidirectionality of parents and children’s behaviors, few studies have focused on parents’ reactions and perceptions to their children’s distress and negative emotions (De Mol et al., 2013; Sameroff, 2000). Finally, most studies examined the relationship between child anxiety and parents’ cognitions or behaviors based on self-reports or observational tasks selected by the researcher and designed to adjust to theoretical constructs, deriving, fitting, and generalizing tendency patterns to overall families of children with anxiety. Most previous works focused on parental behaviors and parenting styles in the context of child expressions of anxiety (e.g., Suveg et al., 2008) but have not directly asked parents about their typical responses to child anxious emotions.

We used an exploratory qualitative research method with the previous limitations in mind (Creswell & Creswell, 2017). It fits our goal perfectly: first, to explore how parents tend to respond to children’s anxiety based on real-life experiences from parents through qualitative exploration, avoiding preconceived or hypothetical scenarios created by researchers. Besides affective (i.e., emotions and their intensity) and behavioral responses (i.e., behaviors), cognitive appraisals were also explored (i.e., cognitions). This integrated link between different types of responses helps us better comprehend the reasons underlying parental reactions and coping strategies to deal with children’s anxiety. We then conducted a qualitative-dominant crossover analysis (Kutscher & Howard, 2022) combining two multivariate techniques (MCA and a two-step cluster analysis) to understand the underlying patterns of responses and the agglomeration of data into groups. As such, the present study had the following objectives: (a) analyze the predominant cognitions, emotions, and behaviors manifested in response to a child’s anxiety, enhanced in a recent real-life situation; (b) identify critical dimensions to discriminate distinct profiles of parents of anxious children; and (c) explore and characterize distinct profiles of parents of anxious children. The derivation of significant dimensions and the cluster analysis helped identify parenting patterns in the context of children’s anxiety, understanding heterogeneity in parenting by identifying subgroups of parents with similar characteristics., and enhancing the effectiveness of parenting programs based on a more tailored and personalized approach.

Method

Participants

Sixty-five parents of 65 anxious children, 50.8% of whom were mothers (n = 33), participated in the study. The parent’s age ranged from 28 to 53 years old (M = 40.56, SD = 5.09). About the parental educational level, 35.9% of parents completed the 9th grade or less (n = 17), 42.7% completed high school or tertiary education (n = 28), and 20.3% finished college or higher graduations (n = 13). All parents were married or lived in common law and were the children’s biological parents, except one stepfather, who has lived with the child since she was born. Concerning socioeconomic status, 30.8% of families were included in low status (n = 20), 55.4% in medium status (n = 36), and 13.8% in medium-high or high (n = 9). Regarding the child’s characteristics, 56.9% were girls (n = 37) aged 8 to 12 (M = 10.02, SD = 0.97). The principal diagnoses of anxious children included generalized anxiety disorder (30.8%; n = 20), social phobia (26.2%; n = 17), specific phobia (29.2%; n = 19), and separation anxiety (13.8%; n = 9). Twenty-three children met the criteria for more than one anxiety diagnosis; one met the criterion for a comorbid secondary mood disorder, and five for a comorbid secondary behavior disorder (oppositional defiant disorder or attention deficit hyperactivity disorder).

Measures

The Child’s Anxiety Diagnosis

The Anxiety Disorders Interview Schedule for Children, Version IV (ADIS-C/P-IV; Silverman & Albano, 1996). This semi-structured interview initially probes for the main criteria of a diagnosis, followed by a more detailed assessment of symptoms, level of fear, avoidance, and interference. We used an adapted procedure that allowed us to interview the child and parent simultaneously, following the same procedure as Khanna and Kendall (2010). Questions were first addressed to the child and, afterward, to the parents. Also, intensity and interference lists were administered separately to parents and children to avoid contamination between responses. A diagnosis was attributed when both parent and child agreed on at least one anxiety disorder with a corresponding significant clinical severity rating (≥4) (see Silverman & Albano, 1996) or clinical judgment based on further questioning if there were discrepancies between informants.

Parental Interview

“Anxious to Know” was a semi-structured, in-depth interview based on an open-ended guide designed by the team members of the more extensive research from which this study was drawn (Beato, 2016). The interview script comprised several thematic groups of questions to explore meanings and perceptions regarding children’s anxiety, parental coping mechanisms to deal with it in daily life situations, and co-parenting. For this study, we have analyzed six questions retrieved from the script to explore the cognitions, emotions, and behaviors related to the child’s anxiety reported by parents. These questions were accompanied by one scheme following the format of functional analysis that assessed parents’ cognitions, emotions, and behaviors/strategies to cope with children’s anxiety. The interviewer wrote the parents’ answers in it. There were also emotion cards to help parents recognize emotions more easily.

The parents were asked to recall the last time their child experienced anxiety, fear, nervousness, or worry about something. In these types of emotional situations, parents think, feel, and behave in specific ways. They may not realize it at that time. We will describe that situation with the most details possible (e.g., when, where, how, and who was there) using this list of questions: 1. What caused your child’s anxious reaction? 2. What did you think in that situation? What came immediately to your mind at that moment? 3. How did you feel? How would you classify the intensity of that emotion on a scale of 0 (nothing) to 8 (a lot)? 4. What did you do in that situation? How did you cope or react? 5. (If parents identify specific coping strategies) Was it an efficient way to deal with your child’s distress? Try to classify that strategy’s degree of efficacy from 0 (not efficient) to 8 (completely efficient). 6. Was it difficult for you to confront yourself with your child’s distress? How much do you think it was on a scale of 0 (not difficult) to 8 (very difficult)?

Procedure

The National Data Protection Commission previously approved this study, the General Administration for Innovation and Curriculum Development from the Portuguese Ministry of Education, and the Ethic Committee of the Faculty of Psychology, University of Lisbon. This study derived from more extensive longitudinal research concerning parental factors in children’s anxiety (Beato, 2016). Our sample was obtained through a multistage process. First, all parents of children 8 to 12 years old from 12 Portuguese public schools, received a flyer about the project’s stages, aims, and ethical considerations and written Informed Consent through their headteachers. Parents were also invited to participate in diverse stages. Those who were interested voluntarily provided their contact information to the first author, expressed in the signed Informed Consent. Second, 987 children authorised to participate completed a questionnaire measuring their anxiety symptoms (SCARED-R) in the classroom administered by the AFB. Seventy-eight at-risk children scored above the cut-off point (i.e., percentile 85). The parents who provided their contacts were informed about these results and were invited to participate in a semi-structured interview (ADIS-C/P-IV) to elaborate a more precise assessment of anxiety status. Sixty-five children had at least one anxiety disorder with high harmful interference, recognized by both parents and children. At the last stage, the fathers and mothers of the children with a confirmed anxiety disorder(s) were invited to a second interview (“Anxious to know”). Verbal informed consent was obtained before the interview. Were included in this phase of the study fathers and mothers of children with an anxiety disorder as the primary diagnosis aware of the significant impairment of anxiety in children’s well-being and daily functioning and living in the same household. The first author individually administered these interviews as a clinical psychologist with expertise in psychological evaluation. The interviews had a medium duration of 45 min and were delivered in rooms with privacy provided by the schools and institutions of the community. For ethical reasons, free psychological treatment was provided to children and families who were interested in the clinical service of the Faculty of Psychology.

Analytic Plan

First, interviews were audio-taped, transcribed, and analyzed. Qualitative content analysis was performed using a hybrid approach, that is, we use both deductive and inductive processes (Zhou & Lin (2016)). The initial codes (e.g., father started to overthink the causes of the child’s distress”; “mother guaranteed that everything was going to be ok to calm the child immediately”) were categorized based on their similarity. Emerging categories were created, followed by pre-established categories (e.g., cognitions, emotions, behaviors, and strategies). The following stage required comparing the analysis’s similarities and differences to agree on the final sub-categories discovered (e.g., reassurance, problem-solving, and rumination). The study team debated, reworked, and bargained over them until the ultimate classifications were determined. Literature supports labels and properties, such as coping strategies adapted from Skinner & Zimmer-Gembeck (2007). The QSR Nvivo10 (2012) software was used to store, analyze, and organize the qualitative material. The first author of our study coded all the data from the respondents. Later, two psychology undergraduate students, blind to the anxiety diagnosis of the child, independently coded 50% of the transcript, respectively. The three coders attain 90% agreement, corresponding to high reliability. Ambiguities and disagreements were discussed and adjusted in team sessions. English translations of a few representative quotations were selected. Translated and revised by a bilingual researcher. Based on a qualitative-dominant crossover analysis (Kutscher & Howard, 2022), two quantitative techniques were performed using IBM SPSS Statistics for Windows, version 26 (IBM et al., New York, USA). Firstly, we performed a multiple correspondence analysis (MCA), part of a broader set of geometry data analysis techniques. MCA describes a multidimensional space characterized by the interdependence of qualitative variables or categories (indicators) with a graphical representation (Roose, 2016.) We investigate the typological structure of this space, enabling us to identify associations among the subcategories that comprised our three broad indicators (i.e., cognitions, emotions, and strategies to deal with the child’s anxiety) to examine the potential presence of subassemblies (homogeneous groups) with specific patterns (Rosa et al., 2016). We used three criteria to define the number of dimensions to retain: (1) eigenvalues larger than their mean, that is, in our case, μ ≥ μα = 1/Q, being Q the number of indicators 1/6 = 0.16, as recommended by Benzecri (1973); (2) inclusion of MCA dimensions with inertia above. 20. and (3) two-dimensional pictures of data were further explored to facilitate the interpretation. As the MCA provided preliminary information about the relations within the data, we subsequently performed a cluster analysis (CA) to group parents according to their similarity on the meaningful MCA dimensions—a two-step CA was employed to identify parent profiles based on MCA dimensions retained. The best cluster solution was chosen based on the log-likelihood distance measure, and Schwarz’s Bayesian Criterion (BIC) was used to select the optimal number of clusters. Overall clustering quality was assessed via the silhouette measure of cohesion and separation (Kaufman & Rousseeuw, 1990), with values of at least 0.20 indicating fair clustering quality. A noise handling of 30% was used as a criterion for outlier removal (Gamito et al., 2016). No cluster could comprise less than 5% of the sample (Costa et al., 2013). Afterward, we performed a second MCA using both the cluster membership (as a supplementary variable) and the indicators we included in the first MCA. Finally, chi-square association tests were performed to examine potential differences in our six leading indicators and the child’s sex and age among clusters. The adjusted standardized residuals were computed to identify the cells with the most significant deviations from expected values, using a cut-off of ±1.96 (Oliveira et al., 2021). All tests of statistical significance were conducted at a “p” value of 0.05.

Results

Qualitative Content Analysis

Before the cognitive-behavioral analysis of parents’ responses to a child’s distress, we explored the type of situation that enhanced children’s anxiety. Most parents identified situations that triggered physical threats (n = 21), such as the fear of the dark or dogs. General everyday worries (n = 17) included various topics that generated preoccupation (e.g., friendships quality, school results, unemployment of an adult relative). Social threat situations (n = 9) included performance anxiety in the classroom or fear of talking to strangers. Separation threat situations (n = 8) were also mentioned and included experiences where children faced or anticipated their parent’s separation (e.g., going with grandparents on vacations, parents picking them up later than usual after school). After that, we explored and organized parents’ responses into the following categories and respective sub-categories.

Content analysis revealed that most parental cognitions could be divided into five sub-categories: (a) thoughts centered on the child, (b) thoughts centered on parental responsibility, (c) ignorance, (d) external factors, and (e) normative reaction. Many parents (n = 27) had thoughts centered on the child’s characteristics, temperament, or symptomology when confronted with the child’s anxious responses in their everyday lives. Some parents justify children’s anxiety based on their perception of the child as fragile, vulnerable, or sensitive. According to one mother, “My son is a very susceptible child. I do not know where all this sensitiveness came from. If something goes wrong, he gets immediately affected (mother, 9-year-old boy). Other parents react more negatively and criticize the child, as evidenced by a mother saying, “He is a completely exaggerated boy. Sometimes it gets almost hysterical” (mother, 9-year-old boy). Taken together, these parental attributions and thoughts seem to reflect a common belief that anxiety is a stable characteristic of the child. Various parents also presented thoughts concerning their parental responsibility and role in the management of the child’s anxiety (n = 15). Many of these parents have feelings of guilt, worry, and anxiety. “She was so in despair with that school result, and I did not know how to help. I was working a lot, and I felt that I was not there with her. I was unavailable to my child and angry with my own life” (mother, 11-year-old girl).

Twenty-three parents presented other possible causes or attributions. Some parents reflect ignorance about what causes anxiety in their child and struggle to comprehend and empathize with the child’s fears and worries (n = 10). For instance, the mother of a nine-year-old boy said: “I cannot find any explanation for all this. It is not enjoyable. I must walk with him all over the house [because of separation anxiety]. Then I asked: ‘Why are you scared?’ However, he did not respond. He never does.s”. Several parents’ thoughts were centered on external factors that might be associated with the child’s anxiety, namely life-negative events such as problems that occurred in school or the peer group and the hospitalization of a brother (n = 8). For instance, one mother of a 12-year-old boy revealed: “He was frightened but has not told us the reason. I immediately considered the hypothesis that some colleagues were hurting him at school”. A minor number of parents interpreted their children’s anxiety as normative reactions adjusted to the situation or their developmental stage (n = 5). A father of a 12-year-old boy said: “I did not get scared. Everyone has bad dreams. It is normal to be afraid of a variety of things. I also have fears”.

Concerning parental emotions and their intensity, most parents identified anxiety and worry as predominant emotions (n = 28). Many also felt impotent or puzzled (n = 18). Ten felt sad or guilty, and eleven felt irritated or angry (n = 11). When parents were asked to classify the intensity of their emotions, most of them said the emotions were strong (n = 42), followed by those who referred to a moderate emotional level (n = 14), and, lastly, the ones with a minimum degree (n = 9).

Asked about the degree of their difficulty in handling a child’s anxiety, the majority found it very difficult (n = 42), followed by nine that found it moderately difficult and 14 that found it relatively easy. As such, most parents struggle with children suffering and negative emotions, and many mentioned that they had to self-regulate their own emotions in these moments. For example, one mother of a 12-year-old girl mentioned that “it is tough for a mother to see her daughter under such pressure, especially when she gets to a limit and cries. It is so difficult to assist and to try not to transmit stress.”.

During anxiety-enhancing situations for children, a considerable range of parental coping strategies were used by parents to respond to and deal with the child’s anxiety. The most commonly used strategy was providing reassurance, which refers to verbalizations intended to calm the child down and sometimes occurs as a response to the child seeking reassurance (n = 17). As indicated by the father of a 9-year-old boy, “I tell him that I and his mother are not okay in our marriage, but I have to guarantee that we will not get the divorce.”. Another common strategy was accommodation. This strategy includes attitudes of compliance with children’s avoidant and dependent behaviors and parental anxious reactions. A 9-year-old boy’s father said, “He asked me to go with him to the toilet because he was afraid, and I simply did.”.

Nine parents reacted with rejection, which manifests adverse reactions such as criticism, devaluation of the child’s feelings and worries, or adverse comments. For instance, the mother of a 9-year-old boy admitted, “I told him: Stop following me! You are just a pain in the ass.”.

Eight parents’ cognitive restructuring, modifying cognitions, distraction, acceptance, and minimizing fears and worries. As mentioned by the father of a 9-year-old girl, “We talk with her about the pros and cons of going to the dentist.”.

Emotional support is a strategy based on empathy, affect expressions, incentive verbal expressions, and praise, as referred to by five parents. “We tell him that he can count on us to help him. We are always here for him, whatever happens,” was an example mentioned by the father of a 9-year-old boy. Also, five parents used problem-solving strategies, including instrumental actions, planning, and developing specific strategies. “We told him that he was starting to sleep alone, and we planned. We first got him a little light, and things started to evolve,” exemplified the father of a 9-year-old boy. Overinvolvement was reported by three and represents controlling, intrusive, and overprotective behaviors toward the child. The mother of a nine-year-old boy provided an example: “I get terrified and preoccupied. Sometimes things just go wrong with them, and we start to protect them because we want the best for them. Put them living in a bubble.”. Non-involvement (n = 4), in the opposite direction, refers to situations where parents do nothing or do not react to the child’s distress (e.g., “I just did nothing and went to bed”; father, 10-year-old girl) or lack of interference associated with reactions of despair, confusion, helplessness, pessimism, or exhaustion e.g., “I was feeling impotent. I think we, as parents, do not know how to transmit serenity. It is a big problem” (mother, 10-year-old boy).

Parents also mentioned some additional responses. For instance, rumination (n = 2) describes a passive and repeated tendency to think about problematic situations. “When she told me the colleague was perturbing her, I wondered if I would ask the teacher. I wanted to help but did not want her to notice I was interfering. I could not stop thinking,” said the mother of a 10-year-old girl. Plus, the reinforcement of bravery reflects parents’ incentive and modeling of confrontation with threatening situations (n = 2). This might include creating small steps for the child to engage in feared behaviors and appropriately exposing them to feared situations. As exemplified by the mother of an 11-year-old boy, “I told him: I am here, and the lights are on. You just must enter the bathroom and put the shirt in the washing machine.”.

Concerning the perceptions of the efficacy of these strategies in response to a child’s anxiety, most parents found them helpful or efficient (n = 32). Some considered them undersized or even inefficient (n = 17), and 18 considered them moderately efficient.

Multiple Correspondence Analysis (MCA)

The characterization of the categories used in the MCA (N = 65) is presented in Table 1.

Table 1 Absolute Frequencies, Percentages, and Nonresponses for Indicators of the Study
Full size table

Concerning the representativeness of indicators’ categories, all presented homogeneous proportions, except for the intensity of emotions, parental coping strategies, and difficulty in dealing with distress, which presented a ratio of approximately 2:1. Whenever possible, we tried to recode the indicators into two categories (high vs. medium/low) to avoid underrepresentation. An exception was made for parental cognitions because we could not aggregate “parental responsibility” with “other causes.”. The limitation to two categories allowed for a clear-cut interpretation of the results because the two categories represent a clear, easy-to-grasp differentiation. Furthermore, fewer categories increase the stability of the results of MCA (Linting & van der Kooij, 2012). The children’s sex and age groups were treated as supplementary variables, that is, zero-weight variables, without interfering with estimating the dimensions in MCA. No non-responses were found. Before MCA, we examined the maximum number of dimensions to retain, following the recommendations of Carvalho (2008). The first solution included the maximum number of dimensions in MCA, seven dimensions (13 active categories minus six indicators), and the variable principal (VPrincipal) normalization method was applied. Only two were retained from the initial seven dimensions, accounting for 56.7% of the total variance, as shown in Fig. 1.

Fig. 1
figure 1

Representativeness of the initial dimensions based on inertia values

Full size image

The MCA has performed once again, but only in two dimensions. The second solution yielded two meaningful dimensions, accounting for a large portion of the adjusted inertia (Table 2). Both dimensions showed eigenvalues larger than their mean and inertia values >0.2, as recommended (Benzecri, 1973). The first dimension, “Emotional Reactivity,” presented an eigenvalue of 2.10, inertia of.36, and Cronbach’s alpha = 0.63 and was derived from parental cognitions, emotions’ intensity, and difficulty dealing with the child’s distress. The second dimension, capacity response,” is described mainly by how parents respond, namely their emotions, coping strategies, and efficacy. This dimension showed an eigenvalue of 1.35, inertia of 2.22, and Cronbach’s alpha coefficient of 28. It should be noted that, despite Cronbach’s alpha being relatively small (less than 60), this represents an acceptable value given the exploratory nature of our study and due to the heterogeneous constructs gathered to capture a two-dimensional representation of the data (Johnson & Wichern, 2007). Indicators with weights greater than 0.3 were considered more significant measures, discriminating between the dimensions. The indicator “Strategy’s efficacy” showed a low discrimination value for both dimensions (<0.10), but it was included in the dimension “Capacity to Response “for further analysis.

Table 2 MCA Discrimination Measures of our Indicators
Full size table

As depicted in Fig. 2, the indicators of the Intensity of emotions and Difficulty in dealing with distress were the most discriminant for “Emotional Reactivity”.

Fig. 2
figure 2

Indicators display and supplementary variables (measures of discrimination)

Full size image

The most discriminant indicators for the “Capacity to Respond” are hierarchically Parental Emotions, Parent coping strategies, and strategy efficacy. As shown, parental cognitions were a relevant indicator for the discrimination of both dimensions. In “Emotional Reactivity” (dimension 1), parental cognitions correlated significantly (transformed variables) with Parental emotions (r(63) 0.40, p < 0.001), Intensity of emotions (r(63) = 0.39, p < 0.001)and difficulty to deal with distress (r(63) = . p < 0.001).In “Capacity to Respond” (dimension 2), parental cognitions correlated significantly with Parental emotions (r(63) = 0.35, p < 0.001) and Difficulty to deal with distress (r(63) = −0.26, p = 0.036); Intensity of emotions correlated with Difficulty to deal with distress (r(63) = 0.66, p < 0.001). No other significant correlations were found. Only correlations >0.30 were considered to have meaningful practical significance. After, a topological schema for parental responses to the child’s distress was performed. Three different types of parental approaches emerge, as shown in Fig. 3.

Fig. 3
figure 3

Multidimensional space between scores for parental responses to the child’s distress and supplementary variables

Full size image

Cluster Analysis (CA)

After a two-step clustering approach with the object, the scores method (obtained via MCA) was performed to identify groups sharing similar characteristics within each of the identified dimensions (emotional reactivity and capacity to respond) and to support the profiles found by MCA. A three-cluster solution was reached (BIC = 61.74) with a ratio size (largest cluster: smallest cluster) of 2.00. According to Kaufman & Rousseeuw (1990), the overall clustering quality was good, with an average silhouette of.60. To validate the 3-clusters solution, the cluster membership was saved. A new MCA was performed using cluster membership as a supplementary variable as the 3 clusters were projected in the multidimensional plan.

The Fig. 4 led to the conclusion that clusters are consistent with the profiles found by MCA, highlighting three different profiles for our sample with different features, described as follows: (1) Reactive and ineffective parents - These parents tend to interpret their child’s anxiety based on diverse factors (e.g., external, normative), react with intense negative affect, use maladaptive strategies to cope with child ́s distress but also be conscious that they are not adequate to solve the problem, and have higher difficulty to deal with the child’s distress; (2) Unreactive and moderately effective parents – Parents that tend to consider that the child’s anxiety reflects child inner characteristics, use both adaptive and maladaptive strategies with moderate efficacy perception; and (3) Anxious and efficient parents – Parents who focus own their responsibility and parental role when the child has distress, reacts with intense anxiety, use both adaptive and maladaptive strategies but believe that they teneffective inffective in managing a child’s anxiety, and have higher difficulty dealing with the child’s distress. After, we examined whether the variables included in the MCA (Parental cognitions, Parental emotions, Intensity of emotions, Parental coping strategies, Strategy’s efficacy, Difficulty in dealing with distress, Children’s sex, and Children’s age group) did not differ significantly among clusters.

Fig. 4
figure 4

Multidimensional space between parental profiles (clusters), scores for parental responses to the child’s distress, and supplementary variables

Full size image

Table 3 shows a significant relationship between parental cognitions and clusters (χ2(4) = 47.71, p < 0.001). Post hoc tests indicated that Cluster 1 (Reactive and ineffective) had a higher percentage of parents with cognitions centered on other causes (Zadj = 2.00) than child characteristics and parental responsibility. Cluster 2 (Unreactive and moderately effective) had a higher percentage of parents with cognitions focused on child characteristics (Zadj = 3.2) than those focused on parental responsibility. (Zadj = −3.60). Conversely, Cluster 3 (anxious and effective) showed a higher percentage of parents who had more cognitions regarding parental responsibility (Zadj = −2.00) than cognitions focused on child characteristics (Zadj = −3.4) and other causes (Zadj = −2.3). Regarding parental emotions, results showed a significant association with clusters (2) = 21.55, p < 0.001. More detailed analysis indicated that Cluster 1 (Reactive and ineffective) had a higher percentage of parents with negative affect (Zadj = 2.1) than anxiety (Zadj = −2.1). On the other hand, we found the opposite pattern for Cluster 3 (Anxious and effective), that is, a higher percentage of anxiety (Zadj = 4.6) than negative affect (Zadj = −4.6). Cluster 2 (Unreactive and moderately effective) showed no significant percentage differences between parental emotions. Results also showed a significant relationship between the intensity of emotions and clusters χ2(4) = 46.16, p < 0.001. Post hoc tests indicated that Cluster 3 (Anxious and effective) had a higher percentage of parents that have high-intensity emotions (Zadj = 4.9) than low-intensity emotions (Zadj = −4.9). A similar pattern was found for Cluster 1 (Reactive and ineffective), with a higher percentage of parents mentioning having high-intensity emotions (Zadj = 2.3) than low-intensity emotions (Zadj = −2.3).

Table 3 Association the between Variables Included in the MCA and Clusters (profiles)
Full size table

On the contrary, Cluster 2 (Unreactive and moderately effective) showed a higher percentage of parents referred to having low-intensity emotions (Zadj = 6.8) than high-intensity emotions (Zadj = −6.8). Concerning parental coping strategies, results revealed a significant association with clusters χ2(2) = 10.83, p = 0.004. More detailed analysis indicated that Cluster 1 (Reactive and ineffective) had a higher percentage of parents with maladaptive strategies (Zadj = 3.3) than with adaptive strategies (Zadj = −3.3). In Cluster 2 (Unreactive and moderately effective), in contrast, more parents had adaptive strategies (Zadj = 2.2) than maladaptive strategies (Zadj = −2.2). Cluster 3 (Reactive and ineffective parents) showed no significant percentage differences between parental coping strategies. Results also indicated a significant relationship between dealing with stress and clusters χ2(4) = 45.99, p < 0.001. Post hoc tests showed that Cluster 1 (Reactive and ineffective) had a higher percentage of parents who reported more difficulties in dealing with distress (Zadj = 4.3) than those who reported low difficulties (Zadj = −4.3). A similar pattern was found for Cluster 3 (Anxious and effective) (Zadj = 3.0 and Zadj = −3.0, respectively). Conversely, Cluster 2 (Unreactive and effective parents) showed a higher percentage of parents who reported low difficulties in dealing with distress (Zadj = 6.8) than those who reported more difficulties (Zadj = −6.8). No other significant associations between variables included in the MCA and clusters were found (p > 0.05).

Discussion

Parent-child behaviors are interactive and bidirectional (Grusec & Kuczynski, 1997). Children’s anxious cognitions and behaviors might be shaped by parental behaviors such as overcontrolling or modeling anxiety (Becker et al., 2012). However, children’s distress might also impact parents’ perceptions, emotions, and behavioral responses. This cognitive-affective style and parents’ strategies when their children are anxious remain unclear (Mills & Rubin, 1990). Within an exploratory qualitative approach, the present study investigated parents’ main cognitions, emotions, and coping strategies parents used in a recent real-life situation that triggered the child’s anxiety.

Parents have identified recent triggering situations that elicited children’s anxious responses in naturalistic settings. Most events were associated with the perception of physical threat (e.g., fear of the darkness) and child’s generalized concerns about different issues in their everyday lives (e.g., friendship quality). In line with previous research, fears and worries, commonly associated with specific phobias and generalized anxiety, are highly present in children (Ollendick & King (1994); Mohammadi et al., 2020). This result also reflects that specific fears are more accessible for parents to identify because they frequently elicit intense, visible, and probably more appellative anxious reactions in children. Children’s various worries might be associated with reassurance-seeking and repeated questioning (Dugas et al., 2001), which increases parents’ awareness about some of those children’s concerns. The less mentioned were threatening social and separation situations. Although there is an elevated prevalence of social anxiety in children and adolescents (Kessler et al., 2012), social situations often occur in non-familial contexts less accessible to parents, such as school, public spaces, or leisure/social events. They frequently are difficult to be detected by others (Thompson et al., 2019). Separation situations might also be rarely mentioned because of family accommodation to the problem, avoiding exposure to separation moments and children’s distress (Weeks et al., 2023).

Findings also suggested that a child’s distress elicited various parental cognitions (i.e., thoughts, interpretations, attributions). According to cognitive-behavioral theory, parents’ interpretations and attributions regarding children’s behavior might predict parental adaptive or maladaptive behavior (Kil et al., 2020). Parents’ emotional socialization practices are influenced by their beliefs about their own and their children’s emotions, known as the parental meta-emotion philosophy (Gottman et al., 1996), which includes attributions, which represent causal beliefs about the reasons for situations or events. Our results reveal various cognitions based on the way parents perceive their child’s anxiety. For many, anxiety was interpreted as a reflex of children’s characteristics, that is, dispositional attributes. Literature has shown that these attributions to internalized emotions are related to children’s internalizing problems and both more dismissive responses to children’s display of unpleasant emotions and more negative parenting responses to children’s strong emotions (Riemens et al., 2023). Some parents perceive anxiety as an expression of external or environmental factors. As stated in previous studies, many parents view behaviors that are thought to be brought on by personality qualities as internal and stable, as opposed to behaviors that are thought to be more transient and brought on by external factors (Steding (2016)). The attributions about their responsibility for their child’s anxiety management and the need for control over the situation were also present. Previous studies have shown that highly anxious parents may have an increased sense of responsibility, believing they should be more in control when their child faces challenges. This inflated responsibility belief may lead to a similar inflated sense of responsibility for their children’s actions and well-being (Apetroaia et al., 2015). Various parents added other attributions, such as anxiety as a normative developmental factor, which might reflect a “positive attribution bias,” when parents attribute negative behaviors to characteristics of the situation rather than the child (Coplan, et al., 2002), lack of impairment in daily functioning or reluctance to perceive the child’s anxiety as problematic.

When confronted with children’s distress, parents’ most frequent emotions were related to anxiety. Many parents also felt impotent or puzzled. In both cases, parents’ feelings seem to be associated with some perception of a lack of control over a child’s distress. Independently of the type of emotions, parents classified them primarily as high. Being exposed to the child’s distress has an intense emotional impact on parents and probably requires emotional self-regulation skills from caregivers. An anxious parent may cope with difficult parenting experiences by avoiding, suppressing, or controlling their child, leading to internal distress. This can result in a lack of trial-and-error learning, leading to the development of self-efficacy and vulnerability to anxiety, as per Tiwari et al. (2008). Parents might have general temperamental characteristics based on high emotional reactivity and variability (see Larsen & Diener, 1987), interfering with the quality of emotional experiences when interacting with their children. Furthermore, when asked to quantify their degree of difficulty seeing their child in distress, most parents revealed that it was a painful experience for them as parents. Both parental feelings and the difficulty in assisting with children’s anxiety might play a significant contribution to how parents perceive the child and themselves and how they cope or react to it. For instance, research has shown that experience avoidance, the inability or reluctance to accept one’s internal distress, is more common among parents of anxious young people, which might trigger controlling, intrusive, or dismissing behaviors to avoid the child’s discomfort and, consequently, the parent’s internal distress (Tiwari et al., 2008).

Parents disclosed various coping strategies to deal with the child’s distress. The most used was reassurance, a usual form of accommodation for children’s anxiety (O’Connor et al., 2020). Parental accommodation, such as providing reassurance, allowing children to skip activities, modifying family routines, and adhering to rules around anxiety-provoking stimuli, can negatively reinforce child anxiety symptoms (Kagan et al., 2017). This behavior allows a child to avoid a feared stimulus, reducing distress and increasing the likelihood of future mental or physical avoidance. However, this accommodation may reinforce the child’s perception that the feared situation warranted their anxious response, causing habituation and increasing the likelihood of heightened anxiety in the future. (O’Connor et al., 2020). As suggested in previous literature, excessive reassurance-seeking represents a transdiagnostic maladaptive emotion regulation strategy on a long-term basis because it does not give the child the necessary resources or active tools to prevent the fears or worries from being repeated (Kane et al., 2018). However, it often has a short-term efficacy, which might explain why parents use it so often (Cougle et al., 2012). Other frequently applied strategies were rejection, cognitive restructuring, and emotional support, reflecting a wide range of behaviors exhibited by parents. The level of perceived efficacy of these strategies was mainly high, which might reflect a sense of control over children’s anxiety. Parental beliefs about parental efficacy reveal how people perceive their efforts as important. As stated in the literature, parents who believe their parental behavior positively impacts a child’s development are likely to invest more effort in their parenting than parents who do not (Simons et al., 1990). After identifying parental responses to the child’s distress in a recent situation, two significant dimensions were identified following cognitive-behavioral assessment and intervention principles. Specifically, parental cognitions, emotions, and coping strategies were used to explore the dimensions that allow organizing and discriminating the predominant characteristics of the participants. One dimension was centered on their capacity to respond, whereas another dimension reflected emotional reactivity, suggesting the importance of coping and emotionality in predicting the different parental profiles.

When these dimensions are considered to determine parental profiles that exist in a naturalistic way among the participants in the study, three main profiles allow for the identification of different groups with specificities. The first profile, reactive and ineffective, included parents with high negativity and reactivity associated with children’s distress. These parents might struggle with emotional socialization because they seem to have difficulty dealing with their emotionality, which impacts the way they cope; that is, they adopt more maladaptive strategies and become more inefficient in the management of a child’s distress. In this case, it is hypothesized that children’s anxiety might be reinforced through two complementary pathways: modeling of anxiety and negative emotions and inadequate rearing behaviors or coping strategies (Emerson et al., 2019). Plus, consistent with the literature, these parents seemed to make different attributions about the children’s anxiety, which might indicate the need for more psychoeducation in interventions with the family (Hiebert-Murphy et al., 2011; Rapee et al., 2000). This maladjusted interaction style might maintain the child’s problematic behavior.

The second profile included parents described as tending to attribute distress to the child’s inner temperament and characteristics. According to Bugental (1987), parents who attribute the causes of distress to the child’s traits and characteristics might react negatively and be controllable by the child (Snarr et al., 2009). In our study, although parents felt an overall negative affect, these emotions were low in intensity. They also perceive themselves as able to deal with children’s distress, which might represent good emotion regulation strategies and higher parenting self-efficacy as protective factors (Hamilton et al. (2015)).

The third profile presented a group of parents with appraisals of self-responsibility and concern about how they could help the child deal with distress, and the most predominant emotion was anxiety/worry. These parents have characteristics from previous research and literature (e.g., Apetroaia et al., 2015). Interestingly, despite these parents’ experience of anxiety symptoms and their cognition evidence of an increased sense of control over their child’s behaviour, most of them adopted adaptive coping strategies, such as problem-solving and emotional support. Consistent with other studies (e.g., Hiebert-Murphy et al., 2011), these parents also showed intense emotional levels and difficulty seeing their children in distress. However, their increased sense of responsibility for children’s distress management might be associated with perceiving them as an active part of the solution to the problem, which might guide this coping style. Hypothetically, this pattern may be guided by beliefs about parenting focused on parental responsibility for children’s well-being, reflecting a way to self-regulate their distress and reduce children’s anxiety in a short time, for instance, through reassurance, or parental own coping and cognitive styles (McLeod et al., 2007). This parent-child interaction pattern can lead to various consequences. Previous findings have shown that maternal anxiety and overcontrolling strategies increase the risk for a child to develop an external locus of control, thereby restricting opportunities for mastery of the environment or instilling a sense of dependence (Becker et al., 2012).

The current study must be conceived based on several limitations. Like cross-sectional design studies, the exposed findings may vary across time, context, and sample. Furthermore, we only analyzed the parental reactions to one recent situation, which might have partially compromised the ecological validity of our results. A longitudinal design using a similar sample would allow us to confirm our findings. Also, recalling past situations could have included memory bias and social desirability from parents. Further, the study had a reduced number of participants. Although the theoretical saturation was obtained for qualitative data, a larger sample would be needed to compare parental profiles according to relevant variables, such as parents’ and children’s gender and age. Our results cannot be extrapolated to the clinical population of children and their respective parents because our findings were based on a small sample recruited from a larger community. Despite the excellent clustering quality in our study, the results may need to be generalizable to other samples. Moreover, external clustering validation metrics could also have been used in addition to the silhouette coefficient as the unique, efficient internal measure for clustering validity assessment in IBM-SPSS. Lastly, parental anxiety was not controlled, which complicated the interpretation of the results, namely the association of specific profiles with the presence of anxiety problems in parents.

Implications and Conclusions

The current study has important strengths. Through the exploratory qualitative method, a realistic overview of parents’ experiences with their children’s anxiety was provided. Also, our findings provided preliminary hypotheses upon which future studies can build. According to previous parents’ cognitive biases, they may be associated with anxious symptoms exhibited by their child (Creswell et al., 2006; Wheatcroft & Creswell, 2007). Furthermore, our study intended to fill a research gap; addressing the link between parental cognitions, emotions, and specific parenting behaviors is needed (Francis & Chorpita, 2009). Even though meta-analysis suggested a moderate association between parenting variables and childhood anxiety, our findings suggest that many parenting variables that might play a more significant role concerning the familial transmission of anxiety remain unexplored (McLeod et al., 2007). The findings raise questions about parents’ perceptions of anxious children, how they feel, and how they behave when the child is distressed. The preliminary data raised here might have important implications for intervention and inspire new studies and reflections. As shown in previous literature, cognitive behavior therapy (CBT) is efficient in the treatment of child and adolescent anxiety disorders, particularly when parents are involved (e.g., James et al., 2013).

A detailed assessment with parents is necessary before implementing family-based interventions for child anxiety. For instance, it is crucial to analyze how parents perceive their children’s anxiety problems, how they feel or how overwhelmed they are when children are distressed, which strategies they use, and how efficient they think they are. Parents should be aware of their frequently unintentional assumptions about their children’s feelings to support their psychosocial adjustment (Riemens et al., 2023). Prevention and early intervention programs for children of anxious parents should then have a rigorous previous evaluation and be adapted to the type of families included in these programs to improve their success and prevent the maintenance of anxiety problems. Moreover, interventions might include the reduction of parental reinforcement of anxiety and avoidance, managing anxiety with children, and using therapeutic techniques such as psychoeducation, contingency management, cognitive restructuring, parental anxiety, negative affect management, collaborative problem-solving, modeling of coping to deal with stressful situations, etc. Given that many of these parents struggle with emotion regulation (personal and from their offspring), supportive emotional socialization might be particularly useful in the treatment, including validation and encouragement of emotional experiences and modeling adaptive emotional regulation techniques (Seddon et al., 2020). However, they should follow a comprehensive assessment and therapeutic approach (Riordan & Singhal, 2018) tailored to each family’s profile. Role-play techniques might represent practical, functional, and applicable procedures to work in parallel sessions or directly with the children.