Obsessional intrusive thoughts in children: An interview based study

Abstract

Objectives

We aimed to investigate the experience of obsessional intrusive thoughts (OITs) in a sample of children aged 8 to 10 years old and to test the main tenets of the cognitive model of OCD. Specifically, we assessed: (1) the prevalence of OITs experienced by young children; (2) their frequency and content; (3) the emotions they evoke; (4) the reasons why they are upsetting; (5) how they are appraised (6) and what control strategies they use.

Methods

Forty‐nine children (28 girls, 21 boys; mean age 9.1 years) from the community completed two self‐report questionnaires assessing anxiety and obsessive–compulsive symptoms. Children were interviewed using the Children's Anxious Thoughts Interview, which assessed their experiences of OITs.

Results

From the 49 participants, 71.43% reported having experienced at least one OIT. The most frequent contents related to harm and doubt. Of the total sample, 28.6% reported having experienced one OIT recently with at least moderate frequency; these participants reported higher anxiety and obsessive–compulsive interference, described feeling nervous and anxious when experiencing their OIT, and rated their OIT as highly important and distracting. The most frequently reported control strategies were cognitive – suppression, distraction and thought replacement.

Conclusions

Almost two thirds of community children reported experiencing OITs. The findings from this study provide preliminary support for the application of the cognitive model of OCD in children. Subtle differences from previous research with adults are discussed.

Practitioner points.

• Data supports the cognitive model of OCD in children.

• Community children appraise obsessional intrusive thoughts as being upsetting and evoking negative emotions.

• Community children use functional and dysfunctional control strategies.

• Future research should explore the usefulness of the CATI in clinical OCD samples.

INTRODUCTION

Obsessive–compulsive disorder (OCD) is characterized by the presence of obsessions and/or compulsions (American Psychiatric Association, 2013). OCD often begins in childhood; between one third and one half of adults with OCD report onset of symptoms in childhood or adolescence, with a mean age of OCD onset of 10.4 years (range 7.7 years – 12.5 years) (Stewart et al., 2004). The incidence of OCD in childhood has been reported to be between 1 and 3% (Stewart et al., 2004). Earlier identification and intervention has been associated with better response and recovery (Albert et al., 2019; Dell'Osso et al., 2010).

Cognitive conceptualizations of OCD in adults, have been shown to be valid for both clinical and non‐clinical samples of individuals (García‐Soriano et al., 2011; García‐Soriano & Belloch, 2013; Radomsky et al., 2014). Cognitive models of OCD purport three key assumptions (Obsessive Compulsive Cognitions Working Group, 1997;Rachman, 1998; Salkovskis, 1985): (1) that obsessions are extreme variants of unwanted obsessional intrusive thoughts (OITs), which are commonly experienced by the general population (Rachman, 1998; Salkovskis, 1985); (2) that the interpretation of OITs is a key factor in their development into obsessions and associated distress and (3) that negative appraisals of OITs will increase the individual's efforts to control those thoughts. The similarities in the presentation of OCD between adults and children (Mataix‐Cols et al., 2008; McKay et al., 2006) have led to the suggestion that psychological interventions derived from a cognitive model would be appropriate for children and adolescents (Barrett et al., 2008). However, only a few studies have specifically tested the key assumptions of the cognitive model with children (Barrett & Healy, 2003; Crye et al., 2010; Farrell & Barrett, 2006; Reynolds & Reeves, 2008; Rizvi et al., 2021).

Previous research has indicated that adults from the general population experience OITs, which are proposed to be on a continuum with obsessions (Berry & Laskey, 2012). This research supports the first assumption of the cognitive model of OCD for adults (García‐Soriano et al., 2011; Rachman & de Silva, 1978; Radomsky et al., 2014). However, the evidence for the validity of this assumption in children is scant. Although some studies have reported the experience of unwanted intrusions in child or adolescent samples (Allsopp & Williams, 1996; Sprung, 2008; Sprung & Harris, 2010), only two studies specifically analyse the presence of OITs (Crye et al., 2010; Igualada et al., 2007). Those two studies were specifically focused on the experience of adolescents, and reported that 90% of 12 and 13 years olds (Igualada et al., 2007) and 77% of 12 and 14 year olds (Crye et al., 2010) experience OITs. The difference in prevalence of OITs between these two samples may be due to differences in methodology. Igualada et al. (2007) used a self‐report questionnaire, whereas Crye et al. (2010) utilized an individual interview to ensure the comprehension and correct identification of intrusive thoughts. Both studies reported that the most frequently reported content of the OITs experienced by adolescents concerned doubts/checking, contamination and harm. There are no known studies assessing OITs in children younger than 12 years old, despite the fact that OCD can emerge in early and middle childhood (Heyman et al., 2001). Moreover, research suggests that children from 8 to 9 years old are aware that they experience unwanted thoughts that provoke unpleasantness, and are able to give examples of these thoughts (Flavell et al., 1998; Harris & Duke, 2006).

Negative appraisals of OITs are based on unhelpful and inaccurate beliefs about these thoughts, such as thought‐action fusion. Different cognitive OCD models emphasize the role of different beliefs in relation to negative appraisals (Clark, 2004; Obsessive Compulsive Cognitions Working Group, 1997, 2001). OITs that are negatively appraised are proposed to become more relevant, emotionally disturbing and interfering to the individual. This negative appraisal assumption of the cognitive model of OCD has received consistent support from research with adults (Belloch et al., 2007; García‐Soriano & Belloch, 2013; Morillo et al., 2007; Moulding et al., 2014; Obsessive Compulsive Cognitions Working Group, 2003, 2005). However, again, research conducted with children is scarce. In their study with non‐clinical adolescents, Crye et al. (2010) assessed how OITs were appraised. The participating adolescents reported that their OITs were moderately distressing (mean = 6, on a 1–10 scale), difficult to dismiss (mean = 6.7), moderately interring (mean = 5.4) and moderately acceptable (mean = 5.7). In a study with children who had an OCD diagnosis (7–13 years old), Barrett and Healy (2003) reported that the children appraised OITs more negatively than non‐clinical samples. When comparing clinical samples of children and adults with a diagnosis of OCD, children appear to report less distress and negative appraisals than adults (Farrell & Barrett, 2006). Different studies have analysed the presence of unhelpful beliefs about thoughts in clinical (Bacow et al., 2009; Coles et al., 2010; Libby et al., 2004) and non‐clinical (Cartwright‐Hatton et al., 2004; Muris et al., 2001) samples of children and adolescents. Those studies support the relevance of unhelpful OCD beliefs in children and adolescents; however, there is no consistency regarding the proposed role of these beliefs, nor if age mediates the relevance of these beliefs (Farrell & Barrett, 2006; Noorian et al., 2015; Reynolds & Reeves, 2008; Rizvi et al., 2021).

The cognitive model assumption that negative appraisals of OITs are associated with efforts to control those thoughts has been proven to be valid for adults with OCD (Belloch et al., 2009; García‐Soriano & Belloch, 2013; Moulding et al., 2014). In relation to children and young people, previous research indicates that non‐clinical adolescents (12–18 years old) use similar thought control strategies to adults, such as social control and reappraisals, although less frequently (Whiting et al., 2014; Wilson & Hall, 2012). Studies with clinical samples of children and adolescents (6–17 years old) similarly suggest that thought suppression (White Bear Suppression Inventory, WBSI) is used less frequently by young people compared to adults (Farrell & Barrett, 2006). However, these previous studies have focused on the use of thought control strategies more generally, and not specifically in relation to the control of OITs.

Although psychological interventions for OCD in children have been based on the cognitive model (Rosa‐Alcázar et al., 2015), there have been very few studies that have been focused on the validity of the assumptions of the cognitive model of OCD in children and adolescents. Most of the studies that have been conducted are limited by predominantly adolescent samples, with no research with younger children. Much of the existing research utilizes adapted adult questionnaires with no child‐specific measurements (Reynolds & Reeves, 2008) and fail to consider the unique features of childhood OCD. Therefore, there is no current evidence‐base for the validity of the cognitive model of OCD for children.

In order to test the key assumptions of the cognitive model of OCD for children, a starting point is to establish whether children in the general population experience OITs. Thus in this study, we aimed to investigate the experience of OITs in a sample of children aged 8 to 10 years old, and specifically test the main tenets of the cognitive model of OCD. Specifically, we assessed: (1) the prevalence of OITs experienced by young children; (2) their frequency and content; (3) the emotions they evoke; (4) the reasons why they are upsetting; (5) how they are appraised and (6) what control strategies they use. Based on the cognitive model of OCD, we hypothesized that: (1) intrusive thoughts with similar contents to obsessions will be reported by this sample of children from the general population; (2) these intrusive thoughts will provoke distress; (3) will be negatively appraised and (4) children will employ thought control strategies in response to those distressing intrusive thoughts.

METHOD

Participants and procedure

Study invitations were sent to the parents/guardians of 159 children attending third and fourth grade at two primary public schools in the metropolitan area of Valencia, Spain (schools selected on basis of convenience). To be eligible to take part, children needed to be aged between 8 and 10 years old. Children with a diagnosis of OCD (reported by parents) were not eligible to take part (exclusion criteria). Parents/guardians provided informed consent for their child's participation in the study, and completed a brief socio‐demographic data questionnaire relating to their child (i.e. gender, age, if their child had any health/mental health conditions and if this was the case to specify). Forty‐nine families provided informed consent. Students with informed consent took part in the Children's Anxious Thoughts Interview (described in the Measures section), which was conducted individually at school during school hours between April–May 2019 and October–November 2019. Before conducting the interview, the interviewer explained the characteristics of the study and asked children for their assent to participate. All children provided assent to take part. The interviewer was a postgraduate Psychology student with knowledge of the OCD cognitive model, and previous experience working with children. She was trained in the administration of the Children's Anxious Thoughts Interview through role‐play with two supervisors (authors), who are experienced researchers and clinicians in childhood OCD. The interviews lasted between 15 and 45 minutes. After the interview, children completed the self‐report questionnaires. Following participation, each child was given an information sheet with contact details of the principal investigator and a sticker as a thank you for their participation. As per the ethical approval for this study, parents/guardians were not given the results from the anxiety and OCD measures. However, general information regarding where to seek help was provided to children taking part. All procedures described were approved by the Humans Research Ethics Committee of the University of Valencia, Spain (registration code H1542364435368).

The final sample of 49 children (28 girls and 21 boys; mean age 9.1 years, range 8–10 years), included a small number of children with attention‐deficit/hyperactivity disorder (n = 2), behaviour disorders (n = 2) and emotional disturbance (n = 1).

Measures

Children's Anxious Thoughts Interview (CATI; see Appendix S1). This structured interview was designed by the authors for the purposes of this study, as there were no existing interviews for this age range and community samples, which focused on obsessional thoughts. The development of this interview was based on previously published self‐report questionnaires (García‐Soriano et al., 2011) and interview schedules (Llorens‐Aguilar et al., 2021; Radomsky et al., 2014), which were developed with similar objectives as the CATI but in adult populations. Importantly, the interview questions and administration procedures were designed specifically for children – with appropriate explanations, examples and use of cartoons to provide illustrative examples and stickers as indicators of item ratings. The interviewer asked each child a series of questions to explore whether they had ever experienced OITs in six content domains, as well as the characteristics of those intrusions. The CATI is composed of two parts.

CATI‐part 1 aimed to determine whether the child has ever experienced an OIT. The interviewer provided a definition and description of OITs with examples. Next, the interviewer provided a description of the six content domains one by one, and asked the child if they had ever had such a thought. The six content domains of OITs were: (1) contamination (i.e. a thought about germs or dirt being on you or around you, and that made you feel uncomfortable); (2) harm/worry; (3) aggression; (4) order; (5) doubt and (6) superstition.

If the child reported having experienced an OIT, then they were asked to describe an example. If the child was uncertain (i.e. they did not know if they had experienced it, or they reported a thought that did not fit the description), then two examples, accompanied by cartoons, were provided, in order to further clarify the question.

For each thought reported by the child, the interviewer asked a set of questions to gather a more detailed description (e.g. the last time they experienced the thought and what they were doing at the time), and in order to determine whether the thought meets criteria to be considered an OIT. Children were asked when was the last time they experienced the thought, and the frequency from ‘hardly ever’ to ‘every day’ (using a visual analogue scale ranging from 0 to 100 mm). Thoughts were considered to be relevant OITs if the participant experienced it recently (in the last 3 months) and frequently (rated at least 50 in a 100 mm scale).

CATI‐part 2 was administered only in relation to each confirmed OIT. If a child reported more than one recent and frequent OIT within a given content domain, then they were asked to choose the most recent, frequent and distressing OIT. If a participant reported no recent and frequent OITs, then the interview concluded. The focus on only recent and frequent OITs for CATI‐part 2 was intended to maximize the child's recall and ensure the thought was analogous to clinical obsessions.

In CATI‐part 2, children first described if they experienced the OIT as an image or verbal thought. Then, children were asked to appraise their OITs, using a visual analogue scale (100 mm), for two domains: (1) emotional reaction (sad, nervous, surprised, confused, angry, scared, upset, happy, excited) and (2) cognitive appraisals (Importance; It will come true; Distracting; and Easy to control). In addition, children were asked open questions about the emotional impact of the OIT and the reasons why they found it upsetting. Finally, children were asked about strategies they used to control the OIT. Participants were presented with a list of control strategies commonly associated with OCD and asked to mark with a sticker the ones they used (ever and most often) (i.e. Suppression; Distraction; Thought substitution; Do something special/compulsion; Tell parents/friends; Avoidance; Do nothing).

Spence Children's Anxiety Inventory (SCAS; Orgilés et al., 2012; Spence, 1998). The SCAS is a 44 item self‐report instrument that assesses symptoms of common anxiety disorders in childhood. Children rated the frequency of each item on a 4‐point scale. The SCAS includes a Total score and six subscales. In our study, the SCAS total score was used; Cronbach's alpha in this sample of children was of .90.

Children Obsessive Compulsive Inventory‐Revised (CHOCI‐R) (Shafran et al., 2003; Uher et al., 2008). The CHOCI‐R is a self‐report instrument that assesses the presence of obsessive–compulsive symptoms in children. There are two sections relating to compulsions and obsessions. There are 10 items in each section, which assess the presence of OCD symptoms (0–2), and offers a total symptoms score, a compulsions score and an obsessions score. In each section, there are additional six items, which assess severity and impairment on a five‐point response scale (0–4). In this study, the total symptoms (range 0–40) and total impairment (range 0–48) scores were analysed. The Spanish version of the instrument (Rosa‐Alcázar & Olivares‐Rodríguez, 2010) was used. Internal consistency in the current sample was α = .89 for the total symptom score, and α = .88 for the total impairment score.

Statistical analyses

All analyses were performed using SPSS software version 26. First, descriptive statistics were calculated for the full sample. Second, the sample was split into three subgroups based on the frequency of their reported OIT. A series of univariate ANOVAs were examined the differences between these sub‐groups on OCD symptoms and interference (CHOCI‐R) and anxiety (SCAS). Partial eta squared was calculated to indicate effect size (Cohen, 1988).

Any data from two open questions (i.e. why the OIT is upsetting; what control strategies were used) was independently coded by two of the authors, neither of whom had interviewed the children. Interrater reliability was calculated using the kappa statistic to assess the reliability of the categorizations, and interpreted following Cohen (1988). Interrater agreement was .87 for the coding of both open questions; disagreements were resolved through discussion until 100% agreement was achieved.

RESULTS

Preliminary analysis

Participants' mean score on the CHOCI‐R Total Impairment scale was 12.02 (SD = 9.37) and on the Total symptom scale 13.31 (SD = 8.25). The SCAS mean Total score was 27.98 (SD = 16.79).

Obsessional intrusive thoughts frequency, content and form

Table 1 provides a breakdown of the number of OITs by content domain. Of the initial 49 participating children, 35 (71.43%) reported at least one OIT. Only four children reported having experienced two types of OIT. No children reported more than two OITs. From the 39 OITs reported, the most frequent related to harm and doubt. All four participants who reported two OITs had OITs relating to doubt; three of them had an OIT relating to harm and one to superstition. Mean OIT frequency was 36.28 (SD = 32.68), ranging from ‘hardly ever’ (0) to ‘every day’ (100).

TABLE 1.

Number of OITs experienced organized by content.

OIT content	Number of OITs reported	Number of recent OITs	Number of OITs experienced with a frequency over 50	Number of recent (appeared in last 3 months) and frequent (over 50 in a 0–100 scale) OITs (over 50 in a 0–100 scale)
Contamination	0	0	0	0
Harm	18	15	7	6
Aggression	1	1	1	1
Order	2	2	1	1
Doubt	14	10	3	3
Superstition	4	4	3	3
Total OITs	39	32	15	14

From the 35 participants who had experienced at least one OIT, 14 (28.6% of total sample) reported having had one OIT in the last 3 months with a frequency of at least 50 (scale 0–100), the most frequent content related to harm. Most of these children (71.42%, n = 10) experienced their OITs as images; four children (28.57%) reported their OITs were in the form of words.

The description of OITs reported by children, as well as their content categorization, is included in Supplementary Material (Table A).

Participants were divided into three groups depending on the frequency of their reported OIT: those with both frequent and recent OITs (n = 14), those who reported OITs but were not frequent and recent (n = 21) and those who did not report any OIT (n = 14). Scores on the OCD (CHOCI‐R) and anxiety (SCAS) measures for each sub‐group are reported in Table 2, alongside between‐group ANOVAs and relevant post‐hoc comparisons. There were significant differences between the three groups on the measures of OCD (symptoms and interference, CHOCI‐R) and anxiety (SCAS). Post‐hoc comparisons indicated significant differences between the Frequent and recent OITs group and the No OITs group on CHOCI Total Interference (p > .001) and SCAS Total (p = .006), but not for CHOCI Total Symptoms (p = .66). Differences between the scores for Frequent and recent OITs group and Not frequent and recent OITs group were significant for CHOCI Total Interference (p = .002), CHOCI Total Symptoms (p = .140) and SCAS Total (p = .004). There were no significant differences between the Not frequent or recent OITs group and No OITs group on any of these scales.

TABLE 2.

Symptom scores on participants depending on their experience of OITs.

Measure	Subgroup			ANOVA
	Frequent and recent OITs n = 14)	Not frequent or recent OITs (n = 21)	No OITs (n = 14)	F (df)	p	ηp 2
CHOCI – Total Impairment	20.36ª (7.23)	10.57b (8.52)	5.86b (6.41)	13.37 (2)	>.001	.368
CHOCI – Total Symptoms	18.64a (6.96)	10.76b (7.42)	11.79a (8.62)	4.84 (2)	.012	.174
SCAS ‐ Total	40.93a (18.29)	23.10b (13.31)	22.36b (13,48)	7.39 (2)	.002	.243

Note: Data offered as mean (SD).

Superscripts a and b represent post hoc comparisons with Bonferroni correction.

Abbreviations: CHOCI, Children Obsessive Compulsive Inventory‐Revised; SCAS, Spence Children's Anxiety Inventory.

Emotional reactions and appraisals of obsessional intrusive thoughts

The following results relate to children who reported frequent and recent OITs (n = 14), and described their emotional and cognitive reactions in CATI‐part 2. The means and standard deviations for each of the emotions and appraisals (visual analogue scales) are reported in Table 3. The most common emotional reaction was getting ‘nervous’, followed by ‘surprised’, ‘scared’ and ‘upset’. When asked to freely describe the emotion elicited by the OIT, ‘anxiety’ was the most commonly reported emotional reaction, with five children describing feelings of ‘frightened’, ‘nervous’ or ‘worried’. Feeling ‘sad’ was reported by three children. The rest of the children described other negative emotions such as ‘overwhelmed’, ‘troubled’, ‘confused’, ‘compelled’, ‘strange’ or ‘bad’.

TABLE 3.

Participants (n = 14) emotional and cognitive reaction to their selected OIT.

	Mean	SD
Emotions
Sad	34.29	37.97
Nervous	75.29	21.03
Surprised	68.57	38.45
Confused	50.71	37.87
Angry	46.07	41.43
Scared	65.36	40.02
Upsetting	64.64	17.67
Happy	11.78	17.72
Excited	8.21	18.56
Appraisals
Important	79.64	32.38
It will come true	51.43	30.10
Easy to control	47.86	39.60
Distracting	71.43	28.04

Note: Data are offered in a scale from 0 to 100.

Regarding appraisals, the children rated their OITs as highly important and distracting. Children were moderately sure the thought could come true, and appraised that that it was moderately easy to control.

When asked why the thought was upsetting (open‐ended question), most children described a reason associated either with the fear of the OITs coming true (n = 8; e.g. ‘Because the building might have burnt down’) or with the emotional consequences of having the OIT (n = 5; e.g. ‘Because of the way it makes me feel’; ‘I want to be calm when I'm cooking’). One child described a reason associated with a compulsion. Answers are included as Supporting Information (Table A).

Control strategies for obsessional intrusive thoughts

Participants described what they did to get rid of the selected OIT; see Supporting Information (Table B). Each participant described one or two control strategies from a total of 18 listed. Most children reported using distraction (n = 6; e.g. go play with friends, do something fun) or replacing the thought with another one (n = 5; think of something else that's happy and not bad). Five children spontaneously described carrying out overt compulsions (e.g. order, checking, touching), and one participant indicated doing nothing to control their OIT. There was one reported strategy about which there was not enough information to determine if it was better considered as a mental compulsion or thought substitution (i.e. think it will not happen to me).

Finally, children were offered a list of control strategies and were asked to indicate the ones that they used (ever and mostly). See Table 4 for the list of the strategies in decreasing order of participant endorsement. Cognitive strategies were used most frequently, including thought suppression, and distraction, followed by replacing the thought with another one. Behavioural strategies (i.e. telling parents/friends; doing something special –compulsion; avoidance) were reported less frequently than cognitive strategies, although still endorsed moderately high.

TABLE 4.

Selected control strategies ordered by frequency (n = 14).

Strategy	Number of ‘Yes’ answers	Number of ‘Used most’ answers	Number (%) of ‘Yes’ or ‘Used most'
Suppression	9	4	13 (92.85%)
Distraction	8	4	12 (85.71%)
Thought substitution	5	5	10 (71.43%)
Do something special/compulsion	7	2	9 (64.29%)
Tell parents/friends	6	2	8 (57.14%)
Avoidance	5	2	7 (50%)
Do nothing	1	0	1 (7.14%)

Note: Participants could indicate more than one option.

Children were asked to rate the degree to which they felt that their strategies worked using a visual analogue scale. Children reported that their strategies were moderately successful (M = 43.84, SD = 34.28). Only three children rated their attempts above 50.

DISCUSSION

This pioneering study is the first to explore the experience of OITs in a sample of children, aged between 8 and 10 years, from the general population. We examined and provided evidence that supports the key assumptions of cognitive models of OCD as they relate to children. Children from a community non‐clinical population took part in an interview about their experiences of OITs (CATI). The reported measures of obsessive–compulsive and anxiety symptoms for these children confirmed non‐clinical levels, with lower obsessive–compulsive symptoms than those previously reported from clinical samples of children (Uher et al., 2008), and similar anxiety levels to those reported in non‐clinical samples of children (Spence, 1998).

Consistent with our first hypothesis, over 70% of children from this community sample reported that they experienced intrusive thoughts with obsessional contents similar to those experienced by adults. Whilst, this rate is lower than that reported using self‐report questionnaires in adolescents (90%) (Igualada et al., 2007), or adult samples (93.6%) (Radomsky et al., 2014), it is comparable to that reported from interviews with adolescents (77%; Crye et al., 2010). This noted difference between findings across the literature suggests that the rate of OITs in children and adolescents may be dependent on the method of assessment (e.g. questionnaire, interview). In this sample of children, the most frequently reported OIT content related to Harm and Doubt, which echoes previous findings with adolescent samples (Crye et al., 2010; Igualada et al., 2007), and indeed findings from adult samples (García‐Soriano et al., 2011; Radomsky et al., 2014). The current sample of children did not report any OITs relating to contamination. This finding contrasts with previous research with adults (García‐Soriano et al., 2011; Radomsky et al., 2014) and adolescents (Crye et al., 2010; Igualada et al., 2007). However, in a literature review of OITs in non‐clinical samples, Berry and Laskey (2012) indicated that contamination is less frequently reported than other OITs. Furthermore, previous research has also indicated that young children with an OCD diagnosis also reported lower contamination obsessions than older youth (Farrell et al., 2006; Højgaard et al., 2016). Within this study, the mean frequency of contamination OITs ranged from hardly ever to every day; this high oscillation was reflected in a high standard deviation. It is worth noting that the prompts relating to contamination thoughts within the CATI referred to ‘germs’ and becoming ‘sick’, and did not include examples associated with disgust or feeling dirty. However, these kind of examples could make it more difficult to distinguish between OITs and egosyntonic and developmentally appropriate disgust reactions. In addition, the data reported here were collected prior to the COVID‐19 pandemic, which may have impacted the rate of contamination thoughts in children more recently.

The current findings also indicated that of those children reporting OITs, most were experienced recently, but a lower percentage were experienced frequently. Specifically, 28.6% of the children reported having experienced OITs in the last three months with moderate frequency, and the most frequent contents were those related with harm. Previous research with adults suggest that OITs relating to harm are commonly reported as the most distressing OITs (Purdon & Clark, 1994). Within this study, many of the reported harm‐related OITs were associated with something bad happening to the child's parents. This finding may indicate a normative theme given the developmental stage of the participating children, but could also be a reflection of separation anxiety.

Thus, in general, the current results indicated that: (i) a high percentage of non‐clinical children experience OITs with contents similar to clinical obsessions, and (ii) a lower percentage of these children experience OITs frequently. These findings are consistent with the key assumptions of the cognitive theory of OCD. Furthermore, those OITs that were experienced more frequently were more analogous to clinical obsessions, and a continuum could be established between intrusions, frequent intrusions and obsessions (García‐Soriano et al., 2011; Morillo et al., 2007).

These findings also indicated that most of the children experienced their frequent OITs in the form of images, and only a quarter as a (verbal) thought. This finding contrasts with previous literature with adults that suggest a higher prevalence of (verbal) thoughts (66.7%) (Llorens‐Aguilar et al., 2021). This finding may relate to a higher reliance on imagery‐based processing (versus verbal) at younger ages (Burnett Heyes et al., 2013). Images (versus verbal form) have been shown to be more powerful in evoking emotions and have a relevant role in maintaining emotional disorders (Holmes & Mathews, 2010). Previous research on OITs in children did not evaluate the form of the OITs (Crye et al., 2010). Thus, the findings from this study indicate a potential key difference in the experience of OITs between children and adults, which is important to note when considering appropriate interventions and supports.

In line with our second and third hypotheses, our results indicated that children endorse negative appraisals of their frequent OITs. This finding is consistent with previous research with adults (Berry & Laskey, 2012) and adolescents (Crye et al., 2010), and with a key supposition of the cognitive model of OCD. Children in this study who reported recent and frequent OITs also experienced higher anxious and OCD symptoms, as well as a higher impairment associated with OCD symptoms than children reporting not recent and frequent OITs. This finding supports the idea that frequent OITs are analogous to clinical obsessions. Children in this sample reported that they feel nervous, scared and upset when experiencing frequent OITs (and not happy or excited). In addition, the current sample of children indicated that they appraise these OITs as important and distracting, with moderate endorsements for the appraisals that the OIT would come true, and was easy to control. Consistently, children also indicated that the OIT was upsetting either because it generated negative emotions or they feared that the thought could come true. Cognitive models of OCD propose that such negative appraisals (e.g. upsetting, interfering or fear it will come true) contribute to the development of OITs into obsessions. Interestingly, those children who report having experienced no OITs did not differ on their scores on the OCD and anxiety measures from those experiencing OITs which were neither recent or frequent. Thus, our data supports the continuum hypothesis in children, as it shows they can experience OITs without them being associated with disturbing effects in terms of OCD symptoms and distress.

Finally, as hypothesized, children in the current sample engaged in control strategies to neutralize the negative emotions or to prevent a dreaded situation associated with their OITs. In fact, data from open and closed questions indicated that only one child indicated doing nothing as a strategy. Moreover, as previously reported for adolescents (Whiting et al., 2014; Wilson & Hall, 2012), the reported strategies were varied and quite similar to those reported by adults but adapted to their age (e.g. play as a distraction, tell parents). In this study, we extended previous research by examining the use of strategies used specifically in relation to the reported OIT, and not general control strategies. The most frequently used strategies reported by the current sample of children were cognitive. When provided with a list of strategies, almost all children chose thought suppression, and more than fifty percent were also implicated in behaviours similar to compulsions. Interestingly, more than half of the participants identified as a strategy telling their parents or friends about the OIT, which may indicate the use of reassurance seeking, a common strategy in children with OCD (Ivarsson & Valderhaug, 2006). Furthermore, the current findings relating to strategies could suggest that shame and concealment are not widespread reactions to OITs in children of this age, at least in the non‐clinical population.

Almost all children in this study also chose distraction and replacing with a positive thought as strategies. These strategies are considered to be adaptive thought control strategies. Previous research shows that these strategies are scarcely employed in samples of adults with OCD (Belloch et al., 2009; Ladouceur et al., 2000); in fact, the reported use of distraction increases after successful interventions (Abramowitz et al., 2003). Children in this study were aware that their control strategies were not always successful, as only three children rated their attempts to control as moderately successful.

Limitations and future directions

The findings reported in this study are based on retrospective reports from the participating children, which may have been affected by potential failures of memory. However, previous research shows that children aged 8–10 years are able to accurately report past events (Pathman et al., 2013), and have the capacity to reflect on their internal experiences (e.g. meta‐cognition) (Harris & Duke, 2006). Furthermore, the CATI questions were designed to collect reliable information on OITs based on recent experiences (last three months), and contextual information and environmental cues were asked about to facilitate the child's recall. Even so, in order to account for the potential bias of retrospective reporting, future research could utilize methods such as ecological momentary assessments to assess the experience of OITs (Russell & Gajos, 2020). As the self‐report questionnaires were completed after the interview, the children's answers may have been biased by their prior experience in the interview. Future studies could control the possible influence of the order of administration through counter‐balancing, or administering measurements on different days. Furthermore, the current dataset is based on a relatively small sample of children, which is further reduced in some of the analyses that are centred on those participants who reported frequent OITs (CATI‐part 2). Despite this limitation, the findings are of relevance as the use of such a young sample and an interview allowed us to accurately identify the experience of OITs, and distinguish them from other negative repetitive thoughts (e.g. worries). Future research should replicate this study with larger samples, and with samples of children with a diagnosis of OCD.

The findings from this study have implications for clinical practice with children. The indicated support for the key assumptions of the cognitive model of OCD (i.e. that young children experience OITs analogous to obsessions) imply that interventions for OCD designed on the principles of this model would be applicable to the experience of children. These interventions may include the main components of cognitive‐behavioural therapy, such as exposure with response prevention, psychoeducation or cognitive restructuring (Rosa‐Alcázar et al., 2015). There were some differences noted between the experiences of the children in this study and those reported by adults in previous studies (e.g. imagery vs. verbal thinking) that could guide specificities in the assessment of intrusions and in the application of CBT to children. However, first these differences require further validation and replication with clinical and larger samples of children. Furthermore, the current findings support the potential use of psychoeducation as a preventative intervention for children at‐risk of developing OCD. Such psychoeducation interventions could help at‐risk children to conceptualize OITs as a universal phenomenon, and understand the impact of negative appraisals and control strategies on these thoughts (Chaves et al., 2021, 2022).

AUTHOR CONTRIBUTIONS

Gemma García‐Soriano: Conceptualization; funding acquisition; investigation; methodology; project administration; supervision; writing – original draft; writing – review and editing. Ángel Carrasco: Conceptualization; data curation; formal analysis; investigation; methodology; writing – original draft; writing – review and editing. Lisa Marie Emerson: Conceptualization; funding acquisition; investigation; methodology; supervision; writing – review and editing.

FUNDING INFORMATION

[Grant RTI2018‐098349‐B‐I00] was provided by Ministerio de Ciencia, Innovación y Universidades together with the Agencia Estatal de Investigación and the European Union European Regional Development Fund A way of making Europe.

CONFLICT OF INTEREST

All authors declare no conflict of interest.

Supporting information

Appendix S1:

García‐Soriano, G. , Carrasco, Á. , & Emerson, L. M. (2023). Obsessional intrusive thoughts in children: An interview based study. Psychology and Psychotherapy: Theory, Research and Practice, 96, 249–262. 10.1111/papt.12437

DATA AVAILABILITY STATEMENT

The data that supports the findings of this study are openly available in roderic.uv.es at http://roderic.uv.es/handle/10550/84321, reference number 1601.