Prevalence of Child Sexual Abuse in Spain: A Survey Study

Introduction

Child sexual abuse (CSA) is one of the more prevalent types of child maltreatment. An illustrative study by Finkelhor et al. (2014), in which telephone surveys were conducted among 2,293 15- to 17-year-old adolescents, revealed a lifetime prevalence rate of 26.6% for girls and 5.1% for boys. A more comprehensive picture of the problem worldwide was provided by Stoltenborgh et al. (2011), who conducted a meta-analysis on the prevalence of this phenomenon using 217 studies published between 1980 and 2008 including 331 independent samples with a total of almost 10 million participants. The results indicated a global prevalence rate of 11.8% for the total population, with again a gender difference from 18.0% in girls versus 7.6% in boys.

The aforementioned meta-analysis by Stoltenborgh et al. (2011) has also indicated that there are quite substantial differences in the prevalence of CSA across various countries and continents. These differences may have to do with cultural characteristics leading to differential disclosure patterns about CSA experiences, but also to real differences in the actual prevalence of this phenomenon. Therefore, it seems important to quantify CSA in specific countries, as this could further clarify the role of the cultural context.

However, one of the difficulties most widely reported by studies is as a matter of fact the high variability of these prevalence rates. One of the main causes of this variability lays in the different conceptualizations and definitions of this kind of abuse. It is important to mention that the definitions of this phenomenon have been quite changing (Rueda et al., 2021). For example, some researchers restrict CSA to sex-related activities involving a child and an adult, while others also include forced sexual actions between a child and a peer. Further, there are marked differences in the severity of forced sexual acts, varying from being touched on private body parts to acts that involve some kind of physical intercourse. Strictly speaking, the term child sexual abuse has never been and probably never will be unequivocally defined (Haugaard, 2000), and we will have to settle for increasingly precise approximations to help us narrow down the magnitude of the problem.

The literature has indicated that CSA is consistently associated with a variety of psychopathological conditions (Hailes et al., 2019), including internalizing (anxiety and depression) and externalizing (oppositional behavior and aggression/delinquency) problems (Lewis et al., 2016; Papalia et al., 2017), but also other negative consequences, such as re-victimization (Hébert et al., 2018), drug use (Lewis et al., 2016; Tonmyr & Shields, 2017), and suicide attempts (Ng et al., 2018), with women being more likely to sustain injuries compared to abused men. Regarding the type of abuse, the evidence revealed that penetrative abuse was linked to severe injuries, while non-contact abuse was associated with less severe injuries (Amado et al., 2015; Merrill et al., 2001). This impact may be deeper than indicated so far, involving also transdiagnostic mechanisms that converge to aggravate the risk of undesirable outcomes such as deficits on emotion regulation, avoidance, or insecure attachment (Noll, 2021). Emotional regulation, together with other transdiagnostic variables such as sensitivity to anxiety or intolerance to uncertainty, is one of the mechanisms most closely linked to the etiology of psychopathology (Pineda, 2018; Pineda et al., 2018).

From the offender’s perspective, there are several causal explanations proposed for child sexual abuse. One proposed explanation is sexual deficit approach. This suggests that the offender has sexual dysfunction or “blockage” and seeks to satisfy his sexual needs through child abuse (Hudson & Ward, 2000). Another suggested explanation is to have more affinity for children than for adults that could be partially linked to sexual deviance approach, which holds that offenders have atypical sexual interests and behaviors that lead them to seek sexual contact with children (Finkelhor, 1984).

There are also theories that point to cognitive and emotional factors, such as the distorted cognitions, which holds that offenders view children as sexual objects and justify their behavior through a series of rationalizations that minimize or deny the harm inflicted on children (Burn & Brown, 2006; Ward, 2000). Other theories include that of child abuse in the offender’s own history, which suggests that offenders may have been sexually abused as children and, in turn, sexually abuse children (Drury et al., 2019).

Finally, some researchers have also highlighted the importance of situational factors, such as availability of children, access to children, and opportunities for contact (Holt & Massey, 2013). These factors may interact with the offender’s personal factors to increase the risk of child sexual assault. It is important to keep in mind that these causal explanations are not mutually exclusive, and that child sexual assault is a complex phenomenon that may have multiple interrelated causes and factors. In fact, meta-analytic reviews have identified more than 700 risk factors that could be summarized in up to 35 domains. Of these, the most relevant have been previous victimization of the child and/or family members, previous victimization of the child distinct from child maltreatment, previous or concurrent forms of child maltreatment in the child’s family environment, and parental history of child maltreatment victimization. Other risks identified were related to parental problems (e.g., intimate partner violence), parenting problems (e.g., poor quality of the parent-child relationship), a non-nuclear family structure, family problems, or child problems such as chronic mental/physical illness (Assink et al., 2019).

In the Spanish context, López et al. (1995) were the first to conduct a study on the occurrence of CSA. These researchers used a stratified proportional random sampling method to create a representative sample of 1,821 adult participants, who were then personally interviewed and assessed by means of a self-report questionnaire. The results indicated prevalence rates of 22% for women and 15% for men. At about the same time, de Paúl et al. (1995) conducted a smaller study using self-report data of a sample of 426 university students in the Basque Country, yielding prevalence rates of 14.8% for women and 9.7% for men. More than a decade later, Pereda and Forns (2007) published the results of another study in 1,033 university students from the University of Barcelona. In this study, it was found that 19.0% of women and 15.5% of men reported they had experienced at least one sexual abuse event during their childhood. It is important to note that the data presented by Pereda and Forns were collected more than 20 years ago (i.e., in 2001 and 2002). Since then, despite the great impact of CSA on individual lives and also more broadly our society (e.g., costs of mental health care), no new studies have been published on the prevalence of the phenomenon in Spain. Note further that the latest studies have been conducted in samples that were recruited in specific regions of Spain (de Paúl et al., 1995 in Basque Country and Pereda & Forn., 2007 in Catalonia) and hence may be less generalizable to the Spanish population as a whole.

Similar to the older study of López et al. (1995), the present study used a sample with participants from all regions of the country. The aim of this work was to provide an update on the prevalence of CSA in Spain. To provide a comprehensive picture of the prevalence of the CSA, we will present data using varying defining criteria such as the prevalence ratio and the contrast of the observed probability of prevalence of CSA categories with a constant by computing the Z score.

Method

Sample

A total of 1,485 participants completed the full survey. This study relied on online proxy/VPN detection software (https://iphub.info) to ensure Spanish respondent localities. Those participants who answered from countries outside of Spain (n = 162) were ruled out from the sample. Most of the participants (70.4%) were studying or had studied at the university level. The final sample consisted of 1,323 participants (confidence level = 95%, margin of error ± 2.694%) of whom 955 were women (72.2%) and 368 were men (27.8%), who had completed questionnaires on paper (n = 168, 12.7%) or on the internet (n = 1155, 87.3%). The age of the participants was between 18 and 70 years (M = 28.40, SD = 10.31).

Measures

Traumatic Life Events Questionnaire (TLEQ; Kubany et al., 2000)

The TLEQ was applied, which assesses exposure to a wide range of potentially traumatic events, including child sexual abuse. The questionnaire evaluates the experiences of a total of 22 traumatic life events as defined by the DSM-IV (American Psychiatric Association [APA, 1994]). With regard to sexual abuse, the questionnaire makes it possible to classify events into three different types according to the actors involved: (1) child sexual abuse I (CSA I) – sexual contact between a child under the age of 13 and another individual 5 years older; (2) child sexual abuse II (CSA II) – sexual contact between a child under 13 and someone of similar age through the use of force, threats, deception, unconsciousness, or abuse of authority; and (3) child sexual abuse III (CSA III) – sexual contact between a 13- to 18-year-old adolescent and another individual, against their wishes or without their consent. It is also possible to classify events into two different types depending on their severity: (a) child sexual abuse (CSA) – sexual contact without oral, anal, or vaginal intercourse; and (b) severe child sexual abuse (SCSA) – sexual contact with oral, anal, or vaginal intercourse. The adoption of this conceptualization of abuse, based on clinical principles, disagrees with the legal concept in the Spanish context, where the age of consent was increased from 13 to 16 years, and does not define a minimum age difference between victim and perpetrator to rule out sexual activity among peers’ difference, but will depend on the proximity in age, degree of development or maturity of the parties (Organic Acts 1/2015). In order to obtain comparable results, we apply the Spanish translated version of the TLEQ (Pereda & Forn., 2007). The authors of the original version reported the degree to which participants responded consistently, calculated with the Cohen’s kappa, obtaining values between .40 and .60 for 11 of the items and more than .60 for the rest, reflecting a moderate and substantial degree of agreement respectively (Kubany et al., 2000; Landis & Koch, 1977).

Procedure

Participants were contacted through social networks or at different university campuses. The procedures conformed to the ethical standards of the responsible human experimentation committee and were in accordance with the guidelines of the World Medical Association and the Declaration of Helsinki and approved by the Institutional Review Board (or Ethics Committee) of Miguel Hernández University (Reference DPS.JPR.03.17).

All the questionnaires were completed anonymously, and the only data collected was the IP from which the questionnaire was completed for the participants in the online modality or the university to which the participant belonged in the case of the pencil and paper questionnaires. All participants read and accepted the conditions of this research, giving their informed consent to participate voluntarily.

Data Analysis

Sociodemographic characteristics of the sample were analyzed. Prevalence and proportion of different types of CSA suffered by sex were also computed. Next, the prevalence ratios (PR) were calculated, which is analogous to the risk ratio (RR) and it is obtained by dividing the prevalence of the presence of CSA by the absence of CSA. The distribution of this effect size is asymmetric, as are the odds ratios, with values greater than 1 indicating that the category is more present in suffering CSA (and the numerical value, the times it is more likely to be more present) and values less than 1, the opposite. The magnitude of PR was interpreted as moderate if PR ≥ 2.47 (if the 95% CI of PR includes 2.47 the size equals 2.47), large if PR ≥ 4.25, and greater than large if PR ≥ 8.82, which roughly correspond to a d of 0.50, 0.80 and 1.20, respectively (Arce et al., 2015). Finally, the study of the prevalence of CSA categories was completed by computing the Z score for the contrast of the observed probability of CSA with a constant (expected probability of CSA defined as unusual prevalence, .01 and trivial prevalence, .05; Fandiño et al., 2021). The magnitude of the effect was interpreted by estimating the effect size with Cohen’s h (h = 0.20, small size; h = 0.50, moderate size; h = 0.80, large size; h = 1.20, more than large size; Arce et al., 2015; Cohen, 2013). The same analyses were calculated to contrast the observed probability of victimization by perpetrator of CSA with an unusual (.01) and trivial effect (.05).

Results

The percentages of the different types of abuse for men and women are shown in Table 1. According to our results, the overall prevalence rate for the milder form of CSA was 18.6%, with percentages being 9.2% in men versus 22.1% in women, reflecting a statistically significant gender difference. More severe sexual contacts that involved some types of intercourse (anal, oral, or vaginal) were less prevalent (3%) and not different between men and women (gender-specific percentages being 2% and 3%, respectively).

Considering the different types of CSA, women were more likely to report experiences of sexual abuse without intercourse as compared to men. For types of CSA that involved oral, anal, or vaginal intercourse, a gender difference was found for CSA III: women more frequently report such a severe form of sexual abuse when they were aged between 13 and 18 years than men. In fact, this was the only comparison that reached a large effect magnitude (PR ≥ 4.25), while the remaining were of moderate magnitude (PR ≥ 2.47), such is the case for CSA III and CSA II without intercourse, and almost moderate magnitude for CSA and CSA I without intercourse.

The occurrence of overall CSA without intercourse is significantly more prevalent than the occurrence of unusual (> .01) and trivial (> .05) events, while the more severe CSA, that sexual contact with oral, anal, or vaginal intercourse, is more prevalent than the probability of encountering an unusual (> .01) event (see Z_.01 and Z_.05 in Table 2). In sum, the probability of the overall occurrence of CSA is significant, i.e., it is more likely than the likelihood of very unlikely events. Additionally, the effect size (i.e., the magnitude with which CSA exceed an unusual and trivial presence, .01 and .05, respectively) ranges from moderate (> 0.50) to large (> 0.80) in CSA without intercourse and small in CSA with intercourse (> 0.20).

Moreover (see Z_.01 and Z_.05 in Table 2), across the different types of CSA, a more than unusual contingency (> .01) was recorded for CSA II and III when sexual contact was under 13 or above, and did not involve intercourse, as well as for CSA III with intercourse. The effect size was moderate (> .50) for CSA II and III without intercourse, lower than small (< 0.20) for CSA III with intercourse. In addition, contingency was more than trivial (> .05) for all types of CSA not involving oral, anal, or vaginal intercourse, with a moderate to large effect size (0.50 < h < 0.80).

Finally, as can be seen in Table 3, the probability that the perpetrator of CSA was a friend was significantly more likely than the fact that this event was trivial (> .05) or unusual (> .01), with a moderate to large effect size (0.50 < h < 0.80), followed by the probability that the perpetrator was a relative, with a small to near moderate effect size (0.50 < h < 0.80), and the probability that the perpetrator was a stranger, which has a small magnitude (> 0.20). The probability of the perpetrator being the couple alone was significantly more likely than the occurrence of unusual events, with a nearly small effect size (< 0.20). Finally, we can see how the abuse perpetrated by the father, mother, or caregiver was not significant in comparison with the occurrence of a random event.

Discussion

It is now more than 25 years ago since López et al. (1995) published the first data on the prevalence of sexual abuse in Spain and about 15 years since the last study published on this topic (Pereda & Forn., 2007). The main objective of the present investigation was to update the data on the prevalence of child sexual abuse in Spain. According with our results, the current prevalence rate is 9% for males and 22% for females. Compared to previous studies, it seems that the prevalence rates have not substantially changed during the past decades (de Paúl et al., 1995; López et al., 1995; Pereda & Forn., 2007). These results are well in line with figures of sexual abuse documented in meta-analytic studies, relying on similar samples in terms of culture and level of socioeconomic development (Assink et al., 2019; Pereda et al., 2009; Stoltenborgh et al., 2011).

Concerning gender differences in reporting experiences of CSA, a consistent finding is that women are at greater risk for unwanted sexual encounters as compared to men (Pereda et al., 2009; Stoltenborgh et al., 2011). However, by disaggregating the abuses by type and severity, we have noted that the likelihood of severe sexual abuse involving some form of intercourse is equally likely for both genders, whereas women are more likely to report instances of sexual abuse without intercourse compared to men. However, we have not found other studies that indicate that there are no differences between men and women in reporting serious sexual abuse. Perhaps this phenomenon being the least frequent and the most serious, this may be the reason for the absence of differences according to the sex of the victim. In short, it should be emphasized that it is important to approach this debate with sensitivity and to recognise that CSA is a very sensitive and distressing issue. Thus, the findings presented in this study are based on reported experiences, and it is essential to recognize that reporting rates can be influenced by several factors, such as social stigma, cultural norms, social expectations, gender roles, or individual differences in disclosure tendencies (Braithwaite et al., 2010; Seematter-Bagnoud et al., 2010). These influencing factors imply that women may be more inclined to report non-penetrative CSA, which could be due to several reasons, such as, for example, that they may feel more comfortable talking about experiences that do not involve intercourse (Lemaigre et al., 2017). However, other studies indicate just the opposite, that there is evidence that in the specific case of men who have been sexually abused, cultural factors could hinder the disclosure (De Jonge & van der Knaap, 2013; Guerra et al., 2021).

Additionally, the results indicate that for the severe forms of abuse involving oral, anal, or vaginal intercourse (referred to as CSA III), women more frequently report such experiences when they were between 13 and 18 years old compared to men. This gender difference was found to have a large effect size, while other comparisons showed moderate effect sizes. These data are consistent with other studies that indicate that men who have been sexually abused, in fact, have more difficulty to reveal CSA (Guerra et al., 2021) and those that report that adolescents are less likely to reveal severe CSA (involving intercourse) (Manay & Collin-Vézina, 2021), due to various cultural factors that could hinder the disclosure.

A relevant and novel approach of this study was to carry out the examination of the prevalence of CSA categories by computing the Z score for the contrast of the observed probability of CSA with a constant (expected probability of CSA defined as unusual prevalence, .01; and trivial prevalence, .05; Fandiño et al., 2021). This constant means that when the observed probability is around .05 or greater and the zeta is significant and positive, the effect is non-trivial, while if it is not significant the effect is trivial. If it is significant and negative, it is contrasted with the constant p < .01, which is an unusual effect, and if it is positive and significant, this indicates that the results mean that the effect is considerable (in this case because it is CSA, as is the case with deaths, where + of .01 is already considerable). Given the above, we discuss the occurrence and significant of different types of CSA based on the presence or absence of intercourse. Thus, it is worth noting that the occurrence and significance of overall CSA without intercourse is more prevalent than trivial or unusual events. Furthermore, severe forms of CSA involving oral, anal, or vaginal intercourse are more prevalent than unusual events. The effect sizes ranged from moderate to large for CSA without intercourse and small for CSA with intercourse. Consistent with previous studies, the incidence is lower in more severe cases (Merrill et al., 2001).

Regarding the perpetrator in CSA cases, in the present study, friends or acquaintances were more likely to be perpetrators, with a notable effect size. Relatives (other than the victim’s parents) showed a moderate association as perpetrators, although the effect size was slightly smaller than for friends. Strangers had a small effect size as perpetrators, indicating a relatively smaller association compared to friends and relatives. The occurrence of the couple alone as perpetrators was significantly more likely than unusual events, with a nearly small effect size. However, the abuse perpetrated by the victim’s father, mother, or caregiver did not show a significant difference compared to a random event. Consequently, the perpetrator appears to be mostly a friend or acquaintance of the victim. These results partially coincide with previous studies, where the victim’s relationship to the perpetrator also appears to be associated with CSA (Australian Bureau of Statistics, 2005; Mwangi et al., 2015). In the example, data reported by the Australian Personal Safety Survey indicated that a substantial amount of abuse is committed by someone known, often related, to the child, being the most common perpetrator of CSA a male relative (other than the victim’s father/stepfather), followed by a family friend, an acquaintance or neighbor, another known person, the victim’s father/ stepfather, and, finally, a stranger (Australian Bureau of Statistics, 2005).

It is surprising that a modern society where the greatest asset should be its children maintains these unacceptable rates of sexual abuse. For example, being a victim of CSA has been associated with a higher risk of becoming an abuse perpetrator (35% of cases versus 11% of non-abused), in what has been called the cycle of abuse (Glasser et al., 2001). Future research should analyze, from the framework of the Life History Theory, the relationship between exposure to negative events and antisocial behavior (Del Giudice et al., 2015; Sousa et al., 2011).

These data should be interpreted with caution due to several limitations. Firstly, as this is a convenience sampling, there may be an over-representation of affected people, since those who have suffered abuse in childhood may be more motivated to participate in this type of study. However, the data in this study are consistent with the main source of evidence in this regard, i.e., international meta-analytic studies, which indicate that between 18.0% and 20.0% of females and around 7.0%-8.0% of males have been sexually abused during childhood (Pereda et al., 2009; Stoltenborgh et al., 2011). Finally, the collection technique, a perfectly validated self-report, which collects information retrospectively, can generate biases due to problems with recall. However, the nature of the phenomenon and the subjective experience of it, discourage any other method of collecting this type of information.

In summary, these data indicate that the rate of sexual abuse remains virtually unchanged despite the supposed evolution of society in terms of literacy, health, and well-being. These data indicate the imperative need to develop active policies to protect children in the first world to prevent child sexual abuse, not to mention the need and relevance in other developing or underdeveloped countries, belonging to the so-called third world, where child protection is negligible. In Spain, on December 28, 2019, the preliminary draft of the Organic Law on the Comprehensive Protection of Children and Adolescents from Violence was approved, aimed at improving the protection response to children and adolescents within the framework of the Convention on the Rights of the Child, and is geared towards meeting the objectives of Agenda 2030, in particular Goal 16 (“Promote just, peaceful, and inclusive societies”) and target 16.2 (“End child abuse, exploitation, trafficking and all forms of violence and torture”). Different agencies, such as Save the Children, CGCOP, etc., insist on the need to process the law urgently.

Despite new and bold laws, campaigns and political actions that seem to show greater concern for the overall well-being of our citizens, especially children, who are the adults of the future, the rates of child sexual abuse are unacceptable and represent a public and general health problem on a par with many others that have greater popularity, attention, and funding. In fact, no organization, not even the WHO or the European Union, considers child sexual abuse among its priorities as a general health issue to track. Therefore, it is time to prevent child sexual abuse in a determined, planned, and prioritized way.