**4. Discussion**

Our research produced the first national and the third European representative survey on adverse childhood experiences in Hungary according to which 25% (n = 293) of the Hungarian general population reported experiencing any childhood adversity before the age of 18 years with no gender difference; 5% (n = 59) of the respondents had four or more ACEs. The most prevalent form of child maltreatment was emotional (5%) and physical abuse (5%, n = 59); sexual abuse (1%, n = 12) was least prevalent. Parental divorce or separation (13%, n = 153), followed by household substance abuse (11%, n = 129) were the most frequent dysfunctional household conditions. The higher prevalence of ACEs among the younges<sup>t</sup> age group of adults may indicate an increasing awareness due to a more open public attitude and changing public opinion.

Our study is the first survey on adverse childhood experiences in a nationally representative adult sample of the Hungarian population; the first survey in any Central–Eastern European country; and the third such survey in developed countries that used a marketing research company for data collection. The European Commission has already established the feasibility of using marketing research/opinion polling agencies in health research: two reports were published on various aspects of the mental health of the population of EU member states in 2006 and 2010 by contracting companies to carry out representative surveys with multistage probability sampling and face-to-face interviews [49,50].

In order to interpret our results, data from the other two representative national surveys cited in the Introduction were considered. As it is shown in Table 7, the occurrence of the most frequent ACEs substantially varies in these countries, with Hungary having the lowest and Germany the highest prevalence.


**Table 7.** Prevalence of some ACEs in representative national samples.

The NESARC and BRFSS surveys have unique features (study design and implementation by public agencies funded by the federal governmen<sup>t</sup> of the US) based on phone interviews, repeated measures that may not be easily copied by other countries.

The strengths of our study include the use of an international standardized screening tool (ACE Score Calculator) in a nationally representative adult sample. Sampling and data collection were carried out by an experienced opinion poll company that used refined and tested sampling methods and had trained interviewers with experience in face-to-face data collection. This not only increased the reliability of data but was also cost-effective.

However, the study has limitations as well. The cross-sectional design and the retrospective nature of data collection limits the scope of interpretation; low awareness of the topic in the country probably increases recall bias, especially among older persons. However, since the prevalence of childhood maltreatment did not change significantly when those above the age of 80 years were removed from the analysis, and since the items of the ACE questionnaire are quite specific, not requiring interpretation, recall bias likely did not influence our results. The interference of dissociative defense mechanisms with recall cannot be excluded, but this bias cannot be avoided by any questionnaires. The conspicuously low frequency of childhood adversity among those with the lowest education merits further investigation.

In order to further probe the comparability of our data, the literature was searched for meta-analyses on the prevalence of child maltreatment and dysfunctional households reported by adults (Table 8). According to Stoltenborgh et al., global estimates of the prevalence in self-report studies were 22.6% for physical abuse, 36.3% for emotional abuse, 12.7% for sexual abuse (7.6% among boys and 18.0% among girls), 16.3% for physical neglect, and 18.4% for emotional neglect. These authors opined that the prevalence of child maltreatment seems to be largely similar across the globe. However, this statement is based mostly on research in western countries, mainly in North America and Europe [52].

Some meta-analyses were identified which focused on the prevalence of child maltreatment and dysfunctional households reported by children (Table 9).

The WHO Regional Office for Europe used 105 prevalence estimates from 50 community surveys to estimate the prevalence of sexual abuse as 9.6% (13.4% in girls and 5.7% in boys), physical abuse 22.9%, and emotional abuse 29.1% with no gender difference in the two latter types of abuse. The few studies that focused on neglect found high prevalence: 16.3% for physical and 18.4% for emotional neglect. As Table 8 shows, there are no differences between global and European prevalence estimates considering the majority of forms of maltreatment—the only exception being female sexual abuse with slightly lower prevalence in Europe [53]. The European report opined that prevalence estimates of child maltreatment would be higher in Eastern Europe. However, Gilbert et al. (2009) reported prevalences with a much greater variability in high-income countries: 3.7–16.3% of children experienced parental violence per year, 10.3% suffered from emotional abuse, and 1.4–15.7% suffered from neglect [27].


### *Int. J. Environ. Res. Public Health* **2019**, *16*, 1048


**Table 9.** A comparison of prevalence rates (estimates and measured data) of child maltreatment across the globe reported by children.

Our population survey measured a considerably lower prevalence of childhood adversity compared to population surveys in Germany or England. This, on one hand, probably reflects underestimation, supported by other data such as the homicide rate under 15 years of age in Hungary that was as high as 0.89 per 100,000 children or the fact that Hungary ranked 23rd out of 27 developed countries based on deaths due to abuse and/or neglect per 100,000 children under the age of 15 [53,59,60]; or that satisfaction with life among young teenagers was the second lowest in Hungary out of 21 developed countries in 2013 [61].

On the other hand, the widely different methods and measurements in various samples (community, clinical, and chance samples) selected by a wide variety of methods severely restricts the comparability of surveys carried out in different countries.

Third, the strong influence of culture, traditions, and religion on the treatment of children including what counts and what does not as maltreatment [62], as well as the possibility of false-negative statements due to psychological motives, must also be taken into account when comparing data on child maltreatment in various countries [32]. The ACE study was seminal in drawing attention to childhood adversity in the US and other developed countries [63], but this topic only recently has commanded attention in Hungary, reflected by the fact that no community-based data collection on childhood adversity had been carried out in the country.

Taking all these points together, the statement of Stoltenborgh and his coauthors (2015) that the prevalence of child maltreatment seems to be largely similar across the globe must be called into question [52]. Moreover, the opposite seems to be likely, which points to the importance of populationor community-based prevalence estimates measured by consistent methodology in each country.
