Prevalence of Psychopathy in the General Adult Population: A Systematic Review and Meta-Analysis. Ana Sanz-García, Clara Gesteira, Jesús Sanz and María Paz García-Vera. Front. Psychol., August 5 2021. https://doi.org/10.3389/fpsyg.2021.661044
Abstract: The main objective of this study was to systematically and meta-analytically review the scientific literature on the prevalence of psychopathy in the general adult population. A search in PsycInfo, MEDLINE, and PSICODOC identified 15 studies published as of June 2021. Altogether, 16 samples of adults totaling 11,497 people were evaluated. Joint prevalence rates were calculated using reverse variance heterogeneity models. Meta-regression analyses were conducted to examine whether the type of instrument, sex, type of sample, and country influenced prevalence. The meta-analytical results obtained allow us to estimate the prevalence rate of psychopathy in the general adult population at 4.5%. That being said, this rate varies depending on the participants' sex (higher in males), the type of sample from the general population (higher in samples from organizations than in community samples or university students), and the type of instrument used to define psychopathy. In fact, using the PCL-R, which is currently considered the “gold standard” for the assessment and definition of psychopathy, the prevalence is only 1.2%. These results are discussed in the context of the different theoretical perspectives and the existing problems when it comes to defining the construct of psychopathy.
Discussion
The main objective of this study was to obtain an estimate of the prevalence of psychopathy in the general adult population and, in this sense, to our knowledge, it is the first systematic or meta-analytic review carried out on this topic. Following a thorough search in the scientific literature, 15 empirical studies were found that had calculated the frequency of psychopathy in samples from the general adult population, including community, organization, and university student samples. These studies used properly described tools and procedures to assess and define psychopathy. After calculating the conjoint mean of their results with meta-analytic procedures, based on a total sample of 11,497 people, it can be estimated that the prevalence of psychopathy in the general adult population is 4.5%.
As could be expected, this prevalence is much lower than that found in samples obtained in forensic or prison contexts. For example, in the meta-analysis of Fox and DeLisi (2019), it was found that the average prevalence of psychopathy among homicide offenders could be estimated at 27.8 or 34.4%, depending on the criterion used to define psychopathy with the PCL-R (cut-off score of 30 vs. 25, respectively). In the second edition of the PCL-R manual (Hare, 2003a), the prevalence of psychopathy, based on a cut-off score of 30, was 15.7% for males (Nicholls et al., 2005) and 10.3% for females (Guay et al., 2018) in the North American normative samples of prisoners.
However, although the average prevalence of psychopathy in the general population is clearly lower than that found in the offender or prison population, the prevalence rates of psychopathy in the general population obtained in the studies reviewed in this work show considerable variation, ranging from a minimum of 0% to a maximum of 21%. In fact, the results obtained in terms of the I2 and Q statistics confirmed that the heterogeneity of the studies was statistically significant.
These variations depend on many factors, such as the role of the type of instrument used to define psychopathy, the participants' sex, the type of sample of the general population, and the participants' country of origin. These factors have been analyzed in this work. In this sense, the results of the present work indicate that the first three factors, but not the country of origin, seem to have a significant impact on the prevalence of psychopathy. Depending on the chosen instrument, the participants' sex or the type of sample selected, prevalence figures can double, triple, or quadruple the figures found with a different instrument or with participants of another sex or from a different subpopulation of the general population. Moreover, the results obtained in terms of the Doi chart and the LFK index indicate that this heterogeneity does not appear to reflect a significant publication bias, but could largely be attributed to these three moderator variables.
In particular, the results of this work indicate that, when using the PCL-R (or any of its versions), an instrument that is currently considered as the gold standard for the evaluation and definition of psychopathy, it can be estimated that the prevalence of psychopathy in the general adult population is only 1.2%. However, if other instruments are used, such as self-reports of psychopathic personality traits like the LSRP (Levenson et al., 1995) or the SRP in their different versions (SRP-II, SRP-III, and SRP-SF; Hare, 1990; Paulhus et al., 2016), the estimate of the prevalence of psychopathy in the general adult population quadruples, reaching 5.4%.
In fact, as virtually all the studies with offenders use the PCL-R or one of its versions, the comparison between the prevalence rates of psychopathy obtained in the general population and in the offender or prison population should primarily focus on studies conducted with the PCL-R. In this sense, the difference in the prevalence rate of psychopathy between the two types of population, general and criminal, is much greater: 1.2%, obtained in the present work for general population, compared to 15.7 and 10.3%, obtained in the normative samples of the PCL-R for male and female prisoners, respectively (Nicholls et al., 2005; Guay et al., 2018), or vs. 27.8%, obtained in Fox and DeLisi (2019) meta-analysis for homicide offenders.
Differences in the prevalence rates as a function of the type of instrument and cut-off point established to identify psychopathy go back to the problems in defining the construct of psychopathy. Those differences also point out a limitation of the present study. We will elaborate on these ideas later in the context of the limitations of this review.
The results of this study also indicate that the prevalence of psychopathy in the general adult population is significantly higher among males than among females. In particular, psychopathy in the general population doubles its prevalence in males compared to females (7.9 vs. 2.9%). This difference is consistent with the results obtained in samples of offenders or incarcerated people, among whom the prevalence of psychopathy is also higher in males than in females (Beryl et al., 2014).
In particular, Beryl et al. (2014) conducted a systematic review of the scientific literature on the prevalence of psychopathy in adult women from within secure settings, which included criminal justice settings, or secure inpatient healthcare settings. They found prevalence rates ranging from 0 to 31% using the PCL-R or one of its versions, although they did not report the average of these rates or the conjoint prevalence. However, from the data they submitted for females in criminal justice settings, it is possible to calculate, for the 13 unique studies that defined psychopathy based on a cut-off score of 30 in the PCL-R or of 18 in the PCL:SV, a weighted average prevalence of 11.9% (Table 3 of Beryl et al., 2014, p. 191). This figure dropped slightly to 11% when also taking into account the data from the 10 unique studies that had evaluated samples of females in secure/inpatient psychiatric settings or mixed samples—secure/inpatient psychiatric and criminal justice settings—(Tables 2, 4, respectively, of Beryl et al., 2014, p. 190, 192). Moreover, these figures hardly varied when only studies using the same instrument, the PCL-R, and the same cut-off score, 30 (12.3 and 11.4%, respectively) were taken into account. Interestingly, these prevalence figures are very similar to those presented by the scales of female prisoners collected in the second edition of the PCL-R manual, which, as noted above, show a prevalence of psychopathy in female prisoners of 10.3% (Guay et al., 2018). In summary, the average prevalence of psychopathy in female offenders or prisoners can be estimated at 10–12%.
In contrast, in male offenders or prisoners, using the PCL or its versions, rates of average prevalence of psychopathy of 15–35% are usually obtained, although the average rates of 15–25% are probably the most adequate (Hare, 1991, 2003a; Guay et al., 2007; Fox and DeLisi, 2019, cited by Nicholls et al., 2005). In the 1991 PCL-R manual, Hare reported that, in a global sample of 1,200 males incarcerated in Canadian prisons, 25% scored 30 or higher on the PCL-R. However, in the second edition of the PCL-R manual, published in 2003 and based on a much larger sample with a total of 5,408 males incarcerated in American prisons, Hare reported that 15.7% of the inmates scored 30 or higher on the PCL-R (Hare, 1991, 2003a; cited by Nicholls et al., 2005). Subsequently, with that same large sample, but eliminating the participants with missing information on some items of the PCL-R (n = 543), Guay et al. (2007) reported that 19% of the remaining 4,865 male inmates scored 30 or higher on the PCL-R. Finally, in the meta-analysis of Fox and DeLisi (2019), it was found that 27.8% of the homicide offenders scored 30 or higher on the PCL-R.
In any case, it seems clear that the prevalence of psychopathy is higher in male offenders or prisoners than in female offenders or prisoners (15–25% vs. 10–12%), and this difference between the sexes is maintained in the general population (7.9 vs. 2.9%), as shown in this meta-analysis.
Another interesting result of this work has to do with the finding of differences in the prevalence of psychopathy between different groups of adults in the general population. In particular, this review has found that the prevalence of psychopathy is significantly higher among workers in some organizations and companies (managers, executives, procurement and supply professionals, advertising workers) than among university students or among people from the general community (12.9 vs. 8.1% and 1.9%, respectively). In turn, the prevalence among university students is significantly higher than among people from the general community (8.1 vs. 1.9%).
The highest prevalence of psychopathy among workers in certain organizations and companies is based on data from only three studies with a total sample of 668 people and should, therefore, be taken with some caution. However, this result is consistent with the scientific literature that proposes that psychopathy is more prevalent in certain professions (e.g., entrepreneurs, managers, politicians, investors, sellers, surgeons, lawyers, telemarketing employees) in which the personality characteristics that define psychopathy could even facilitate their success in these professions (Hare, 2003b; Dutton, 2012; Babiak and Hare, 2019; Fritzon et al., 2020).
More surprising may be the result that among university students, there is a higher prevalence of psychopathy than among people in the community. Following the previous argument, it could be assumed that among university students of certain professions there could be more people with psychopathic traits (e.g., students in business administration and management, marketing), but it could also be assumed that among university students from other professions, there could be more people with less psychopathic traits and characterized, on the contrary, by high levels of empathy, altruism, candor, trust, humility, and responsibility (e.g., students from health professions, social work, and other professions closely linked to helping). In fact, in a study of Hassall et al. (2015), it was found that business university students, in comparison to university students of psychology, showed significantly higher levels in the four psychopathy factors measured by the SRP-III (Paulhus et al., 2016). Unfortunately, this work did not provide data on the prevalence of psychopathy in the two groups of university students. In addition, in the study of Dutton (2012), mentioned in the Introduction, among the 10 professions with higher levels of psychopathic traits, there were some that require a university degree (e.g., lawyer, surgeon, journalist) and, likewise, among the 10 professions with lower levels of psychopathic traits, there were also several that require a university degree (e.g., nurse, teacher, doctor).
Therefore, future research with university students should examine whether there are significant differences in psychopathy among students of different careers. This implies that, not only among university students of certain careers may there be a higher prevalence of psychopathy than in the general population, but that among university students of other careers, there may be a similar prevalence. It could even be that among university students of certain careers, there may be a lower prevalence of psychopathy than in the general population.
Research on differences in psychopathy between people of different professions or between university students of different careers departs from the traditional application of the construct of psychopathy to the forensic and prison area. That research intertwines, as discussed in the Introduction, with the most recent interest in the presence of psychopathy in everyday life (Dutton, 2012; Babiak and Hare, 2019; Fritzon et al., 2020), in the definition of psychopathy in terms of normal personality models such as the Big Five model (Lynam and Miller, 2019), and in the concept of successful or integrated psychopathy (Dutton, 2012; Lilienfeld et al., 2015). The fact that, as found in this review, most studies on the prevalence of psychopathy in the general population were published in the twenty-first century, especially in the last 10 years, is also consistent with those most recent interests far from the area of forensic and prison psychology.
Finally, no significant differences in the prevalence of psychopathy in the general population were found in this work as a function of the country of origin of the evaluated people. This absence of differences is not consistent with the results of the scientific literature on criminal and prison populations, which show the existence of differences between countries, especially between North American and European countries, in terms of the prevalence and levels of psychopathy in this type of population. For example, in the review of Beryl et al. (2014), a trend was found of lower rates of prevalence of psychopathy in European samples of women in prison or in prison hospitals than in American samples. Consistently, in the meta-analysis of Fox and DeLisi (2019), and after discarding the extreme values from samples composed exclusively of homicides with psychosis or psychopathy, significantly higher levels of PCL-measured psychopathy were found in homicide offenders from the USA and Canada than in homicide offenders from Finland, Sweden, and Germany.
Although these two reviews have reported that psychopathy prevalence is higher in North American male and female offenders and prisoners than in European male and female offenders and prisoners, the reasons for these differences are unclear. Beryl et al. (2014) suggest that the reason is “that the PCL instruments are designed to test the construct of ‘psychopathy’ as manifested in North American (male) offenders, and are less well-suited to identifying ‘psychopathy’ as manifested in European offenders” (p. 190). However, following the cultural facilitation model and Cooke et al.'s (2005) suggestions, an alternative reason is that complex social processes, such as socialization and enculturation, can suppress the development of certain aspects of psychopathy and facilitate the development of others. Therefore, it may be that socialization and enculturation in European countries suppress the development of certain psychopathic personality traits, or that those social processes in North American countries facilitate the development of certain psychopathic personality traits. There is also the possibility that both explanations are valid.
In any case, the results of the present review suggest that those differences between countries in the prevalence of psychopathy are unique to the prison or criminal population, but do not extend to the general population.
However, studies using samples from the general population of many different countries around the world have found cultural differences in the levels of different psychopathic traits. For example, in the study of Neumann et al. (2012) with 33,016 people (19,183 women) from 58 countries belonging to 11 world regions, significant differences were found between these regions in terms of the levels of different psychopathic traits (interpersonal, affective, antisocial, and lifestyle), as measured by one of the brief versions of the Hare SRP (SRP-E).
To further complicate the scenario of empirical results on the relationships between psychopathy and culture, the differences found in some studies with samples from the general population sometimes go in the opposite direction to those found in offender or prisoner populations. Thus, in the study of Lilienfeld et al. (2014), mentioned in the Introduction, in which they analyzed the responses of 3,338 people to the PPI-R-SF applied online, the Europeans showed higher levels of psychopathic traits than the Americans.
As a result, future research should address whether differences between countries in psychopathy only appear in terms of levels of certain psychopathic traits, but not in terms of the prevalence of psychopathy. When speaking about prevalence of psychopathy, we refer to it as defined by the presence of a clear set of psychopathic traits and with a certain level of intensity of such traits and/or a certain degree of impairment caused by such traits. It should also be examined whether such differences translate into a pattern of consistent differences between North American and European countries.
The results obtained in this work and the conclusions that have been reached should be assessed taking into account some of the limitations of the review itself. The most important limitations concern the high variability of the characteristics of the reviewed studies and the prevalence rates found, the small number of studies conducted to date that can help control such variability, and the methods assessing psychopathy in the reviewed studies. As already mentioned, prevalence rates vary greatly depending on factors such as the type of instrument used to define psychopathy, the participants' sex, and the type of sample from the general population. Given the small number of studies that currently constitute the scientific literature on the prevalence of psychopathy in the general population and the great heterogeneity of these studies in terms of their characteristics, it is very difficult to examine the effects of one of its factors while controlling the effect of the remaining factors. In fact, in this work, the number of subsamples/samples to examine gender prevalence was smaller than for calculating the overall prevalence. Therefore, in that smaller set, factors such as the type of instrument or sample did not reach statistical significance, thus preventing a more statistically potent analysis of the effect of gender after controlling the effects of these two factors and vice versa.
Among the factors that affected the variability of the prevalence of psychopathy, it is worth highlighting the type of instrument used to define psychopathy, since this factor points out important issue underlying this review. There is a high heterogeneity in the methods used to assess psychopathy in the reviewed studies. In addition, some of these method are more susceptible to criticisms related to their reliability and validity than others (e.g., the methods used in Hagnell et al., 1994; Gustafson and Ritzer, 1995; Pethman and Erlandsson, 2002). That heterogeneity and these criticisms go back to the problems in defining the construct of psychopathy. The different theoretical perspectives for this purpose which characterize the research of this construct are also an issue, and have already been discussed in the Introduction. In this sense, for example, an interesting exchange of views has recently been published on the debate over what components are essential to, or constitute part of psychopathy. It has also been discussed whether those components are necessary and/or sufficient (Brislin and Patrick, 2020a,b; Lynam, 2020; Marcus and Nagel, 2020). Consequently, one of the most important challenges that research in the area of psychopathy has to face is to achieve a valid and consensual definition of the construct of psychopathy and, related to this, to decide which instrument or instruments are the most valid and reliable to measure this construct. These needs are most evident when studying psychopathy in the general population because, as mentioned above, virtually all studies on psychopathy in the population of offenders or prisoners use the PCL-R or one of its versions (see the reviews of Beryl et al., 2014, and of Fox and DeLisi, 2019).
On the other hand, future research should also focus on the prevalence of the components of psychopathy, especially on the prevalence of psychopathic traits. Moreover, future research should also be conducted on the prevalence of the other personality constructs that are included under the Dark Triad label: Machiavellianism and narcissism.
Despite the above-mentioned limitations, the obtained results reflect relatively strong trends in the data that at least deserve to be the subject of future research and the formation of hypotheses to be taken into account in such research. In short, these trends allow the following conclusions to be drawn:
1) The prevalence of psychopathy in the general adult population can be estimated at 4.5%.
2) This prevalence is much lower than that found in the offender or prison population, which usually ranges between 10 and 35% (Nicholls et al., 2005; Guay et al., 2018; Fox and DeLisi, 2019).
3) The prevalence rates of psychopathy in the general population show considerable variation as a function of the type of instrument used to define psychopathy, the participants' sex, and the type of sample from the general population.
4) Using the PCL-R (or any of its versions), lower psychopathy prevalence rates are obtained than if self-reports of psychopathic personality traits are used.
5) As the PCL-R is currently considered the “gold standard” for the assessment and definition of psychopathy, the prevalence of psychopathy in the general population may be only 1.2% and, therefore, the difference with the prevalence of the offender or prison population may be even greater.
6) As is often the case in the offender and prison population, the prevalence of psychopathy in the general adult population is significantly higher among males than among females.
7) The prevalence of psychopathy is significantly higher among workers in some organizations and companies (e.g., managers, executives, procurement and supply professionals, advertising workers) than among university students or people from the general community. In turn, the prevalence of psychopathy among university students is significantly higher than among people from the general community, although the latter result could be due to the type of career that university students are pursuing (e.g., company careers vs. helping careers).
Posts
