Emerging research on people with pedohebephilic interests suggest disparities in neurological, psychological, social, and behavioural characteristic between people with pedohebephilia who have sexually offended against children versus those who have not (e.g., Beier et al., 2009; Gerwinn et al., 2018; Kärgel et al., 2015, 2017; Konrad et al., 2017; Massau et al., 2017; Schiffer et al., 2017; Stephens et al., 2024). For example, compared to non-justice involved MSICs, MSICs with CSA histories tend to display diminished functional activity in neural networks implicated in motivational and socio-emotional processes, socio-cognition (e.g., deciphering social cues, theory of mind, moral judgements, sexual disinhibition), and impulse inhibition (Kärgel et al., 2015, 2017; Massau et al., 2017; Schiffer et al., 2017). Gerwinn and colleagues (2018) also found associations with experience of CSA as a factor associated with perpetration of sexual offending against children more broadly, as well as high prevalence (45% to 58%) of mental illness relative to non-justice involved persons. Thus, MSICs who have perpetrated CSA represent a heterogeneous population compared to those with no criminal justice involvement.

Recent research examining detected and undetected MSICs by Stephens and colleagues (2024) suggested that clinical and risk factors such as additional paraphilias, hypersexuality, dark personality traits, and features of their pedohebephilic sexual interests and history of adverse childhood experiences differentiated those MSIC with CSA offences from those without detected CSA histories. Therefore, psychological risk factors may be particularly relevant for distinguishing MSICs who perpetrate CSA from those who do not. Furthermore, people with sexual interests in children tend to experience psychological distress and stigmatization, even if they have never perpetrated CSA (e.g., Beier et al., 2009; Gerwinn et al., 2018; Konrad et al., 2017; Lievesley et al., 2020). Unfortunately, this social and internalized stigma are deterring factors for help-seeking among MSIC populations, regardless of detection status (Elchuk et al., 2022; Houtepen et al., 2016; Jahnke, Imhoff, et al., 2015; Jahnke, Schmidt, et al., 2015; Lasher & Stinson, 2017; Levenson et al., 2017). A lack of access to professional resources may exacerbate existing psychological vulnerability and thereby increase an individual’s risk for perpetrating CSA and decrease their quality of life (B4U-ACT, 2011; Farmer et al., 2012).

Researchers have attempted to understand the diversity among men who perpetrate CSA by developing psychometric typologies to represent subtypes within the larger group of known offenders using risk factors and offence history variables (e.g., Robertiello & Terry, 2007). These classification systems are then used to guide risk assessment and case management to reduce offending risk and guide prevention (Knight & Prentky, 1990; Mandeville-Norden & Beech, 2009; Martínez-Catena et al., 2017; Robertiello & Terry, 2007; Ward & Siegert, 2002). Typologies also offer insight about motives driving CSA, clinical presentations, needs for intervention, etiological mechanisms, and offence trajectories. However, how well these offender-based typologies apply to persons who have sexual interest in children but never act on them is unclear. Rather than over-rely on risk-criminal based typologies, some researchers have called for typologies drawing from a clinical health-based perspective to better support individuals with these interests who are not justice-involved (Lievesley et al., 2020). Developing a typology of non-forensic MSICs in the community may clarify targets for interventions in such contexts to increase their psychological health, while also identifying avenues to prevent victimization of children that could be applied to the non-offending population.

Multifactorial theories of CSA have indicated that a variety of developmental and biopsychosocial-sexual risk and protective factors function in dimensional, interactive, and mitigating ways to contribute to the perpetration of sexual offences against children, or lack-thereof (Hanson & Morton-Bourgon, 2005; Mann et al., 2010; Stephens et al., 2024; Ward & Beech, 2008). Integrating findings from theory and research, CSA risk factors rest across five biopsychosocial domains: 1) developmental factors (i.e., adverse childhood experiences, adult attachment style); 2) self-dysregulation (i.e., sexual preoccupation, impulsivity, affect regulation); 3) sexual interests (i.e., pedohebephilia, other paraphilic interests, sexual history); 4) distorted cognitions (i.e., pro-offending cognitions, emotional congruence with children); and 5) socio-affective deficits (i.e., mental health, distress, personality, loneliness).

Developmental risk factors reflecting adverse experiences during childhood and insecure attachment during adulthood can catalyze neurological and relational disturbances that may predispose men to other CSA risk factors. People with CSA offences have higher rates of sexual and/or physical victimization in childhood than people with sexual offences against adults and the general population (Bailey et al., 2016; Gerwinn et al., 2018; Levenson et al., 2016; Maniglio, 2011; Seto, 2018; Whitaker et al., 2008). People with CSA offences are more apt to have insecure attachment styles in relationships, which contributes to difficulties in interpersonal functioning and intimacy-related schemas (Gunst et al., 2017; Maniglio, 2011; Wood & Riggs, 2009).

Self-regulatory problems also are strong predictors of any recidivism among sexual offenders (Hanson & Morton-Bourgon, 2005). Inhibitory processing deficits may impair control of inappropriate sexual arousal, execution of sexual behaviours, and cognitive reappraisals of arousal to problematic stimuli (Mann et al., 2010). Compared to the general population, inhibitory control functions are significantly impaired among people who perpetrate CSA (Joyal et al., 2014). Impairments in neurobiological systems subsequently affect capacities for self-regulation within other sexual, psychosocial, cognitive, and behavioural systems that, in turn, perpetuate an individual’s risk for CSA.

Sexual deviancy, including attraction to children, has emerged as one of the strongest CSA risk factors (Hanson & Morton-Bourgon, 2005; Whitaker et al., 2008). The presence of pedohebephilic sexual interest is especially relevant as it offers additional pathways to interact with other biopsychosocial and situational risk factors to contribute to greater motivation and risk to perpetrate CSA. For example, exclusive pedophilic interests are associated with CSA perpetration and higher rates of sexual recidivism (Bailey et al., 2016). Additionally, people with multiple paraphilias (e.g., exhibitionism, voyeurism, sexual masochism, fetishism; APA, 2013) pose higher risk of sexual recidivism (Hanson & Morton-Bourgon, 2005).

Forensic literature recognizes antisocial cognitions as prominent risk factors for general offending, and for sexual offending specifically (Bonta & Andrews, 2024; Whitaker et al., 2008). Pro-offending cognitions foster beliefs that support or justify CSA (Ward & Keenan, 1999; Wood & Riggs, 2009). Likewise, experiencing higher emotional congruence with children, defined as “an exaggerated affective and cognitive affiliation with children and childhood… including emotional attachment and dependency needs … more likely met by interacting with children than with adults” (McPhail et al., 2013, p. 737), may interfere with forming healthy relationships with other adults or motivate desires for sexual and/or romantic experiences with children (Houtepen et al., 2016; Mann et al., 2010). The presence of both pedohebephilic sexual attraction and a wish to emotionally connect with children may contribute to increased sexual deviance and elevated risk of sexual behaviour with children (Konrad et al., 2018).

Socio-affective deficits increase vulnerability to dysregulation and criminal behaviours when one lacks capacity to adaptively manage their distress or navigate relationships (Hanson & Morton-Bourgon, 2005; Whitaker et al., 2008). Features related to mental health issues, personality traits, and loneliness are common among people who perpetrate CSA and may directly or indirectly have a role in the perpetration of CSA offences, as well as implications with barriers to help-seeking (e.g., stigma; Gunst et al., 2017; Hanson & Morton-Bourgon, 2005; Jahnke, Imhoff, et al., 2015; Jahnke, Schmidt, et al., 2015; Konrad et al., 2017; Mann et al., 2010; Stephens et al., 2024; Whitaker et al., 2008). In contrast, one’s ability to form intimate, supportive relationships with other adults can be protective against CSA (de Vries Robbé et al., 2015).

Despite the wealth of CSA criminal risk information, there is limited empirical evidence examining whether these same forensic population identified risk characteristics are relevant to “non-forensic” (i.e., no known sexual offending) MSICs living in the community. Applying these frameworks may still provide insight to understand heterogeneity among community MSICs as well as potential needs for preventative interventions. Since sexual interest in children is presumed to render men more at-risk for committing CSA, examining how community MSICs compare to groups of men with other, non-child oriented paraphilic interests can elucidate risk and protective factors when pedohebephilic interest is present.

There exists a subset of MSICs who have not, and some will not, sexually offend against children. Whereas some community MSICs report no issues with managing their sexual interests to refrain from CSA, others may be struggling to remain offence-free (Bailey et al., 2016). Thus, some MSICs may present elevated risk of committing CSA and some may experience significant distress and stigmatization regardless of offence status (Konrad et al., 2017). Research identifying differences between MSICs who do and do not sexually offend is in its infancy (Gerwinn et al., 2018). Extending MSIC research from forensic to non-forensic community samples may identify areas of need to inform preventative intervention against victimization. By expanding beyond risk-related priorities to include clinical health perspectives, development of services to focus on mental health and well-being of vulnerable persons may meet dual aims of personal wellness as well as by-product impacts toward prevention of risk and harm (Lievesley et al., 2020). Although paraphilic interest in general may elevate motivation for sexual offending, extant literature is clear in denoting etiological pathways to CSA that implicate increased vulnerability among some men with pedohebephilic interests. As such, the current research used a comparison group of men with other non-pedohebephilic paraphilic interests to examine how they compare on biopsychosocial-sexual factors with a heterogeneous group of non-forensic MSICs. Inclusion of such a comparison group also may facilitate identification of features that are transdiagnostic versus MSIC-specific targets for prevention efforts, as well as explore how characteristics may be unique to men who specifically identify a degree of sexual interest in children compared to men with paraphilias in general.

The primary goal of the current study was to explore the potential presence of clusters of community-based MSICs who report never criminally engaging in CSA. Men who endorsed sexual interest in children were recruited online to ensure anonymity and promote honest reporting. Extending theory and research on risk and protective factors for CSA, two major objectives were pursued. First, to identify underlying clusters of characteristics (e.g., developmental, self-regulatory, sexual, cognition, and socio-affective) among self-identified members of the non-criminal justice involved MSIC population to develop a psychometric-based typology. It was hypothesized that MSICs would be meaningfully grouped into at least two clusters reflecting meaningful variations in biopsychosocial-sexual characteristics. Second, to examine variations across biopsychosocial-sexual variables that best characterize and distinguish MSIC clusters. This included determining how profiles vary in risk/protective factors compared to MSICs with detected CSA and community men with other paraphilias, with variations expected in severity of clinical presentation.

Participants were screened based on inclusion criteria: 1) their sex/gender (i.e., male/men); 2) age category (≥18 years); 3) self-reported “non-traditional” or “atypical” (i.e., paraphilic) sexual interest of any type; 4) literacy in English reading/writing; 5) not having previously done the survey; and 6) informed consent to participate. Development of the typology only included men who reported sexual interests in children who have not been formally involved in the criminal justice system related to these interests (no reported charges or convictions for CSA offences). Men with other paraphilic interests and MSICs with historical CSA offences were used as comparison groups for Research Questions 2.

A priori power analysis required ≥132 participants to detect a medium effect size with adequate (i.e., 80%) statistical power with a .05 p-value for proposed analyses (Dziak et al., 2014; Tabachnick & Fidell, 2013). The present study aimed for a sample size of at least 250 MSICs to improve generalizability of results and account for unequal group sizes and oversampled (N ≈ 1,000) to ensure sufficient sample size for statistical power in the analyses (Dziak et al., 2014). A summary of steps taken for data cleaning and conditioning is described in Figure 1 (detailed overview available in Stewart, 2022). The final dataset contained 833 cases from the initial 2,386 with any survey interaction, and consisted of both MSIC participants (73.1%, n = 609; of whom 584 denied detected CSA offences) and other paraphilia participants (26.9%, n = 224).

The current study was approved by the Human Ethics Research Board of the authors’ affiliated institution. Online recruitment proceeded by posting recruitment advertisements across a variety of media including posts on social media, support groups and online forums, educative non-profit and prevention organizations, listservs for sex researchers and clinicians, and crowdsourcing research recruitment websites. The survey was hosted through SoSci Survey, a Germany-based online survey platform that administered and compiled data to ensure anonymity and confidentiality (Leiner, 2019). No identifiable participant data was collected. To foster honest participant engagement in a safe research environment (e.g., Klein et al., 2018; Ray et al., 2010), site moderators were contacted for permission to share the recruitment ad with their online communities.

After reviewing informed consent, eligible participants could voluntarily choose to complete the survey (~30 minutes in length). For questionnaires specific to cognitions (i.e., ABCS + CASA) or emotional congruence with children (CS-ECWC), participants were provided explicit definitions which stated that references to “child” or “children” referred to persons aged 14 years or younger, to align with definitional criteria used in this research. Participants recruited through MTurk (n = 300) were compensated $3.00 for their engagement as per expectations with that platform. Participants recruited more broadly across the Internet were offered entry into a draw to win one of ten $50 gift cards. Survey content was identical across both mediums.

Prior to conducting statistical analyses, data were screened and conditioned. An overview summarizing each step of data conditioning is summarized in Figure 1 (for detailed description of conditioning process, see Stewart, 2022).

Research Question 1 identified underlying a posteriori clusters from biopsychosocial-sexual factors of MSICs in the community who had no previous criminal justice system involvement for CSA (n = 584). Using MPlus 6.12, latent cluster analysis (LCA) identified latent subgroups of MSICs based on scores from measures of developmental (i.e., adverse experiences [ACES-R total score], adult attachment [ECR-S subscale scores]), self-regulatory (i.e., sexual preoccupation [HBI total score], impulsivity and disinhibition [SUPPS-P subscale scores]), sexual (i.e., paraphilic interests [PIS-E total number of paraphilic interest categories]), cognitive (i.e., pro-offending attitudes [ABCS + CASA total score], emotional congruence with children [CS-ECWC total score]), and socio-affective (i.e., mental health [DASS-21 total score], personality [PID-5-BF total score], loneliness [SELSA-S subscale scores]).

Using the latent groups from Research Question 1, Research Question 2 used descriptive analyses (i.e., Chi Square, frequencies, means, standard deviations) and MANOVA to examine which biopsychosocial-sexual factors significantly differ between MSIC clusters and men who endorse other paraphilic interests (i.e., Other Paraphilia; n = 225). The small subsample of community MSICs who self-reported prior charges or convictions for sexual offences against a child under the age of 16 years (i.e., MSIC-CSA; n = 25) were compared on mean scores across factors. Due to small sample size, analyses of MSIC-CSA group are a preliminary and exploratory effort to compare offender profiles with community MSICs with no detected CSA.

LCA used 17 continuous measures of developmental, self-regulatory, sexual, cognitive, and socio-affective functioning to examine 2- to 10-class cluster solutions based on previous literature (i.e., forensic CSA typologies proposed by Beech, 1998; Groth et al., 1982; Henry et al., 2010; Knight & Prentky, 1990; Lanning, 1992; Mandeville-Norden & Beech, 2009; Ward & Siegert, 2002). The results of 2- to 6-class cluster solutions are summarized in Table 1. Maximum likelihood parameter estimates and observed scores were used to examine class membership probability for each k-class solution to minimize within-cluster variation, while also maximizing between cluster variation (Bakk & Vermunt, 2016; Vermunt & Magidson, 2002).

Based on recommended BIC, LMR, BLRT, entropy fit indices, and combined with nuanced interpretability and theoretical parsimony (Contractor et al., 2018; DiStefano & Kamphaus, 2006; Dziak et al., 2014; Nylund et al., 2007), a 3-class solution was selected as the optimal model for “undetected” community MSICs. Estimated membership probability for 3-class solution was 92.5% for Cluster 1, 95.0% for Cluster 2, and 90.4% for Cluster 3. Probability of misclassification was 7.5% for Cluster 1, 5.0% for Cluster 2, and 9.6% for Cluster 3. A graphical representation of z-score means for the 3-class solution is depicted in Figure 2.

Latent class membership differences for all measured characteristics included in the LCA are described in Table 2. Latent clusters were conceptually labelled according to significant patterns of biopsychosocial-sexual characteristics endorsed by members of those classes. Cluster labels were assigned based on major trends observed across characteristics and may not fully capture the nuances, heterogeneity, and complexity within clusters. Overall, the three clusters appear to represent different severity levels of vulnerability factors for community MSICs.

Participants classified in Cluster 1 (n = 165, 28.3%) were characterized by lower levels of developmental and biopsychosocial-sexual vulnerability factors, thus labelled Low Vulnerability. Members of Cluster 1 endorsed significantly lower adverse childhood experiences, attachment anxiety, attachment avoidance, sexual preoccupation, negative urgency, lack of premeditation, positive urgency, pro-offending attitudes, emotional congruence with children, mental health disturbance, mildly problematic personality traits, romantic loneliness, familial loneliness, and social loneliness compared to participants in other clusters. Participants in Cluster 1 also had significantly lower scores on number of paraphilic interests compared to men in Cluster 3.

MSICs classified in Cluster 2 (n = 270, 46.2%) demonstrated intermediate endorsement of all measured vulnerability factors and labelled Moderate Vulnerability. Cluster 2 MSICs showed higher endorsement than Cluster 1 and lower endorsement than Cluster 3 on measures of adverse childhood experiences, attachment anxiety, sexual preoccupation, negative urgency, lack of premeditation, positive urgency, pro-offending attitudes, emotional congruence with children, mental health disturbance, problematic personality traits, family loneliness. Men in Cluster 2 also had significantly lower scores on number of paraphilic interests compared to Cluster 3 and significantly higher scores of romantic loneliness and social loneliness compared to Cluster 1.

Finally, Cluster 3 (n = 149, 25.5%) MSICs were distinguished by higher endorsement of most biopsychosocial-sexual vulnerability factors and was named High Vulnerability. Of note, average endorsement of adverse childhood experiences, sexual preoccupation, and mental health disturbance were above recommended cut-offs of clinical severity for participants in Cluster 3.

Compared to participants in Cluster 1 and Cluster 2, MSICs in Cluster 3 endorsed significantly higher adverse childhood experiences (clinically significant; see Materials section), attachment anxiety, sexual preoccupation (clinically significant; see Materials section), negative urgency, lack of premeditation, positive urgency, pro-offending attitudes, emotional congruence with children, mental health disturbance (clinically significant; see Materials section), moderately high problematic personality traits, and family loneliness. Higher attachment avoidance, sensation seeking, total number of paraphilic interests, romantic loneliness, and social loneliness also were observed for participants in Cluster 3 compared to those in Cluster 1.

Descriptive analyses of individual characteristics explored differences between the MSIC subtypes and Other Paraphilia groups. Results are summarized in Table 3. Although dispersed internationally, most participants across all groups were similar in being white North American men who were educated and employed. On average, they ranged in age from their late-twenties to early/mid-thirties. The full sample showed little propensity for antisocial behaviour as per self-reported criminal history. For sexual history, MSIC groups tended to report younger ages of onset but concurrently were less sexually experienced than Other Paraphilia group. This finding aligns with forensic literature suggesting that people with pedohebephilic interests or histories of CSA offences begin sexual behaviours at younger ages (e.g., Gerwinn et al., 2018; Houtepen et al., 2016; Levenson et al., 2017). Among Moderate Vulnerability MSICs, limited sexual and relationship history suggests challenges with intimacy and social aptitude.

Moderate and High Vulnerability groups had relatively lower academic achievement and higher unemployment than the Low Vulnerability and Other Paraphilia groups. The MSIC groups more often identified heterodivergent orientation and reported being single, suggesting difficulties with adult intimate relationships particularly among the Moderate and High Vulnerability groups.

Next, five multivariate analysis of variance (MANOVA) examined which variables best differentiated between the identified MSIC groups, MSIC-CSA men, and men with other paraphilias (Cramer, 2003; Mertler & Vannatta, 2002; Tabachnick & Fidell, 2013). MANOVAs used participants’ self-reported scores on measures’ subscales as dependent variables to identify groups differences. Follow-up univariate ANOVA and Tukey’s post hoc analyses (p-value ≤ .05) are summarized for all MANOVAs in Table 2.

The emergent Vulnerability Typology revealed that clusters of men endorsing sexual interest in children with no detected CSA histories were primarily distinguished based on self-reported intensity of biopsychosocial-sexual risk factors associated with perpetration of CSA. The Low Vulnerability MSIC cluster (28.2%) reported normal, unimpaired levels of measured vulnerability characteristics. This profile demonstrated more social stability and aptitude, better capacity for self-regulation, and lower developmental predisposition to risk factors despite their sexual interest in children. These strengths may offer protective effects against perpetrating CSA. In contrast, the largest proportion of MSICs were classified in the Moderate Vulnerability cluster (46.2%). Overall, this second profile was characterized by intermediate impairment across domains, including indications of elevated sociosexual challenges. Although the Moderate Vulnerability cluster may not be as “globally” distressed as the High Vulnerability group, they may not derive benefit from protective factors to the same degree as the Low Vulnerability group. The High Vulnerability cluster (25.5%) of the MSICs group and was characterized by global functional deficits across all biopsychosocial domains of interest, which is indicative of broad clinical and criminogenic needs relevant to risk factors associated with CSA. The presence of numerous dysfunctional mechanisms among High Vulnerability MSICs may render them more susceptible to multiple etiological pathways leading to sexual offending. Recent literature examining MSIC populations with detected and undetected CSA histories found that those with CSA histories tend to score highest on measured biopsychosocial-sexual factors, suggesting that the High Vulnerability group may especially benefit from interventions to prevent perpetration of CSA (Stephens et al., 2024).

Aligning with existing forensic typologies and theoretical models, vulnerability emerges as a multifaceted phenomenon, involving proximal and distal factors that may increase risk for problematic functional outcomes, whether legal or otherwise (e.g., Hanson & Morton-Bourgon, 2005; Mann et al., 2010). The dynamic nature of identified vulnerability factors offers intervention targets to reduce risk for criminal behaviour, increase wellness, and build good lives. Whereas external or internal stressors may destabilize and increase one’s vulnerability to problematic consequences, it is possible that well-matched, proactive intervention informed by evidence-based clinical models may shift MSICs to lower vulnerability status. By extending the forensic literature, the Vulnerability Typology provides a framework to guide application and prioritization of therapeutic and risk management targets for community MSICs.

Congruent with the Risk-Need-Responsivity (RNR) model of criminal behaviour and crime prevention (Bonta & Andrews, 2024), the Vulnerability Typology suggests merit in having service intensity graded to match the level of criminogenic and clinical need reflected in different MSIC profiles. Overall, results of the current study suggest that the needs of Low Vulnerability profiles may best be addressed with low intensity intervention, or may not require any intervention if their protective features are sufficient to support ongoing prosocial self-management. Moderate-intensity intervention would be suitable for MSICs with Moderate Vulnerability profiles, adopting a skill-building approach to increase effective strategies for managing deviant sexual interests and navigating life stressors. Finally, the globally elevated level of need characterizing High Vulnerability profiles represents multiple areas to target in intervention, which can warrant longer-term, higher-intensity treatment efforts.

Overall, the current findings show that men’s vulnerability for CSA perpetration is not necessarily contingent on a single problematic risk factor (e.g., pedohebephilia), but more likely emerges from the combined effects of several co-occurring risk and destabilizing factors. Facets of psychosocial wellness may exert destabilizing effects that elevate susceptibility to other criminogenic factors, potentially moving closer to crossing legal sexual lines in parallel with heightened subjective distress. Establishing secondary prevention programs may improve accessibility of therapeutic resources for MSICs to support them in remaining offence-free while concurrently addressing a variety of general mental health issues and criminogenic needs affecting their daily functioning and risk potential (Beier et al., 2009; Lasher & Stinson, 2017; Stephens & McPhail, 2019). The current study supports the notion that, regardless of their offence status, community MSICs do report challenges across an array of dynamic factors, many of which may be intervention targets. These findings further imply that help-seeking MSICs may not necessarily present at clinical service agencies for concerns explicitly related to their sexual interests, but nevertheless may benefit from risk-informed intervention services. For example, MSICs in the community may present to treatment for concerns such as coping with internalized stigma or general mental health, in addition to seeking help for their pedohebephilic interests (Stephens et al., 2024). In combination with research showing high rates of mental health disorders and related concerns (e.g., suicidality, substance misuse, internalized stigma) among community MSICs (Elchuk et al., 2022; McPhail & Olver, 2020; Stephens & McPhail, 2019; Stephens et al., 2024), this research highlights the benefit of addressing specific mental health and intimacy needs among this population. CSA Efforts to prevent CSA would best serve at-risk MSICs by addressing both criminogenic and possibly destabilizing psychological needs.

Although not explicitly assessed in the current study, it is likely that shame and stigma influence subjective experiences and present additional psychological need targets to address in prevention and intervention efforts (Elchuk et al., 2022; Jahnke, Imhoff, et al., 2015; Jahnke, Schmidt, et al., 2015). Indeed, associations exist between internalized stigma, loneliness, and psychological distress among populations with sexual interest in children (Elchuk et al., 2022; Stephens et al., 2024). Integration of principles prescribed by RNR and Good Lives Model (Ward & Gannon, 2006; Ward & Laws, 2010; Ward & Marshall, 2004; Ward & Stewart, 2003), in addition to practices adopted in humanistic, person-centered, and trauma-informed orientations (Levenson & Grady, 2019; Lievesley et al., 2020; Walker, 2021), may help match intervention intensity to equip MSICs with internal and external resources to develop capacity, skills, and resources to attain goals in non-offending ways to live a better life. Preventative intervention with MSICs can balance dual aims of personal wellness and public safety by promoting stability (i.e., management of impulses and emotions), accountability (i.e., safe decision making, supportive prosocial support network), and plans (i.e., engagement in fulfilling activities and prosocial goals) toward a life that is socially acceptable, personally meaningful, and subjectively worth living. Some preventative intervention efforts that have emerged to address the dual foci of targeting criminogenic and destabilizing psychological needs among MSICs have included Germany’s Project Dunkelfeld, Canada’s Talking for Change, and online peer-support forums such as Virtuous Pedophiles, B4U-ACT, and HelpWanted.

A strength of the current study was the substantial effort made to maximize recruitment via safeguards to assure participant anonymity, confidentiality, and honest responding among a stigmatized and hard-to-reach population (Klein et al., 2018; Ray et al., 2010). We strove to maintain transparency by communicating directly with group representatives and moderators for permission to share the study with community members, as well as recruiting participants from other broader online mediums. These approaches resulted in a sufficiently large sample to investigate research hypotheses as proposed. In addition, we were comprehensive in assessing many constructs relevant to understanding biopsychosocial-sexual MSIC profiles using psychometrically appropriate measures.

Notwithstanding the strengths of the current study, results must be interpreted considering several limitations. Most significantly, self-report questionnaires present potential for bias in subjective self-evaluations. In addition, online data collection from a hidden population presents benefits and pitfalls. Online research methodologies can encourage self-disclosure and increase reach to diverse participant pools but presents challenges in verifying participant eligibility and self-selection bias. Thus, participants completed pre-screening items as part of the informed consent to access the survey. Low-quality data were removed during conditioning processes. Questionnaires were disseminated across a variety of websites to reduce sampling bias, but cultural norms (e.g., mandatory reporting, ease of accessibility), forums’ philosophical attitudes, and perceived acceptability of the research’s rationale could have impacted which sites permitted recruitment ads and who chose to participate (Jahnke, Imhoff, et al., 2015; Ó Ciardha et al., 2022). Likewise, self-reported data on paraphilic sexual arousal may be susceptible to social desirability and impression management (Pedneault et al., 2021) and participants may have under-reported symptoms and arousal due to stigmatization or concerns regarding detection (Lasher & Stinson, 2017; Ray et al., 2010). Self-reported paraphilic sexual arousal may be conservatively interpreted as a minimum or “floor level” of sexual attraction (i.e., MSICs under-reporting intensity of pedohebephilic arousal versus over-endorsing attractions).

Another element of self-report methods of data collection is that, due to underreporting of sexual offending (Cotter, 2021), it is possible that some non-justice system involved MSICs may have engaged in CSA that had not yet come to the attention of law enforcement (Stephens & McPhail, 2019). Thus, the present study refrains from explicitly labelling this sample as “non-offending,” though such might reasonably be inferred. Accurate measurement of undetected CSA is challenging due to ethical responsibilities to ensure participant safety and honest responding while adhering to mandatory reporting protocols. Nevertheless, distinguishing community MSICs by detection status offers opportunity to examine differences between those whose behaviour to-date has not resulted in formal justice involvement, versus those have been “caught” for CSA and have since returned to community.

Finally, data collection occurred in the early weeks of the COVID-19 pandemic. Emerging research on the psychological impact of COVID-19 in North America finds that the pandemic correlated with elevated levels of distress, including depression, anxiety, general lifestyle disturbance, destabilization, and isolation (Best et al., 2021; Panchal et al., 2021). Participants may have reported data during a time of increased mental health disturbance wherein existing vulnerabilities may have been negatively exacerbated by the pandemic.

The current findings offer multiple directions for future research. These results, including the typological profiles of community MSICs, should be replicated to promote credibility, generalizability, and reproducibility. Studies should test the robustness of the typology using alternative design methodologies (e.g., in-person data collection, indirect measures of sexual interest) or applying new analytical approaches. Accessing a larger comparison sample of MSIC-CSA participants may elucidate similarities and differences between men who have already crossed sexual boundaries, versus MSICs who have not acted (and may never act) on their sexual interests toward children. Such an examination would serve to further differentiate between sexual interests and sexual offence behaviours.

Future research also should extrapolate from the typological model to inform evidence-based preventative interventions that meet the dual needs of subjective wellness and objective risk mitigation. The emergence of distinct Low, Moderate, and High Vulnerability groups in community MSICs suggests that clients may present with different need and protective profiles. Therapeutic targets may be prioritized differently in effective secondary intervention approaches. Examining interactions between factors within and across domains, including whether additive or combined factors affect vulnerability and risk profiles, also presents an area for future study.

Recognition of typological groups may facilitate investigations into help-seeking behaviours among community MSICs. The Vulnerability Typology may provide a framework to understand which groups are most likely to seek or utilize services, what issues most compel help-seeking (e.g., mental health or lifestyle concerns, versus sexual interest in children), and what barriers to services are most relevant for each group. In turn, future research can explore ways to enhance service accessibility and availability for community MSICs.