A Network Approach to Hypersexuality: Insights and Clinical Implications

Abstract

Background In spite of a growing interest in research on hypersexuality, consensus about its etiology and best treatment strategy has not been achieved. Aim To further the empirical and clinical understanding of hypersexuality by exploring the structure of its symptoms using a network analytic approach. Methods In 2014, an online survey advertised as focusing on Internet pornography, sexual health, and relationships was carried out among Croatian men and women aged 18–60 years (Mage = 31.1 years, SD = 9.67). In a sample of 3,028 participants, we applied a network analytic approach to explore the structure of hypersexuality symptoms. In the network, nodes represented hypersexuality symptoms and associated sexual behaviors, while their connections were operationalized as partial correlations. 4 Research questions were addressed: (1) does the hypersexuality network differ between genders; (2) which symptoms are centrally positioned; (3) what is the topological location of pornography use; and (4) are there distinct clusters (“communities”) of symptoms in the network? Outcomes We estimated and plotted hypersexuality networks by gender using items from the Hypersexual Disorder Screening Inventory and the Hypersexual Behavioral Consequences Scale, as well as indicators of sexual desire, pornography use, sexual intercourse, and masturbation frequency. Results The structure of the hypersexuality network was surprisingly similar in women and men, both in terms of symptom centrality and the clustering of symptoms. Psychological distress and negative emotions triggered by sexual fantasies and/or behaviors, together with a loss of control over sexual feelings, occupied central positions in the networks. Pornography use was located peripherally in both the men’s and women’s hypersexuality networks. Clinical Translation Psychological distress and negative emotions triggered by sexual fantasies and/or behaviors constituted the core of the hypersexuality network, which makes them potential prime targets for clinical intervention and calls for normalization of (presumably self-stigmatized) sexual expression through affirmative therapy and interventions that enhance self-care, self-compassion, and adaptive coping mechanisms. Strengths & Limitations This is the first network analytic approach to hypersexuality. Apart from its novel insights about the structure of hypersexuality, the study employed several methods to assure reliability and robustness of findings. Considering that networks were estimated in a convenience-based community sample, the findings might not generalize to clinically distressed individuals. Conclusion Our results demonstrate the usefulness of network analytics to hypersexuality in a non-clinical sample and we encourage future clinical and longitudinal explorations of hypersexuality using this novel approach.

A Randomized Controlled Study of Group-Administered Cognitive Behavioral Therapy for Hypersexual Disorder in Men

“…HD and its symptoms have been proposed to be structured as a core network between psychological distress (i.e., depression, anxiety, stress) and negative emotions triggered by excessive and uncontrolled sexual behaviors (Werner, Stulhofer, Waldorp, & Jurin, 2018). The network perspective on hypersexuality suggests that HD emerges as a result of a dynamic system of various symptoms and components that interact (Werner et al, 2018). This theoretical network approach to the structure of HD, is in line with the definition of psychiatric disorders as mechanistic property clusters (MPC) proposed by Kendler, Zachar, and Craver (2011).…”

“…• Development and validation of measures of specification of sexual behaviors involved in CSBD As it has been proposed that HD (and by extension also CSBD) is an umbrella term, covering symptoms related to difficulty controlling sexual behaviors, there may be a multitude of possible causes that alone or in combination result in HD symptomatology (Kafka, 2010;Montgomery-Graham, 2016). Indeed this is what have been indicated in the network approach to hypersexuality (Werner et al, 2018). Even if the relevance of the HD criteria has been validated for a particular group of patients, a broader knowledge of subcategories based on patient characteristics is warranted.…”

Comparing Gaussian graphical models with the posterior predictive distribution and Bayesian model selection.

“…Recently, the focus has shifted from exclusively estimating a network from one group, to comparing those estimated from di erent sub-populations (Fried et al, 2018). For example, group di erences have been examined in depression networks (e.g., good vs. poor depression prognosis; Beard et al, 2016), as well as gender di erences in hyper-sexuality networks (Werner, Štulhofer, Waldorp, & Jurin, 2018). These comparisons have sometimes been speculative, for example based visual inspection, or with a re-sampling approach that was recently introduced to psychology (Van Borkulo et al, 2014).…”

Bayesian Hypothesis Testing for Gaussian Graphical Models:Conditional Independence and Order Constraints

“…Exploring these relations has shed new light upon a variety of research areas including personality (e.g. Costantini et al, 2015) and psychopathology (e.g., Di Pierro et al, 2019;Werner et al, 2018).…”