Walton, Cantor, Bhullar, and Lykins (2017) provided a comprehensive review of hypersexuality and in so doing presented myriad issues, such as definitional properties of the construct, prevalence rates, common criticisms surrounding attempts at pathologizing hypersexual behavior, as well as commonly studied features and correlates. Other comprehensive reviews of hypersexuality have been conducted in the last two decades (Gold & Heffner, 1998; Kaplan & Krueger, 2010; Kingston, 2016; Kingston & Firestone, 2008; Montgomery-Graham, 2017). In addition to their general overview, however, Walton et al. introduced the “sexhavior cycle,” a descriptive model outlining the initiation and maintenance of hypersexual behavior. In this Commentary, I focus on the definition and conceptualization of hypersexuality and then I apply core epistemic values in the evaluation of their proposed sexhavior cycle.

Defining and Conceptualizing Hypersexuality

Hypersexuality has proven to be an elusive concept to define and measure despite considerable attention being devoted to this issue. Walton et al. specifically define hypersexuality as a “pattern of recurrent, intense, and excessive preoccupation with sexual fantasies, urges, and behavior that individuals struggle to control” along with associated consequences. Similar to the definition of mental disorders used in the DSM-5 and ICD-10, the aforementioned definition includes two essential components: a set of symptoms (observable and subjective) and impairment.

Such a general and descriptive definition is well suited for a conceptual review, but it will need to be more specific for clinical use. For example, clinicians will need to conduct an appraisal of the negative consequences associated with the observable and subjective symptoms. In addition, it is not always easy to determine the level of distress or impairment (and the relevant domains of such distress or impairment) that is required for a diagnosis. Note that these problems are not unique to the proposed definition of hypersexual disorder; they are relevant for most, if not all, psychological disorders.

Additionally, we need to determine what is sufficiently intense or frequent to warrant concern and to suggest dysfunction. Knowledge of norms of desire and behaviors for the reference group of the individual is required but is rarely available. Moreover, among those who demonstrated markedly increased sexual behaviors, a baseline level of desire and behavior would need to be determined, something that is not always easy to do when individuals consult at peak distress. Even with good group norms or individual baseline data, a decision needs to be made with respect to the appropriate cutoff. Some have suggested a cut-point near the 90th percentile (e.g., Långström & Hanson, 2006) to specify extreme urges or behaviors. However, prior to selecting some arbitrary cutoff point, it is first important to determine whether the construct itself is best represented dimensionally or categorically at the latent construct level. In other words, does hypersexuality characterize a non-arbitrary class of individuals or does the construct simply characterize individuals at the high end of a continuum of sexual urges and behaviors?

Only a few studies have been conducted that specifically examined the structure of hypersexuality with particular emphasis placed on whether hypersexuality represents a distinct category or whether it is better represented along a continuum of high sexual drive. Despite the fact that researchers and clinicians typically conceptualize hypersexuality as a categorical entity, recent research employing taxometric procedures suggests that it is likely something that differs in degree rather than kind (Graham, Walters, Harris, & Knight, 2016; Walters, Knight, & Långström, 2011). Kingston et al. (2017b) had further replicated these aforementioned taxometric findings in a large sample of student as well as community-based samples and found further supportive evidence for a dimensional latent structure. Of course, this does not mean that cutoff scores are meaningless, but rather they need to be specified along the construct continuum, which would then correspond to particular clinical decisions and consequences.

Aside from definitional properties, considerable attention has been directed toward how best to conceptualize hypersexual disorder. The most common etiological models reviewed include sexual compulsivity, sexual impulsivity, sexual addiction, and, more recently, neurobiological models, including the principles outlined in the dual control model (Bancroft & Janssen, 2000). Walton et al. surmised that hypersexuality is the likely result of multiple etiological pathways and that any one of the aforementioned conceptual models is, in isolation, likely an insufficient explanation. This idea is consistent with what I had suggested in my earlier review papers (Kingston, 2016; Kingston & Firestone, 2008). Indeed, I have argued that adopting any one conceptual model and applying it to all individuals presenting with hypersexuality is insufficient to address the underlying heterogeneity. I have underscored the importance of dysphoric mood states and emotional dysregulatory processes as a predominant explanatory mechanism for hypersexuality. However, more recently, we posited that hypersexuality and impulsivity may be more directly linked (Kingston, Graham, & Knight, 2017a; Reid, Berlin, & Kingston, 2015). This relationship is underscored by research showing consistent covariation between high sexualization and self-centered impulsivity (Kastner & Sellbom, 2012) and the fact that childhood maltreatment may negatively impact areas of the brain related to both emotional regulation and cognitive control (Teicher, Tomoda, & Andersen, 2006; Tottenham et al., 2010).

Evaluating the Sexhavior Cycle

The significant heterogeneity among predictors of hypersexual behavior is what led Walton et al. to propose an “alternate conceptualization” of hypersexuality. The sexhavior cycle proposes four distinct and sequential stages: sexual urge, sexual behavior, sexual satiation, and post-sexual satiation. The cycle is further discussed within the context of observed variability in the frequency and intensity of sexual arousal as well as the role of cognitive processing (i.e., cognitive abeyance) and sexually incongruent behavior. Although not conceptualized as such in their article, the sexhavior cycle can readily be described as a descriptive model of the hypersexual process. Descriptive modelsFootnote 1 emphasize proximal, as opposed to distal, etiological factors and specify the cognitive, behavioral, affective, and contextual factors that culminate in the criterion behavior.

A number of researchers have outlined core epistemic values that are useful in determining the relative strengths and weaknesses of a particular theory (Ward, Polaschek, & Beech, 2006). Briefly, these include: (1) empirical adequacy and scope (does the theory account for the observed phenomena and existing findings?); (2) internal coherence (does the theory contain contradictions or gaps?); (3) unifying power and external consistency (is past theory integrated in a meaningful way?); (4) explanatory depth (can the theory describe underlying mechanisms and processes?); and (5) fertility (does the theory provide new predictions and avenues of inquiry?).

Based on the aforementioned epistemic values, the sexhavior cycle presents with a number of strengths. First, as reflected in their comprehensive review, the sexhavior cycle incorporates a number of relevant theories and processes in a clinically meaningful way (unifying power and external consistency). Specifically, Walton et al. introduce the concepts of cognitive abeyance and sexual incongruence to partly explain the underlying neuropsychological processes inherent in the initiation and maintenance of hypersexual behavior. Another obvious strength of the sexhavior cycle is the underlying fertility or heuristic value. Walton et al. indicate that they are currently conducting an online study validating this cycle in a sample of self-identified “sexual addicts.” It will be particularly interesting to see the extent to which cognitive abeyance and sexual incongruence facilitate hypersexual behavior, as has been described in earlier clinical samples. Also, from a clinical perspective, this theory may lead to some important insights into the treatment of hypersexuality although not much attention was devoted by Walton et al. to this avenue of future research. Nevertheless, it is clear how perhaps providing some psychoeducation on the underlying features of this model with the overarching goal of improved self-monitoring and cognitive restructuring may be beneficial for some clients.

Despite these strengths, there are several areas in which the sexhavior cycle may need further development or clarification. Many of these concerns revolve around empirical adequacy and scope, which is not surprising given that the theory is new and much of the previous hypersexual literature was based on clinical anecdote rather than empirical data.

The sexhavior cycle denotes a predominantly linear process whereby individuals transition from one stage to the next. However, the process may be more dynamic whereby some may transition back and forth, particularly between sexual behavior and increased sexual urges. Similarly, in their description of the theory, the mechanisms underlying why some individuals are able to employ effective coping, while others may fail to do so, are not yet clear. Relatedly, Walton et al. discuss the concepts of guilt and shame as emotions that can temporarily inhibit sexual behavior. A number of studies have shown that negative emotions can exert the somewhat paradoxical effect in hypersexual individuals of increased sexual behavior. Indeed, the classic description of hypersexuality highlights a cyclical pattern whereby someone first experiences negative affect that is perceived as unbearable, and then engages in sexual behavior in order to temporarily relieve these intense, negative emotions. There is also some evidence that the type of attributions one makes can affect the outcome. For example, shame often results from an attribution that behavior resulted from an internal and uncontrollable factor (e.g., lack of ability) which, in turn, leads to a lack of effort to avoid engaging in the criterion behavior. In contrast, if the attribution is to a controllable factor, such as lack of effort, then there may be guilt rather than shame and renewed effort to avoid relapse. Such finer descriptions could be embedded in the theoretical account. Finally, it was not clear to me as to when guilt and shame occur in the cycle. Is it only during post-sexual satiation or can individuals experience these emotions at other stages as well?

Lastly, Walton et al. introduce the concept of cognitive abeyance as a state of “inactivity, deferment, suspension, or diminution of logical cognitive processing.” This condition is hypothesized to occur during particularly heightened sexual arousal and is theorized to reduce volitional control. There are clearly instances whereby individuals exhibit reduced cognitive functioning when in a heightened state of sexual arousal and for whom the state, itself, reduces their cognitive functioning. However, some people presenting with hypersexuality have to engage in at least some degree of planning to execute their goals of sexual gratification (e.g., arranging sexual encounters). Moreover, many individuals with hypersexuality have clear objectives for sexual activity and are able to conceal their sexual activities from loved ones (at least for a period of time), suggesting that they have some degree of cognitive control in certain domains. Again, further description and refinement would be beneficial, particularly with regard to improving the internal coherence of the proposed cycle.