Treatment seeking for problematic pornography use among women (2017)

Background and aims

Methods

Results

Discussion

Conclusion

Figure 1. Path analysis of the extended model showing standardized path coefficients tested with the use of 95% bias-corrected confidence intervals (**p ≤ .001; *p < .05). Values in brackets are standardized coefficients for the direct effects before accounting for indirect pathways. Bolded arrows represent the relationships related to our main hypothesis. The rest of the paths represents secondary hypotheses. PU in the name of the variable stands for pornography use. Dashed lines indicate paths that were excluded from the final version of the model for females. Sample sizes for each variable are listed in Table 1

To the best of our knowledge, this is one of the very limited number of studies on females seeking treatment for problematic PU and the first one investigating factors related to treatment-seeking behaviors. Because of the lack of such studies on females, we used our previous studies on male samples as the reference point for our analysis. The results of this study show both similarities and clear differences between results for female problematic PU and previous studies on this subject for males (Gola et al., 2016; Kraus, Martino, et al., 2016). First, our analysis showed that females seeking treatment for problematic PU have higher levels of negative symptoms associated with PU and higher amounts of pornography consumption than non-treatment seekers. This particular result is not surprising, considering the results obtained in previous studies (Gola et al., 2016; Kraus, Martino, et al., 2016). However, more interestingly, our analysis showed that treatment-seeking women may be prone to periods of disinhibition (decidedly higher maximum number of masturbations during 1 day and longer episodes of non-stop pornography watching). In the available literature, we can find evidence that rigid social norms in some cases may contribute to problematic PU, because they promote periods of abstaining from pornography, followed by a period of disinhibition and excessive PU (Carnes, 1983; Kraus, Martino, et al., 2016; Wordecha, Wilk, Kowalewska, Skorko, & Gola, 2017). Initial evidence confirming this interpretation can be found in the differences in religiosity between women seeking and not seeking treatment. The treatment-seeking group reported higher values for both subjective religiosity and the average amount of religious practices during a week. We elaborate on the possible role of social norms and religiosity in female problematic PU below, discussing it together with the results of other recent studies.

The second part of our analysis was based on a statistical model of relations among variables related to treatment seeking and problematic PU. In line with many previous results showing sex-related differences in sexual functioning, the results obtained in this study on a female sample differ from previous studies on male samples. Before summarizing our findings from the current analysis on a female sample, we would like to remind the main conclusion from our previous study on males (Gola et al., 2016). We showed that: (a) the mere amount of PU is a very weak predictor of treatment seeking but (b) it is related to the severity of negative symptoms (measured by the SAST-R), and this factor explains treatment-seeking behavior. Besides that, (c) among males, age is not related to the amount of PU and (d) the onset of PU does not predict the severity of negative symptoms associated with PU. Similarly, (e) the amount of religious practices neither predict treatment seeking nor the severity of negative symptoms associated with PU (Gola et al., 2016).

As we hypothesized, for females, the mere amount of PU was more strongly related to the treatment seeking for problematic PU. The amount of PU was also related to the severity of associated negative symptoms (Figure 1), and the severity of associated symptoms was related to treatment seeking. The latter relation was much weaker among males (ad. b). Moreover, differently from our analysis for males, the relationship between the amount of PU and treatment seeking among females remained significant, even when accounting for mediation through the severity of negative symptoms. This interesting result shows that women with problematic PU perhaps seek treatment not only because of the negative impact of PU on their life but also because of the sheer amount of PU (while in previous studies focusing on male samples, the latter factor is non-significant). This raises a question about the possible explanation of why the mere fact of frequent PU can be perceived as a problem among females. The most probable reason is that regular PU may be perceived by the majority of females as less normative behavior than it is among males. Among males, weekly PU seems to be a normative behavior (about 70%–80% of males between ages of 18–30), whereas among females, less than 20% use pornography on a weekly basis (as shown in large Danish and Scandinavian studies: Hald, 2006; Kvalem et al., 2014). This difference may shape the belief (among women) that frequent PU is some kind of deviant behavior in contrast to males, among whom the same behavior can be perceived as normative. Thus, the mere fact of regular PU may cause a subjective impression that certain women differ from the majority of women, which can result in the interpretation of regular PU as a problematic behavior that needs treatment. If this interpretation is correct, the subjective sense of experiencing problems connected with PU among females could be amplified by moral or religious beliefs about pornography and masturbation. Recent studies on general populations showed that religiosity may be related to higher tendency for self-perceived “pornography addiction” (Grubbs et al., 2016) or reported negative consequences of frequent sexual activity (Štulhofer et al., 2016). We tested if religiosity can also be related to treatment seeking (Figure 2) (ad. e) by including the amount of religious practices as a predictor of treatment seeking, while also investigating its relation to the amount of PU. Indeed, the amount of religious practices is the strongest predictor of treatment-seeking behavior among females with problematic PU (while it was non-significant in a corresponding analysis for men; Gola et al., 2016). Moreover, our analysis showed that after introducing religious practices into the model, the relationship between the mere amount of PU and treatment seeking lost its significance (Figure 2). Such a finding is in line with numerous studies showing that female sexuality is usually more related to cultural and social aspects than among males (Adams & Turner, 1985; Barry & Schlegel, 1984; Baumeister, 2000; Christensen & Carpenter, 1962; Earle & Perricone, 1986; Ford & Norris, 1993). Here, we could perhaps say that these cultural aspects contribute to the subjective interpretation of regular PU as problematic and leads to treatment seeking.

In our model, the amount of religious practices was also positively related to pornography consumption (estimate = 0.55). However, this association turned out to be significant only for treatment seekers, and not significant in the non-treatment-seeking group. This indicates that such relation seems to be a characteristic of the clinical group and is not necessarily present in the general population. In addition, it is worth noting that the amount of consumed pornography and religious practices (reflecting an importance of religious norms) was higher among treatment seekers. One possible interpretation for these results is that for some treatment-seeking individuals, behavioral engagement in the behavior-supporting religious norms (religious practices) can be a tool to regulate negative emotions caused by previous engagement in behavior violating these norms (pornography consumption). Another potential mechanism that can be proposed is that both pornography consumption and engagement in religious practices can be seen as a consequence of increased strength of impulses to watch pornography among treatment seekers. Thus, pornography consumption can simply be a sign of yielding to one’s impulses, and religious practices can be viewed as a way of dealing with them. If this scenario is true, both amount of PU and religious practices would be positively correlated, although this relationship would be determined by an underlying factor such as craving for PU.

Another possible interpretation for the high correlation between PU and religious practices among treatment-seeking individuals can be made in terms of ironic processes of mental control theory (Wegner, 1994). Higher and more rigid religious norms can lead to higher levels of inhibition for behavior (or thoughts related to behavior) that is seen as not convergent with these norms (e.g., watching pornography). However, as shown in numerous cognitive studies (see Abramowitz, Tolin, & Street, 2001 for a review) in some cases, inhibition can have a paradoxical effect, leading to higher frequency of behaviors that violate the norm. This can make the norm itself more salient and in turn elevate the level of behaviors that support the norm – in this case – religious practices. Thus, any kind of behavior enforcing rigid religious norms, and behaviors violating this norm can become mutually supportive, even when conscious intention of an individual was aimed at entirely contrary effects. Although previous studies on paradoxical effects of suppression focused mostly on thought suppression (Abramowitz et al., 2001), we have some evidence that emotion suppression can lead to similar, ironic effects (Webb, Miles, & Sheeran, 2012). In addition, some researchers suggest the role of paradoxical effects of suppression in the development of psychological disorders such as obsessive–compulsive disorder (OCD; Purdon, 2004), and many clinicians point to similarities between CSBs and OCD (see Gola, 2016; Kor, Fogel, Reid, & Potenza, 2013 for a review). All mechanisms described above are hypothetical and cannot be verified on the basis of our data alone. However, we believe that they are worth investigating in future studies that will be aimed at clarifying the nature of the relationship between religiosity and pornography consumption among treatment seekers for problematic PU.

In addition, our analysis expands on findings of previous studies regarding the relationship between religiosity and the severity of experienced negative symptoms (Grubbs et al., 2016; Štulhofer et al., 2016). When considering only the bivariate relationship between these two variables, our results confirm conclusions from the previous studies and indicate that the relationship in question is positive and significant (r = .25 for religious practices and r = .09 for subjective religiosity; Table 2). However, when amount of PU is included as an additional predictor of negative symptoms, religiosity is no longer related to the latter variable, while remaining a powerful predictor of treatment seeking (Figure 2).

Results concerning the relation of religiosity to negative symptoms and treatment seeking for problematic PU are especially interesting in the broader context of the relationship between religiosity and other forms of psychopathology. In previous research, higher level of religiosity was shown to be positively correlated with psychological well-being (Dilmaghani, 2017; Ismail & Desmukh, 2012; Joshi, Kumari, & Jain, 2008), life satisfaction (Pfeifer & Waelty, 1995), and inversely related to psychopathology in clinical patients (Gupta, Avasthi, & Kumar, 2011; Sharma et al., 2017). On the other hand, some research (McConnell, Pergament, Ellison, & Flannelly, 2006) suggests that higher degree of spiritual struggles can be positively correlated with some dimensions of psychopathology (anxiety, phobic anxiety, depression, paranoid ideation, obsessive–compulsiveness, and somatization). In addition, we have evidenced that at least some of the religious denominations can be associated with higher levels of OCD symptoms (Abramowitz, Deacon, Woods, & Tolin, 2004; Gonsalvez, Hains, & Stoyles, 2010). This indicates that the impact of religious convictions on psychopathology can be moderated by psychopathology type and characteristics of religious belief. In addition, as we have shown in our final model, in the specific case of problematic PU among women, religiosity seems to be related to treatment seeking rather than to psychopathological symptoms. Here, our results are in line with previous studies showing that strength of religious convictions and amount of religious practices are positively related to mental health service use (Pickard, 2006).

Interestingly, for females, age plays a significant role in PU; this includes both the age of the subject (ad. c) and age of the onset of PU (ad. d), while none of these variables were significant in our previous study on men (Gola et al., 2016). Younger women declared using pornography more often than older individuals, and those who started using pornography at a younger age tended to report the higher severity of negative symptoms related to PU. The explanation of this finding definitely deserves further investigations. Such investigations could address two interesting questions: (Q1) Does the popularity of PU increase among younger generations of females? (Q2) Is the female brain more vulnerable to the conditioning of a certain type of sexual stimuli than the male brain?

(Q1) According to our knowledge, there are no longitudinal data allowing us to address this question. Interestingly, recent survey data from the UK (Opinium Research, 2014) show that by the age of 18, pornography viewing was common and typical for 98% of boys and girls. Such a result may suggest that PU among girls has increased over the past years (perhaps due to Internet availability) and equalized among boys, as older studies indicated sex-related differences in PU. For example, Sabina et al. (2008) reported that among American college students, 93.2% of males and 62.1% of females watched Internet pornography by the age of 18, whereas Træen, Spitznogle, and Beverfjord (2004) reported that among a representative sample of Norwegians, during their whole life, 87.9% of males and 62.9% of females had seen a pornographic magazine, 77.2% versus 55% watched a pornographic movie, and only 36.6% versus 8.9% watched pornography on the Internet. Other data suggest that a profile of hypersexual activity among women may have also changed over the last decade. Briken, Habermann, Berner, and Hill (2007) reported that the most dominant sexual behavior among treatment-seeking women was risky casual sex (among men, it was PU and masturbation), whereas the team Klein et al. (2014) reported PU as the most common behavior among women obtaining high scores in the Hypersexual Behavior Inventory (Reid, Garos, & Carpenter, 2011). In our opinion, the hypothesis about an increasing ratio of female pornography users deserves careful study. It would also be interesting to examine how the patterns of dominant forms of sexual activity change among treatment-seeking women.

(Q2) In the numerous studies on substance use (Grant & Dawson, 1998), the onset of use is an important factor related to the severity of symptoms. In our studies on males (Gola et al., 2016), we expected to see such a relation with the onset of PU. Surprisingly, we did not. But among females, the onset of PU is significantly related to both the severity of associated negative symptoms and to the amount of PU. It is possible that women’s sexuality is more susceptible to learning (Baumeister, 2000). If so, then the question about increasing popularity of PU among young females (Q1) would be even more important to study.

Besides the above-discussed effects, we also noticed a large disproportion in the ratio of males and females seeking treatment for problematic PU. Our recruitment procedure was exactly the same for males and females. In the case of males, it took us 12 months to recruit 132 treatment-seeking individuals, whereas among females, we needed 18 months to find 39 subjects. This shows that males are seeking treatment because of problematic PU 5.07 times more often than females. This result provides empirical verification of the 5:1 ratio previously estimated by Kuzma and Black (2008), and is in line with previous studies showing a 4:1 ratio (Briken et al., 2007).