Viewing Sexual Stimuli Associated with Greater Sexual Responsiveness, Not Erectile Dysfunction: A Comment by Richard A. Isenberg MD

Upon review of Drs. Prause and Pfaus’ manuscript, “Viewing Sexual Stimuli Associated with Greater Sexual Responsiveness, Not Erectile Dysfunction,” I feel obliged to question the authors’ presentation of their data, analysis, and far-ranging conclusions. The vast availability of pornographic videos through the Internet has led to a host of social, physical, interpersonal, and emotional problems for our patients [1–5]. Drs. Prause and Pfaus are right in focusing the medical community’s attention on one reported physical complication: pornography-induced erectile dysfunction. Unfortunately, I do not see how this study advances our understanding of the phenomenon.

The authors’ description of the study population was seriously incomplete. The authors cite four published studies as the source of the study population (see Table 1) [6–9]; however, upon examination of those studies, I can only source 234 of the 280 men assessed in this investigation. Forty-six men are unaccounted for. The authors provide no accounting of the origin of their study population, nor characterization or identification of the source of the subpopulations selected for assessment of the outcome measures. For instance, in only one study [6] were subjects assessed for erectile dysfunction through use of the International Index of Erectile Function (IIEF). The 2013 Prause paper reports on IIEF results from 47 men, yet the authors of the present study report on IIEF findings in 133 men. Were these 86 additional subjects excluded from analysis in the 2013 study, or were they from some other uncharacterized database? Multiple other discrepancies are found between the manuscript and the cited sources:

1. Discrepancies identified in manuscript

IIEF = International Index of Erectile Function.

The apparent inclusion of these subjects from the 2013 Prause study [6] in the analysis of sexual arousal and sexual desire raises further concerns. While this investigation was designed to address sexual arousal and desire in the laboratory setting in response to the viewing of pornographic films, the 47 men in the 2013 Prause study were shown still photographs rather than films. It seems unlikely that the viewing of still photographs generates a level of arousal comparable with explicit pornographic videos [10]. The authors offer no justification for the inclusion of data from these subjects, nor any accounting to indicate that these subjects were excluded from their analysis of sexual arousal and desire. Moreover, it is clear from the published manuscripts that the other three studies providing subjects for this investigation [7–9] used videos of inconsistent duration (20 seconds to 3 minutes). Without uniformity of the erotic stimulus, the legitimacy of pooling data from the various sources is questionable.

It is disturbing that the authors do not provide descriptive statistics about the study’s central parameter: the hours of pornography viewed. While the authors report that they have clustered the data into three bins (none, less than 2 hours, more than 2 hours), they do not provide basic population statistics such as the mean, standard deviation, median, or range for hours of pornography viewing for the overall population or any subpopulation. Without understanding the populations in terms of the critical parameter, the reader cannot translate the study findings to the care of his/her individual patients.

The hours-viewed parameter itself is poorly defined. We are not told if the self-report of hours referenced the preceding week, the average over the last year, or was entirely left to subject interpretation. Were there subjects who were new porn users who had not had enough exposure to develop erectile or other sexual issues? Were there subjects who were previously heavy users who had recently cut down or eliminated their pornography viewing? Absent a well-defined and consistent referent, the porn use data are uninterpretable.

Furthermore, the authors do not report on relevant viewing parameters such as total pornography usage, age of onset, presence of escalation, and extent of sexual activity with partner which may have bearing on male sexual functioning [11,12]. In addition, the exclusion of hypersexual men (the men who generally complain about pornography-induced erectile dysfunction) raises questions about the relevance and generalizability of the study’s erectile function findings.

Even more disturbing is the total omission of statistical findings for the erectile function outcome measure. The statistical tests that the authors used are not identified, although the reader is told that there were “several.” No statistical results whatsoever are provided. Instead, the authors ask the reader to simply believe their unsubstantiated statement that there was no association between hours of pornography viewed and erectile function. Given the authors’ conflicting assertion that erectile function with a partner may actually be improved by viewing pornography (with fruit fly studies cited for support), and their boastful prepublication promotion of their findings on Twitter (https://twitter.com/NicolePrause/status/552862571485605890), the absence of statistical analysis is most egregious.

The authors clearly devoted much time and energy to their research project. It is unfortunate that they have not provided the reader with sufficient information about the population studied or the statistical analyses to justify their conclusion that pornography is unlikely to negatively impact erectile functioning. While there is some indication in the data that nonporn-addicted men watching brief pornography films may have increased sexual arousal and desire, this is hardly a novel finding.