Archives of Sexual Behavior, VoL 20, No. 2, 1991

Psychophysiological and Endocrine Responses to Sexual Arousal in Women Julia R. Heiman, Ph.D., 1,5 David L. Rowland, Ph.D., 2 John P. Hatch, Ph.D., 3 and Brian A. Gladue, Ph.D., 4

The unexplored possibility that a sexually induced endocrine response might prime further sexual arousal in women guided the current investigation. Healthy, premenopausal, heterosexual women in the follicular phase of their menstrual cycle were randomly assigned to either an experimental or a control group. The experimental group was exposed to a sexually "explicit videotape, while the control group saw a nonerotic videotape. Ninety minutes later both groups saw a sexually explicit videotape. Vaginal vasocongestion and hormones (cortisol, prolactin, luteinizing hormone, testosterone) were measured continuously and subjective responses were sampled at 20-min intervals. Compared to controls, experimental subjects showed a greater amplitude and longer duration vaginal response to the second videotape. Subjective measures showed greater sexual response to the second erotic videotape compared to the first, an effect that was not mediated by the hormones measured here. Prolactin decreased significantly across the session for both group~, and several behavioral and affective responses were significantly correlated with hormonal levels. Commonalities and divergence with results of prior research point to the complexity and subtlety of endocrine interactions with sexual re.sponse as well as likely sex differences in hormone-behavioral interactions. KEY WORDS: psychophysiology; endocrine response; female sexual arousal.

1Department of Psychiatry and Behavioral Sciences, University of Washington, Seattle, Washington. 2Department of Psychology, Valparaiso University, Valparaiso, Indiana. 3Department of Psychiatry, The University of Texas Health Science Center at San Antonio, San Antonio, Texas. 4Department of Psychology, North Dakota State University, Fargo, North Dakota. 5To whom all correspondence should be addressed at University of Washington Medical School, Harborview CMHC ZA-31, 326 Ninth Avenue, Seattle, Washington 98104. 171 0004-0002/91/0400-0171506.50/0 © 1991PlenumPublishingCorporation

172

Heiman et al.

INTRODUCTION Conditions that facilitate or diminish sexual arousal provide important information about the nature of sexual response and the treatment of sexual problems. For example, numerous studies document the fact that repeated exposure to erotic stimulation results in a significant reduction in sexual arousal in both men and women (Howard et al., 1970; Heiman, 1977; Schaefer and Colgan, 1977; Rubin et al., 1979; O'Donohue and Geer, 1985). In a typical experimental protocol, subjects are presented with an erotic (sexually explicit) stimulus during which genital and/or subjective sexual arousal is measured. Subsequent erotic stimuli are presented at various intervals and, with further exposure to these materials, genital and subjective response show a gradual decline. This arousal decline has been referred to as habituation and appears to occur to different stimulus modalities (slides, audio and visual tapes) as well as constant and varied stimuli. A few studies have not demonstrated habituation in men (Rosen, 1973; Heiman, 1977; Julien and Over, 1984), raising further questions as to how long and over what experimental conditions habituation develops. The conditions for facilitation of sexual response are rarely examined. In a study using sexually functional men, we (Rowland et al., 1987) demonstrated that exposure to a sexual stimulus (videotape) actually enhanced sexual arousal to a second sexual stimulus presented 80 to 90 rain later. While the mechanism underlying this effect is not yet fully understood, considerable evidence suggests that sexual arousal and/or behavior may induce phasic hormonal changes in males of many species, including humans (Stearns et al., 1973; Pirke et aL, 1974; LaFerla et al., 1978; Hellhammer et aL, 1985; Rowland et al., 1987), and it may be that under such circumstances, these hormonal changes (e.g., luteinizing hormone, cortisol) facilitate further sexual response to repeated stimulation. In women, however, the possibility that a sexually induced endocrine response could prime further sexual arousal remains unexplored. In most subprimate mammalian females there is a precise relationship between hormonal events and sexual behavior, such that changes in the endocrine state surrounding ovulation affect the expression of the behavior. In some instances, as with reflex ovulators, the inverse of this relationship has also been demonstrated-specific components of sexual behavior can produce significant changes in luteinizing hormone (LH) which in return result in ovulation (for a review, see Jochle, 1975). In the human female the relationship between endocrine state and sexual arousal and/or behavior has been only minimally explored. Some studies point to a tonic

Sexual Arousal in Women

173

facilitatory role for estrogen (Adams et al., 1978) and testosterone (Bancroft et al., 1983; Shel-win and Gelfand, 1987) on sexual interest in women. More indirect evidence has shown that changes in self-reported and physiological (vaginal) arousal occur over the menstrual cycle and during menopause (Schreiner-Engel et al., 1981; Mateo and Rissman, 1984; Morrell et al., 1984). Behaviorally induced phasic changes in gonadotropic or ovarian hormones have not been demonstrated experimentally in women, the one notable exception being a report by Stearns et al. (1973) which found a prolactin (PRL) response following intercourse in women. Additional evidence is suggested by studies that have found an inverse relationship between coital frequency and variability on menstrual cycle length (Cutler et al., 1979) and an intercourse-induced acceleration of basal body temperature shift and the onset of menses (Stanislaw and Rice, 1987). Overall, while there are no animal studies that have found exposure to sexual stimuli causes endocrine changes in nonhuman females, there are strongly suggestive indications that sexual behavior influences ovulation and/or hormonal events in human females. This fact in combination with the outcome of a prior study with men (Rowland et aL, 1987) prompted the present investigation of women. In the present study, the relationship between sexual arousal, endocrine events, and facilitation of further sexual arousal was investigated in women during the follicular phase. This experiment was designed to determine (i) whether initial exposure to an erotic stimulus might enhance sexual response to an erotic stimulus presented 90 rain later rather than show evidence of habituation; (ii) whether such enhancement might be related to a phasic testosterone (T), LH, or PRL response; (iii) whether cortisol (CORT), a hormone that appears to affect sexual response in men (Rowland et al., 1987), is related to genital or self-reported sexual arousal in women; and (iv) whether predictable relationships exist among hormones, genital response, and self-reported arousal.

METHOD

Subjects Subjects were 13 premenopausal women volunteers between the ages of 21 and 45 (x = 32 years) who were recruited from the community through newspaper advertisements. None of the women reported having a sexual dysfunction, taking any medications, or having a physical (vascular, neurological) or psychological (psychotic or major affective) disorder that

174

Ileiman et al.

was considered likely to affect the data collected. All the women were predominantly heterosexual, had prior experience with sexual intercourse, masturbation, and oral sex, and had seen explicitly sexual books or films. Three of the women were married, three were single, three were divorced or separated from their spouses, and four were living with someone or had other living arrangements. Eleven of the women reported having a steady sexual partner at the time they were interviewed. Subjects were within 6 to 11 days (x = 8) since the beginning of their last menstrual period, and all reported that they had not engaged in sexual activity on the day of testing.

Apparatus and Materials

Questionnaires. Subjects c o m p l e t e d several forms prior to the laboratory session: (i) the SCL-90, a standardized, self-report scale measure of a variety of psychological symptomatology (Derogatis, 1977); (ii) a Personal History Questionnaire (PHQ) and Sexual History Form (SHF) to provide information on sexual practices, orientation, functioning, and basic demographics (Heiman and Rowland, 1983); and (iii) a Medical History Form to record physical history and functioning including medication and substance use. During the laboratory session subjects completed Subjective Rating Scales (Heiman, 1977; Heiman and Rowland, 1983) following each videotape and at 20-min intervals, coinciding with every other blood sampiing for hormone assay. These rating scales consisted of a series of cognitive and affective items (i.e., sexual arousal, aggression, anger, relaxation, worry) rated on 7-point Likert-type scales. Stimulus Materials. The stimuli were three videotapes, each about 18 rain in length. Two of the videotapes contained explicit sexual scenes involving heterosexual intercourse; the third videotape, which served as a control stimulus, was a documentary that contained no explicit sexual content. The erotic videotapes were nonviolent, were pretested for equivalent arousal values, and were the same tapes used in the Rowland et al. (1987) study of males. Physiological Recording. Physiological sexual arousal was determined with a vaginal photoplethysmograph, which measures changes in vaginal vasocongestion as confirmed in other research (Geer et al., 1974; Sintchak and Geer, 1975). Each woman placed the photoplethysmograph, the size of a tampon (6 cm x 1.5 cm) in the outer third, 2 cm beyond the introitus, of the vagina. The AC signal, measuring vaginal pulse amplitude (VPA) was a m p l i f i e d by a G r a s s ( M o d e l 7P1, 0.1 sec, time c o n s t a n t )

Sexual Arousal in Women

175

preamplifier, and recorded on paper. The AC signal represents relative blood pulse changes and is apparently not affected by the drift and temperature confounds of the DC signal (Beck et al., 1983). VPA was sampled manually by averaging responses over 20-sec intervals, taken every 10 min to correspond with blood collection sampling, throughout the 2-hr procedure. Maximum arousal was derived from the VPA measure by selecting the 20-sec interval of greatest VPA amplitude during a given 10-rain interval.

Procedure and Setting To provide privacy, the laboratory consisted of subject and experimenter rooms, separated by a door and connected by an intercom. Subjects sat at approximately a 30 ° angle in a reclining chair and watched a 19-inch color television monitor. All apparatus was controlled from the experimenter's room. During a telephone screening, the experimental procedure was explained in detail to subjects, including the use of erotic videotapes and venipuncture for blood sampling. To minimize a sense of coercion, subjects then came to the laboratory only to view the environment, be interviewed, and make a tentative decision about participation. At this time, informed consent was obtained, and subjects were randomized to either an experimental (n = 7) or control (n = 6) group. All contacts were with a female experimenter. At the second session, the laboratory procedure began at 11 AM. Each woman again heard the procedure along with careful instructions as to the placement of the vaginal photoplethysmograph. Subjects remained fully clothed for the procedure. After placement of the plethysmograph, a female physician entered the subject's room and inserted a butterfly needle into the subject's left arm for blood collection. When adequate blood flow was established, usually within 2 to 5 rain, the subject was left alone, and all further communication with her was via intercom. A 30-min baseline was followed by the first 1g-rain videotape (erotic for E group, nonerotic for C group). For the next 80 rain, the woman remained seated quietly, at times filling out various forms. A second videotape (erotic for both groups) was then shown, followed by another 20 rain of sitting quietly. Videotapes were counterbalanced in the E group, which viewed two different erotic stimuli during the session; the C group saw, in counterbalanced order, 1 of the 2 erotic tapes.

176

I t e i m a n et al.

Blood Sampling and Hormone Assay Blood for hormone assay was drawn continuously from the subject's left arm by means of a heparinized tube and an indwelling butterfly needle (Cormed Thromboresistant Needle and Tubing Set). Blood flow approximated 0.6 ml/min (100 ml over the 170-rain session). A fraction collector automatically shifted to the next collection tube at 10-rain intervals, with each sample thus representing a different 10-min period of the laboratory session. Blood collected for assay was allowed to clot at room temperature for 30 min, refrigerated (4°C) for 1 hr, and centrifuged to separate serum from cells. Serum was stored frozen (-20°C) for later analysis. All samples were assayed in duplicate, and where possible, samples from a given subject were assayed for a given hormone within a single run to minimize interassay variability. Additionally, all of the testosterone and cortisol assays were conducted in a single run. For PRL and LH, half of the assays were conducted in a single run (i.e., either 6 or 7 subjects' assays). Conisol. Samples were assayed for serum cortisol using a direct modified double-antibody method based on that reported by Farmer and Pierce (1974), using a commercially available kit (Pantex, Santa Monica, CA). Briefly, 200 pl of sera is incubated with antibody highly specific for cortisol (cross-reactivity with ll-deoxycortisol < 15%, corticosterone < 2%, all other steroid and protein hormones < 0.1%). A second antiserum (goat) precipitates the cortisol-antibody complex separating bound from free steroid. Performance characteristics of this assay method are sensitivity of 50 pg per sample; intraassay and interassay coefficients of variation are 10.0 and 10.9%, respectively. Prolactin. Serum PRL was assayed using a direct modified double-antibody method based on that reported by Hwang et al. (1971) and commercially available as a kit (Immuno-nuclear Corporation, Stillwater, MN). Serum (100 pl) is incubated with rabbit antihuman PRL antibody and 125 [-PRL. A second antibody (goat antirabbit) precipitates the antiPRL complex into bound and free steroid. Performance characteristics are cross-reactivity of anti-PRL to other protein hormones .05.

RESULTS

Differences Between Groups on Demographic and Psychological Data (SCL-90), Sexual Functioning, and Sexual Experience (PHO and SHF) The experimental and control groups did not differ on any of the demographic, SCL-90, sexual experience or sexual functioning variables.

Behavioral Responses

Genital Responding The two erotic tapes did not produce significantly different levels of VPA response in either group, suggesting that the two erotic tapes elicited approximately equal levels of sexual arousal. Therefore, erotic tape order was collapsed and omitted as a factor in subsequent analyses. Baseline VPA level also did not differ between the two groups. A N O V A of the VPA measures revealed statistically significant effects for Groups, F (1, 11) = 15.85, p = 0.002, indicating that there were overall differences in genital responding betwden the E and C subjects when collapsed over the session. In addition, there was a significant Sample effect, F ( 1 5 , 165) = 3.95, p = 0.0002, such that differences in VPA response resulted over the course of the session. The interaction factor, however, was not significant. These effects are displayed in Fig. 1. Post hoc contrasts by Bonferroni adjusted t tests revealed that the E group showed greater VPA than the C group during exposure to the first, t(11) = 3.05, p < 0.05, as well as the second, t ( l l ) = 2.75, p < 0.05, videotape. There was also a statistically significant difference between the

Sexual Arousal in Women

179

,=~

12-

6

*", lo0.

E E C

8-

6-

= ~

< e~ N

••••=

~Control J I

2

I

o

'

~

l

I

I

8'o

Experimental

J

, I

[

12o

I

!I.

i

14o

Time (min) Fig. 1. Vaginal pulse amplitude (mean and SE) responses across the session for experimental and control groups.

VPA of the two groups at the combined sample points following the termination of the second videotape, t (11) -- 3.76, p < 0.05. Figure 1 shows that this effect was due to a more rapid reduction in the C group VPA following the termination of the second videotape. VPA was not significantly different between the groups at other points during the procedure. The E group appeared to respond more intensely to the second than the first videotape. However, this difference did not achieve significance, t(6) = 1.18, p > 0.05.

Subjective Responding Seven-point Likert-scale questions asking subjects to subjectively rate how "sexually aroused," "mentally sexually aroused," and "physically sexually aroused" they were prior to the first viedotape, during the first videotape, four times at 20-rain intervals between the two videotapes, and during the second videotape were submitted to two-way ANOVAs. These ANOVAs yielded statistically significant overall Group effects for general sexual arousal F ( 1 , 11) = 5.28, p = 0.04, mental, F ( 1 , 11) = 5.35, p = 0.04), and physical, F (1, 11) = 6.01, p = 0.03, sexual arousal, with the experimental group reporting a greater degree of sexual arousal than the control group. Figure 2 shows the results for "sexual arousal."

180

I l e i m a n et

ai.

13) 8e" t~ re i

6-

m (n o

._. 0

2-

,D 0

I

~ I 2()

menta, Control

I 40

I 60

80

100

1:~0

' 140

' 160

Time (min) Fig. 2. Subjective report (mean and SE) of arousal across the session for experimental and

control groups.

The effects for sampling interval were also statistically significant for sexual arousal, F (6, 66) = 26.35, p < 0.0001, mental, F (6, 66) = 20.76, p < 0.0001, and physical, F (6, 66) = 24.13, p < 0.0001, sexual arousal. In addition, the interaction effects were statistically significant in the case of sexual arousal, F (6, 66) = 4.95, p < 0.001, and physical sexual arousal, F (6, 66) = 4.30, p = 0.002. Post hoc contrasts revealed that the two groups differed significantly in their subjective responses to the first but not the second videotape. However, the experimental group did report significantly greater subjective sexual arousal, t(6) = 3.87, p = 0.018, mental, t(6) = 3.24, p = 0.018, sexual arousal, and physical, t (6) = 4.58, p = 0.004, sexual arousal during the second videotape compared to the first videotape. Hormonal Responses A N O V A showed that there were no statistically significant differences between the two groups on any of the four hormones measured at baseline or during the session. The only statistically significant simple effect occurred for PRL, F (13, 143) = 4.94, p = 0.014. Analysis of orthogonal polynomial components suggested that this variance was best described by a decreasing linear trend from the beginning to the end of the experiment, F (1, 11) = 6.65, p = 0.026. Values for each hormone appear in Table I.

Sexual Arousal in Women

181

CD

~¢'~1

t'--I

c-4 t"4

¢'-4

o~c~

% O

L~

c---~

>-~

©

©

E tf~

~+~ tf~c~

Psychophysiological and endocrine responses to sexual arousal in women.

The unexplored possibility that a sexually induced endocrine response might prime further sexual arousal in women guided the current investigation. He...
917KB Sizes 0 Downloads 0 Views