Sexual behavior over the past year of 32 outpatients with Bipolar disorder is compared to that of 44 Comparison patients that had never had an episode of affective illness. Subjects were outpatients treated with drugs and psychotherapy in routine office practice. Differences in sexual behavior between the two groups as a whole were minimal, but meaningful differences emerged when subgroups were compared. Compared to control men, Bipolar men had had more partners in the last year and were more likely to have had sex without condoms. Compared to Bipolar females, Bipolar males had more sex partners, had more sex with strangers, and were more likely to have engaged in homosexual behavior. Even so, some patients in the Comparison group also had engaged in risky sexual behavior. They had failed to use condoms and had had sex with strangers and prostitutes during the previous year.
This paper discusses a study of sexuality in Bipolar outpatients treated in private practice in comparison to outpatients without affective disorder or schizophrenia.
The evidence that sexuality is often increased in adults with Bipolar disorder during hypomanic or manic episodes includes controlled and uncontrolled studies and clinical case reports [
The study we discuss below was designed to assess a dimension of sexuality not included in most research to date. We were interested in sexuality as part of the person’s temperamental trait behavior, rather than simply being influenced by affective state. We sought to assess whether sexuality was different among individuals with Bipolar disorder even when they are well treated and stable compared with those without affective disorder. As part of the preparation for the study, our group reviewed the literature on sexuality in Bipolar disorder, which we now discuss.
In a retrospective chart review study of 1,000 Bipolar II patients, Akiskal [
Of the uncontrolled studies, Jamison et al.’s study [
Recognizing the value of this approach, we were prompted to investigate diverse aspects of sexuality in relatively stable Bipolar individuals. We also wished to study a group of Bipolar patients underreported in the literature—those treated in routine office practice.
We compared a Bipolar patient group to other psychiatric patients with no affective disorder seen in a private practice setting, with the goal of increasing knowledge about their sexual behavior. We expected Bipolar patients to have more sexual thoughts and behaviors than other patients treated by private practitioners, even when Bipolar patients were euthymic and on medication.
One purpose in carrying out this investigation was to utilize a patient population different from that usually reported in scientific studies. Psychiatrists in private practice often obtain valuable information that is not routinely published. We collected such information by systematically querying each psychiatrist in our group about the characteristics of patients that he/she knew well, having treated them for months and even years. The data collected would therefore supplement that usually reported in studies of more conventional design. Participating psychiatrists were all psychodynamically trained and provided both psychotherapy and psychotropic medication to the patients studied. They knew the patients in depth. This was an advantage since subjects may be reticent to report sexual information in questionnaire studies or to impart such information to researchers with whom they have no relationship. Since data were coded and only reported in aggregate, no individual patient could be identified. All study procedures were approved by the New York State Psychiatric Institute Institutional Review Board. Participants gave written informed consent.
Patients (
Psychiatrists had practices in the Far West (1), Rocky Mountain states (1), and Northeastern United States (5). All Bipolar patients and 59 percent of Comparison patients were taking at least one psychotropic medication (see below for details).
In order to be reported on, patients had to be 18 years of age or older, to have attended at least four appointments in the 12 months before the study, and to have been seen during the last five business days before the study data collection period. Each psychiatrist reported on
Psychiatrists filled out a questionnaire for each patient, based on what they knew about him or her (see below). Although they did not further query any patient in order to answer the questions, they consulted patients’ charts as needed. We felt it would be intrusive for psychiatrists to suddenly inquire about specific aspects of their sexual behavior.
The study questionnaire, titled
All DSM IV-TR diagnoses were listed for each patient [
IBM SPSS was used for data analysis. In preliminary analyses we examined the distribution, means, and standard deviations of the demographic and other continuous measures, as well as the distributions of categorical measures. All but three continuous measures were found to be normally distributed (see below). There was a series of questions regarding specific sexual practices (Q43–51) for which “Don’t know” was the most common response. These data are not presented.
Seven questions (Q36–42) address the “Nature of Partners In The Past Year,” ranging from primary partners to strangers to paid phone/internet sex. The affirmative answers to these questions were summed and considered a partner risk measure. Items Q55, 57, and 58 (substance use, sex with strangers, and anal/vaginal intercourse without condoms with nonmonogamous partners) were also summed as a risk-taking measure.
Continuous measures were compared across the two patients groups using the
Psychiatric diagnoses are presented first since diagnosis (plus the treatment characteristics noted above) was the criterion by which patients were selected.
The Bipolar group consisted of individuals diagnosed with a primary diagnosis of one of the Bipolar disorders according to DSM-IV-R criteria. Total
The Comparison group of psychiatric outpatients consisted of individuals who never had an episode of affective illness, either Bipolar disorder or unipolar depression. Total
A number of patients in each study group had more than one Axis I disorder. The Bipolar group of
Some patients also had Axis II disorders. Of the Bipolar group, 8 patients had received diagnosis of diverse Axis II conditions.
Of the Comparison group, 16 patients had received Axis II diagnoses.
Subject demographics are listed in Table
Demographics: statistics are the
Demographic measures | Bipolar disorders | Comparison | |||||
---|---|---|---|---|---|---|---|
|
|
Group effect | |||||
Mean | sd | Mean | sd |
|
df |
| |
Age | 41.6 | (13.2) | 42.2 | (14.3) | 0.17 | 74 | 0.869 |
Hollingshead occupation code | 7.4 | (1.5) | 8.1 | (1.2) | 2.10 | 67 | 0.040 |
Hollingshead SES | 55.6 | (8.6) | 60.1 | (6.7) | 2.43 | 67 | 0.018 |
|
|||||||
% |
|
% |
|
Fisher’s Exact | |||
|
|||||||
Gender (female % and |
50 | (16) | 41 | (18) | 0.488 | ||
Education level | |||||||
High school/GED | 3 | (1) | 0 | (0) | |||
Some college | 9 | (3) | 7 | (3) | |||
B.A./B.S. degree | 47 | (15) | 36 | (16) | |||
Graduate school | 41 | (13) | 57 | (25) | 0.364 | ||
Current employment status | |||||||
Retired | 9 | (3) | 5 | (2) | 0.644 | ||
Homemaker | 16 | (5) | 2 | (1) | 0.077 | ||
Unemployed | 16 | (5) | 9 | (4) | 0.480 | ||
Student | 9 | (3) | 14 | (6) | 0.726 | ||
Race | |||||||
White | 88 | (28) | 96 | (42) | |||
Black | 3 | (1) | 0 | (0) | |||
Asian/PAC. ISL | 3 | (5) | 5 | (2) | |||
Other | 6 | (2) | 0 | (0) | 0.215 | ||
Hispanic | 13 | (4) | 2 | (1) | 0.233 | ||
Marital status | |||||||
Married | 38 | (12) | 48 | (21) | |||
Single | 53 | (17) | 46 | (20) | |||
Widowed/divorced | 9 | (3) | 7 | (3) | 0.678 | ||
Live-in partner | 56 | (18) | 57 | (25) | 1.0 | ||
Religion | |||||||
Protestant | 28 | (9) | 30 | (13) | |||
Catholic | 19 | (6) | 14 | (6) | |||
Jewish | 28 | (9) | 43 | (19) | |||
Other | 6 | (2) | 7 | (3) | |||
None | 19 | (6) | 7 | (3) | 0.460 |
Educational level for the patients in both groups was high. The majority of patients were college graduates or had a graduate degree. The Comparison group had a higher occupational status according to the Hollingshead Scale [
Psychiatric history and clinical characteristics.
Clinical measures | Bipolar disorders | Comparison | |||||
---|---|---|---|---|---|---|---|
|
|
Group effect | |||||
Mean | sd | Mean | sd |
|
df |
| |
Age in years @ onset 1st |
15.7 | (5.8) | 21.3 | (11.7) | 2.72 | 67 |
0.008 |
Global assessment of function | |||||||
Average over the past year | 68.7 | (8.7) | 72.6 | (9.6) | 1.80 | 74 | 0.076 |
Lowest week during last year | 54.8 | (13.7) | 64.9 | (12.1) | 3.41 | 74 | 0.001 |
Duration of current treatment (months) | 77.3 | (107.1) | 46.6 | (67.4) | 1.43 | 74 |
0.158 |
Hypomanic and manic symptoms | 5.7 | (1.5) | 0.14 | (.50) | 18.16 | 46 |
<0.001 |
|
|||||||
% |
|
% |
|
Fisher’s Exact | |||
|
|||||||
Episodes during the past year | |||||||
Manic episodes | 6 | (2) | 0 | (0) | 0.174 | ||
Depressive episodes | 31 | (10) | 0 | (0) | 0.000 | ||
Mixed episodes | 50 | (16) | 0 | (0) | 0.000 | ||
Hospitalizations (lifetime) | |||||||
0 | 63 | (20) | 98 | (41) | |||
1 | 19 | (6) | 2 | (1) | |||
2 or more | 19 | (6) | 0 | (0) | <0.0001 | ||
Current psychotropic medications | |||||||
Antipsychotics | 44 | (14) | 7 | (3) | <0.001 | ||
SSRI | 47 | (14) | 32 | (14) | 0.228 | ||
Antidepressants | 23 | (7) | 14 | (6) | 0.363 | ||
Mood stabilizers | 72 | (23) | 0 | (0) | <0.001 | ||
Anxiolytics | 48 | (15) | 41 | (18) | 0.638 | ||
Stimulants | 13 | (4) | 9 | (4) | 0.711 | ||
Sleep | 19 | (6) | 18 | (8) | 1.00 | ||
Phosphodiesterase | 10 | (3) | 2 | (1) | 0.300 | ||
Other | 16 | (5) | 5 | (2) | 0.129 | ||
Number of medications prescribed |
2.8 | (1.3) | 1.3 | (1.3) | 5.20 | 74 | <0.001 |
Bipolar family history |
53 | (16) | 5 | (2) | <0.001 | ||
Significant medical illness (during the last year) | 34 | (11) | 30 | (13) | 0.803 |
Other measures of psychopathology also showed the Bipolar group to be generally more severely affected. Thus, mean age of onset of psychiatric symptoms in the Bipolar group was 15.7 years while the Comparison group had onset at 21.3 years. Duration of present treatment differed by more than 30 months on average with Bipolar individuals having been in treatment longer. This was not a significant difference due to large standard deviations. More patients with Bipolar disorder were reported to be taking psychiatric medications than Comparison patients (100% versus 59%; Yates
Patients with Bipolar disorder were more likely to be taking antipsychotics and mood stabilizers. There were no differences between the patient groups in percentages taking SSRI antidepressants, other antidepressants, anxiolytics, stimulants, sleep medications, and phosphodiesterase inhibitors.
Psychiatrists were asked whether the patient’s parents, siblings, or children had definite or probable Bipolar disorder. Over 50% of B patients but only 4.5% of C patients had at least one Bipolar relative related by blood, a highly significant difference.
The Bipolar patients were more likely to report sexual experiences that may be described as nontraditional than were the Comparison subjects. Although this might not have been revealed in comparisons of individual items because of relatively small numbers of subjects, the
In other ways the groups were similar in sexual behavior over the previous year. For instance, patients in both groups reported an average of “moderate enjoyment” of their sex lives (Q32). Mean frequency of sexual activity with another person over the last year was “sometimes” defined as two to eight times/month (Q34).
Questions 43–51 and 53-54 concerned the frequency of certain sexual behaviors/experiences—masturbation, orgasm, sexual desire, and novelty-seeking sexual behavior—for the year and for the most intense week of the year. For these items there were too much missing data for analyses to be meaningful, and the results are not presented here. This was a result of research design. The practicing clinicians did not always have readily available details about the patients’ most intense week in the last year. If the questionnaire had been administered to patients directly, this information would have been straightforward to elicit.
Another type of question assessed diagnosis of a DSM-IV Sexual Dysfunction or Paraphilia (Q52). Very few patients in either group had received one of these diagnoses (Bipolar
Sexual risk-taking was assessed with six questions (Q55–60). We found no significant differences between Bipolar and Comparison groups for any of the items. The first four were frequency with which they engaged in recreational substance use with sex (Q55), substances they used if any (Q56), sex with strangers (Q57), and receptive anal or vaginal intercourse without a condom with partners of unknown HIV status (Q58). One person in each group was known to have positive HIV status, not a significant difference. There was also no difference in how many patients had been diagnosed with any other STD in the last year.
Sexual orientation identity and gender identity were assessed with four questions (Q61–64). Sexual
A separate question asked was for the psychiatrists to rate patients’ sexual
Sexual orientation self-identity in the last year (Q63) was assessed separately from the Kinsey scale questions and was found not to differ significantly between groups. Most patients were reported by their psychiatrists as having a “straight” identity. The percentages with a self-identity as gay or lesbian were about the same for both groups—16 percent versus 18 percent (Bipolar versus Comparison). One Bipolar person was reported to have a bisexual identity and one to have “other.” Gender identity in the last year was exclusively male or female in all cases except for the one Bipolar patient reported as transgender, male to female.
Special circumstances which might affect sexual functioning were assessed with five questions (65–69). Hormonal status was assessed for women in both groups and did not differ. Most women were premenopausal.
When the entire Bipolar and Comparison groups, both males and females, were compared, there were no special circumstances over the last year reported that might affect sexual functioning (falling in love, pregnancy, ill health, and sexual abuse) that differed between groups.
A question (70) assessed bipolarity in the patients’ relationships. One patient in each group was reported to definitely or probably have a Bipolar partner.
Additional information sometimes emerges when statistical tests are performed on subgroups of interest. We looked at two additional comparisons for sexual history items. We compared patients by gender so that Bipolar males were compared to Comparison males, and Bipolar females were compared to Comparison females. We also compared Bipolar males with Bipolar females and Comparison males with Comparison females.
For the comparison between men, more Bipolar patients were reported to have had additional lovers (
Finally, meaningful differences in the sexual behavior of the men versus women in each group were examined. Bipolar males differed from Bipolar females on two items only. More Bipolar males than females had had two or more sexual partners
Comparisons of Comparison males to females showed that Comparison males had engaged in more frequent sexual activity during the most intense week over the last year (
Another way to look at the sexual risk-taking among this outpatient sample is to note how many people had engaged in these behaviors
This naturalistic comparison study of the sexual experience and activity of Bipolar patients had unique design features. We hoped that our research approach might demonstrate the usefulness of systematic investigation of patients not usually studied and of harvesting data that is often not shared in the professional community. In our view the method we devised could be adapted for the investigation of diverse types of patients other than those with Bipolar disorders and of diverse forms of behavior other than sexuality.
To the best of our knowledge there has been no previous controlled study of sexuality in Bipolar outpatients carried out in such depth and detail. Indeed studies of sexuality in psychiatric patients have been scant. In the case of our own study, the patients received both psychotherapy and medication from their doctors and had largely durable and positive working alliances with them. This is clear from the mean duration of treatment with the psychiatrists—almost four years for the Comparison group and over six years for the Bipolar group. To learn about sexuality, the topic of the questionnaire discussed in this report, the trusting therapeutic relationships that our respondents had with these patients proved crucial.
The psychiatrists who provided study data were all members of the Human Sexuality Committee of Group for the Advancement of Psychiatry. Based on an inclusive review of the literature in Bipolar disorder, the group had constructed a questionnaire, which they subsequently filled out about the patients. Respondents had no hypothesis about which specific aspects of sexual experience and activity would be increased.
A crucial question about the sample reported on in this study is whether their psychiatric disorders were accurately diagnosed. The usual indices of diagnostic accuracy are reliability and validity. No effort was made to establish diagnostic reliability in our study using commonly accepted methods. Psychiatrist respondents were experienced in using the DSM-IV-TR and carefully reviewed diagnostic criteria for each subject selected.
A number of measures indicate that the diagnosis of the Bipolar patients was
Given that individuals with schizophrenia and mood disorders were excluded from the Comparison group, it is not surprising that the Bipolar group had had more episodes of severe illness during the course of their lives than the Comparison group. Over one third of Bipolar patients (
Although the current GAF score between the two groups was similar, those in the Bipolar group were likely to have been more impaired during the worst week of the past year (mean score of 54.7 or “serious impairment”), while mean score for Comparison patients during the worst week was 64.8 or “moderate symptoms.”
Mean age of onset of psychiatric symptoms in the Bipolar group was significantly younger than in the Comparison group (15.6 versus 21.2 years), compatible with the overall trend of psychopathology in the sample.
In keeping with what has been reported elsewhere [
The fact that the Bipolar patients and Comparison patients were well-balanced with respect to use of SSRI medications is particularly important given the well-known sexual inhibitory effects of these drugs [
It is important for diagnostic accuracy that no patient with unipolar depression was included in the Comparison group. In two surveys of members of the National Depressive and Mani-Depressive Association (DMDA), most Bipolar patients were initially misdiagnosed [
Of the Bipolar patients in our study, nine (28 percent) had more than one Axis I disorder and eight (25 percent) had Axis II disorders. The Comparison patients were similar in this regard, with nine (20 percent) having more than one Axis I disorder and 16 (36 percent) having an Axis II disorder.
These findings are consistent with what might be expected from published studies showing significant comorbidity in Bipolar individuals [
Our group of psychiatrists had discussed the literature on Bipolar disorder and sexuality among ourselves and expected our Bipolar patients to be more sexual in some sense than controls. This expectation was only modestly supported by the pooled data, however.
The group of Bipolar patients had more sexual partners than did Comparison patients and increased frequency of risky sexual behavior.
In addition, Bipolar males had more sexual partners during the past year than Bipolar females and more Bipolar males had engaged in homosexual activity. Bipolar males were also somewhat more likely than Comparison males to have engaged in receptive intercourse without condoms during the past year.
We note that risky sexual behavior has been reported to be increased in psychiatric patients generally [
Our investigation did find certain risky behaviors to be increased particularly in the males who were Bipolar, but these findings were modest. One reason for this may well have been that every patient with Bipolar disorder in this investigation was medicated. Although the patients were treated with psychotherapy as well, we conjecture that the dimensions of sexual motivation that were embedded in hypomanic or manic affective tendencies were diminished by pharmacological agents. It might also be that the particular patients reported on in this sample, although clearly Bipolar, were not drawn from the most extreme segment of the illness spectrum, particularly at the time of data collection.
An astonishing and clinically important finding of this investigation was the frequency of unconventional and risky sexual behavior among the Comparison patients. This included sex with strangers, failure to use condoms with partners not known to be monogamous, and sex with prostitutes. This indicates the importance of the sexual history in the general assessment of psychiatric patients, not just Bipolar individuals.
Our study had a number of limitations. The convenience group of patients that the therapists reported on was not representative of Bipolar patients generally. The fact that the patients were largely Caucasian and well-educated and able to pay for private treatment limited generalizability of our findings. The greatest strengths of our design were also its greatest weaknesses. Therapists with preconceived beliefs about the questions being studied provided data about their patients. Obviously this might have biased responses. Interestingly, however, the therapists’ expectations were frequently not confirmed. The advantage of using the therapeutic alliance to study sexuality seemed apparent to our group. In order to protect the alliance, closed ended questions about sexuality were not part of our design. Therapists provided only historical data that they had acquired from their psychotherapeutic work with patients. We judged that the truthfulness of patient responses, however, outweighed the limitations of this approach. Due to practical limitations we focused only on the past year. The sexual history of patients prior to that was not explored. The longer arc of sexuality in Bipolar individuals over the life cycle remains for others to explore. Another limitation is that we did not use independent standardized measures to document possible temperamental differences between the two groups and to provide additional mood data. This would be straightforward to do in a future study. Similarly, it would be straightforward to administer the questionnaire directly to Bipolar and control individuals in a research clinic, thus obtaining much more information about the fine details of sexual variations (e.g., during the most sexual week) over the time period studied. In conclusion we stress that our study might be thought of at two levels: as research about sexuality in Bipolar patients and as a feasibility study of carrying out research in a private practice setting.
All authors were members of the Human Sexuality Committee of the Group for the Advancement of Psychiatry (GAP). Jennifer Downey and Richard C. Friedman were Chairpersons of the committee.
The authors declare that they have no competing interests.
The authors acknowledge the insight, advice, and steadfast help of Drs. Jean Endicott, Ph.D., and Ray Goetz, Ph.D., of the Clinical Phenomenology Division, New York State Psychiatric Institute/Department of Psychiatry of Columbia University College of Physician & Surgeons, and of Dr. Stewart Adelson, M.D., of the Departments of Psychiatry, Columbia University and Weill-Cornell School of Medicine. The authors appreciate the careful reading of an earlier version of the paper by the Publications Committee of the Group for the Advancement of Psychiatry.