THE LIKELIHOOD OF infection with sexually transmitted diseases (STD) is affected by behavioral risk factors and risk markers. Multiple dimensions of sexual behavior, such as age at first intercourse, the number of lifetime sex partners, partner recruitment patterns, frequency and timing of sexual intercourse, the prevalence of certain sexual practices, and prophylactic and hygienic practices including vaginal douching, tampon use, and contraception influence the risk for STD and their aftermath.1 Physiological changes in women during the reproductive years also are thought to affect the risk of STD. Such changes are related primarily to the menstrual cycle, pregnancy, and contraceptive use.2 The risk of upper tract infection in women appears to be influenced by the menstrual cycle. Evidence from clinical studies suggests that symptomatic gonococcal or chlamydial pelvic inflammatory disease occurs most frequently during the first week of the menstrual cycle.3,4 One reason for the increased risk for infection appears to be the relative penetrability of the cervical mucous during menses; another reason is the reflux of potentially contaminated blood into the fallopian tubes during menstrual uterine contractions.2 For gonococcal infection, two other factors may be responsible: iron, which is abundant in menstrual blood, may promote gonococcal growth, and the type of gonococcus that causes tubal infection may proliferate at the cervix during menstruation.5
The normal flora of the genital tract consists of organisms that colonize the mucosa of normal hosts and play a significant role in defense against infection by genital pathogens. These organisms are often highly susceptible to the local environment and hormonal influence. For instance, the symptoms of trichomoniasis have been noted to appear or escalate more frequently during or immediately after the menstrual period,6 suggesting that the vaginal microenvironment affects the pathogenicity of trichomonads,7 and that the microenvironment changes from time to time in the same person. Further, other factors in the vaginal microenvironment, including menstrual blood, contribute to the expression of pathogenicity.8
Chlamydia trachomatis and mycoplasma hominis are important etiological agents in acute and chronic pelvic inflammatory disease, and these organisms have been demonstrated in the cervix, the endometrium, and the fallopian tubes.9 Retrograde menstrual blood flow from the uterine cavity through the fallopian tubes has been demonstrated in the great majority of women with patent tubes. Hence, microorganisms might spread with such blood, contributing to the development of pelvic inflammatory disease.4 The menstrual cycle also has been shown to effect phenotypic changes in gonococci that may alter gonococcal virulence; therefore, intercourse during the menstrual period may expedite the spread of gonococci from the cervix to the endometrium and fallopian tubes.10 Further, associations with the menstrual period have been suggested for gonococcal and trichomonal infections of the lower genital tract.5–8
More recently, the presence of blood during vaginal intercourse has emerged as a risk factor for heterosexual transmission of human immunodeficiency virus (HIV) in some studies.11–13 Because HIV has been found in menstrual fluids, intercourse during menses places male partners at increased risk for acquiring HIV through heterosexual intercourse.11
If intercourse during menstruation is a risk factor for STD, health interventions may help reduce the incidence of transmission of STD by means of this practice. Knowledge of the prevalence of such risk behaviors and their distribution in the population is necessary for effective planning of prevention strategies. This article reports on the prevalence and distribution in the United States of sexual intercourse during menses, the net effects of individual characteristics on the distribution of this practice across population subgroups, and the empirical association between sexual behavior during menses and self-reported STD history. To our knowledge, these are the first population-based estimates of this behavior.
The data used in this article are from the 1991 National Survey of Women, conducted for the Battelle Memorial Institute by the Institute for Survey Research at Temple University (Philadelphia, Pennsylvania) with a representative sample of 20- to 37-year-old women in the contiguous United States.‡ The total sample of the NSW consists of 1669 women; 728 black and 941 nonblack. The sampling was based on a multistage, stratified, clustered, area probability design, and the black population was oversampled to ensure statistically adequate representation. All interviews were conducted in person by trained professional interviewers. The survey response rate was 71%.‡ Item nonresponse rates were trivial; approximately 0.5% for the questions on sex during the last menstrual period (0.54%) and usual coital behavior during menstruation (0.36%). The final sample was weighted to account for differential selection probabilities, oversampling, and nonresponse. Sample weighting adjusts for the effects of stratification, clustering, disproportionate area sampling, oversampling of blacks, and for the effects of differential nonresponse. Weighted data permit generalizations from the sample survey data to the United States population represented by the sample, subject to sampling and nonsampling errors.
Two outcome variables were used in the analysis, the usual practice of vaginal intercourse during menstruation and vaginal intercourse during the last menses. Also included in the analyses were age, race, relationship status, religion, and region of residence as social and demographic correlates, and education, income, and public assistance receipt as indicators of socioeconomic status. In addition, coital frequency and the life-time number of sex partners were included in the analyses to examine their association with the practice of sexual intercourse during menses and to control for their independent effects on the likelihood of STD infection. The STD outcome is based on self-reports of whether the respondent ever had any sexually transmitted infections. During the interview, women were shown a flash card with a list of six common STD (AIDS-HIV, chlamydia, genital herpes, genital warts, gonorrhea, and syphilis),§ and they were asked to indicate whether they had ever had any one of the diseases shown on the card or any other STD that was not shown on the card. The “prior STD experience” variable is obtained from the self-reports of incidence of the diseases shown on the card, as well as those not shown but specified by the women. In addition to the six diseases shown on the card,§ various other STD (e.g., trichomonas, Gardnerella vaginalis, hepatitis B, chancroid) were mentioned by a small number of women and these were included in the prior STD experience variable.∥
Data analysis was confined to women who had ever had vaginal intercourse. We used the chi-square statistic to test the statistical significance of the differences in the proportion of women who reported having vaginal intercourse (either during their last menstrual period or usually) across categories of social and demographic characteristics and sexual behavior. We used multivariate analysis to control for the effects of other variables and to examine the independent effects of social and demographic characteristics and sexual behavior on vaginal intercourse during menses. Logistic regression was used to estimate the multivariate models.14 Initially, three nested equations were estimated to examine the direct and indirect effects of the covariates on the outcome variables. The first equation included only the background characteristics, the second model added the SES measures to the first model, and the third model added the sexual behavior variables to the second equation. Only the reduced models are presented here.
In 1991, 16% of the women in the sample reported that they had vaginal intercourse during their last menstrual period (Table 1). This practice was reported by a significantly greater proportion of white women than of black or Hispanic women. There was no age pattern with respect to intercourse during the last menstrual period. Reported vaginal intercourse during the last menstrual period was almost twice as high among women with at least some college education as among those with no college education, and it was most frequent among women who were cohabiting and least frequent among those with no regular sex partner. The proportion of women reporting this practice was lowest among women who were affiliated with conservative Protestant denominations and most common among non-Christians and those with no religious affiliation. Practice of vaginal intercourse during the last menstrual period also was higher among women with higher income, women living in the West, women who had intercourse more than twice a week, and among women who had seven or more lifetime sex partners.
The proportion of women who reported having vaginal intercourse during menses “usually” was 26%. Patterns of usual practice were similar to patterns of recent practice. White women, women of Hispanic origin, women with education beyond high school, women who were cohabiting or who had a regular sex partner, non-Christians, women without religious affiliation, and Catholic women were most likely to report usually having sexual intercourse during menses. Like vaginal intercourse during the last menses, the usual practice of intercourse during menstruation also was higher among women with higher income, women living in the West, women who had coitus more than twice a week, and women who had seven or more lifetime sex partners (Table 1).
The multivariate analysis largely confirmed the bivariate associations (Table 2). After controlling for the effects of all other variables, the likelihood of having had sexual intercourse during the last menses was significantly higher among white women, married and cohabiting women, women who had a regular sex partner, and women who lived in the western United States. The likelihood of sex during menses significantly increased with education, income, coital frequency, and the number of sex partners. Women who, on the average, had intercourse more than once a week and women who had a lifetime number of seven or more sex partners were considerably more likely to have had intercourse during their last menstrual period than women who had sex less frequently (once a week or less) and women who had fewer than seven lifetime partners. There was no significant effect of age on whether women had intercourse during the last menses. The significant bivariate association observed between religion and sex during menstruation all but disappeared in the multivariate context, especially after the sexual behavior variables were included in the model.
The results from the multivariate analysis of the usual behavior patterns of women regarding sex during menstruation were similar to those obtained from the analysis of the most recent behavior. After controlling for the effects of other variables, higher education, frequent intercourse, and multiple sex partners significantly increased the likelihood of intercourse during the menstrual period. No statistically significant effects were observed for race-ethnicity, religion, region, or income. In contrast to its effect on recent behavior, age was strongly related to the usual coital behavior during menstruation. Women younger than 35 years were significantly more likely to have intercourse during the menstrual period than women 35 years of age and older. In fact, the practice of this behavior seemed to increase with age until approximately age 35 years and to decline thereafter. Finally, when the effects of other factors were controlled for, married or cohabiting women were less likely to report usually having sexual intercourse during their menstrual period than were single women.
Last, we examined the relative risk of self-reported prior STD experience according to sexual behavior during menses.∥ These data show that the cumulative prevalence of self-reported STD experience was significantly higher among those who have sexual intercourse during their menses, as well as among those women who had intercourse during their last menstrual period (Table 3). The odds of a prior STD experience is nearly three times as high among those who usually have intercourse during their menstrual period as among those who do not. The relative risks remain significantly different in the two groups even after we controlled for the effects of age, race, ethnicity, religion, marital and relationship status, education, income, welfare status, lifetime number of sex partners, and coital frequency. Although neither the usual nor the most recent behavior of the women can be causally related to prior incidence of STD with this data set, the results confirm the postulated hypothesis that there is an association between intercourse during menses and STD infection.
These findings have interesting implications regarding sexual behavior and the prevention of sexually transmitted infections. Although only approximately one fourth of sexually active women in the United States represented by this sample generally engage in coital activity during their menstrual period, this practice was significantly more common among certain population groups. Notably, black women, older women, and less educated women were more likely to refrain from intercourse when they had menstrual bleeding. A number of cultural and normative factors regarding sexual hygiene and personal cleanliness may account for these behavioral differences. This practice also appears to be more common among Catholic women than would be expected. It might be that, because the Catholic Church proscribes all birth control except natural family planning, sexual intercourse during the menstrual period is more common among Catholic women exactly because it conforms with the church doctrine on family planning. The low prevalence among conservative Protestants might be attributed to their strict adherence to the canons of the Old Testament, which specifically prohibit this practice. This is reflected in the regional variation of this practice as well, with lower prevalence in the south and midwest regions of the United States, where affiliation with conservative Protestant doctrines is more common. Further, intercourse during menstruation is associated closely with regular and frequent sexual intercourse and large number of sex partners-two behaviors that are among the principal components of the transmission dynamics of STD.15
It also is apparent from these data that there is a strong association between the practice of sexual intercourse during menstruation and STD history. However, although at the individual level the data demonstrate a strong and statistically significant positive association between sexual intercourse during menstruation and selfreported STD history, in the aggregate this practice was more common among subgroups marked by low STD prevalence and incidence.16 Further, this behavior is less prevalent in the south and midwest regions, which are marked by higher STD rates, than it is in regions in which the STD rates are relatively lower.16 There are two possible reasons for this seeming anomaly: First, it may be that the relationship between STD and sexual intercourse during menstruation is a spurious one caused by the correlation between coital frequency, partner accumulation, and sex during menses. Yet, the multivariate results do not support this conclusion. After controlling for the potentially confounding effects of these two sexual behavior variables, intercourse during menstruation still is associated with more than a twofold increase in self-reported STD history. Second, it is possible that the positive relationship between STD history and sexual intercourse during menstruation results, in part, from a reporting bias. We rely on self-reports of sexual behavior and STD history, and it is commonly accepted that such reporting is often subject to a social desirability bias. Our data suggest that nonconservative women who are open minded about their sexuality were more likely to report sex during menses than were more conservative women. In addition, perhaps these same women were more likely to report their past STD experiences. Further, it is likely that women among whom sex during menses was more common were also more likely to recognize STD symptoms and more likely to have had regular health care visits during which such symptoms were diagnosed.
Like most survey data, these data have certain limitations. Despite these limitations, our data provide the first population-based evidence on the distribution of usual and recent sexual intercourse during menstruation by demographic, social, and behavioral characteristics. Moreover, such limitations notwithstanding, the data demonstrate a distinct association between sexual intercourse during menses and STD experience. Unless future studies are able to show that this practice does not increase the risk for any sexually transmitted pathogens, prevention strategies should at least inform sexually active persons about the risks of having unprotected vaginal intercourse during menstruation. Such prevention efforts should be focused particularly on persons who have large numbers of sex partners, who have sexual intercourse frequently, and on young women, white women, and on women with relatively higher levels of education and income.
1. Aral SO. Sexual behavior as a risk factor for sexually transmitted disease. In: Germain A, et al, eds. Reproductive tract infections. New York: Plenum Press, 1992:185–198.
2. Ehrhardt AA, Wasserheit JN. Age, gender, and sexual risk behaviors for sexually transmitted diseases in the United States. In: Wasserheit JN, Aral SO, Holmes KK, eds. Research issues in human behavior and sexually transmitted diseases in the AIDS era. Washington, DC: American Society for Microbiology, 1991:97–121.
3. Eschenbach DA, Harnisch JP, Holmes KK. Pathogenesis of acute pelvic inflammatory disease: Role of contraception and other risk factors. Am J Obstet Gynecol 1977; 128:838–850.
4. Sweet RL, Blankfort-Doyle M, Robbie MO, Schacter J. The occurrence of chlamydial and gonococcal salpingitis during the menstrual cycle. JAMA 1986; 255:2062–2064.
5. James J, Swanson J. Color opacity colonial variants of Neisseria gonorrhoea
and their relationship to the menstrual cycle. In: Brooks GF, Gotschlich EC, Holmes KK, Sawyer WD, Young FE, eds. Immunobiology of Neisseria gonorrhoea.
Washington, DC: American Society for Microbiology, 1978:338–343.
6. Catterall RD. Trichomonal infection of the genital tract. Med Clin North Am 1972; 56:1203–1205.
7. Rein MF, Muller M. Trichomonas vaginalis and trichomoniasis. In: Holmes KK, Mard PA, Sparling PF, et al, eds. Sexually transmitted diseases. 2nd ed. New York: McGraw-Hill, 1990:481–492.
8. Cohen MS, Weber RD, Mardh P-A. Genitourinary mucosal defenses. In: Holmes KK, Mard PA, Sparling PF, et al, eds. Sexually transmitted diseases. 2nd ed. New York: McGraw-Hill, 1990:117–127.
9. Westrom L, Mard PA. Acute pelvic inflammatory disease (PID). In: Holmes KK, Mard PA, Sparling, PF, et al, eds. Sexually transmitted diseases. 2nd ed. New York: McGraw-Hill, 1990:593–613.
10. Brunham RC, Holmes KK, Embree JE. Sexually transmitted diseases during pregnancy. In: Holmes KK, Mard PA, Sparling PF, et al, eds. Sexually transmitted diseases. 2nd ed. New York: McGraw-Hill, 1990:771–801.
11. Alexander N. Sexual transmission of human immunodeficiency virus: Virus entry into the female genital tract. Fertil Steril 1990; 54:1–18.
12. De Vincenzi I. Longitudinal study of human immunodeficiency virus transmission by heterosexual partners. N Engl J Med 1994; 331:341–346.
13. Cohen JB, et al. Sexual and other practices and risk of HIV infection in a cohort of 450 sexually active women in San Francisco. Abstract Book, III International Conference on AIDS. Washington, DC; 1987; WP.57.
14. Hanushek EA, Jackson JE. Statistical methods for social scientists. New York: Academic Press, 1977.
15. Anderson RM. The transmission dynamics of sexually transmitted diseases: The behavioral component. In: Wasserheit JN, Aral SO, Holmes KK, et al, eds. Research issues in human behavior and sexually transmitted diseases in the AIDS era. Washington, DC: American Society for Microbiology; 1991;38–60.
16. Division of STD/HIV Prevention. Sexually transmitted disease surveillance, 1993. US Department of Health and Human Services, Public Health Service. Atlanta: Centers for Disease Control, December 1994.