Throughout much of the United States, rates of human immunodeficiency virus (HIV) and other sexually transmitted infections (STIs) are dramatically higher among blacks than whites. The reasons for these marked racial disparities remain poorly defined.
Sexual networks and patterns of partner mixing play a critical role in the spread of sexually transmitted infections (STIs) throughout a population. 1–6
One pattern that can accelerate the transmission of STIs is concurrent (overlapping) sexual partnerships. 7 Compared with sequential monogamy, concurrent relationships permit more rapid spread of an STI through a population because an individual becoming infected by one partner already has other partners to infect. In addition, an individual’s partners who entered the concurrent partnership earlier are put at risk for infections from that individual’s subsequent partners. 8 This is in contrast to serial monogamy. Mathematical models demonstrate that concurrent partnerships can dramatically accelerate spread of an STI among partner networks immediately after the pathogen enters the population. 9 Thus, a higher prevalence of concurrent partnerships among blacks could be a factor in their higher incidence of heterosexually acquired HIV. We hypothesize that the prevalence of concurrency is higher among blacks than among U.S. non-Hispanic whites due to blacks’ much lower marriage prevalence. The lower marriage prevalence is likely a consequence of their adverse socioeconomic context and much lower sex ratio (ratio of men to women). 10–12
We analyzed data from Cycle V of the 1995 National Survey of Family Growth (NSFG) to determine the prevalence, distribution, and correlates of concurrent partnerships among women of childbearing age in the United States. The NSFG is conducted periodically by the National Center for Health Statistics of the Centers for Disease Control and Prevention to collect data on factors affecting pregnancy and women’s health in the United States. 13 Data collection for Cycle V took place during January through October 1995. Interviews were conducted by trained interviewers using computer assisted personal interview technology. To enhance recall ability for dates, respondents constructed a calendar showing the month, year, and respondent’s age from birth through the date of the interview of important life events.
The sampling frame for the 1995 NSFG consisted of all 25,534 civilian, non-institutionalized U.S. women age 15–44 years who lived in 21,168 households that responded to the 1993 National Health Interview Survey (NHIS). 14 The NSFG sample contained one woman from each of the 5,341 NHIS households with an apparently eligible Hispanic or black woman and one woman from each of about 55% of the NHIS households with eligible women who were neither Hispanic nor black. The 79% response rate yielded 10,847 completed cases. Sample weights were computed to adjust for subsampling, nonlocation, and nonresponse, and further adjusted to match the Census estimates of the U.S. population by age, race, marital status, and parity. Weighted analyses yield estimates for the approximately 60,201,000 civilian, non-institutionalized women in the United States in 1995. 14
Dates of Sexual Partnerships
The NSFG interview asked each respondent the month in which she first had sexual intercourse with: her first partner, spouse, each man with whom she cohabited, and any other sexual partners she had since January 1991. Cohabitation was defined as living with a man with whom the respondent was having a sexual relationship; “sexual relationship” in this and other contexts was defined as a relationship involving penile-vaginal intercourse. The month of last sexual intercourse was also asked for non-marital partners who were not regarded as “current.”
Definition of Concurrent Partnerships
We defined concurrent partnerships as two or more partnerships that were “current” or where first sexual intercourse with one partner occurred before the month of last sexual intercourse with another partner. Partnerships that ended by January 1991, the start of the study period, were ignored. Because the NSFG contains limited information about the continuity of sexual relations between dates of first and last sex, we treated all partnerships as continuous from the month of first sex to the month of last sex. For about 15% of the women classified as “concurrent” (and included in the analyses), their only concurrent partnership(s) began before January 1991; thus some of these women may have had no actual overlap in partners during the study period.
We operationalized the above definition by using computer routines to classify partnership histories. To verify that the computer algorithms were performing correctly, we inspected the records for 100 women who had been computer-coded as “concurrent” (99 classifications were confirmed) and 50 women reporting more than one partner since January 1991 but computer-coded as “not concurrent” (49 classifications were confirmed). When the month of first sexual intercourse with a partner was not provided, we substituted (where possible) the earlier of the month the respondent and her partner began cohabiting or married. Similarly, when the month of last intercourse was not available, we used (when available) the month of final separation, divorce, or death of the partner, whichever was earliest. If concurrency status could not be determined due to missing partnership dates, then the woman was dropped from concurrency analyses.
Imputation and Visual Review
The problem with the above imputation procedure was that in a significant number of cases information for one kind of partnership (eg, spouse) appeared to have been incorrectly placed in a record location reserved for another kind of partnership (eg, casual partner), creating the appearance of two different, overlapping partnerships (one missing the month of first and last intercourse). Imputation in such a situation produces spurious concurrency. Because 35% of records ultimately classified as having a concurrent partnership required imputation, we visually reviewed all records where imputation created an appearance of concurrency. We also reviewed all records in which two partnerships had identical start months and end months, as these records might simply contain the same partnership data entered twice. In all, over 1,500 observations, including the majority of those deemed “concurrent,” were classified following visual review.
Analyses with Covariables
We examined the association of concurrent partnerships with the following variables: ethnicity, age at interview, education, marital status, lifetime number of sexual partners, number of partners since January 1991, age at first voluntary vaginal intercourse, work status, and receipt of Aid to Families with Dependent Children (AFDC), as well as combined family income from all sources in the 12 months prior to the survey (as a percent of the 1994 poverty line), which we grouped into four categories. 15 Data on education and income were used only for women over age 21. 13 The 10 American Indians were omitted from analyses that controlled for ethnicity.
We used stratified analyses and multiple logistic regression models with concurrency as the dependent variable and ethnicity as the “exposure” of primary interest to assess residual associations, after inclusion of potential explanatory variables of a priori interest that were associated with concurrency. (Poverty, receipt of AFDC, and education were assessed as indicators of socioeconomic status. Age and marital status are associated with sexual partner number 16; early age of sexual debut is associated with sexual risk behavior later in life. 17) Collinearity was not a problem because correlation coefficients for all pairs of variables used in the same model were below 0.50, and no principal component had a variance proportion at or below 0.8 on more than one variable. 18 Neither the estimated coefficients nor the standard errors were inflated, suggesting absence of numerical problems. 19 We examined models with various subsets of explanatory variables in the complete data set and in a subset chosen for greater homogeneity (women age 22 and older, with age at first intercourse between 12 and 24 years, and sexually active for at least 5 years). Variables with multiple levels (age, age at first intercourse, education, income) were analyzed as categorical variables in some models and as continuous variables in others.
SAS (versions 6.12 and 8.01, SAS Institute, Inc.) was used for data management and descriptive analyses. SUDAAN (version 7.5.2, Research Triangle Institute) was used for statistical analyses to account for the complex sampling design and weighting. Design effects for prevalence estimates were mostly below 1.5.
Distributions of ethnicity and other demographic characteristics of women in the NSFG are shown in Table 1. Marked, though familiar, ethnic disparities in poverty are evident: blacks and Hispanics were most likely to report annual incomes of less than 150% of the poverty line (41% and 40%, respectively) and to receive AFDC (23% and 15%, respectively); whites were least likely to have incomes below 150% of the poverty line (16%) and to receive AFDC (4%).
Sexual Relationship History
African American women were much less likely to be married at the time of interview (25%) than were whites (54%), Hispanics (47%), or Asian Americans or Pacific Islanders (49%) (Table 2). The differences in percentages across the ethnic groups were essentially unchanged by the classification of unmarried cohabiting women as married.
Median age at first sexual intercourse was 16 or 17 years. Black women tended to initiate intercourse earlier than other groups (38% before age 16); Asian American and Pacific Islander women were older at first intercourse (19% before age 16) than white and Hispanic women.
Black women were more likely than white women to report having had more than one sex partner (ever): (79%vs 68%), but only slightly more likely to have had more than five (32%vs 30%). Hispanic and Asian American/Pacific Islander women were much less likely to report having had more than one sex partner.
Approximately 12% of women were classified as having had concurrent partnerships since January 1991 (Table 3). As expected, concurrency prevalence was strongly associated with number of sexual partners. Overall, concurrency was much less likely to have occurred among women who were currently married (4%) and most likely to occur among women who were formerly married (22%), had combined family income below 150% of the poverty line (17%), received AFDC (23%), were age 18–24 years at interview (23%), first had sexual intercourse at age 12 or 13 (35%), or reported at least 11 lifetime sexual partners (41%). Among black women, educational attainment was inversely related to crude concurrency prevalence, but in other ethnic groups no association was apparent.
Marked ethnic differences in concurrency were evident. Concurrency prevalence was highest among black women (21%), lowest among Asian American and Pacific Islanders (6%) and Hispanics (8%), and intermediate among whites (11%). These ethnic patterns persisted in stratified analyses across all variables, except for women who had been formerly married, where concurrency prevalence was the same (24%) among blacks and whites. Even among women reporting the same number of partners since January 1991, concurrency prevalence for black women was considerably higher than for the other ethnic groups. Concurrency was moderately related to poverty among black and white women but not among the other racial/ethnic groups.
Multiple Logistic Analysis
The crude association of concurrency prevalence with younger age, earlier sexual debut, and being unmarried remained in multiple logistic models, as did the inverse association with Hispanic ethnicity. Terms for Asian American ethnicity and black ethnicity were included in all models, although the regression coefficient Asian Americans was small and for blacks it diminished substantially with the inclusion of marital status and age at first intercourse. The association of concurrency with education, somewhat inconsistent in the stratified analyses in Table 3, became monotonic and only weakly positive (no further details presented). Associations with income and AFDC were apparently explained by other variables, and so income and AFDC were omitted from the final models. Coefficients for the final model (with all class variables) are presented in Table 4.
Concurrent sexual partnerships can efficiently spread HIV and other STIs throughout a population. 8,9,20,21 Participation in concurrent sexual partnerships has been associated with increased risk of STI transmission. 22–25 Some authors have implicated concurrent partnerships in the spread of HIV infection in Sub-Saharan Africa and other populations with extensive heterosexual HIV transmission. 26–28
To determine the prevalence of concurrency and its predictors among women in the United States, we analyzed sexual partnership histories reported in the 1995 NSFG. About 26% of women of reproductive age had two or more sexual partnerships during the approximately 54 month-year study period, and 46% of these women had concurrent partnerships. Substantial ethnic differences were noted in the crude prevalence of concurrency, which was most common among black women (21%), lowest among Asians (6%) and Hispanics (8%), and intermediate among whites (11%). Concurrency prevalence was strongly related to age, age at first intercourse, marital status, and receipt of AFDC, and less strongly associated with education, income, and work status. The multiple logistic models suggested that the relationship with income, work status, and AFDC could be explained by other variables. Age, age at first intercourse, and marital status remained strongly associated with concurrency. Hispanic ethnicity was significantly inversely associated, and black ethnicity retained at most a weak residual positive association.
Previous studies of concurrency have dealt with special populations, such as adolescents with STD exposure and college students. Among adolescents at a public STD clinic in San Francisco who had at least one main sexual partner during the preceding 6 months, 31% reported a concurrent partnership during a main relationship. 22 Coital diaries of 82 adolescent females who either had or were exposed to an STD revealed concurrent partnerships among 13% during a 21-month period. 29 A small survey of college students found similar levels of concurrency among African American (47%) and white (50%) men but higher prevalences among white (38%) than African American (19%) women. 30
Several studies based on national surveys have documented reported numbers of sexual partnerships in the past year, 31–35 but the extent of concurrency is unclear. Among sexually active U.S. women studied in the 1988 NSFG, approximately 3% reported at least two partners in the previous 3 months. 36 In that sample, unmarried women of all age and ethnic groups were equally likely to report having multiple partners. 35 Given the brevity of the time span, many of these partnerships are likely to have overlapped, although concurrency was not specifically examined. In contrast, the current study specifically examines the prevalence of concurrent partnerships and their correlates.
In the United States, explanations for the marked ethnic disparity in rates of HIV and other STIs have remained elusive. Causes are likely multifactorial and include differential access to healthcare, poverty, and prevalence of infection in the pool of sexual partners. But a pattern of increased sexual risk behavior (eg, partner number, condom use, sexual practices) has not been clearly established for black women. Consistent with the findings of previous investigations, 17, 35 our data showed that similar proportions of black and white women reported more than five partners during the preceding 5 years.
In contrast, differences between blacks and whites have recently been found in the nature of sexual networks; among blacks, more sexual mixing occurs between those with fewer partners and those with many partners, a pattern that efficiently transmits STIs throughout a population. 37 Our study suggests that concurrent sexual partnerships may represent another network pattern that contributes to the high rates of HIV and other STIs among blacks in the United States.
Black women’s greater prevalence of concurrency is unlikely to be due to attitudes about sexual relationships, as some studies report similarity between blacks and whites in such attitudes. 38–40 More likely explanations, suggested by our analyses, are younger age at sexual debut and marital status. Nevertheless, the potential role of these factors as determinants of concurrency does not diminish the likelihood that concurrency is a determinant of higher STD rates in blacks.
Early age of sexual debut predicts larger partner number later in life 16,41 and has been associated with extramarital intercourse among men. 42 However, the nature of the link between concurrency and age of first sexual intercourse is unclear.
In contrast, the pathway between marital status and concurrency is more direct. Because marriage usually constrains sexual interactions with other partners, strong associations between marital status and numbers of sexual partners are expected. Married persons are far less likely than those who are unmarried to have had multiple partners during the past year. 16,41 In 1997, only 34% of African American adults were married and living with a spouse, a percentage that was substantially lower than for whites (59%), Asians and Pacific Islanders (57%), and Hispanics (51%). 43,44
A possible explanation for the high percentage of black women who are unmarried is the relative scarcity of black men. The low sex ratio (ratio of men to women) among blacks results largely from higher mortality rates among black males due to disease, violence, and the disproportionate incarceration of black men. 45 This “male shortage” depresses marriage rates overall and may also raise divorce rates. Moreover, men’s scarcity affords them greater opportunity to maintain simultaneous relationships with different women. 12 The effect of the low sex ratio is compounded by economic marginality, a feature of life for many blacks that has a well-known deleterious effect on marital stability. 46,47
A major strength of this study is its use of a large, high-quality, nationally representative data set with oversampling of blacks and Hispanics. A significant limitation, however, is the lack of data on men. This lack prevented our investigating whether concurrency among men may be responsible for the apparent anomaly of Latinos having lower concurrency prevalence than white women despite having higher STD rates.
Our measure of concurrent sexual partnerships relies on respondents’ self report and their recalling months of first and last sexual intercourse with various partners going back nearly 5 years, a cognitively difficult recall operation, 48 although the NSFG did employ several techniques to improve accuracy of recall. Nevertheless, our ethnic comparisons would be seriously affected by recall errors only if the errors were differentially distributed by ethnicity.
In conclusion, this study revealed ethnic differences among U.S. women of childbearing age in the reported prevalence of concurrent partnerships. Differences in marital status and age at first sexual intercourse appeared to mediate the higher concurrency prevalence among African Americans. Greater involvement in concurrent partnerships may play a role in the longstanding ethnic disparities in rates of HIV and other STIs experienced by African Americans. Further research should seek to understand the phenomenon of concurrency in women and men and the factors that promote this sexual pattern and discourage long-term monogamy. In particular it is important to re-assess long-standing national policies regarding economics, employment, substance use, and law enforcement that help destabilize relationships in the black community. 49
We are grateful to Sevgi Aral, Anjani Chandra, Myron Cohen, and Paul Godley for their insightful comments during the preparation of this manuscript.
1. Ghani AC, Swinton J, Garnett GP. The role of sexual partnership networks in the epidemiology of gonorrhea. Sex Transm Dis 1997; 24 (1): 45–56.
2. Rothenberg R, Narramore J. The relevance of social network concepts to sexually transmitted disease control. Sex Transm Dis 1996; 23 (1): 24–29.
3. Garnett GP, Hughes JP, Anderson RM, et al.
Sexual mixing patterns of patients attending sexually transmitted diseases clinics. Sex Transm Dis 1996; 23 (3): 248–257.
4. Rothenberg RB, Sterk C, Toomey KE, et al.
Using social network and ethnographic tools to evaluate syphilis transmission. Sex Transm Dis 1998; 25 (3): 154–160.
5. Aral SO, Hughes J, Stoner B, et al.
Sexual mixing patterns in the spread of gonococcal and chlamydial infections. Am J Public Health 1999; 89: 825–833.
6. Aral SO. Sexual network patterns as determinants of STD rates: paradigm shift in the behavioral epidemiology of STDs made visible. Sex Transm Dis 1999; 26: 262–264.
7. Aral S, Holmes K. Social and behavioral determinants of the epidemiology of STDs: Industrialized and developing countries. In: Holmes K, Per-Anders M, PF S, Lemon S, Stamm W, Piot P, Wasserheit J, eds. Sexually Transmitted Diseases. New York: McGraw-Hill, 1999; 39–76.
8. Morris M, Kretzschmar M. Concurrent partnerships
and transmission dynamics in networks. Social Networks 1995; 17: 299–318.
9. Watts CH, May RM. The influence of concurrent partnerships
on the dynamics of HIV
/AIDS. Math Biosci 1992; 108 (1): 89–104.
10. Adimora A, Schoenbach V, Martinson F, et al.
Social Context of Sexual Relationships among Rural African Americans
. Sex Transm Dis 2001; 28: 69–76.
11. Geronimus A, Bound J, Waidmann T, Hillemeier M, Burns P. Excess mortality among blacks and whites in the United States. N Engl J Med 1996; 335: 1552–1558.
12. Guttentag M, Secord P. Too Many Women: The Sex Ratio Question. Beverly Hills: Sage, 1983.
13. Abma J, Chandra A, Mosher W, Peterson L, Piccinino L. Fertility, family planning, and women’s health: New data from the 1999 National Survey of Family Growth. Vital Health Stat 1997; 23: 1–114.
14. Potter F, Iannachione V. Sample design, sampling weights, imputation, and variance estimation in the 1995 National Survey of Family Growth. Vital Health Stat 1998; 2: 1–63.
15. US. Bureau of the Census. Income, poverty, and valuation of noncash benefits: 1994. Current Population Reports
1996;Series P-60(No. 189).
16. Laumann EO, Gagnon JH, Michael RT, Michaels S. The Social Organization of Sexuality. Chicago: The University of Chicago Press, 1994.
17. Seidman SN, Mosher WD, Aral SO. Predictors of high-risk behavior in unmarried American women: adolescent environment as risk factor. J Adolesc Health 1994; 15 (2): 126–132.
18. Kleinbaum D, Kupper L, Mueller K. Applied Regression Analysis and Other Multivariable Methods. 2nd ed. Belmont, CA: Duxbury, 1987.
19. Hosmer D, Lemeshow S. Applied Logistic Regression. New York: Wiley, 1989.
20. Kretzschmar M, Morris M. Measures of concurrency in networks and the spread of infectious disease. Math Biosci 1996; 133 (2): 165–195.
21. Morris M, Kretzschmar M. Concurrent partnerships
and the spread of HIV
. AIDS 1997; 11 (5): 641–648.
22. Rosenberg MD, Gurvey JE, Adler N, Dunlop MB, Ellen JM. Concurrent sex partners and risk for sexually transmitted diseases among adolescents. Sex Transm Dis 1999; 26 (4): 208–212.
23. Daker-White G, Barlow D. Heterosexual gonorrhoea at St Thomas’–II: Sexual behaviour and sources of infection. Int J STD AIDS 1997; 8 (2): 102–108.
24. Koumans E, Farley T, Gibson J, et al.
Characteristics of persons with syphilis in areas of persisting syphilis in the United States: Sustained transmission associated with concurrent partnerships
. Sex Transm Dis 2001; 28: 497–503.
25. Potterat J, Zimmerman-Rogers H, Muth S, et al.
Chlamydia transmission: Concurrency, reproduction number, and the epidemic trajectory. Am J Epidemiol 1999; 150: 1331–1339.
26. Hudson CP. AIDS in rural Africa: a paradigm for HIV
-1 prevention [editorial]. Int J STD AIDS 1996; 7 (4): 236–243.
27. Hudson CP. Concurrent partnerships
could cause AIDS epidemics. Int J STD AIDS 1993; 4 (5): 249–253.
28. Garnett GP, Johnson AM. Coining a new term in epidemiology: concurrency and HIV
. AIDS 1997; 11 (5): 681–683.
29. Howard MM, Fortenberry JD, Blythe MJ, Zimet GD, Orr DP. Patterns of sexual partnerships among adolescent females. J Adolesc Health 1999; 24 (5): 300–303.
30. Johnson EH, Jackson LA, Hinkle Y, et al.
What is the significance of black-white differences in risky sexual behavior
? J Natl Med Assoc 1994; 86 (10): 745–759.
31. Peterson JL, Catania JA, Dolcini MM, Faigeles B. III. Multiple sexual partners among blacks in high-risk cities. Fam Plann Perspect 1993; 25 (6): 263–267.
32. Binson D, Dolcini MM, Pollack LM, Catania JA. IV. Multiple sexual partners among young adults in high-risk cities. Fam Plann Perspect 1993; 25 (6): 268–272.
33. Dolcini MM, Catania JA, Coates TJ, et al.
Demographic characteristics of heterosexuals with multiple partners: the National AIDS Behavioral Surveys. Fam Plann Perspect 1993; 25 (5): 208–214.
34. Choi K-H, Catania J. Changes in multiple sexual partnerships, HIV
testing, and condom use among US heterosexuals 18 to 49 years of age, 1990 and 1992. Am J Publ Health 1996; 86: 554–556.
35. Kost K, Forrest JD. American women’s sexual behavior
and exposure to risk of sexually transmitted diseases. Fam Plann Perspect 1992; 24 (6): 244–254.
36. Seidman SN, Mosher WD, Aral SO. Women with multiple sexual partners: United States, 1988. Am J Public Health 1992; 82 (10): 1388–1394.
37. Laumann EO, Youm Y. Racial/ethnic group differences in their prevalence of sexually transmitted diseases in the United States: A network explanation. Sex Transm Dis 1999; 26: 250–261.
38. Tucker MBT, Mitchell-Kernan CM. Trends in African. American family formation: A theoretical and statistical overview. In: Tucker MBT, Mitchell-Kernan CM, eds. The Decline in Marriage Among African Americans
: Causes, Consequences and Policy Implications. New York: Russell Sage Foundation, 1995.
39. Scanzoni J. The Black Family in Modern Society: Patterns of Stability and Security. Chicago: The University of Chicago Press, 1977.
40. Heiss J. Values Regarding Marriage and the Family From a Woman’s Perspective. In: McAdoo HP, ed. Black Families. Thousand Oaks, California: Sage Publications, Inc., 1997; 284–300.
41. Johnson A, Wadsworth J. Heterosexual Partnerships. In: Johnson A, Wadsworth J, Wellings K, Field J, eds. Sexual Attitudes and Lifestyles. London: Oxford Blackwell Scientific Publications, 1994; 110–144.
42. White R, Cleland J, Carael M. Links between premarital sexual behaviour and extramarital intercourse: a multi-site analysis. AIDS 2000; 14: 2323–2331.
43. US Census Bureau CPS. Selected Characteristics of the Population by Race: March 1997. Vol. 2000 US Census Bureau, 1997.
44. US Census Bureau CPS. Selected Characteristics of the Population by Hispanic Origin: March 1997. Vol. 2000 US Census Bureau, 1997.
45. Cherlin A. Marriage, Divorce, Remarriage. Cambridge, MA: Harvard University Press, 1992.
46. Ross H, Sawhill I. Time of Transition: The Growth of Families Headed by Women. Washington, DC: The Urban Institute, 1975.
47. Hoffman S, Holmes J. Husbands, Wives, and Divorce. In: Duncan G, Morgan J, eds. Five Thousand American Families - Patterns of Economic Progress. Ann Arbor, Michigan: Institute for Social Research, 1976: 23–75.
48. Warnecke RB, Sudman S, Johnson TP, et al.
Cognitive aspects of recalling and reporting health-related events: Papanicolaou smears, clinical breast examinations, and mammograms. Am J Epidemiol 1997; 146 (11): 982–992.
49. Wilson WJ. Truly Disadvantaged: The Inner City, The Underclass, and Public Policy. Chicago: The University of Chicago Press, 1987.