Co-parenting and Sexual Partner Concurrency among White, Black, and Hispanic Men in the United States

Abstract

Background

Concurrent sexual partnerships (partnerships that overlap in time) increase the spread of infection through a network. Different patterns of concurrent partnerships may be associated with varying STI risk depending on the partnership type (primary vs. non-primary) and the likelihood of condom use with each concurrent partner. We sought to evaluate co-parenting concurrency, overlapping partnerships in which at least one concurrent partner is a co-parent with the respondent, which may promote the spread of sexually transmitted infections (STIs).

Methods

We examined sexual partnership dates and fertility history of 4928 male respondents in the 2002 National Survey of Family Growth. We calculated co-parenting concurrency prevalence and examined correlates using Poisson regression.

Results

Among men with ≥1 pair of concurrent partnerships, 18% involved a co-parent. 33% of black men involved in co-parenting concurrency were < 25 years, compared to 23% of Hispanics and 6% of whites. Young black men (age 15–24) were more likely to engage in co-parenting concurrency than white men, adjusting for socio-demographic characteristics, sexual and other high-risk behaviors, and relationship quality. Compared to white men age 15–24, black and Hispanic men were 4.60 (95% CI 1.10, 19.25) and 3.45 (95% CI 0.64, 18.43) times as likely to engage in co-parenting concurrency.

Conclusion

Almost one in five men engaging in concurrent sexual partnerships in the past year was a co-parent with at least one of the concurrent partners. Understanding the context in which different types of concurrency occur will provide a foundation on which to develop interventions to prevent STIs.

Background

Concurrent sexual partnerships (relationships that overlap in time), have been associated with the transmission of sexually transmitted infections (STI) including syphilis [1], chlamydial infection [2], and heterosexually acquired HIV infection [3]. While the rate of partner acquisition may be similar in concurrent compared to serially monogamous partnerships, the overlap of sexual partnerships can lead to faster spread and establishment of STIs in a population [3–6].

Research concerning the socio-cultural factors that influence the occurrence of concurrency has begun to emerge for some populations, such as the relationship between acculturation and sexual behavior among Hispanic youth [3, 7, 8]. In addition, qualitative research has identified different concurrency patterns that may be associated with varying STI risk depending on partnership type (primary vs. non-primary) and the likelihood of condom use with each concurrent partner [9]. One pattern potentially associated with high STI risk involves concurrency in the context of a co-parenting relationship [9]. Co-parenting concurrency involves engaging in sexual intercourse with a co-parent while in another sexual partnership. Black, unmarried fathers report difficulty with ending a sexual relationship with the mother of their children despite not being in a mutually monogamous relationship with her. Furthermore, women in main partnerships with unmarried fathers reported sexual activity outside the relationship as more acceptable if it occurs with a co-parent [9, 10].

To date, no published study has quantitatively examined co-parenting in the context of concurrent sexual partnerships. We used data from male respondents in Cycle 6 of the National Survey of Family Growth (NSFG) to: (1) calculate the overall and race-specific- prevalence of co-parenting concurrency; (2) describe co-parenting concurrency patterns, and (3) determine demographic and behavioral correlates of co-parenting concurrency.

Methods

The NSFG is a cross-sectional survey conducted by the National Center for Health Statistics designed to examine trends in contraception, marriage, divorce, sexual activity and fertility[11]. Cycle 6 of the NSFG was conducted in 2002 and was the first cycle to include men. Men and women aged 15–44 years in the US household population were targeted, and teens (aged 15–19), African Americans, and Hispanics were oversampled [12]. The survey collected data about demographic, socio-economic, and behavioral characteristics and was administered by female interviewers using computer-assisted personal interviewing (CAPI). More sensitive questions were administered using audio computer-assisted self-interviewing (ACASI) [12]. Seventy-eight percent of males sampled completed the interview, yielding a total of 4928 male respondents [12]. We excluded 274 men who reported a race/ethnicity other than white, non-Hispanic black or Hispanic from all analyses because only 12 men in this group engaged in concurrency resulting in a final sample of 4654 men.

Concurrent Sexual Partnerships

Concurrency with female partners was determined, as in previous research [3, 13–15], by examining dates of first and last intercourse with each partner discussed during the interview. Reported dates of first sexual intercourse for up to four (current wife/partner and three most recent) sexual partners were ordered sequentially. Partnerships that ended 12 months before the interview were excluded. The dates of first and last sex for all partnerships were compared for men who provided information on two or more sexual partners.

For each partnership pair, the month of first sexual intercourse with the later partner was compared with the month of last sexual intercourse with the earlier partner. If the month of first sex with the later partner occurred before the month of last sex with the earlier partner, the partnership was considered concurrent. Co-parents were defined as a man and woman who are the joint biological parents of a child. For each sexual partner, a respondent was asked questions about children he co-parented with the partner, including biological, foster, adopted, and step children. Only biological children were included in our co-parenting definition, and biological children from other partnerships that ended more than 12 months before the interview were not included. A concurrent partnership pair was classified as co-parenting concurrency if the respondent had a biological child with at least one of the concurrent sexual partners.

Additional Measures

A conceptual model for the association between co-parenting and concurrency was used to identify potential correlates of co-parenting concurrency. Socio-demographic characteristics included age, race, educational attainment, and household income as a percent of the 2000 US poverty line. Sexual behaviors that affect the risk of STIs included the respondent’s number of sexual partners (lifetime and in the past 12 months), frequency of condom use, and age at first sexual intercourse. Each respondent was asked about relationship characteristics, sexual activity, and fertility in relation to his reported sexual partners. We categorized incarceration for at least 24 hours as never, within the past 12 months and greater than 12 months ago. Cohabitation status at the time of the child’s birth and average relationship duration were used as proxy measurements for relationship quality.

Analysis

All variables were coded as dichotomous or nominal categorical variables. All analyses were conducted using Stata version 10 (Stata Statistical Software: Release 10. College Station, TX) and incorporated the NCHS-provided sample weights and sampling design variables [12]. We examined demographic, socio-economic, fertility, and sexual behavior characteristics among all male respondents (N=4654), all fathers (N=1653), and all men with overlapping partnerships with women in the past 12 months (N=430). We calculated the prevalence of co-parenting concurrency, with 95% confidence intervals (95% CI) overall and by racial/ethnic group. We calculated chi-square statistics for bivariable associations of co-parenting concurrency with socio-demographic and behavioral and relationship characteristics. Effect measure modification by race/ethnicity and age was examined using a product interaction model and a Wald test at the p<0.20 significance level. Prevalence ratios and 95% CIs were calculated using a multivariable Poisson regression model including all covariates of interest and a race by age interaction term.

Results

Differences between men engaging in concurrent partnerships and the entire NSFG sample have been described in detail in previous analyses [3]. Approximately 18.0% of concurrent sexual partnerships among US men involved a co-parent, and the overall prevalence varied slightly by race/ethnicity (Table 1). Black and Hispanic men who engaged in co-parenting concurrency were considerably younger than white men who engaged in co-parenting concurrency. Slightly more than a third of black men involved in co-parenting concurrency were younger than 25 years, compared to 23% of Hispanic men and only 6% of white men (Figure 1). The Wald p-value for the interaction between race/ethnicity was 0.06 indicating PR modification by race/ethnicity and age.

TABLE 1.

Co-parenting Concurrency Prevalence among US Men Reporting Concurrency in the Past 12 Months (N=430), 2002 National Survey of Family Growth

	Co-Parenting Concurrency
	Unweighted N	Weighted %*
Overall	59	18.0
Age at Interview (Years)
15–19	7	7.6
20–24	10	5.8
25–29	12	25.6
30–34	12	22.6
35–39	11	19.7
40–45	7	29.5
Race/Ethnicity
White	13	14.7
Black	30	17.8
Hispanic	16	18.2
Education§
< High School	17	47.1
High School/GED	18	16.9
Some College	10	13.4
Bachelor’s Degree or Higher	4	5.3
Household income as a percent of 2000 poverty line§
<150%	22	39.7
150%–249%	5	28.7
250–399%	12	17.4
≥400%	10	8.4
Current Marital Status
Married	18	76.9
Cohabiting	9	36.7
Previously Married†	8	7.1
Never Married	24	6.9
Number of Biological Children#
0	0	0
1	28	42.6
2	12	44.1
3	13	72.1
≥4	6	61.6
Number of Children Born Outside Marriage#‡
0	13	41.9
1	27	55.1
2	11	59.6
≥3	8	56.6
Cohabitation at Child’s Birth‡
Non-Cohabiting Only	16	35.2
Cohabiting Only	31	51.5
Both Cohabiting and non-Cohabiting	12	70.5
Multiple Partner Fertility‡§
No	41	12.5
Yes	18	51.8
Age at First Sexual Intercourse (Years)
≥18	10	25.2
16–17	8	8.2
14–15	22	36.4
≤13	19	19.5
Number of Lifetime Sexual Partners
0	0	0
1–2	0	0
3–5	9	12.6
6–10	14	10.9
≥11	36	15.0
Number of Sexual Partners in the Past 12 Months
0	0	0
1	0	0
2	29	28.9
3	19	13.1
≥4	11	4.8
Condom Use During the Last Month
None of the time	21	23.4
Some of the time	13	24.6
All of the time	18	7.3
Incarceration for ≥24 hours
Never	38	18.9
>12 months ago	14	11.8
Within past 12 months	7	14.3

^*Weighted to account for stratification, clustering, and unequal selection probabilities yielding nationally representative estimates. Percents may not sum to zero due to rounding.

^§Among men aged 22 years and older (n=309)

^†Includes separated, divorced, and widowed

^#Includes children fathered with the respondent’s current wife/cohabiting partner or 3 most recent partners in the 12 months prior to the interview

^‡Among men who have a biological child (n=136)

^§Includes children fathered with the respondent’s current wife/cohabiting partner, 3 most recent partners in the 12 months prior to the interview, former wives, first premarital cohabiting partner, or other biological children fathered with women who were not discussed in other sections of the interview

FIGURE 1.

Age Distribution of Co-parenting Concurrency by and Race/Ethnicity ^

^N=430 white, non-Hispanic black, and Hispanic men reporting concurrency in the past 12 months

In a previous analysis of these data, it was estimated that 11% of the men had concurrent partnerships [3]. Among this subset of 430 men, the prevalence of co-parenting concurrency was highest among men with less than a high school education and decreased with increasing education (Table 1). The prevalence of co-parenting concurrency among men with the lowest household incomes was almost five times the prevalence among men with the highest household incomes (39.7% vs. 8.4%). Co-parenting concurrency prevalence was slightly higher among men who had children born outside marriage compared to men who did not but did not vary depending on the number of children born outside marriage (Table 1). Co-parenting concurrency was more prevalent among fathers who had children with multiple partners (51.8%) than among fathers who did not have multiple partner fertility (12.5%).

Based on unadjusted prevalence ratios (PR) and Wald tests (Table 2), age at interview, education, household income, condom use during the last month, cohabitation at the time of the child’s birth, and average relationship duration were associated with co-parenting concurrency. Among men who engaged in concurrent partnerships, those with an average relationship duration of 3–5 years were 5 times as likely to be involved in co-parenting concurrency [PR 5.23 (1.98, 18.83)] as those whose average relationship lasted less than 1 year. The association was even stronger for average relationship duration of 6 years or more compared to less than 1 year [PR 13.79 (5.58, 34.10)].

TABLE 2.

Correlates of Co-parenting Concurrency among Men Who Had Concurrent Partnerships in the Past 12 Months^{^}

	Co-parenting Concurrency in the past 12 months
	Yes		No
	N	%	N	%	Bivariable Model PR (95% CI)		Wald p-value	Multivariable Model PR (95% CI)
Total	59	16.3	371	83.7
Age at Interview (Years)
15–24	17	18.0	176	50.9	0.26	(0.11, 0.57)	0.0002	3.51	(1.86 6.62)
25–34	24	43.1	98	26.6	0.95	(0.41, 2.21)		1.60	(0.84 3.03)
≥35	18	39.0	97	22.6	1.0			1.0
Education
< High School	22	47.7	78	19.9	1.0		0.01	1.0
High School/GED	20	24.2	120	30.0	0.43	(0.20, 0.93)		0.74	(0.45 1.23)
Some College	13	23.7	115	35.2	0.56	(0.28, 1.15)		0.81	(0.41 1.62)
Bachelor’s Degree or Higher	4	4.4	58	15.4	0.16	(0.04, 0.58)		0.27	(0.12 0.58)
Household income as a percent of 2000 poverty line
<150	25	38.7	70	17.4	4.00	(1.97, 8.14)	0.007	2.36	(1.30 4.28)
150–249	7	21.7	72	17.2	2.14	(0.58, 7.81)		0.90	(0.49 1.64)
250–399	14	18.1	76	22.7	1.43	(0.50, 4.05)		2.14	(1.20 3.83)
≥400	13	25.6	153	42.8	1.0			1.0
Age at First Sexual Intercourse (Years)
≥18	10	26.4	47	15.2	1.0		0.4	1.0
16–17	8	13.9	104	29.1	0.37	(0.10, 1.32)		0.66	(0.31 1.39)
14–15	22	36.4	148	36.9	0.83	(0.32, 2.20)		0.49	(0.28 0.86)
<14	19	23.3	72	18.8	0.78	(0.27, 2.24)		0.59	(0.28 1.24)
Race/Ethnicity
White	13	47.5	152	53.4	1.0		0.9	1.0
Black	30	29.1	136	26.2	1.21	(0.52, 2.80)		1.21	(0.65 2.23)
Hispanic	16	23.4	83	20.5	1.23	(0.49, 3.09)		1.22	(0.74 2.02)
Condom Use During the Last Month
None of the time	21	60.1	101	39.0	2.50	(1.14, 5.48)	0.002	1.88	(1.13 3.12)
Some of the time	13	20.9	41	12.7	2.59	(1.14, 5.87)		1.65	(0.90 3.02)
All of the time	18	19.0	166	48.3	1.0			1.0
Cohabitation at Child’s Birth *
Non-Cohabiting Only	16	19.2	28	38.7	0.62	(0.34, 1.12)	0.07	0.86	(0.46 1.59)
Cohabiting Only	31	42.7	34	43.9	1.0			1.0
Both Cohabiting and non-Cohabiting	12	38.1	15	17.4	1.29	(0.85, 1.98)		0.89	(0.55 1.45)
Average Relationship Duration
<1 year	7	8.3	141	36.4	1.0		<0.001	1.0
1–2 years	8	12.4	109	27.3	1.96	(0.52, 7.39)		0.67	(0.31 1.45)
3–5 years	20	28.5	89	27.8	5.23	(1.98, 13.84)		1.85	(0.95 3.60)
≥6 years	24	50.9	32	8.4	13.79	(5.58, 34.10)		3.43	(1.86 6.31)
Incarceration for ≥24 hours
Never	38	65.8	209	54.8	1.0		0.5
>12 months ago	14	18.2	99	26.5	1.07	(0.54, 2.09)		0.60	(0.35 1.03)
Within past 12 months	7	16.1	63	18.7	0.75	(0.27, 2.06)		0.54	(0.34 0.85)

^{^}N=430 men reporting concurrency in the past 12 months

^*Among men who have a biological child (n=139)

The associations of co-parenting concurrency with poverty, condom use, average relationship duration, and incarceration history persisted in the final, multivariable model (Table 2). Lower household income and increased relationship duration were associated with an increased likelihood of co-parenting concurrency, with PRs increasing as household income decreased. Men who never used a condom were more likely to have engaged in co-parenting concurrency in the past 12 months compared to men who always used a condom [PR 1.88 (1.13, 3.12)]. Having a history of incarceration, particularly incarceration within the past 12 months, was associated with a decreased likelihood of co-parenting concurrency [PR 0.54 (0.34, 0.85)].

Young black men (age 15–24) were more likely to engage in co-parenting concurrency than white men, adjusting for socio-demographic characteristics, sexual and other high-risk behaviors, and relationship quality (Table 3). The largest racial differences in co-parenting concurrency prevalence were observed among men age 15–24. Compared to white men age 15–24, black and Hispanic men were 4.60 (95% CI 1.10, 19.25) and 3.45 (95% CI 0.64, 18.43) times as likely to engage in co-parenting concurrency. White men age ≥35 were slightly more likely than black and Hispanic men to engage in co-parenting concurrency.

TABLE 3.

Adjusted Prevalence Ratios (PR) and 95% Confidence Intervals (95% CI) for Co-Parenting Concurrency by Age and Race/Ethnicity, 2002 National Survey of Family Growth ^*

	Co-Parenting Concurrency			Prevalence	PR	95% CI
Age at Interview (Years)	No	Yes	Total	weighted %
15–24
White	76	3	79	1.8	1.0
Black	55	9	64	14.3	4.60	(1.10, 19.25)
Hispanic	45	5	50	8.5	3.45	(0.64, 18.43)
25–34
White	30	6	36	28.9	1.0
Black	40	13	53	22.1	0.99	(0.37, 2.65)
Hispanic	28	5	33	17.2	1.83	(0.69, 4.87)
≥35
White	46	4	50	22.2	1.0
Black	41	8	49	18.3	0.86	(0.36, 2.10)
Hispanic	10	6	16	53.0	0.90	(0.49, 1.66)

^*Estimates calculated using multivariable Poisson regression for survey data adjusted for socio-demographic characteristics (age, education, household income as a % of the 2000 US poverty line), sexual behaviors (age at first sexual intercourse, condom use), relationship quality (average relationship duration, cohabitation at the time of the child’s birth), and other high-risk behaviors (incarceration history); N=430 men who engaged in concurrent sexual partnerships in the past 12 months

Discussion

This study is the first to explore quantitatively the role of co-parenting relationships in concurrent sexual partnerships. Almost one in five men engaging in concurrent sexual partnerships with women in the past 12 months had a biological child with at least one of his concurrent partners. Research among US men estimated concurrency was three and two times as likely among non-Hispanic blacks and Hispanics, respectively, compared to non-Hispanic whites [14]. Data from our analyses do not suggest racial/ethnic differences in the overall prevalence of co-parenting among men engaging in concurrency, though co-parenting concurrency did vary considerably when examined jointly by race/ethnicity and age. The largest racial/ethnic disparities in co-parenting concurrency prevalence were observed among men aged 15–24 with blacks and Hispanics being four to five times as likely to engage in co-parenting concurrency as their white counterparts.

Results from this analysis highlight a potential population for which STI/HIV prevention messages could be developed. Young people (age 15–24), including young parents, have been found to engage in a variety of risk behaviors, such as having multiple and concurrent sexual partners, unprotected intercourse, drug or alcohol use, and needle sharing [16–18]. Inconsistent condom use was almost four times as likely among adolescent couples with a child compared to those without a child [19]. Furthermore, young parents in relationships were generally unaware of their intimate partner’s HIV testing history [20].

Co-parenting is generally discussed in the context of married couples, though it can occur in a number of different scenarios [21]. Approximately 40% of all births in the US in 2007 were to unmarried women, and the proportion of births to unmarried non-Hispanic black women (71.6%) was approximately 2.5 times as high as the proportion of births to non-Hispanic white women (27.8%) [22]. Relationships between unmarried parents are often unstable and characterized by repeated break-ups and reunions, [23, 24] creating an environment conducive to concurrency. In our analyses, births outside marriage were reported by over three quarters (76.3%) of men engaging in co-parenting concurrent partnerships, supporting the idea of increased concurrency among unmarried parents.

The term nonresident father includes a wide variety of men (e.g. divorced men who may or may not be remarried) but has more recently been used in research targeting non-resident fathers, regardless of marital status [25–27]. Nonresident fathers’ involvement with their children differs by race/ethnicity, and this difference can be partially explained by the status of the mother-father relationship [25]. Specifically, minority nonresident fathers were more likely to maintain romantic relationships with their child’s mother than white fathers, while mothers who had children with white men were more likely to re-partner [25]. Thus, it is possible that the co-parenting relationship, particularly among unmarried racial/ethnic minorities, could impact the formation and persistence of concurrent sexual partnerships.

The trend toward co-parenting concurrency’s increased prevalence among white men age 35 and older compared to black and Hispanic men of the same age further highlights the importance of considering social contexts surrounding concurrent partnerships. Racial/ethnic differences in marriage and cohabitation may explain some of the observed differences in co-parenting concurrency. Blacks in the US are less likely than other racial/ethnic groups to marry[28], and among both men and women, unmarried individuals are much more likely to engage in concurrent partnerships than married individuals [13, 28]. Thus, one possible explanation is that co-parenting concurrency’s occurrence among young black and Hispanic men results from continued sexual activity with the co-parent after initiation of a new, perhaps main, relationship. Conversely, among older white men, co-parenting concurrency could be occurring in the context of an extramarital affair.

The cross-sectional nature of the data prohibited us from drawing causal inferences and must be acknowledged as a limitation. We were also not able to examine the contexts surrounding transitions into and out of sexual partnerships. Information on partnerships and children conceived in them was available for at most four sexual partners and only partnerships active during the past year. Men who had other partners could have had concurrent partnerships and children that were undetected. Additionally, sexual partnership dates were reported by month and year which could have introduced some ambiguity in determining concurrency status. For example, a sexual partnership that appeared to span two years could actually have consisted of one sexual act with a woman during one month and a second sexual act with the same woman two years later. Finally, the limited number of outcomes and a significant age by race/ethnicity interaction resulted in small cell counts which decreased the precision of our effect estimates.

We defined co-parents as a man and woman who are the joint biological parents of a child. This definition was more restrictive than that proposed in the sociology and child development literature, which includes co-parents regardless of their sexual orientation or biological linkage to the child [21]. Though some instances of co-parenting could have been missed by our more specific definition, the significance of a biological child as a continuing manifestation of earlier sexual intimacy argues for differentiating adoptive and biological children in examining co-parenting concurrency.

Accuracy of self-report in this study depends on both recall and willingness to disclose sensitive information. The NSFG 2002 utilizes a life calendar approach to assist respondents in recalling information, but the potential for misreporting partnerships and/or dates remains. Self-report of sexual behaviors varies depending on the mode in which the survey is administered [29], and the use of ACASI likely improved the completeness of self reported sensitive and high-risk behaviors [30–32]. We have no evidence that reporting of sexual behaviors differed according to concurrency status.

Although the contextual factors that promote concurrency are not yet clear, they are likely to include a combination of imbalanced sex ratios, low marriage rates, economic differentials, media influences, and community and cultural norms. Our results show that the prevalence of co-parenting concurrency differs by race/ethnicity and age and that this concurrency pattern is most prevalent among young black and Hispanic men. A comprehensive understanding of the types of concurrent sexual partnerships and the contexts in which they occur should provide a basis for more effective prevention interventions and public messages.

Co-parenting relationships are complex and have profound implications for child health and development. Concurrent sexual partnerships add an additional layer of complexity to co-parenting relationships, which can affect the health of the co-parents, their other partners, and their community. Thus, future research should examine whether there is a link between co-parenting concurrency and STI transmission. The concept of co-parenting concurrency could be incorporated into STI studies by including questions on fertility histories and dates of sexual intercourse into data collection instruments. The importance of co-parenting relationships in sexual networks warrants further investigation and could be beneficial in determining the role these relationships play in population-level STI dissemination.

Summary.

A study of US men found that almost one in five men engaging in concurrent partnerships was a co-parent with at least one concurrent partner, and co-parenting concurrency was most common among young black men.