Union Formation Among Youth in Sub-Saharan Africa: Does Early Sexual Debut Matter?

Abstract

Is early sexual debut a key determinant of union formation among young adults in sub-Saharan Africa (SSA)? Using a life-course perspective, this article analysed data from Demographic and Health Surveys of selected countries to examine the relationship between sexual behaviour measured by age at first sex and the likelihood of cohabitation or marriage. These associations were examined using binary and multinomial logistic regressions. Findings suggest that the decision for males to cohabit, get married, or remain single differs by age at sexual debut, although there is no similar evidence for females. While an understanding of the patterns revealed is significant for policy formulation on union formation of young people, it can help to reduce the reproductive health challenges among youth in SSA.

Keywords

age sexual debut union formation family Africa

The last few decades have been characterized by changes in union formation in sub-Saharan Africa (SSA; Lloyd, 2005; Mensch, Singh, & Casterline, 2005). Modernization and diffusion of ideas are changing family structure, with competing strains of social regeneration and economic constraints intensifying the process. Consequently, the experience of early adulthood is increasingly diverse (Frech, 2014): some marry early, whereas others cohabit or remain single. Early marriage has been associated with early motherhood, with negative consequences on health of mothers (Boden, Fergusson, & John Horwood, 2008; Lee & Gramotnev, 2006), infants (Hoffman, Bann, Higgins, Vohr, Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network, 2015; Morinis, Carson, & Quigley, 2013), and both their futures. Especially, the breakdown in family structure is transforming the role of marriage as the normative forum for sexual behaviour. Premarital sex is more common, and a growing number of young people are having sex outside serious relationships (Stephenson, Simon, & Finneran, 2014). This is leading to increased cohabitation, a practice that has been associated with unprotected sexual activity (Rosenbaum, Zenilman, Rose, Wingood, & DiClemente, 2016) and intimate sexual violence (Abramsky et al., 2011). Although studies are documenting the changes in family formation among older women, few studies exist on young adults. This study adapts the life-course perspective to examine the relationship between sexual behavior and union formation among young adults.

Based on the life-course framework that emphasizes timing and sequencing of important events as determinant of behavior (Clarkberg, 1999; Kleinepier & de Valk, 2016), studies have examined how early childhood experiences influence adult health. Raley, Crissey, and Muller (2007) state that a life-course approach acknowledges that events in one stage are a result of what occurred in the previous stage. Using data from multiple waves of the National Longitudinal Survey of Adolescent Health, Raley et al. (2007) conclude that romantic relationship experiences at the end of high school increased the likelihood of marriage, while nonromantic sexual experiences predicted later cohabitation. Focusing on males and females aged 15–49, Coltabiano and Castiglioni (2008) found that while boys have earlier sexual debut compared to girls, cohabitation increased among females. Based on a sample of “ever married” Cameroonian women aged 19–34, Subaiya and Johnson (2008) concluded that education is associated with later sexual debut and later union formation. Youth who experienced early sexual debut but remain in school are more likely to remain unmarried compared to dropouts. While these studies provide important insights, there are no recent studies on the topic.

This study examines the relationship between sexual behavior measured as age at sexual debut and union formation in SSA. The study is significant because it utilizes nationally representative sample of youth and focuses on both marriage and cohabitation. With high prevalence of early marriage in the subcontinent, the study examines the effect of structural changes on the union formation of a demographic group that is at the center of development discourse and policy and research interests.

Method

Data Source

We used data from the Demographic and Health Surveys (DHS) which is a representative cross-sectional study conducted about once every 5 years. The DHS is a reliable source of information on population health topics in developing countries. The selected countries have had a DHS from 2012 and beyond, as well as high levels of child marriage based on literature (Table A1 in the Appendix shows countries and regional grouping).

Respondents were aged 15–24 years and must have experienced sexual debut. Because of gender differences in sexual behavior (Odimegwu & Somefun, 2017), we conducted a separate analysis for females (N = 54,328) and males (N = 20,742; also see Table A1).

Measures

Outcome variable

We used two outcomes: age at cohabitation and union formation. While age at cohabitation was measured as “early union” and “late union,” early union was defined as union that occurred before 18 years and coded 1 if union occurred before the age of 18 and “0” if otherwise (Nour, 2006; Fedd et al., 2015), based on considerations of child protection laws and level of physical, psychological, and biological development.

For the second outcome, the DHS asked a question on current marital status of the respondents. Responses were never in union, married, living with partner, widowed, divorced, and separated. Those widowed, divorced, and separated were coded as “previously married.” Union formation was operationalized as 1 = never married, 2 = cohabiting, 3 = married, and 4 = previously married.

Independent variables

The key independent variable is age at sexual debut, measured as the age at which respondents initiate sex for the first time. We used this as a continuous variable because it is a more powerful approach (Altman & Royston, 2006) as it allows us to examine a nonlinear relationship and presents a more accurate measure of behavior. Other explanatory variables associated with youth sexual behaviors and included were age, place of residence, educational status, household wealth status, sex of household head, and exposure to mass media (Calvès, 2016; Handa et al., 2015).

Results

Analysis Plan

The mean and standard deviation of the dependent and independent variables were calculated. Binomial and multinomial logistic regressions were used: The former was presented as odds ratios (OR) while the latter was presented as relative risk ratios. Analysis was conducted using Stata (Version 14). The adjusted model has been adjusted for covariates. The model was run separately for each region and for males and females (see detailed results in Table A2 –A5). To account for undersampling and oversampling in individual countries, weighting factors were applied at various levels of analysis. In the descriptive analysis, we presented the weighted percentage distribution, while robust standard errors were computed for the multivariate regression models to take into consideration the stratification and clustering of the sample design (Cameron & Miller, 2015).

Descriptive Statistics

The results in Tables 1 and 2 show that the mean age at sexual debut was 16 years for males and 18 years for females. More than half of the females in all the regions were married before 18 years except in East Africa (44%). Males in Central Africa also had the highest percentage of early marriage (27%) compared to other regions and regional average (17%). By union formation, males (9%) and females (28%) in Central Africa also had the highest percentage of young adults cohabiting. More than half of the females in Southern Africa (57%) and West Africa (58%) were married.

Table 1.

Socioeconomic and Demographic Profile of Young Males in Sub-Saharan Africa.

Characteristics	Central Africa = 4,871 (23%)	East Africa = 4,728 (23%)	Southern Africa = 7,446 (35%)	West Africa = 3,695 (17%)	Total = 20,742
Age at sexual debut (mean) and (SD)	15.7 (5.3)	15.8 (4.8)	16.4 (5.0)	17.8 (7.8)	16.2 (5.7)
Cohabitation
Married before 18	26.96	11.32	13.68	21.14	17.29
Married after 18	73.04	88.68	86.32	78.86	82.71
Union status
Never married	83.33	82.31	78.15	78.93	80.45
Cohabiting	9.08	5.65	1.08	4.31	4.57
Married	4.96	10.07	18.61	15.47	12.90
Previously in union	2.63	1.97	2.16	1.29	2.07
Age
15–17	20.89	16.20	19.51	10.38	17.46
18–19	23.55	21.02	22.56	18.67	21.75
20–24	55.56	62.78	57.92	70.95	60.80
Place of residence
Urban	64.07	37.45	33.63	47.40	44.10
Rural	35.93	62.55	66.37	52.60	55.90
Educational attainment
No education	1.73	2.52	1.77	14.07	4.12
Primary	15.76	49.65	39.90	14.73	31.97
Secondary	75.63	39.27	54.45	61.05	57.14
Higher	6.88	8.56	3.89	10.15	6.77
Work status
No	49.17	26.25	36.28	35.38	36.85
Yes	50.83	73.75	63.72	64.62	63.15
Household wealth status
Poor	33.85	31.91	32.63	28.40	32.00
Middle	21.34	20.61	21.88	19.82	21.10
Rich	44.81	47.48	45.49	51.79	49.91
Sex household head
Male	76.62	74.64	73.56	81.37	75.92
Female	23.38	25.36	26.44	18.63	24.08
Media exposure
No	75.54	17.53	50.22	36.67	41.72
Yes	24.46	82.47	49.78	53.33	58.58

Note. Percentage distributions were weighted to account for clustering and the complex design of the Demographic and Health Survey for each country.

Table 2.

Socioeconomic and Demographic Profile of Young Females in Sub-Saharan Africa.

Characteristics	Central Africa = 11,624 (21%)	East Africa = 13,217 (24%)	Southern Africa = 14,248 (26%)	West Africa = 15,237 (28%)	Total = 54,328
Age at sexual debut (mean) and (SD)	17.8 (13.1)	18.1 (11)	18. (13)	19.5 (17)	18.5 (14)
Cohabitation
Married before 18	60.20	43.94	57.06	73.45	59.64
Married after 18	39.80	56.06	42.94	26.55	40.36
Union status
Never married	44.31	40.49	30.64	30.83	36.01
Cohabiting	28.28	14.43	3.91	8.23	12.90
Married	19.39	38.09	56.87	58.28	44.68
Previously in union	80.2	6.99	8.59	2.65	6.41
Age
15–17	18.60	12.38	14.79	17.80	15.87
18–19	21.69	19.70	22.00	21.59	21.26
20–24	59.71	67.92	63.20	60.61	62.88
Place of residence
Urban	58.97	38.46	27.68	38.43	40.01
Rural	41.03	61.54	72.32	61.57	59.99
Educational attainment
No education	6.73	6.22	4.22	40.70	15.48
Primary	26.19	53.71	52.00	18.69	37.55
Secondary	62.55	33.92	41.38	36.33	42.68
Higher	4.54	6.15	2.40	4.28	4.30
Work status
No	53.36	39.15	54.61	56.53	52.04
Yes	46.64	60.85	45.39	43.47	47.96
Household wealth status
Poor	37.54	36.06	41.66	40.17	39.00
Middle	20.98	18.49	19.09	19.54	19.48
Rich	41.48	45.45	39.25	40.29	41.53
Sex household head
Male	73.43	70.83	70.85	80.69	74.16
Female	26.57	29.17	29.15	19.31	25.84
Media exposure
No	80.02	30.31	61.60	54.25	57.84
Yes	19.98	69.69	38.40	45.75	42.16

Note. Percentage distributions were weighted to account for clustering and the complex design of the Demographic and Health Survey for each country.

Results in Tables 3 and 4 show the unadjusted association between age at sexual debut and age at early marriage among males and females.

Table 3.

Unadjusted and Adjusted Association Between Age at Sexual Debut and Age at Early Union Formation Among Males.

Characteristics	Central Africa	East Africa	Southern Africa	West Africa	Sub-Saharan Africa
Unadjusted panel	OR (CI)	OR (CI)	OR (CI)	OR (CI)	OR (CI)
Age at sexual debut	0.84 [0.79, 0.90]***	0.84 [0.79, 0.90]***	0.96 [0.93, 1.00]	0.63 [0.58, 0.69]	0.86 [0.84, 0.89]***
Adjusted panel
Age at sexual debut	0.77 [0.70, 0.85]***	0.83 [0.78, 0.90]***	0.98 [0.94, 1.01]	0.66 [0.60, 0.72]***	0.86 [0.83, 0.89]***
Place of residence
Urban	1.03 [0.55, 1.920	0.71 [0.40, 1.27]	0.52 [0.34, 0.81]***	1.03 [0.55, 1.90]	0.80 [0.62, 1.04]
Rural
Educational attainment
No education	0.75 [0.25, 2.21]	0.34 [0.17, 0.66]***	0.45 [0.24, 0.83]***	0.76 [0.44, 1.31]	0.65 [0.49, 0.86]***
Primary	0.66 [0.23, 1.84]	0.19 [0.08, 0.44]***	0.14 [0.07, 0.28]***	0.70 [0.40, 1.19]	0.49 [0.36, 0.66]***
Secondary and higher
Work status
No	0.68 [0.36, 1.27]	0.99 [0.30, 3.23]	1.49 [0.94, 2.38]	0.74 [0.41, 1.33]	0.89 [0.67, 1.18]
Yes
Household wealth status
Poor	0.73 [0.43, 1.25]	0.64 [0.35, 1.17]	0.85 [0.57, 1.26]	0.42 [0.21, 0.81]*	0.75 [0.58, 0.96]*
Middle	0.81 [0.42, 1.58]	0.43 [0.21, 0.87]**	0.68 [0.42, 1.09]	1.04 [0.56, 1.92]	0.70 [0.53, 0.93]*
Rich
Sex household head
Male	1.24 [0.58, 2.64]	1.86 [1.07, 3.21]*	1.41 [0.92, 2.15]	0.68 [0.33, 1.42]	1.27 [0.97, 1.67]
Female
Media exposure
No	1.10 [0.64, 1.90]	1.38 [0.78, 2.44]	0.77 [0.57, 1.04]	0.87 [0.57, 1.33]	0.74 [0.62, 0.89]***
Yes

Note. CI = confidence interval; OR = odds ratio.

*p < .1 (significant at 10%). **p < .05 (significant at 5%). ***p < .01 (significant at 1%).

Table 4.

Unadjusted and Adjusted Association Between Age at Sexual Debut and Age at Early Union Formation Among Females.

Characteristics	Central Africa	East Africa	Southern Africa	West Africa	Sub-Saharan Africa
Unadjusted panel	OR (CI)	OR (CI)	OR (CI)	OR (CI)	OR (CI)
Age at sexual debut	0.99 [0.99, 0.99]*	0.98 [0.98, 0.99]*	0.99 [0.99, 0.99]*	0.99 [0.99, 0.99]	0.99 [0.99, 0.99]
Adjusted panel
Age at sexual debut	0.99 [0.99, 0.99]*	0.98 [0.97, 0.98]*	0.99 [0.99, 0.99]	0.99 [0.99, 0.99]	0.99 [0.99, 0.99]
Place of residence
Urban	1.08 [0.95, 1.23]	0.89 [0.76, 1.03]	1.11 [0.99, 1.25]	1.24 [1.10, 1.39]***	1.02 [0.96, 1.08]
Rural
Educational attainment
No education	0.94 [0.79, 1.12]	0.48 [0.40, 0.58]***	0.88 [0.72, 1.08]	0.57 [0.50, 0.65]***	0.45 [0.42, 0.48]***
Primary	0.61 [0.51, 0.72]***	0.13 [0.10, 0.17]***	0.29 [0.24, 0.36]***	0.24 [0.22, 0.28]***	0.23 [0.21, 0.24]***
Secondary and higher
Work status
No	1.09 [0.98, 1.21]	0.81[0.72, 0.92]***	0.99 [0.91, 1.08]	1.00 [0.91, 1.10]	0.93 [0.88, 0.97]**
Yes
Household wealth status
Poor	1.03 [0.90, 1.19]	0.77 [0.66, 0.90]***	1.04 [0.92, 1.17]	0.75 [0.66, 0.85]***	0.45 [0.42, 0.48]***
Middle	0.85 [0.74, 0.98]*	0.64 [0.55, 0.75]***	0.86 [0.76, 0.97]*	0.60 [0.52, 0.68]***	0.23 [0.21, 0.24]***
Rich
Sex household head
Male	1.18 [1.04, 1.35]*	1.04 [0.90, 1.20]	1.10 [1.00, 1.22]	0.79 [0.68, 0.91]***	0.98 [0.92, 1.04]
Female
Media exposure
No	0.82 [0.71, 0.96]*	1.09 [0.96, 1.23]	0.90 [0.82, 0.98]	0.85 [0.77, 0.94]***	0.74 [0.71, 0.78]***
Yes

Note. CI = confidence interval; OR = odds ratio.

*p < .1 (significant at 10%). **p < .05 (significant at 5%). ***p < .01 (significant at 1%).

Adjusted Association Between Age at Sexual Debut and Age at Early Marriage

After controlling for other covariates in Table 4 in the adjusted panel, age at sexual debut remained significantly associated with early marriage among males in Central Africa (OR = 0.77), East Africa (OR = 0.83), and West Africa (OR = 0.66). There was no difference in the likelihood of early marriage for females by sexual debut. Some covariates were associated with early marriage among males and females, although direction of association differed. For example, by place of residence, males in rural Southern Africa had lower odds (OR = 0.52) of early marriage compared to males in urban areas. Females in rural Southern Africa were 11% more likely to be married before the age of 18, although this association was insignificant. However, females in rural West Africa (OR = 1.24) were significantly more likely to marry before the age of 18.

By socioeconomic characteristics, secondary and higher education reduced the odds of male marriage before the age of 18 in East Africa (OR = 0.19) and Southern Africa (OR = 0.14) and reduced early marriage among females in Central Africa (OR = 0.61), East Africa (OR = 0.13), Southern Africa (OR = 0.29), and West Africa (OR = 0.24). While work status was not associated with early marriage among males in all regions, females in East Africa (OR = 0.81) had significantly lower odds of being married before the age of 18. Household wealth status was associated with lower odds of child marriage for females in SSA (OR = 0.23), but this was only significant for males in East Africa (OR = 0.43).

By sex of the household head, males from female-headed households in East Africa (OR = 1.86) and females in Central Africa (OR = 1.18) were significantly more likely to be married before age 18. On the other hand, female household headship reduced the odds of early marriage among females in West Africa (OR = 0.79).

Results in Table 5 show the unadjusted association between age at sexual debut and age at union formation among males and females.

Table 5.

Unadjusted and Adjusted Relative Risk for Age at First Sex and Union Status for Each Region of Africa.

	Cohabiting Versus Never Married	Married Versus Never Married	Cohabiting Versus Never Married	Married Versus Never Married
	Unadjusted RRR (CI)	Unadjusted RRR (CI)	Adjusted RRR (CI)	Adjusted RRR (CI)
Young males
Central Africa	1.01 [0.98, 1.04]	1.03 [1.01, 1.06]***	1.03 [0.98, 1.09]	1.05 [1.00, 1.11]
East Africa	1.17 [1.12, 1.21]***	1.15 [1.11, 1.19]	1.16 [1.12, 1.21]***	1.14 [1.10, 1.19]***
Southern Africa	1.11 [1.03, 1.20]***	1.10 [1.02, 1.18]***	1.14 [1.04, 1.25]***	1.13 [1.04, 1.23]***
West Africa	0.88 [0.83, 0.95]***	1.02 [1.01, 1.03]***	0.89 [0.83, 0.96]***	1.01 [1.00, 1.02]***
Young females
Central Africa	1.00 [1.00, 1.01]***	1.01 [1.01, 1.01]***	1.00 [1.00, 1.01]***	1.01 [1.01, 1.02]***
East Africa	1.01 [1.00, 1.01]***	1.01 [1.00, 1.01]***	1.01 [1.00, 1.02]	1.02 [1.01, 1.02]
Southern Africa	1.02 [1.01, 1.03]***	1.01 [1.01, 1.02]***	1.03 [1.02, 1.04]***	1.02 [1.01, 1.03]***
West Africa	1.01 [1.00, 1.01]***	1.01 [1.02, 1.02]***	1.01 [1.00, 1.02]***	1.02 [1.01, 1.03]***

Note. CI = confidence interval.

*p < .1 (significant at 10%). **p < .05 (significant at 5%). ***p < .01 (significant at 1%).

Adjusted Association Between Age at Sexual Debut and Age at Union Formation

After controlling for covariates, an increase in age at first sex was significantly associated with higher odds of cohabiting among males in East Africa, relative risk ratios (RRR) = 1.16) and Southern Africa (RRR = 1.14) but a lower odds of cohabiting among West African males (RRR = 0.89). Similarly, increase in age at first sex was significantly associated with higher odds of being married among males in East Africa (RRR = 1.14) and Southern Africa (RRR = 1.13; see Table 3).

There was no association between place of residence and the odds of cohabiting among males but males in rural Central Africa (RRR = 2.23) and West Africa (RRR = 1.97) had higher odds of being married compare to those in urban area, while males in rural East Africa (RRR = 0.61) were significantly less likely to be married. Results differed for females because rural residency was a protective factor for cohabitation in Central Africa (RRR = 0.83) but a risk factor for females in West Africa (RRR = 1.56). Females in rural Central Africa (RRR = 1.33) had higher odds of being married, while those in East Africa had lower odds (RRR = 0.77) of marriage.

By socioeconomic characteristics, secondary and higher education among males reduced the odds of cohabitation in East Africa (RRR = 0.17) and Southern Africa (RRR = 0.30) and lowered the odds of being married in East Africa (RRR = 0.32), Southern Africa (RRR = 0.30), and West Africa (RRR = 0.17). Secondary and higher education were protective factors for cohabitation and marriage among females in all regions. Males and females who were working had higher odds of cohabiting or being in a union in all regions. Also, males and females from rich households had lower odds of cohabiting or being in a union. Meanwhile, males from female-headed households had lower odds of cohabiting or getting married in all regions. This association was similar for females except in Southern Africa (RRR = 2.54) where females from female-headed household had significantly higher odds of being married. Finally, males exposed to mass media in Southern Africa had significantly higher odds of cohabiting (RRR = 3.48) or getting married (RRR = 1.36). Exposure of females to mass media was significantly associated with high likelihood of cohabiting in Southern Africa (RRR = 1.47) and West Africa (RRR = 1.31).

Discussion

This research found that age at first sex is an important determinant of union formation among male youth in SSA. Although social norms influence over- and underreporting of sexual activity among males and females (Plummer & Wight, 2011), respectively, our finding on the mean age at sexual debut is hardly distinguishable from previous research (Doyle, Mavedzenge, Plummer, & Ross, 2012; Stephenson et al., 2014).

A larger number of male youth in Central Africa married at an early age than males in other regions. This result differed for females, about three quarters of whom married early compared to their counterparts in other regions. These are consistent with other studies on age at marriage in Central Africa and West Africa (Fedd, Edmeades, Lantos, & Onovo, 2015; Meekers & Gage, 2017). As with Ayiga and Rampagane (2013) who attributed the regional variations in age at first marriage to cultural difference and stage of demographic transition, our findings highlight the contextual effects of region on youth behaviors in SSA. Also, while no support exists in the literature for the higher occurrence of cohabitation among males and females in Central Africa, higher youth poverty level and poor access to education may be important factors. Decades of conflict and social and political unrests in the region may also influence girls’ decision to move in with partners.

We also found that delaying sexual debut delayed early union formation among males in all regions except Southern Africa. Once again, these differences could be attributable to cultural differences between regions. Better data on cultural norms could help future research to highlight the mechanism behind these differences. While these associations were not evident for females, factors like place of residence and socioeconomic characteristics may be responsible for early union formation among them. In all the regions, education (secondary and higher) and household wealth status delayed early union formation. This results is similar with findings in developed (Cantalini, 2017) and developing countries (Kamal, Hassan, Alam, & Ying, 2015), including SSA (Manda & Meyer, 2005). It suggests that improving the socioeconomic status and education of young girls could be an efficient way of delaying union formation, and their negative effects, among this population.

Most crucially, while early sexual debut significantly influences the kind of first union of male youths, similar results was not observed for females. This gender difference can be attributable to the differences in union status at the first sex. In fact, it more likely that male sexual debut influences their union type, for it may have happened before union, while because of cultural norms, female sexual debut occurs when union has been formed (Molla, Berhane, & Lindtjørn, 2008). However, the associations differed by region for males which can be attributed to the different cultural norms in each region.

Further results showed differences in union formation patterns by place of residence. For males, in Central Africa and West Africa, residing in a rural area was associated with higher odds of being married compared to being single. Rural residence has been generally associated with early marriage because of its agricultural economic structure and limited educational opportunities (Kuépié, Shapiro, & Tenikue, 2015; Orazem & King, 2007). However, different results were seen in East Africa where males living in rural areas were less likely to be married. More detailed date with variable on contextual norms and cultural practices could help to understand this difference.

Limitations

The age at first sex collected in the DHS is auto-declared, which could cause error in measurement because of recall bias and taboo around early sex especially for women. Also, reporting of age at union might be inaccurate also due to recall bias and may be more prevalent in the rural areas where literacy level is low. Because of sociocultural norms in some areas, cohabiting youths may report being married.

Conclusion

The study supports the hypothesis that the type of unions young adults form is determined by their age at sexual debut. From a policy intervention perspective, our results demonstrate the complexity of developing gender and region-wide strategies for addressing youth sexual behavior and union formation.

Footnotes

Appendix

Table A5.

Unadjusted and Adjusted Relative Risk for Age at First Sex and Union Status Among Male and Female Youth in West Africa.

	Males		Females
	Cohabiting Versus Never Married	Married Versus Never Married	Cohabiting Versus Never Married	Married Versus Never Married
Sample Characteristics	Adjusted RRR (CI)	Adjusted RRR (CI)	Adjusted RRR (CI)	Adjusted RRR (CI)
Place of residence
Urban	1.34 [0.84, 2.15]	1.97 [1.35, 2.87]***	1.56 [1.29, 1.89]***	1.21 [0.83, 1.74]
Rural
Education
No education	2.60 [1.12, 6.02]***	0.27 [0.18, 0.40]***	0.67 [0.51, 0.88]***	0.26 [0.18, 0.38]***
Primary	1.86 [0.83, 4.17]	0.17 [0.12, 0.24]***	0.24 [0.18, 0.32]***	0.10 [0.06, 0.14]***
Secondary and higher
Working
No	3.96 [2.31, 6.79]***	6.90 [4.72, 10.09]***	1.65 [1.41, 1.95]***	1.77 [1.34, 2.32]***
Yes
Wealth status
Poor	0.39 [0.23, 0.64]***	0.51 [0.36, 0.73]***	0.86 [0.69, 1.07]	0.87 [0.62, 1.21]
Middle	0.32 [0.19, 0.53]***	0.43 [0.29, 0.62]***	0.90 [0.71, 1.12]	0.46 [0.30, 0.71]***
Rich
Sex household head
Male	1.07 [0.68, 1.68]	0.20 [0.11, 0.37]***	0.44 [0.37, 0.53]***	1.20 [0.89, 1.62]
Female
Media exposure
No	1.43 [0.93, 2.19]	1.04 [0.79, 1.37]	1.31 [1.10, 1.56]***	1.05 [0.80, 1.39]
Yes

Note. CI = confidence interval.

*p < .1 (significant at 10%). **p < .05 (significant at 5%). ***p < .01 (significant at 1%).

Acknowledgment

This research was supported by the Consortium for Advanced Research Training in Africa (CARTA). CARTA is jointly led by the African Population and Health Research Center and the University of the Witwatersrand and funded by the Carnegie Corporation of New York (Grant No--B 8606.R02), Sida (Grant No:54100029), the DELTAS Africa Initiative (Grant No: 107768/Z/15/Z). The DELTAS Africa Initiative is an independent funding scheme of the African Academy of Sciences (AAS)’s Alliance for Accelerating Excellence in Science in Africa (AESA) and supported by the New Partnership for Africa’s Development Planning and Coordinating Agency (NEPAD Agency) with funding from the Wellcome Trust (Grant No: 107768/Z/15/Z) and the UK government, The statements made and views expressed are solely the responsibility of the fellow. We also acknowledge Nancy Luke for giving feedback that helped improved the manuscript.

Author Contributions

Oluwaseyi Dolapo Somefun contributed to conception, design, acquisition, analysis, and interpretation; drafted the manuscript; critically revised the manuscript; gave final approval; and agrees to be accountable for all aspects of work ensuring integrity and accuracy. Clifford Odimegwu contributed to conception, critically revised manuscript, gave final approval, and agrees to be accountable for all aspects of work ensuring integrity and accuracy. Arlette Simo Fotso contributed to conception, design, and interpretation; critically revised the manuscript; gave final approval; and agrees to be accountable for all aspects of work ensuring integrity and accuracy. Kudus Adebayo contributed to conception and interpretation, critically revised manuscript, gave final approval, and agrees to be accountable for all aspects of work ensuring integrity and accuracy.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Oluwaseyi Dolapo Somefun

Open Practice

All data and materials have been made publicly available via the Open Science Framework and can be accessed at https://dhsprogram.com/data/available-datasets.cfm. The design and analysis plans for the experiments were not preregistered. The complete Open Practices Disclosure for this article can be found at . This article has received the badges for Open Data.

References

Abramsky

Watts

C. H.

Garcia-Moreno

Devries

Kiss

Ellsberg

… Heise

(2011). What factors are associated with recent intimate partner violence? Findings from the WHO multi-country study on women’s health and domestic violence. BMC Public Health, 11, 109.

Altman

D. G.

Royston

(2006). The cost of dichotomising continuous variables. British Medical Journal, 332, 1080.

Ayiga

Rampagane

(2013). Determinants of age at first marriage in sub-Saharan Africa: A comparative study of Uganda and South Africa. Journal of Social Development in Africa, 28, 9–36.

Boden

J. M.

Fergusson

D. M.

John Horwood

(2008). Early motherhood and subsequent life outcomes. Journal of Child Psychology and Psychiatry, 49, 151–160.

Calvès

A. E.

(2016). First union formation in urban Burkina Faso: Competing relationship transitions to marriage or cohabitation. Demographic Research, 34, 421.

Cameron

A. C.

Miller

D. L.

(2015). A practitioner’s guide to cluster-robust inference. Journal of Human Resources, 50, 317–372.

Cantalini

(2017). Does education affect the timing or probability of family formation? An analysis of educational attainment and first union in Italy. Research in Social Stratification and Mobility, 49, 1–10.

Clarkberg

(1999). The price of partnering: The role of economic well-being in young adults’ first union experiences. Social Forces, 77, 945–968.

Coltabiano

Castiglioni

(2008). Changing family formation in Nepal: Marriage, cohabitation and first sexual intercourse. International Family Planning Perspectives, 34, 30–39.

10.

Doyle

A. M.

Mavedzenge

S. N.

Plummer

M. L.

Ross

D. A.

(2012). The sexual behaviour of adolescents in sub-Saharan Africa: Patterns and trends from national surveys. Tropical Medicine & International Health, 17, 796–807.

11.

Fedd

N. S.

Edmeades

Lantos

Onovo

(2015). Child marriage adolescent pregnancy and family formation in West and Central Africa: Patterns trends and drivers of change. Retrieved from https://www.unicef.org/wcaro/english/Child_Mariage_Adolescent_Pregnancy_and_Family_Formation_in_WCA.pdf

12.

Frech

(2014). Pathways to adulthood and changes in health-promoting behaviors. Advances in Life Course Research, 19, 40–49. doi:10.1016/j.alcr.2013.12.002

13.

Handa

Peterman

Huang

Halpern

Pettifor

Thirumurthy

(2015). Impact of the Kenya cash transfer for orphans and vulnerable children on early pregnancy and marriage of adolescent girls. Social Science & Medicine, 141, 36–45.

14.

Hoffman

Bann

Higgins

Vohr

, Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network. (2015). Developmental outcomes of extremely preterm infants born to adolescent mothers. Pediatrics, 135:1082–1092.

15.

Kamal

S. M. M.

Hassan

C. H.

Alam

G. M.

Ying

(2015). Child marriage in Bangladesh: Trends and determinants. Journal of Biosocial Science, 47, 120–139.

16.

Kleinepier

de Valk

(2016). Ethnic differences in family trajectories of young adult women in the Netherlands: Timing and sequencing of events. Demographic Research, 35, 671–710.

17.

Kuépié

Shapiro

Tenikue

(2015). Access to schooling and staying in school in selected sub-Saharan African countries. African Development Review, 27, 403–414.

18.

Lee

Gramotnev

(2006). Predictors and outcomes of early motherhood in the Australian Longitudinal Study on Women’s Health. Psychology, Health & Medicine, 11, 29–47.

19.

Lloyd

C. B.

(2005). Growing up global: The changing transitions to adulthood in developing countries. Washington, DC: National Academies Press.

20.

Manda

Meyer

(2005). Age at first marriage in Malawi: A Bayesian multilevel analysis using a discrete time-to-event model. Journal of the Royal Statistical Society: Series A (Statistics in Society), 168, 439–455.

21.

Meekers

Gage

A. J.

(2017). Marriage patterns and the demographic dividend. In Africa’s population: In search of a demographic dividend (pp. 251–265). Berlin, Germany: Springer. Retrieved from https://link.springer.com/chapter/10.1007/978-3-319-46889-1_16

22.

Mensch

B. S.

Singh

Casterline

J. B.

(2005). Trends in the timing of first marriage among men and women in the developing world. In The changing transitions to adulthood in developing countries: Selected studies (pp. 118–171).

23.

Molla

Berhane

Lindtjørn

(2008). Traditional values of virginity and sexual behaviour in rural Ethiopian youth: Results from a cross-sectional study. BMC Public Health, 8, 9.

24.

Morinis

Carson

Quigley

M. A.

(2013). Effect of teenage motherhood on cognitive outcomes in children: A population-based cohort study. Archives of Disease in Childhood, 98, 959–964.

25.

Nour

N. M.

(2006). Health consequences of child marriage in Africa. Emerging Infectious Diseases, 12, 1644.

26.

Odimegwu

Somefun

O. D.

(2017). Ethnicity, gender and risky sexual behaviour among Nigerian youth: An alternative explanation. Reproductive Health, 14, 16.

27.

Orazem

P. F.

King

E. M.

(2007). Schooling in developing countries: The roles of supply, demand and government policy. Handbook of Development Economics, 4, 3475–3559.

28.

Plummer

M. L.

Wight

(2011). Young People’s Lives and Sexual Relationships in Rural Africa: Findings from a Large Qualitative Study in Tanzania, ISBN 9780739135785. Lanham, MD: Lexington Books.

29.

Raley

R. K.

Crissey

Muller

(2007). Of sex and romance: Late adolescent relationships and young adult union formation. Journal of Marriage and Family, 69, 1210–1226.

30.

Rosenbaum

J. E.

Zenilman

Rose

Wingood

DiClemente

(2016). Predicting unprotected sex and unplanned pregnancy among urban African-American adolescent girls using the theory of gender and power. Journal of Urban Health, 93, 493–510.

31.

Stephenson

Simon

Finneran

(2014). Community factors shaping early age at first sex among adolescents in Burkina Faso, Ghana, Malawi, and Uganda. Journal of Health, Population, and Nutrition, 32, 161.

32.

Subaiya

Johnson

(2008). What’s in the gap? Factors associated with the interval between age at first sex and age at first marriage in Cameroon and their implications for reproductive health and women’s empowerment. Retrieved from https://dhsprogram.com/pubs/pdf/WP43/WP43.pdf