“‘He supported me 100%’: Mexican-immigrant fathers, daughters, and adolescent sexual health”

Abstract

Objective:

First and second generation Mexican-origin adolescents in the U.S. face social and economic disadvantage and sexual health disparities. Although fathers can support child and adolescent development, the literature has portrayed Mexican-origin immigrant fathers as emotionally distant and sexist. This study aims to treat migration as a social determinant of health to examine father-daughter relationships and adolescent sexual health in Mexican-origin immigrant families.

Methods:

Integrating qualitative data from life history interviews with 21 Mexican-origin young women in immigrant families with quantitative data on first and second generation Mexican-origin young women in the National Longitudinal Study of Adolescent to Adult Health, this study describes father-daughter relationships, examines the association between father-daughter relationships and daughters’ early sexual initiation, and considers the impact of migration on the father-daughter relationship and sexual health among Mexican-origin young women.

Results:

Qualitative data identify four types of father-daughter relationships: “good,” hostile, distant, and conflicted. Supporting the qualitative patterns, quantitative data find that positive or “good” father-daughter relationship quality is significantly associated with reduced risk of early sexual initiation. Importantly, father-daughter separation across borders and economic inequality facing immigrant families is associated with hostile or distant father-daughter relationship quality and increased risk of early sexual initiation.

Conclusions:

Reports of good father-daughter relationships are common and may protect against early sexual initiation in Mexican-origin immigrant families. Policies that keep families together and reduce economic inequality among immigrants may also reduce sexual health disparities among immigrant adolescents.

Introduction

Mexican-origin adolescents make up the largest proportion of immigrant adolescents in the U.S. (Ennis, Rios-Vargas, and Albert 2011; US Census Bureau 2012). This fast growing population faces economic and health disparities, including high rates of adolescent birth and sexually transmitted infections (STIs) (Martin et al. 2015; CDC 2014), both associated with early sexual initiation (Sandfort et al. 2008). Supportive parenting can support child and adolescent development and is associated with reduced risky sexual behaviors (Rodgers 1999). However, there is a dearth of literature on father-child relationships across race/ethnic groups; the majority of research on parenting and adolescent sexual health has focused on mother-child relationships (Samari and Seltzer 2016; Merten and Henry 2011; Trejos-Castillo and Vazsonyi 2008).

Even fewer studies on parenting and adolescent sexual health specifically focus on Mexican-origin immigrant families (Killoren et al. 2011; Tyrell et al. 2014; Updegraff, Delgado and Wheeler 2009). And Latino and Mexican-origin fathers as often described as controlling, sexist, or providing economic, not emotional support (Crockett et al. 2007; Russell and Lee 2004). Moreover, prior research on parenting and adolescent outcomes, such as early sexual initiation, focus on cultural aspects of parenting and overlook the process of migration as a social determinant of health. Yet immigrant parents face both socioeconomic disadvantage and discrimination (Castañeda et al. 2015). Thus, extending the literature, we examine migration as a social determinant of health to explore the role of Mexican-immigrant fathers on the sexual health of their daughters, rather than focusing only on cultural parenting practices.

Father-child relationships

The handful of studies on Latino and Mexican-origin fathers often portray them as sexist and emotionally distant and explain parenting behavior through a patriarchal “Latino culture,” reinforcing stereotypes (Crockett et al. 2007; Russell and Lee 2004). Other studies found that Latino and Mexican-origin adolescents may have more warm or trusting relationships with mothers than fathers (Crockett et al. 2007; Killoren et al. 2011; Updegraff, Delgado, and Wheeler 2009). However, father-daughter relationships among Latino and Mexican-origin immigrant families have not been well described and even less is known about father-daughter relationships and adolescent sexual health.

Fathers and adolescent sexual health

Risky behavior among adolescents, such as early sexual initiation, may have long-term negative consequences including unintended pregnancy and STIs (Sandfort et al. 2008). Although there is a large body of research demonstrating supportive two-parent families reduce risk of early sexual initiation (Merten and Henry 2011), growing evidence suggests that fathers’ high-quality involvement is also beneficial to children’s well-being and development (Lamb 2000). Children raised in father-absent households have earlier ages of first sexual intercourse than those raised in father-present households (Mendle et al. 2009). Father’s relationship quality has been found to be more beneficial when the father lives with the adolescent (Carlson 2006), but father’s relationships with children may be protective even when provided by a nonresident father (Amato and Gilbreth 1999).

The quality of paternal care also influences sexual behavior independent of other family stressors. Some studies find that close father-daughter relationships protect against risky behavior among college students (Rostad, Silverman, and McDonald 2014), and may be associated with college versus motherhood as a first transition to adulthood (Amato et al. 2008). However, one study found that neither maternal nor paternal support was associated with sexual risk behaviors among Latino adolescents (Killoren and Deutsch 2013). More research is needed on father-daughter relationship quality and sexual health among Mexican-immigrant U.S. families.

Fathers, adolescent sexual health, and migration

Examination of parenting and adolescent sexual health behaviors among Latinos and immigrants across the globe focus on cultural parenting practices and often overlook migration as a social determinant of health. Latino and Mexican-origin immigrant families are not a homogenous group and immigrant parents face a socio-political context that many native-born parents do not face (Castañeda et al. 2015). First, Latino families differ by country of origin and immigrant generation. Children from immigrant families are either foreign-born themselves (first generation) or have at least one foreign-born parent (second generation). Family characteristics and sexual health outcomes differ by immigrant generation. For example, second generation adolescents are more likely to live with two biological parents than the first (Landale, Thomas, and Van Hook 2011), and first generation Latino and Mexican-origin adolescents are less likely to have sex and use contraception compared to second generation adolescents (McDonald, Manlove, and Ikramullah 2009). Protection against sexual risk conferred by mother-child relationships is conflicting; it may vary by immigrant generation (Killoren et al. 2011) or protect across all generations (Samari and Seltzer 2016; Trejos-Castillo and Vazsonyi 2008). However, less is known about relationship quality with fathers and adolescent sexual behavior across immigrant generations.

Second, the migration process itself may influence parenting and adolescent health outcomes. Families are often separated during the migration process because of the cost of migration and deportation (Suarez-Orozco, Todorova, and Louie 2002). They also have limited access to public assistance and migrant workers are often exploited in low-wage jobs with no protections such as paid time off (Brown and Patten 2014; Child Trends Data Bank 2014; Fussell 2011). Immigrant parents may be predisposed to “survival parenting,” working long and taxing hours that limits the time they have to spend time with their children (Vasquez 2011). Thus, cultural parenting practices, including gendered treatment of daughters, and parent-child relationships are influenced by the socio-political context of migration (González-López 2004; Castañeda et al. 2015; Coleman-Minahan 2017).

Conceptual framework

In order to address these gaps in the literature by examining father-daughter relationships and adolescent sexual health within a socio-political context of migration, we draw on a conceptual framework that integrates segmented assimilation theory with the theory of gender and power. Segmented assimilation theory has been used to examine adolescent economic, academic, and health risk outcomes (Portes, Fernández-Kelly, and Haller 2009; Eitle, Wahl, and Aranda 2009). It identifies three pathways of socioeconomic integration of children of immigrants (second generation) that is influenced by both parental attributes and the host country: (1) upward mobility and assimilation into the U.S. middle class while compromising ethnic identity; (2) upward mobility through embracing both ethnic identity and U.S. values; and (3) downward mobility by engaging in risk behaviors that lead to poor educational and economic outcomes, including adolescent childbearing (Portes and Zhou 1993). The theory suggests that social capital, particularly a stable two-parent family with supportive parenting, may buffer social and economic disadvantage and discrimination facing immigrant families and reduce adolescent risk behaviors (Portes and Zhou 1993). Growing up in a two-parent family is associated with upward economic mobility in second generation immigrants (Portes, Fernández-Kelly, and Haller 2009), greater educational attainment, and decreased risk of early sexual initiation (Lammers et al. 2000; Merten and Henry 2011). Thus, supportive father-daughter relationships may also reduce risk of early sexual initiation among adolescents in immigrant families. However, this theory does not address the role of gender in immigrant families.

We build on the work of social scientists who use the theory of gender and power to examine sexual health and family relationships among immigrant families (Hirsch 2003; González-López 2004; García 2012). The theory explains gender inequality through three structures, the gendered division of labor and power in the home and workplace, and cathexis, or social and cultural expectations (Connell 1987). Application of this theory reminds us to examine gender inequality at multiple levels, rather than individual relationships alone. In the household, immigrant parents, specifically fathers, often treat daughters and sons differently and have stricter controls for daughters (Ayala 2006; Foner 2009; Wilson, Dalberth, and Koo 2010), which could constrain father-daughter relationships. At the economic and workforce level, father-daughter relationships may be particularly vulnerable to the father’s position in the migrant labor force and his ability to provide for the family because fathers, in their male roles, are often considered the economic supporters of the family (Crockett et al. 2007; Killoren et al. 2011; Enriquez 2015). Thus, combining segmented assimilation theory and theory of gender and power allows us to more thoroughly describe father-daughter relationships, their impact on adolescent sexual health, and how the socio-political context of migration, particularly economic disadvantage, impacts father-daughter relationships and sexual health.

We use this conceptual framing and integrate qualitative and quantitative data to address three aims: (1) describe father-daughter relationship quality among Mexican-origin immigrant families; (2) examine the association between father-daughter relationship quality and early sexual initiation among Mexican-origin young women; and (3) investigate how the socio-political context (i.e. immigrant generation, father-daughter separation, economic disadvantage) facing immigrant families in the U.S. impacts father-daughter relationship quality and sexual initiation.

Methods

We draw on a larger mixed methods study that examines migration, family structure and socioeconomic status (SES), gender and sexual expectations, and the sexual health of Mexican-origin young women (Coleman-Minahan 2017). The study is a partially mixed concurrent equal status design (Leech and Onwuegbuzie 2007). We used an iterative process to integrate qualitative life history interviews with first and second generation Mexican-origin women in the Denver metro area with data from first and second generation Mexican-origin women in the National Longitudinal Study of Adolescent to Adult Health (Add Health). Quantitative analysis and qualitative data collection and analysis occurred separately, but simultaneously, with preliminary findings from one method being used to adapt data collection or analysis of the other. For example, in order to draw inferences between samples, inclusion criteria for qualitative participants were chosen so participants are in the same birth cohort as the Add Health participants. During the study, the association between father-daughter relationship quality and sexual health behaviors of young women arose thematically during interviews, so we conducted additional quantitative analysis to test our qualitative hypotheses for generalizability. Qualitative and quantitative data were then integrated to answer the research questions.

Qualitative data provide rich description of father-daughter relationships (Aim 1) and immigrant families’ experiences with economic disadvantage, as well as how these factors impact young women’s early sexual initiation (Aims 2 & 3). Data from first and second generation Mexican-origin young women in Add Health provide greater generalizability by allowing us to examine how father-daughter relationship quality is associated with early sexual initiation (Aim 2) in a nationally representative sample and to examine migration, specifically immigrant generation and economic disadvantage, as a social determinant of early sexual initiation (Aim 3). The study was approved by the Colorado Multiple Institutional Review Board.

Qualitative Data

Between March 2013 and April 2014, the lead author conducted qualitative semi-structured life history interviews with 11 first generation and 10 second generation Mexican-origin women in the metropolitan area of Denver, Colorado, U.S.A. Inclusion criteria included: age 27–41, identified as female, migrated to the U.S. from Mexico prior to age 18 (one participant migrated shortly after marrying at 19 years old) or born in the U.S. with at least one parent born in Mexico. Two first generation and two second generation participants were pairs of sisters. The women were in the same birth cohort as the Add Health participants. All first generation women spent the majority of their time in the U.S. in the Metro Denver area and all but one second generation woman spent much of their childhoods in the Metro Denver area.

Women were sampled purposively through active and passive recruitment at community organizations in the Metro Denver area. The lead author conducted each semi-structured, focused life history interview in the language of the participant’s choice, lasting between one and two hours. Interview topics were chosen based on our conceptual framework, literature on sexual and reproductive health among adolescents, and Hirsch’s (2003) use of life history interviews to examine gender, relationships, and sexual health among migrants. Topics included childhood, family relationships, parental expectations, gendered division of labor and power in the household, migration experiences, romantic relationships, and family planning. All women were compensated for their time.

Qualitative Analysis

We conducted a content analysis guided by segmented assimilation theory and the theory of gender and power using a five step process; reading, open and closed coding, displaying, reducing, and interpreting (Strauss and Corbin 1990; Ulin 2002). All interview transcriptions were edited in both English and Spanish, and discrepancies in translation were discussed and clarified with a bilingual, bicultural research assistant. The analysis process was reflexive as the lead author recorded thoughts and biases that arose in a journal and Atlas-ti qualitative software was used to provide organization and transparency.

We analyzed father-daughter relationships during childhood and adolescence. Two women described relationships with both a stepfather and a biological father. They are included twice, once for each father-daughter relationship. We systematically compared relationships between the first and second generation and those who had sex prior to age 17 (17.2 was the mean age of sexual initiation for the sample). We also analyzed associations among relationship quality and sexual health behaviors and family SES, migration experiences, and household gender equality. The authors compared main findings to data displays and direct quotes. Finally, results were verified with several community organizations serving Mexican-origin immigrant families.

Quantitative Data

We used data from all four waves (1994–2008) of Add Health, which began in 1994 with 20,475 7–12 graders, and participants were interviewed again in 1996 (Wave II), 2001–2002 (Wave III), and 2007–2008 (Wave IV). Add Health survey methodology has been described elsewhere (Harris et al. 2009). We excluded three participants missing on ever having sexual intercourse, one participant missing on the presence of a resident father, three participants with an age at first sex prior to 11, and participants who were missing a sample weight. The final sample included first and second generation female Mexican-origin participants (N=498).

The outcome of age at sexual initiation was coded based on participants’ responses about month and year of first [penis-vagina] sexual intercourse in Waves I and II and reported age of first intercourse in Waves III and IV.

The key independent variable, father-daughter relationship quality, was measured at Wave I. We created a factor variable based on five survey questions/statements: (1) “How close do you feel to your father?”; (2) “How much do you think he cares about you?; (3) “Most of the time, your father is warm and loving toward you”; (4) “You are satisfied with the way your father and you communicate with each other”; and (5) “Overall, you are satisfied with your relationship with your father.” The responses were all coded on a scale from 1 (not at all/strongly disagree) to 5 (very much/strongly agree). The Eigenvalue was 2.88 and all variables loaded onto one factor with loadings above 0.55. Cronbach’s alpha reliability for the five items was 0.88. To include comparisons between the influence of relationship quality and not having a resident father, we created an additional categorical variable based on the factor variable that included father absence, the lower 75^th percentile of relationship quality, and the highest 25^th percentile of relationship quality.

Control variables included immigrant generation, presence of a resident mother (1 = ‘Yes’ 0 = ‘No’), father’s education, and family income. Immigrant generation was coded categorically as first generation (born in Mexico to parents born in Mexico) and second generation (U.S.-born to at least one Mexican born parent). Father’s education was taken from the parent survey with the participant in-home survey being used for those missing on the parent survey. Father’s highest level of education was coded categorically as less than a high school education, a high school education, and greater than a high school education, and a continuous variable for father’s years of education is included in the interaction models.

Quantitative Analysis

The longitudinal design of Add Health allowed us to use survival models to examine how family factors in early adolescence are associated with risk of sexual initiation over time. Age at sexual initiation was modeled using complementary log-log discrete time hazard models because age was measured in one year intervals and was not truly continuous. A person-year dataset was constructed with participants right censored at age 20 because first sex after age 20 cannot result in adolescent pregnancy. Age and age squared are included because of the concave relationship between age and first sex.

We examined sociodemographics and sexual health outcomes of Add Health and qualitative and participants. Multivariate models included: (1) a model examining the hazard of sexual initiation with relationship quality (categorical), immigrant generation, and a resident mother; (2) the baseline model with relationship quality (continuous), controlling for immigrant generation, father’s education, and a resident mother; (3) the baseline model adding an interaction between relationship quality and immigrant generation to examine if relationship quality influences risk of sexual initiation differently across immigrant generation; and (4) the baseline model adding an interaction between relationship quality and father’s education to examine if relationship quality influences risk of sexual initiation differently by SES. Aside from family income, all variables were missing less than 5% of data. We used ten multiple imputed data sets created by chained equations to address missing data (Little and Rubin 2002). All analyses are weighted to adjust for Add Health’s sampling frame.

Results

We identify and describe four types of father-daughter relationships elucidated by qualitative data. We then explain associations between father-daughter relationships and sexual initiation from patterns and context in the qualitative data that we confirm with quantitative data. Finally, we integrate qualitative and quantitative data to describe how the socio-political context of migration (i.e. immigrant generation, father-daughter separation, and economic disadvantage) are associated with father-daughter relationships.

Table 1 presents weighted demographic characteristics for Add Health and qualitative participants.

Table 1.

Percentages and Mean (SE) of Quantitative (Weighted) and Frequencies of Qualitative Samples^α

	Add Health Participants (1994–2008)		Qualitative Participants
Demographic and Family Characteristics at Wave I	First Generation	Second Generation	First Generation	Second Generation
Age in 2013	35.5	34.9*	32.9	30.7
Age at migration (Mean (SE))	8.0 (0.45)		12.2
Resident Father	74.1	82.6	5	9
Resident Father is Biological	79.3	89.0
Resident Mother	90.5	95.8*	9	10
Family income	$16,177.30	$29,637.90*
Father’s Education
<High School Grad (ref)	87.6	73.2	6	0
HS Grad/GED/Vocational	5.7	12.7	3	6
>High School Grad	6.7	14.2†	2	4
Father-daughter Relationship Qualityβ (continuous) (Mean (SE))	−0.42 (0.12)	−0.22 (0.10)
Father-daughter Relationship Quality (categorical)
No resident father (ref)	26.4	17.4
Lowest 75% relationship quality	58.5	55.8
Highest 25% relationship quality	15.1	26.8†
Age at First Sex (Mean (SE))	17.4 (0.41)	17.3 (0.38)	17 (0.58)	17.6 (0.65)
Experienced an Adolescent Birth	20.7	18.0	6	3
Number of Observations	195	303	11	10

^αp-values refer to significant differences among Add Health participants only

^βThe greater the value, the better the relationship quality

^*p<0.05

^**p<0.01

^***p<0.001

^†p<0.10

Percentages unless labeled as Mean (SE)

Among the Add Health participants, first generation adolescents are significantly less likely than second generation adolescents to have a mother in the home (90.5% vs. 95.8%) and have significantly lower family incomes ($16,177 vs. $29,638) and less educated fathers than second generation adolescents.

Father-daughter relationships

Instead of identifying relationship quality on a numeric scale, qualitative data elucidated four types of father-daughter relationships: (1) “good” relationships, characterized by support or closeness; (2) hostile relationships, characterized by a “mean” or violent father; (3) emotionally distant relationships; and (4) a conflicted relationship, characterized as a relationship with both supportive attributes and frustrations with gender inequality.

Good relationships.

Ten of the 21 participants (four first generation, five second generation) recalled that relationships with their fathers (8 biological, 1 step-father) were generally “good” and were the most common relationship type. Providing additional detail, participants described three attributes of good father-daughter relationships: (1) fathers “sacrificing” for their families, (2) fathers encouraging academic success, and (3) fathers providing emotional support and affection.

Analyzing both participants’ descriptions of their relationships with their fathers and their parents’ migration processes, six young women described positive attributes of their fathers as both sacrificing for their children by migrating to the U.S. and spending time with children despite their long hours and intense physical labor in migrant work. Additionally, three young women specifically mentioned how their fathers took them out to dinner or on vacation despite money often being very tight. First generation Isabel recalled,

My dad used to work construction and that was his only employment, and so he would get home at 7 or 8 at night and I can’t imagine going home after a long day of laborious work and still have the energy to play with your kids. (laughs) …. He would play with us when he got home. Not all the time but he would. He was very unselfish and he joined the weekends when he was at home, he would, you know, going out to have dinner and having four kids it’s expensive, you know, it’s expensive.

Fathers’ encouragement of their daughters’ education was another positive attribute described by participants. A number of fathers helped with homework and checked report cards, often being more fluent in English than mothers. Academic expectations of finishing high school and attending college were linked with fathers’ (and mothers’) migration sacrifices. For example, second generation Julieta said, “He would always tell us those stories [about his migration] because he wanted us to be better and he wanted us to do things better and just thank God for what we had here because he didn’t have that out there [Mexico] and stuff like that.”

Some participants recalled they were closer to their mothers than their fathers. However, emotional support and affection from fathers were far from rare. Young women stated they received a lot of love, cariño, or affection, from their fathers, and for two young women, they were closer to their fathers than their mothers. Second generation Cristina described, “He was just so affectionate. He was understanding and if you wanted to talk he would listen… I would feel more comfortable with him [than my mother], yes.” Similar to Cristina, other women also recalled that their fathers gave them support and advice. Second generation Teresa said, “He supported me 100% in a lot of things.” Indeed, three participants, all from different families, described themselves as “daddy’s little girl.”

Hostile father-daughter relationships.

Six participants described hostile relationships with a father. Four were biological fathers and two were stepfathers. All six women were separated from their biological father at some point in their lives, demonstrating a potential association between family instability and hostile fathering. All described mean and violent fathers. Violence ranged from “sometimes he would beat my mom” to ripping a phone out of a wall out of jealousy. For example, first generation Juana recalled, “My dad was really bossy… He was really mean and powerful, you couldn’t answer back or say anything because he will beat you up. I mean relationship with him? No, not at all.” Other women also mentioned that “he wasn’t loveable,” and none of these women described affection or emotional support from their fathers.

Hostile father-daughter relationships occurred more often in households with extremely gendered divisions of labor and power. For example, in addition to violence, Juana’s father did not help with household chores and treated her brothers differently than her sisters. Additionally, some mothers who migrated to the U.S. without a domestic partner (with or without children) became economically dependent upon male partners, often in unhealthy relationships. These step-fathers sometimes reinforced gender inequality and were violent or emotionally abusive.

Emotionally distant father-daughter relationships.

Although hostile father-daughter relationships implied emotional distance, three other young women described emotionally distant relationships with no emotional abuse or violence. All three of these young women mentioned they did not talk much or spend much time with their fathers. Second generation Ana said, “His job was mostly to take care of us [financially], he didn’t really talk to me much.” Although two of these women were closer to their mothers, second generation Jessica said communication and “expressing feelings” wasn’t present with either parent. She felt she and her parents had a “good relationship” but attributed her parents’ distant relationship with her and sister to “being more focused on work and providing for the family…”

A conflicted or ambiguous father-daughter relationship.

One young woman was conflicted when describing her relationship with her father. First generation Eva, who immigrated at age eight, described her relationship with her father who remained in Mexico, “We’d go and visit [and] when you only see someone a few times a year maybe you do kind of go all out and go out of your way to give them everything… because to me I thought it was always funner to be with my dad.” Later she described, “My relationship with him [father], we’re really close, however, culturally too he doesn’t talk about politics or anything like that with me because I’m a woman.” Eva described how gender inequality negatively impacted their relationship.

Participants who described good relationships also had frustrations with fathers during adolescence. For example, Julieta described how she was not allowed out of the home without her brother or father because her father felt it was too dangerous for a girl; she said, “I hated it. I really did. I didn’t like it at all. I think I just learned how to live with it and I was just like yeah whatever. Once I stopped fighting him for it he was like, ‘okay you can go out.’” She did not link this gender unequal treatment to her description of her relationship with her father, and unlike Eva, she did not appear conflicted, she reported a loving and supportive relationship with her father. Additionally, at least one participant within each type of father-daughter relationship described how their fathers spent more time with their brothers or gave their brothers more freedom.

Supporting qualitative data, quantitative data demonstrate adolescents’ reports of father-daughter relationship quality is skewed to the left (not shown) suggesting the majority of Add Health participants also report positive father-daughter relationship quality.

Father-daughter relationships and early sexual initiation

Qualitative data suggest positive father-daughter relationship quality is associated with reduced risk of early sexual initiation. Of the four participants who initiated sex early and had a father figure (prior to age 17, the mean age of the sample), only one had a good father-daughter relationship, and her father was deported when she was 11. Comparatively, of the 14 who delayed sex, nine had a good father-daughter relationship (one also had a hostile step-father). Three of the four participants who had early sex had a hostile father-daughter relationship while only three of the 14 participants who delayed sex had a hostile relationship. These frequencies are not intended to be interpreted as statistically significant, but they identify patterns that are supported by individual experiences. Perhaps the characteristics of good father-daughter relationships, sacrificing for the family, encouraging academic success, and providing emotional support help adolescents delay sex. For example, second generation and college graduate Carla, who had a good relationship with her father, said the fear of disappointing her dad enabled her to delay sex and pregnancy and finish college, “…maybe having that strict figure in my life and having someone to push me to do further than what I possibly think I can is a big thing.”

Quantitative data support these patterns among a national sample. Table 2 shows that the hazard of first sex for those in the highest 25% of father-daughter relationship quality is 62% lower than those without a resident father (p<0.001), and the hazard for those in the lowest 75% of father-daughter relationship quality is 40% lower than those without a resident father (p<0.05) (Model 1). Father absence increases the hazard of sexual initiation more so than both high and low quality father-daughter relationships. As father-daughter relationship quality increases, the hazard of sexual initiation decreases (HR 0.89, p=0.081) (Model 2).

Table 2:

Discrete time hazard models for sexual initiation and father-daughter relationship quality. Mexican-origin young women Add Health (1994–2008)

	Model 1	Model 2	Model 3	Model 4
Second generation	1.33	1.10	1.25	1.12
	(0.87 – 2.03)	(0.77 – 1.58)	(0.84 – 1.86)	(0.77 – 1.63)
First generation	Ref.	Ref.	Ref.	Ref.
Father-daughter relationship quality (categorical)
No resident father	Ref.
Lowest 75% relationship quality	0.60*
	(0.38 – 0.94)
Highest 25% relationship quality	0.38***
	(0.22 – 0.65)
Father’s education
Less than high school education		Ref.	Ref.	Ref.
High school education		1.07	1.10	0.82
		(0.54 – 2.09)	(0.55 – 2.21)	(0.36 – 1.88)
More than high school education		1.79*	1.73*	1.69†
		(1.02 – 3.14)	(1.00 – 2.99)	(0.92 – 3.13)
Father-daughter relationship quality		0.89†	0.73***	0.93
		(0.78 – 1.01)	(0.64 – 0.83)	(0.81 – 1.06)
Immigrant generation X relationship quality			1.32**
			(1.09 – 1.60)
High school educated father X relationship quality				0.72
				(0.39 – 1.33)
Higher educated father X relationship quality				0.81
				(0.46 – 1.43)
Has a resident mother	0.56**	0.51†	0.48*	0.51†
	(0.37 – 0.84)	(0.24 – 1.12)	(0.23 – 1.00)	(0.24 – 1.07)
Age	108.19***	123.13***	121.51***	123.15***
	(20.75 – 564.12)	(17.81 – 851.43)	(17.78 – 830.52)	(17.86 – 849.40)
Age X Age	0.87***	0.87***	0.87***	0.87***
	(0.83 – 0.92)	(0.82 – 0.92)	(0.82 – 0.92)	(0.82 – 0.92)
Constant	0.00***	0.00***	0.00***	0.00***
	(0.00 – 0.00)	(0.00 – 0.00)	(0.00 – 0.00)	(0.00 – 0.00)
Person-years	3,623	3,007	3,007	3,007

^***p<0.001

^**p<0.01

^*p<0.05

^†p<0.10

^αFactor variable. The greater the value, the higher the reported father-daughter relationship quality.

Socio-political context of migration

Immigrant generation

Among qualitative participants there are no substantial differences in father-daughter relationship quality by immigrant generation. However, second generation participants were more descriptive about their relationships with fathers and described more open communication than first generation participants. Among Add Health participants in the quantitative sample, there is no significant bivariate difference in father-daughter relationship quality and immigrant generation with the continuous father-daughter relationship quality variable (Table 1). However, the categorical variable of father-daughter relationship quality does show differences by immigrant generation. In contrast to the qualitative data, 15% of first generation and 27% of second generation adolescents are in the highest 25% of reported relationship quality with fathers, suggesting second generation adolescents report higher relationship quality than first generation adolescents (p<0.10). Moreover, Table 2, Model 3 show higher father-daughter relationship quality is more protective for the first generation (HR 0.73, p<0.001) than the second (HR 0.97, p<0.05 [1.32*0.73=0.96]).

Father-daughter separation

In addition to participants interpreting their fathers’ love and support through migration “sacrifices” as described above, qualitative data elucidate that the socio-political context of migration shapes also father-daughter relationships and adolescent sexual health through both father-daughter separation and economic disadvantage.

Second generation Mariela, who married at 15 years old and had her first child at 17 said that her biological father, “was an awesome dad. I don’t have any bad memories growing up. He was just the typical dad. He would try to give us as much as he could.” However, her father was deported when she was 11 years old. Her mother remarried to economically support her four children and her stepfather was controlling and jealous. The loss of her supportive biological father and addition of a controlling stepfather during adolescence were directly related to her sexual health decisions. Mariela said, “It was kind of hard, my mom and my stepdad were kind of strict with me. I think that’s one of the reasons is that I kind of got married early was not just because, obviously because I love my husband, but I also wanted to get out of the house.” Mariela initiated sex at age 14, left the house to marry her husband at 15, and had her first child at 17.

Comparatively, a supportive stepfather during adolescence may buffer a hostile relationship with a biological father. First generation Yesinia reported her biological father was violent and ran the house as he did in the Mexican military. However, her stepfather, whom her mother married in the U.S. when Yesinia was 10, was supportive and loving. Although there are other factors related to sexual initiation, a good relationship with her step-father may have buffered some of the hostile influence of her biological father because she delayed sex until age 18. Finally, all ten participants who were separated from their fathers for an extended period of time were separated across a border. Of these ten participants, only Eva maintained a close, but conflicted, relationship with her father by talking on the phone and visiting him in Mexico a few times a year.

Economic disadvantage

Participants with hostile father-daughter relationships more often had fathers who worked in unskilled labor, such as cleaning office buildings. Of the six fathers of young women with hostile father-daughter relationships, four worked in unskilled labor compared to three of the ten fathers of young women with good father-daughter relationships. Supporting this pattern, in families with hostile father-daughter relationships, fathers often worked in unskilled migrant labor and were more frequently described by participants as being very poor and struggling for basic necessities, like food, than families with other types of father-daughter relationships. Indeed, second generation Carla thinks growing up in a single parent Latino family home increases the risk for adolescent pregnancy. Similarly, first generation Isabel believes a reason her younger sister had an adolescent pregnancy was because their father passed away when her younger sister was only ten. When asked about solutions to reducing adolescent pregnancy, she said,

…more than anything else, I think it starts with the family. And for those girls who don’t have that, for those, you know the whole deporting people, usually the males get deported before the females, the dads get deported. So what’s going to happen with these girls? Is that going to make the stats [adolescent births] triple?

Both Carla and Isabel highlight the intersection between the socio-political context of migration, father-daughter relationship quality, and adolescent sexual health behavior.

Looking at economic disadvantage quantitatively, Table 2, Model 4 shows the effect of father-daughter relationship quality was null when fathers were less educated, HR=0.93 (0.81 – 1.06), and was not significantly different when fathers were more educated. Thus, relationship quality did not reduce the risk of early sexual initiation in any level of education. However, from Models 2 and 3, father-daughter relationship quality was protective for early sexual initiation given father’s education and other covariates.

Discussion

First and second generation Mexican-origin adolescents make up a large proportion of the adolescent population in the U.S. (US Census Bureau 2012) and they face both social and economic disadvantage and sexual health disparities (CDC 2014; Martin et al. 2015). This innovative mixed methods study advances the literature by considering migration as a social determinant of father-daughter relationship quality and early sexual initiation among Mexican-origin young women in immigrant families.

Father-daughter relationships

The quality of father-daughter relationships in Mexican-origin immigrant families has been understudied and underestimated. Extending Crockett and colleagues’ (2007) who concluded that the 19 Mexican-origin youth who participated in focus groups, had less open communication with fathers than mothers, that fathers were rarely affectionate, and that daughters often desired more emotional support, our quantitative and qualitative data suggest good father-daughter relationships are far from rare. Among qualitative participants, it was the most common relationship type and moving beyond simple survey questions, multiple participants described themselves as “daddy’s little girl.” Attributes of a good father-daughter relationship include fathers sacrificing for the family by working hard long hours, encouraging and being engaged in their daughter’s education, and providing love and affection. Although participants’ good relationships include varying levels of frustration and disappointment, these findings dismantle common stereotypes that Mexican-origin fathers are rarely close and affectionate and are often sexist or possessive of daughters. Supporting our findings, Updegraff et al. (2009) conducted interviews with multiple family members within 246 Mexican-origin families to examine socialization processes and found that fathers of Mexican-origin adolescents were involved, supportive, and knowledgeable about adolescent daily activities. Similarly, González-López’s (2004) ethnographic work with Mexican immigrants found that fathers from urban areas in Mexico sought to protect their daughters from early sex and pregnancy so they could mature emotionally and attend college. In fact, they cared more about protecting their daughters from harm and increasing chances of economic success than virginity in and of itself. These results do not imply that young women have no frustration or conflict with their immigrant fathers, nor do they imply the participants did not desire more emotional support from their fathers. However, contrary to findings that focus on emotionally distant or sexist Mexican fathers, these daughters highlight positive aspects of their relationships with fathers.

Father-daughter relationships and sexual initiation

Supporting segmented assimilation theory’s assumption that parental social capital reduces risk behaviors among second generation adolescents, both qualitative and quantitative data suggest that positive father-daughter relationship quality is associated with lower risk of early sexual initiation. This aligns with work that shows high quality father involvement benefits the health and development of children (Amato and Gilbreth 1999; Carlson 2006; Lamb 2000). However, our results differ from Killoren and Deutsch (2013) who found no association between paternal support and adolescent sexual risk among Latino adolescents in a national sample. They used a single survey question to measure paternal support compared with our factor relationship quality measure and qualitative data. They also used a composite sexual risk variable and did not disaggregate by country of origin. More research specific to Mexican-origin immigrant fathers and the sexual health outcomes of their daughters is needed.

Quantitative results indicate that father-daughter relationship quality is more protective for first generation adolescents than second generation adolescents, although they had similar father-daughter relationship quality scores. The qualitative data and prior research suggest that first generation adolescents are more disadvantaged than second generation adolescents; they are poorer, experience more family instability, may not have U.S. citizenship, and have more difficulty with the English language (Vasquez 2011). Perhaps having a high quality father-daughter relationship is more protective for first generation adolescents because of their social and economic disadvantage.

The socio-political context of migration

Advancing prior literature that focuses exclusively on cultural parenting practices of immigrant families, we argue that the socio-political constraints of migration influence father-daughter relationships. Similar to Crockett and colleagues (2007), participants considered “sacrificing” for family to be an attribute of good fathers. Expanding Crockett and colleagues’ (2007) finding that fathers are important providers of economic and indirect emotional support, we find that attributes of a positive father-daughter relationship quality are complex. Specifically, a positive relationship with fathers might mean a father is “sacrificing” for the family through hard, low paying migrant labor, but also could mean fathers are making an effort to spend time with or money on children despite these barriers. Fathers may verbalize their sacrifice as a motivator for their daughters to succeed.

Similar to Dreby’s (2012) ethnographic work on children’s experiences in Mexican-origin households, our ten qualitative participants who were separated from their fathers, were separated across a border, making it difficult to maintain a close relationship with a nonresident father. These participants may be at higher risk for sexual risk behaviors than adolescents in U.S.-born families who have a strong tie to a nonresident father (Booth, Scott, and King 2010). However, these participants also offer insights into experiences of daughters in families that encounter deportation. Our results elucidate an important consideration regarding adolescent sexual health and the adolescent’s age when she is first separated from a supportive father and her age when her mother remarries. Future research examining father-daughter relationship quality and timing of parental separation may help identify highly vulnerable periods for families.

Deportation has significant negative impacts on child and adolescent health and development beyond the consequences of parental separation experienced by native U.S. families (Chaudry et al. 2010; Dreby 2012), as Mariela’s experience with her father’s deportation illustrates. Only two participants described the impact of having a parent deported, likely because many of the participants’ parents eventually received legal residence through the Immigration Reform and Control Act of 1986. However, deportations have more than tripled since 2002 and now exponentially more adolescents are being separated from their fathers (Golash-Boza and Hondagneu-Sotelo 2013), not to mention the detrimental effects for both parents and children of living in fear of deportation (Enriquez 2015). Our results that young women in immigrant families who are separated from their fathers may be at increased risk of early sex, supports immigration reform that keeps families together.

The quantitative results do not identify a significant interaction between father-daughter relationship quality and father’s education. Father-daughter relationship quality does not vary in its ability to reduce the risk of early sexual initiation at any level of father’s education. However, when considering father’s education and other covariates, father-daughter relationship quality is protective for early sexual initiation. Moreover, among qualitative participants, hostile father-daughter relationships more often occurred in relatively poorer families where fathers worked in unskilled labor. Additionally, two young women who described emotionally distant relationships with fathers felt their fathers spent more time economically providing for the family than spending time with them. This type of “survival parenting” (Vasquez 2011) may be more profound among immigrant families who lack access to economic resources and work in exploitive migrant labor (Fussell 2011). Moreover, Enriquez’s (2015) qualitative interviews with undocumented parents highlights that immigrant fathers, not mothers, were likely to feel like bad parents because the harsh socio-political context renders a father’s gender role of economic provider precarious. This further reinforces our argument that father-daughter relationships among immigrant families must be analyzed within the context of migration. Improving the quality of father involvement among nonresident fathers may reduce risky sexual behaviors among adolescents, but preventing father absence across borders by supporting immigration reform that keeps families together and paying a living wage that reduces economic strain on families and allows parents to spend more time with their children may have a much larger impact on adolescent sexual health outcomes.

Finally, the influence of gender was threaded throughout our qualitative findings. Similar to prior research (Ayala 2006; Foner 2009; Wilson, Dalberth, and Koo 2010), participants’ described differential treatment between sons and daughters. Extending prior research, qualitative data reveal an association between hostile father-daughter relationships, including violence by fathers, and early sexual initiation. This may support Killoren and Deutsch’s (2013) finding that having a strict father increases sexual risk behavior and Updegraff and colleagues’ (2009) finding that father-daughter conflict is associated with risky behavior among Mexican adolescents. Relationships with conflict or strict fathers may have similar attributes to hostile relationships. Finally, families of qualitative participants who described more hostile father-daughter tended to have lower SES and greater household gender inequality. The association between gender inequality, lower SES, and early sexual initiation has been documented previously (Coleman-Minahan 2017), but it is important to reinforce that gender inequality is not rooted in cultural processes alone, but is highly dependent upon social and economic disadvantage often perpetuated by immigration policy (González-López 2004; Gutmann 1996; Hirsch 2003).

Limitations

There are several limitations to this study. We made inferences between a nationally representative sample and a local purposive sample. However, both samples were from the same birth cohort and the purpose of the study was not to generalize between samples but to build on method strengths. Recall bias is present in life history interviews and with surveys, although the longitudinal design of the Add Health decreases recall bias. While we did not have the sample size in the qualitative or quantitative data to consider age at migration, the age at which girls migrate to the US may affect their relationships with fathers and risk of early sexual initiation. Interviewing young adult women provides information on their outcomes and the influence of the father-daughter relationship on their decisions as adolescents. Additionally, Add Health is the only longitudinal dataset that allows us to examine the second generation. There are other family, peer, and partner dynamics influencing sexual health that we did not examine. Finally, we did not interview fathers, although daughters’ perspectives are first step in examining father-daughter relationships.

Despite these limitations, this innovative mixed method study examines migration as a social determinant of father-daughter relationships and early sexual initiation among Mexican-origin young women. Qualitative data allow us to describe the lived experiences of these young women that are not captured with survey data alone. Moreover, quantitative data allow for comparisons of direct associations and interactions, such as how father-daughter relationship quality impacts early sexual initiation differently across immigrant generations, which is beyond the scope of qualitative data. Finally, this study contributes to the dearth of research on father-daughter relationships among Mexican-origin immigrant families in the U.S. by suggesting good father-daughter relationships are common and may reduce early sexual initiation among adolescent daughters.