AIDS and Behavior

, Volume 16, Issue 6, pp 1570–1583

Early Adolescent Romantic Relationships and Maternal Approval among Inner City Latino Families

  • Alida Bouris
  • Vincent Guilamo-Ramos
  • James Jaccard
  • Michelle Ballan
  • Catherine A. Lesesne
  • Bernardo Gonzalez
Original Paper

DOI: 10.1007/s10461-011-0034-8

Cite this article as:
Bouris, A., Guilamo-Ramos, V., Jaccard, J. et al. AIDS Behav (2012) 16: 1570. doi:10.1007/s10461-011-0034-8


Latino youth are at high risk for acquiring HIV during adolescence. The present study documented the nature of adolescent romantic relationships among 702 Latino eighth grade students and their mothers in the Bronx, NY. The study examined adolescent romantic relationships, the association between participation in such relationships and intentions to engage in sexual risk behavior, and maternal influences on adolescent’s involvement in intimate behaviors in romantic relationships. Almost 50% of youth had been in a romantic relationship, which typically lasted 3–3.5 months. Mothers tended to approve of intimate behaviors and sexual activity in romantic relationships more so for males than females. Latino youth tended to underestimate maternal disapproval of a range of intimate behaviors, and the correlations between perceived and actual maternal approval were generally low in magnitude. Finally, maternal orientations towards their adolescent engaging in romantic relationships were associated with their child’s intentions to have sexual intercourse in the future.


Romantic relationships Early adolescence Latinos Maternal approval 


Los jóvenes latinos están en alto riesgo de contraer el VIH durante la adolescencia. El presente estudio ha documentado la naturaleza de las relaciones románticas entre 702 adolescentes latinos del octavo grado y sus madres en el Bronx, NY. El estudio examinó las relaciones románticas entre adolescentes, la asociación entre la participación en estos tipos de relaciones e intenciones de participar en el comportamiento de riesgo sexual, y las influencias de los madres en la participación de los adolescentes en los comportamientos íntimos de las relaciones románticas. Casi 50% de los jóvenes habían estado en una relación romántica, que normalmente duró de 3 a 3.5 meses. Las madres tendían a aprobar los comportamientos íntimos y la actividad sexual en las relaciones románticas más para los jóvenes masculinos que para las femeninas. Los jóvenes latinos tendían a subestimar la desaprobación materna de una serie de comportamientos íntimos, y las correlaciones entre la aprobación materna percibida y la real eran generalmente de baja magnitud. Por último, las orientaciones de los madres hacia la participación de las relaciones románticas entre los adolescentes se asociaron con las intenciones de sus hijos tener penetración sexual en el futuro.


Sexual risk behavior among Latino adolescents remains a serious public health concern. According to the 2009 Youth Risk Behavior Survey, approximately 49% of Latino high school students reported having ever had vaginal sexual intercourse [1]. Among Latino youth who had transitioned to sexual activity, 6.7% reported sexual debut before age 13 and 14.2% reported having four or more sexual partners in their lifetime [1]. In comparison, 3.4% of White youth reported first sex before age 13 and 10.5% reported having four or more sexual partners [1]. Compared with their non-Latino peers, Latino adolescents are less likely to report contraceptive use at first sex [2] and condom use at their most recent act of vaginal sexual intercourse [1]. Epidemiological surveillance data indicate that Latino adolescents are at high risk for experiencing the potentially negative health consequences of sexual risk behavior, including HIV, sexually transmitted infections (STIs), and unintended pregnancies [3, 4, 5, 6]. In 2006, the rate of HIV diagnoses among Latino adolescents aged 15–19 years old was 7.2 per 100,000 compared to 1.6 per 100,000 for White youth [4]. Similarly, Latino youth have higher rates of STIs, including chlamydia, gonorrhea, and syphilis, than their non-Latino, White peers [4]. Finally, despite an overall decline in the teen pregnancy rate, Latina adolescents continue to have the highest teen birth rate in the United States (U.S.) and significant increases in southeastern states have been observed, with some rates nearly double the national average of 81.7 births per 1,000 Latino adolescent girls [5, 6]. Taken together, these findings highlight the urgent need for additional research on Latino adolescent sexual behavior.

The Importance of Studying Adolescent Romantic Relationships

Most theorizing on adolescent sexual risk behavior has focused on social-cognition models that emphasize the beliefs, attitudes, and perceptions of sexual risk behavior on the part of individuals. However, adolescent sexual risk behaviors are dyadic in nature, suggesting that research is needed to help us understand the dyadic contexts in which sexual risk behavior occurs, namely romantic relationships. Romantic relationships are a normative aspect of adolescent development [7] and research indicates that adolescents strongly value these relationships. For example, a study of sixth through eighth graders in the U.S. observed high levels of interest in having a boyfriend or girlfriend across all grade levels [8]. Not surprisingly, romantic relationships become increasingly common as adolescents get older. Data from the National Longitudinal Study of Adolescent Health indicates that over 40% of 13 year olds, over 60% of 16 year olds, and over 70% of 18 year olds have been in a romantic relationship in the previous 18 months [9]. Among youth aged 13–15 year old, 42% have been on a date and 50% have been in a romantic relationship in the previous 18 months [10].

In addition to their increasing prevalence, experiences in romantic relationships are important because they can have long-term consequences for adolescent’s emotional health [11], social and academic development [12], and self-esteem [13, 14]. Numerous scientists have argued that the seeds of healthy adult relationships are sewn in adolescent romantic relationships [15, 16, 17, 18]. Furthermore, romantic relationships are of key interest to prevention scientists because of their strong association with sexual activity, contraceptive use, teen pregnancy, STIs and HIV [19, 20, 21, 22]. Nationwide, the majority of sexually active adolescents report that sexual debut occurred in the context of a romantic relationship [2, 19], with sexual intercourse occurring early in a relationship. For instance, among youth reporting that first sex occurred in a romantic relationship, 24% reported intercourse in month 1, 38% within 1–3 months, and 38% within four or more months of the relationship starting [19].

A number of studies have suggested that adolescents may view romantic relationships as conferring special social status and as legitimizing sexual intercourse. Feldman et al. [23] found that initiation of new sexual activities was considered more appropriate when it occurred within a serious romantic relationship. In a national survey of 12–19 year olds, Albert [24] found that approximately two-thirds of youth strongly agreed that sex should occur only within such a relationship. In a separate study with eighth grade youth, the two most common reasons cited for wanting to be in a romantic relationship were experiencing personal growth such as feeling close to, connected to, and loved by a partner, and for attaining increased popularity and social attention [25]. Similarly, Rodgers [26] found that the modal reason for first intercourse among youth was “to have the partner love them more.”

Parental Influences on Adolescent Romantic Relationships and Sexual Behavior

Although a body of knowledge is emerging about the importance of adolescent relationships, we know relatively little about the nature of romantic relationships in young, inner city Latino adolescents. In addition, even less is known about how the mothers of Latino adolescents view romantic relationships and what they approve or disapprove of with respect to intimate behaviors in the context of these relationships. To date, most research on parental influences has focused on White adolescents and has examined how the quality of parent’s romantic relationships [27, 28] or parent–child attachments [29, 30] influence the quality of adolescent’s romantic relationships. Across these studies, perceptions of warmth, reciprocity, and intimacy in parent’s romantic relationships and within the parent–child relationship have been positively associated with satisfaction, relatedness, and feelings of love and intimacy in youth’s romantic relationships in both adolescence and early adulthood [31, 32, 33].

In contrast, less research has explored how parenting practices specific to adolescent’s romantic relationships may influence adolescent sexual behavior. The dearth of empirical work in this area is surprising given the large body of research indicating that parents can invoke a number of specific practices that are associated with delayed coital debut and decreased sexual risk taking [34, 35, 36]. Parenting practices identified as having a robust relationship with reduced sexual risk behavior on the part of youth include parent–adolescent communication about sex [37], parental attitudes and values that may discourage sexual behavior [38], and parental monitoring, control and supervision [36, 39]. These practices are thought to shape adolescent’s sexual behavior by socializing youth to parent’s values and expectations, communicating to and teaching youth the knowledge and skills necessary to make lower-risk decisions, limiting perceived and real opportunities through which adolescents might be able to engage in risky sexual behaviors, and by enacting appropriate disciplinary strategies when adolescents transgress parental rules [40, 41].

Within the small body of research examining parenting practices specific to adolescent romantic relationships, similar types of parental socializing, monitoring and communication practices have been identified, including rules about teen dating [33], beliefs about ideal partner characteristics [42], concerns about the potential negative consequences of dating [43], monitoring, supervising, and limiting of dating [40, 42, 43] and encouragement and support of romantic relationships [26, 42, 43]. To date, the majority of research in this area has explored parent and adolescent variations in dating-specific practices and the association between these practices and affective dimensions of adolescent’s romantic relationships [43]. Consistent with previous research on parent–adolescent communication about sex, parental disapproval of sex, and parental monitoring [37, 38], parent and youth reports of dating-specific practices tend to be discrepant, with parents reporting higher levels of involvement in adolescent dating than their children [42]. A number of studies have also found differences in dating-specific practices as a function of adolescent gender [33, 43]. In a sample of majority White adolescents, Kan et al. [43] found that parents enacted stronger dating restrictions (e.g., limiting time with romantic partners) for daughters than for sons and provided greater autonomy in dating (e.g., telling teens that choosing a boy/girlfriend is their choice and not the parent’s) for sons than for daughters. In one of the few studies to explore dating practices among diverse youth, Latino daughters reported stronger expectations that mothers would be consulted to help solve dating-related problems than did sons, an association that was not observed for African American or White youth [42]. In addition, Latino mother–daughter dyads had similar expectations of maternal involvement in dating [42], possibly reflecting greater levels of communication about dating and romantic relationships in mother–daughter dyads.

In one of the few studies examining the relationship between dating-specific practices and adolescent sexual behavior, Longmore et al. [40] found that dating-specific practices were associated with adolescent sexual behavior above and beyond a number of important adolescent- and parent-level factors. Specifically, parental attitudes that youth should delay sexual debut until after age 18 were negatively associated with coital debut 1 year later while the frequency of parent–teen disagreements about dating (e.g., how often parents and teens argued about dating) was positively associated with the onset of sex [40]. Interestingly, youth reports of parental limits on autonomous decision-making in dating, parent–adolescent communication about sex, and parental monitoring were not correlated with sexual debut [40].

Taken together, the extant research suggests that the parent–child relationship is an important socialization context and that dating-specific practices may influence youth’s involvement and sexual behavior in romantic relationships [33, 40, 42, 43]. To date, we know of no studies that have investigated how dating-specific practices are associated with orientations to have sexual intercourse among inner city Latino youth. The lack of research on young Latino adolescents is unfortunate. Although romantic relationships and sexual behavior are normative aspects of adolescent development, epidemiological data indicate that Latino youth are disproportionately affected by HIV, STIs and unplanned pregnancies relative to their White peers [2, 4, 6]. Thus, having a better understanding of how Latino mothers can utilize dating-specific parenting practices to support reduced sexual risk taking has important implications for strengthening parent-based interventions to prevent unplanned pregnancies, STIs and HIV among Latino youth.

Study Goals and Hypotheses

The present study sought to explore models of maternal influences on early adolescent’s involvement in romantic relationships and orientations toward sexual behavior in a large sample of inner city Latino eighth graders who vary by gender and Latino ethnicity (e.g., Dominican, Mexican, and Puerto Rican). First, we document the nature of romantic relationships by describing (1) the percent of youth involved in romantic relationships, (2) how long these relationships tend to last, and (3) the reported intensity of feelings in those relationships. Second, we explore the association between being in a romantic relationship and adolescents’ intentions to engage in sexual intercourse in the near future. We hypothesized that youth in romantic relationships would report stronger behavioral intentions to engage in future sexual intercourse than youth who were not so involved [18, 19]. We then describe the extent to which Latino mothers were perceived by their adolescents as approving or disapproving of them having a boyfriend/girlfriend and the extent to which Latino youth accurately perceived maternal orientations about romantic relationships. Next, we document the extent to which Latino adolescents perceived their mothers as approving or disapproving of a variety of specific intimate behaviors in the context of romantic relationships. Based on previous research with older adolescents, we hypothesized that young Latino adolescents would underestimate maternal orientations that oppose intimate behaviors in romantic relationships and that mothers would report higher levels of disapproval of intimate behaviors for daughters than for sons, a double standard suggested by research with non-Latino populations [33, 42, 43]. Finally, we examine if maternal approval–disapproval of relationship-related behaviors that are intimate but less explicitly sexual (e.g., holding hands, kissing in private) are associated with adolescent’s intentions to engage in sexual intercourse over and above maternal approval–disapproval of relationship-related behaviors that are more sexual in nature.

The present study adds to the literature in several ways. First, to the best of our knowledge, this is one of the first studies to explore how dating-specific practices influence orientations to engage in sexual behavior in a diverse sample of early Latino adolescents. To date, we know of no research that has focused explicitly on Latino families or has examined potential ethnic subgroup differences in early adolescent’s romantic relationships and perceptions of maternal approval of dating and intimate behaviors. The few studies on dating-specific parenting have focused on older [40] or on White adolescents [33, 43], with limited attention on maternal orientations toward intimate behaviors surrounding sexual activity in romantic relationships during early adolescence. In doing so, we also consider the perspectives of both boys and girls to explore the extent to which gender differences in romantic relationships and orientations toward sexual activity may be present in early adolescence.


Respondents and Procedures

Respondents were drawn from a sample of 702 mother–adolescent dyads that were recruited from the eighth grade at ten middle schools in the South Bronx community of New York City. The sample was comprised of Dominican (n = 265), Puerto Rican (n = 232) and Mexican (n = 205) mother–adolescent dyads. As New York City’s poorest borough, the Bronx has a median household income well below the state average [44]. Additionally, youth living in the Bronx experience disproportionately high rates of HIV, STIs, and pregnancy [45].

A random sample of eighth grade students was generated from participating schools. Invitations to participate in the study were carried out by bilingual recruiters who called parents or legal guardians to invite the resident mother and adolescent to attend a data collection event at the adolescent’s school. In total, 83% of all mother–adolescent dyads who were contacted were recruited into the study. Among this group of recruited dyads, 89% attended a school-based data collection event. Numerous scheduling attempts were made for the school-based events. In-home interviews were conducted with 11% of families, which represented those families who were unable to attend a school event despite multiple attempts. During the recruitment process, a refusal bias survey was administered to obtain key demographic variables. No refusal bias was observed between families who refused and families who agreed to participate in the study.

Mothers completed consent and permission forms and adolescents completed assent forms. Consent forms were provided in both Spanish and English and were reviewed with mothers and adolescents by project staff. Each dyad was compensated $25.00 for participating in the study. Institutional Review Board approval was obtained.

We interviewed the mothers of adolescents, as opposed to fathers, for several reasons. First, numerous studies examining parent–adolescent communication about sex have indicated that mothers are more likely to be the primary source of information and are more influential on adolescent sexual activity than are fathers [46, 47]. Second, as suggested by Costigan and Cox [48], there may be non-trivial selection effects among fathers who agree to participate in research studies. Finally, as indicated by our previous research with the target sample, the majority of homes were single female head of households. Thus, interviewing mothers only was determined to be the best option, given that statistical modeling is complicated by the presence of non-random missing data, in this case due to single mother households.

Data collection was carried out through self-administered questionnaires. To ensure truthful responding, project staff informed adolescents that their answers were confidential and explained to both mothers and adolescents that names could not be linked to a survey. Adolescents were notified to skip any question they felt uncomfortable answering. Adolescent males and females received the same survey, with surveys reflecting gender appropriate pronouns. Before participants started the survey, they were asked to identify their language preference (e.g., English or Spanish). The Spanish version of the survey was translated into Spanish using methods described in Marín and Van Oss Marín [49], and was found by translators to be linguistically equivalent to the English version. The measures presented here reflect the female version of the instrument.


Latino Ethnicity

Adolescents and their mothers indicated their ethnicity using multiple questions. First, they were asked if they were of Hispanic or Latino origin. If they answered yes, they were asked their Latino background. Only adolescents responding affirmatively to the first item and only those self-identifying as Puerto Rican, Mexican or Dominican are included in the analysis.

Sexual Behavior and Intentions to Engage in Sexual Intercourse

Adolescents were provided with a developmentally appropriate definition of vaginal sexual intercourse and then were asked if they had ever engaged in it. Responses were scored dichotomously to indicate if the adolescent had ever engaged in vaginal sexual intercourse. Because there are base rate issues with behavioral measures of sexual intercourse among middle school youth, behavioral intentions were used as proxy indicators for future behavior. In addition, research suggests that behavioral intentions are reliable predictors of future behavior [50, 51, 52, 53]. Four items, scored on a scale from 1 to 5, assessed adolescent intentions to engage in sexual intercourse: (1) “If I had the opportunity and it was with a boy I liked a lot, I would engage in sexual intercourse;” (2) “I think I am ready to have sexual intercourse;” (3) “I would have sexual intercourse now if I had a boy who would do it with me;” and (4) “I plan on having sexual intercourse in the next 6 months.” The alpha for the scale was 0.86 and a single composite item reflecting adolescent’s overall intentions toward sexual intercourse was created.

Maternal Disapproval of Romantic Relationships

Adolescents were asked to respond to a series of items assessing maternal disapproval of having a boyfriend and a variety of intimate behaviors in the context of relationships. Adolescents were instructed that it did not matter if they had done any of the activities with a boy or if they had a boyfriend, but to respond how much they thought their mother would approve or disapprove of them doing these activities. The activities were: (1) holding hands with a boyfriend; (2) telling a boyfriend you love him; (3) a boyfriend and her calling themselves a couple; (4) kissing a boyfriend in private; (5) a boyfriend and her touching each other under their clothing; (6) a boyfriend and her touching each other with no clothes on; and (7) having sexual intercourse with a boyfriend. The items were scored on a scale from 1 to 5 and coded so that higher scores reflected higher levels of perceived maternal approval. Mothers also indicated the extent to which they approved or disapproved of these activities using the same scale format.

Adolescent Relationship Status and the Nature of the Relationship

Adolescents were asked to indicate if they had ever been involved in a romantic relationship, if they had been in a romantic relationship in the past 6 months, if they were currently in a romantic relationship, and the total number of romantic relationships they had been in. They were also asked to indicate, in months, how long their most recent relationship lasted.

Adolescents were asked to think of their current or most recent romantic relationship if they were not currently in one, and to indicate the age of their partner, how strong their feelings were for their partner (using a five point scale ranging from 1 = not at all, 2 = a little, 3 = somewhat, 4 = moderately, 5 = very strong), whether their feelings were stronger than their partner’s feelings, whether they thought the partner was “the one,” if they had thought about having a family with the partner, if the partner had pressured them to have sexual intercourse, and if they thought that having sexual intercourse would make them closer to the partner, all answered on a “yes” or “no” metric.


Preliminary Analyses

Missing data were minimal, amounting to fewer than 2% of the cases on a given variable. Missing values were imputed using the expectation–maximization method with importance resampling [54]. All analyses were preceded by checks for outliers using graphical methods, standardized dfbetas, and the analysis of residuals. No outliers were evident. In general, there was non-normality in several variables that could interfere with the analyses based on traditional regression and structural equation modeling (SEM). For this reason, all regressions and SEM were conducted using bootstrapping with 2,000 bootstrap replicates coupled with bias-corrected P values as implemented in M Plus. Descriptive statistics for the sample are presented in Table 1. The intraclass correlation for key outcomes as a function of schools was consistently negligible (less than 0.01), so no adjustments for clustering as a function of schools were introduced. Analyses that took into account clustering yielded comparable conclusions.
Table 1

Descriptive statistics for sample (N = 702)



Mean age of adolescent


Maternal mean age


Percent male


Adolescent Latino ethnicity

 Percent Puerto Rican


 Percent Dominican


 Percent Mexican


Language adolescent speaks at home

 Percent mostly Spanish


 Percent equal Spanish and English


 Percent mostly English


Percent adolescents born on US mainland


Percent adolescents who are Catholic


Language mother speaks at home

 Percent mostly Spanish


 Percent equal Spanish and English


 Percent mostly English


Percent mothers born on US mainland


Percent mothers who are Catholic


Characteristics of Urban, Latino Early Adolescent Romantic Relationships

Table 2 presents descriptive information about the nature of Latino adolescent romantic relationships. In general, there was little variability in the parameter estimates as a function of Latino ethnicity (only one result as a function of ethnicity was statistically significant, and it could be attributed to chance given the number of contrasts), so results are presented as a function of gender, for which there were non-trivial differences. Just over 60% of the youth (who are about 13 year olds) reported having already been involved in a romantic relationship at some time in their life, with Latino males reporting rates that were about 10–15 percentage points higher than Latino females. About 25%, or one in four of the youth, were currently involved in a romantic relationship. In general, romantic relationships were short lived, lasting an average of just 3–3.5 months. Latino males reported experiencing stronger feelings for their partners than Latino females, but both genders tended to characterize their feelings as being “moderately strong.” Latino females were more likely to report having an older partner (about half had an older partner). This is important because prior research has shown that having an older partner substantially increases the risk of unintended pregnancies [10]. Almost 20% of the youth (1 in 5), males and females alike, reported thinking about having a family with their romantic partner. In addition, Latino males were more likely than Latino females to state that having sex would make them feel closer to their partner. Overall, 11% of the youth had already engaged in sexual intercourse, with Latino males being more likely to have had sex than females (see Table 2).
Table 2

Descriptive statistics about Latino relationships




Critical value

Ever had boyfriend/girlfriend




Had boyfriend/girlfriend in past 6 months




Currently have a boyfriend/girlfriend




How many relationships have had




Months relationships have lasted




Your feelings stronger than partner




How strong are your feelings




Was your boyfriend/girlfriend older than you




Age of partner in last relationshipa




Did you think partner was “the one”




Did you think about having a family with




Did partner pressure you to have sex




Did you think sex would make you closer




Percent have had sexual intercoursea




Critical value is a z-value when contrasting percents and a t-value when contrasting means, in all cases for the contrast between males and females

aContrast is statistically significant, P < 0.05

bContrast was significant without correction for familywise error but not significant when using the Holm modified Bonferroni correction

Romantic Relationships and Intentions to Engage in Sexual Intercourse

We next examined the relationship between intentions to engage in sexual intercourse and involvement in a romantic relationship (past or present) using a 3 × 2 × 2 factorial design where the dependent variable was the intention to engage in sexual intercourse at this time in one’s life, and the three factors were Latino ethnicity, gender, and whether or not the adolescent had ever had a boyfriend or girlfriend. All factors were between-subjects in nature. There were no statistically significant differences in intentions to engage in sexual intercourse as a function of Latino ethnicity, nor did Latino ethnicity interact with gender or relationship status. We therefore focus on gender and relationship differences, with relevant cell means appearing in Table 3. For descriptive purposes, rows 1 and 2 of Table 3 present the percentage of youth in the relevant cells whose overall mean intention on the intention scale was above the mid-point category, suggesting a positive intention to engage in sexual intercourse in the near future. Overall, about 24% of youth intended to have sexual intercourse in the near future.
Table 3

Adolescent relationships and the intention to have sexual intercourse




Had relationship

No relationship

Had relationship

No relationship

Mean intent to have sex





% plan to have sex





Mean items are scored on a scale from 1 to 5 where higher scores reflect higher levels of intention. See text for significance tests

Single degree of freedom contrasts revealed a direct association with relationship status such that those youth who had been in a romantic relationship were more likely to intend to engage in sexual intercourse in the near future than those who had not been in a romantic relationship (mean intention of 2.11 vs. 1.60, t(654) = 6.74, P < 0.05). In addition, Latino males were more likely to intend to have sexual intercourse at this time in their lives than Latino females (mean intention of 2.08 vs. 1.63, t(654) = 5.96, P < 0.05). The association between relationship status and intentions to engage in sexual intercourse varied as a function of gender, which was indicated by a statistically significant interaction contrast between gender and relationship status (t(654) = 2.14, P < 0.05). For Latino males, the association between being in a relationship and intentions to engage in sexual intercourse in the future was stronger than it was for Latino females (male means for had and not had relationship = 2.41 vs. 1.74; for females, the corresponding means were 1.80 vs. 1.46).

In sum, the above analyses show an association between being involved in romantic relationships and intentions to engage in sexual intercourse in the future among middle school Latino youth. This relationship holds across both genders and all Latino ethnicities, though it is somewhat stronger for boys than girls. The data underscore the central role that romantic relationships have for future sexual activity of young adolescents.

Maternal Approval/Disapproval of Specific Relationship Behaviors

We next examined adolescent characterizations of the extent to which their mothers would approve or disapprove of them performing each of seven intimate behaviors while in a relationship as well as the mothers’ self reports of their own degree of approval or disapproval. This was analyzed in a 3 × 2 × 2 × 7 factorial design where the first two factors (Latino ethnicity and gender) were between-subject in nature and the second two factors were within-subjects in nature (maternal report of maternal approval vs. adolescent perception of maternal approval and the seven behaviors). We refer to the third factor as “source” and the last factor as the “type of behavior.” Non-pooled error terms were used for single degree of freedom contrasts for the within-subject factors, to avoid problems with the assumption of sphericity.

The only statistically significant between-subject main effect was gender, with boys, on average, experiencing higher overall levels of approval for the behaviors than females (2.94 vs. 2.48, t(562) = 8.94, P < 0.05), which is consistent with a gender-based double standard that sanctions sexual-related behaviors more heavily for girls than boys. The relevant means for each behavior as a function of gender are reported in Table 4.
Table 4

Average maternal disapproval/approval of specific relationship behaviors collapsing across reporting source


Female (%)

Male (%)

Critical value

Holding hands with a boyfriend

3.45 (20)

4.00 (9)


Telling a boyfriend you love him

3.40 (22)

3.93 (11)


Calling themselves a couple

3.11 (33)

3.61 (18)


Kissing a boyfriend in private

3.07 (34)

3.52 (20)


Touching each other under their clothing

1.52 (83)

1.98 (65)


Touching each other with no clothes on

1.43 (84)

1.77 (70)


Having sexual intercourse

1.39 (86)

1.73 (72)


Critical value is a t-value from a single degree of freedom contrast for the gender difference in the larger factorial analysis of variance (see text). All contrasts are statistically significant for P < 0.05 with or without a Holm-modified Bonferroni control for familywise errors. Values in parentheses are percent of cases where the value was below the scale midpoint of 3, implying disapproval

There was a statistically significant overall main effect of source, with the mean perceived maternal approval of the behaviors being larger than actual maternal approval (2.78 vs. 2.63, t(562) = 3.06, P < 0.05). This suggests that adolescents, overall, see their mothers as being more approving of engaging in the behaviors than the mothers actually are.

Multiple single degree of freedom contrasts for the type of behavior were statistically significant, with the dominant and obvious trend being that the mean approval of the behaviors of holding hands, telling a partner you love him or her, calling yourself a couple and kissing in public were statistically significantly larger (P < 0.05, indicating more approval of these less intimate behaviors) than touching each other under one’s clothes, touching each other with no clothes on, and having sexual intercourse (see Table 4). This was true for all Latino ethnicities and both genders, although there were some nuanced between-group mean differences within a given behavior which we do not report here in the interest of space.

Accuracy of Adolescent Perceptions of Maternal Approval

We correlated the perceived and actual maternal approval ratings for each of the seven behaviors to document the extent to which adolescents accurately perceive the orientations of their mothers. The correlations for each behavior were: (1) holding hands with a boyfriend–girlfriend, r = 0.12, (t(695) = 3.19, P < 0.05); (2) telling a boyfriend–girlfriend you love him/her, r = 0.06, (t(695) = 1.59, ns); (3) a boyfriend–girlfriend calling themselves a couple, r = 0.06, (t(695) = 1.59, ns); (4) kissing a boyfriend–girlfriend in private, r = 0.06, (t(695) = 1.59, ns); (5) a boyfriend–girlfriend touching each other under their clothing, r = 0.10, (t(695) = 2.65, P < 0.05; (6) a boyfriend–girlfriend touching each other with no clothes on, r = 0.06, (t(695) = 1.59, ns); and (7) having sexual intercourse with a boyfriend–girlfriend, r = 0.13, (t(695) = 3.46, P < 0.05). The correlations were uniformly low and suggest little correspondence between what mothers actually approve or disapprove of and adolescent perceptions of those orientations.

Perceived Maternal Approval of Behaviors and Intentions to Engage in Sexual Intercourse

The seven specific relationship behaviors clustered into two groups, those behaviors that are intimate but less sexual in nature (e.g., holding hands, telling a partner you love him/her, calling themselves a couple, kissing a partner in private) and those behaviors that are intimate and sexual in nature (e.g., touching each other under clothing, touching each other with no clothes, having sexual intercourse). We tested the relationship between maternal approval of these behaviors and intentions to have sexual intercourse using the latent variable SEM structural model shown in Fig. 1. The solid arrows were hypothesized a priori and the dashed arrows were added to the model based on fit diagnostics, as described below. The model posits two latent variables for adolescent perceptions of maternal approval for engaging in the two clusters of behavior, as well as two corresponding latent variables focused on maternal actual approval (as measured from mothers). The maternal actual approval latent variables were assumed to impact the corresponding perceived maternal approval latent variables, and all four of these latent constructs were assumed to influence a latent endogenous variable reflecting the intention to engage in sexual intercourse in the future. The latter latent variable had four indicators representing the four measures of behavioral intent. The residual terms for the latent perceived maternal approval variables were assumed to be correlated to reflect factors other than the latent actual approval variables that influence latent perceived maternal approval. Although not shown in Fig. 1, gender and Latino ethnicity were included as covariates for all endogenous variables in the model. Missing data was handled using full information maximum likelihood.
Fig. 1

SEM for maternal approval. PA perceived approval, AA actual approval, path coefficients are standardized. For structural model, unstandardized path coefficients are in parenthesis. Gender and ethnicity covaried from all endogenous variables, *P < 0.05

The initial model yielded unsatisfactory global fit indices. Inspection of modification indices suggested the need to add a small cross loading for the behavior of “touching each other under clothing” with the latent variables for intimate behaviors that are not sexual (see dashed arrows in Fig. 1). This makes conceptual sense, because the behavior is intermediate between being non-sexual and sexual. In addition, modification indices suggested the need for three correlated errors between indicators, none of which were large, nor did their presence or absence alter in a substantive way the magnitude and significance of the more theoretically interesting and hypothesized path coefficients between the latent variables in Fig. 1. These correlated errors were included in the final model to protect against left-out-variable error [55].

Figure 1 presents the relevant coefficients for the final model, which included all parameters indicated by solid or dashed arrows in Fig. 1. Unstandardized coefficients are in parentheses, while standardized coefficients are not in parentheses. Curved double headed arrows reflect correlations. All residuals are standardized. Both the measurement model and the structural model were well behaved. The overall fit indices showed a satisfactory Comparative Fit Index (CFI = 0.97), a satisfactory standardized RMR (SRMR = 0.03), a satisfactory Root Mean Square Residual (RMSEA = 0.049, with 90% CI = 0.042–0.055), and a non-significant P value for close fit (P < 0.63).

Of primary interest in the analysis are the four path coefficients linking the latent variables reflecting perceived and actual maternal approval for the two clusters of romantic specific behaviors to the latent intention to engage in sexual intercourse in the near future. Only two of the four path coefficients were statistically significant and both focused on the more sexually charged romantic behavior cluster (see Fig. 1). The coefficients linking maternal approval of the intimate but less sexual behaviors to intentions to have sex were not statistically significant, which suggests that mothers approving of more innocuous behaviors, such as holding hands and kissing, is not associated appreciably with intentions to have sexual intercourse over and beyond the influence of approval of the more sexually charged behaviors.

We tested for differences in structural coefficients and factor loadings as a function of gender and Latino ethnicity using multi-group SEM strategies with robust estimators, but found no evidence of such group differences. We also tested for two-way interaction effects between the four latent predictors using latent variable interaction models described by Marsh et al. [56], but found no evidence for such interactions. Finally, we tested if structural coefficients and factor loadings differed as a function of whether an adolescent had ever been in a romantic relationship. We observed one statistically significant effect. Specifically, a statistically significant single degree of freedom (scaled) chi-square (χ2 = 6.6, P < 0.05) was observed for group differences in the path linking perceived maternal approval of the intimate but less sexual behaviors to intentions to have sex. The unstandardized path coefficient for adolescents who had been in a relationship (coefficient = 0.24, z = 3.99, P < 0.05) was significantly stronger than the corresponding path for those who had never been in a relationship (coefficient = 0.06, z = 1.11, ns). This should be interpreted with caution because of its post hoc character and the fact that it occurred in the context of multiple contrasts.


As Latino youth become interested in forming romantic relationships, parents face numerous challenges. On the one hand, parents need to help their adolescent sons and daughters develop the relational skills and competencies to form healthy romantic relationships that may lay the foundation for future relationships in adult life [15, 17, 18]. On the other hand, parents need to ensure that their children do not engage in early sexual activity or sexual risk behaviors that may put their child at risk of HIV, STIs, and unintended pregnancies during adolescence. Although dating and romantic relationships are a normative aspect of adolescent development, few would argue that 13 year old boys and girls should be having sexual intercourse. Indeed, research suggests that sexual debut before age 14 is associated with an increased risk of pregnancy and STIs [10]. However, forming romantic bonds, holding hands, and kissing in private represent non-sexually intimate behaviors that reasonably emerge for many youth at this age. The challenge for parents is to ensure that their children behave in responsible ways that are consistent with maternal values and enable adolescents to safely explore their developing sexuality without transitioning to sex at too early an age. Few studies have been conducted on the nature of romantic relationships during early adolescence in inner city Latino youth. Even fewer studies have been conducted on maternal approval and disapproval of Latino adolescents having boyfriends/girlfriends and engaging in relationship-related behaviors. The present study addressed these issues and observed a number of interesting findings.

First, we found that romantic relationships in 13 year old, inner city Latino adolescents are frequent but brief. At least half of the youth had been in a romantic relationship by the eighth grade, with relationships generally lasting between 3 and 4 months in duration. Despite their brief duration, the romantic feelings in these relationships were characterized as being moderately strong, with over a third of the girls thinking that their partner is “the one.” About one in five of the youth entertained thoughts about having a family with their partner, a finding that takes on increasing importance when one considers that approximately 50% of Latino girls will become pregnant by age 20 [57]. This suggests that pregnancy intentions may begin in early adolescence and should be addressed as an important component of sexual risk reduction interventions for middle school-aged Latino youth.

Whereas about 25% of the youth were currently involved in a romantic relationship, a sizeable percent of the youth had already been in romantic relationships, a result that signals that issues of romantic relationships and sexuality are salient to youth in the eighth grade. Boys tended to report stronger feelings than did girls and a significantly greater percentage of boys (27%) than girls (13%) endorsed the belief that sex would make them feel closer to their partner. This gender difference may reflect the finding that eighth grade boys were more likely than their female peers to have been in romantic relationships and to have already transitioned to vaginal sexual intercourse. Indeed, we found that being in a romantic relationship was linked to intentions to engage in future sexual behavior: both Latino boys and girls who had been in a romantic relationship had stronger intentions to engage in sexual intercourse at this time in their lives than Latino youth who had not been in a romantic relationship. Although this association tended to be more pronounced for Latino boys than girls, it was evident in both genders.

Overall, the data point to the central role that romantic relationships have for young, inner city Latino youth and the salience that sexuality may take on in those relationships. This is true for both boys and girls and for Puerto Rican, Dominican, and Mexican youth. Our research underscores the need to supplement models of individual decision-making about sexual risk taking to incorporate the relationship context within which such decisions are made. The large literature on relationship dynamics and sexual risk taking with young adults and cohabitating youth likely is of limited value for understanding adolescent romantic relationships, as the nature of relationships is qualitatively different for these different groups. Early adolescents are still developing cognitively, emotionally, socially, morally, and physically. They are experiencing, or have only just experienced, substantive pubertal and physical changes and are in the early stages of becoming tangibly sexualized. Their romantic relationships are brief, characterized by strong feelings, and are often with older partners, at a time when an age difference of only 1 year can mean a great deal [10]. Across a number of nationally representative surveys, it has been found that sexual intercourse is much more likely to occur when youth date partners who are two or more years older [10], a finding that may be especially be true for Latina youth [58, 59].

It is against this backdrop that Latino parents must help their sons and daughters develop the tools to engage in healthy and safe relationships. About 30% to 35% of Latino mothers approved of their eighth grade daughters having a boyfriend and about 40–50% approved of their eighth grade sons having a girlfriend. This represents a fairly substantial minority of parents who condone romantic relationships for 13 year olds. Consistent with previous studies, we found that girls were more likely to date older partners than were boys. Although we did not ask youth and mothers about maternal approval of dating older partners, research has found that a third of Latino adults believe it is acceptable for Latino teens to date a partner who is three or more years older [60]. Moreover, sizeable gender differences have been observed, with 53% of Latina adolescents and 37% of Latino adolescents reporting that it is okay to date an older partner [60]. Our data also documented a fairly consistent double standard for Latino girls and boys, with mothers being more approving of boys being in romantic relationships and engaging in intimate behaviors in those relationships. Our findings add to previous research documenting gender differences in parental, adolescent and societal attitudes about adolescent dating and sexual behavior, with girls endorsing and experiencing more conservative attitudes and behaviors [61, 62]. A recent meta-analysis of gender differences in sexuality indicated that gender disparities in the embracement of a sexual double standard may diminish as youth become older [62], suggesting that adolescence may be an optimal time to intervene. In addition, some researchers have suggested that the double standard may pose a unique health risk for men, as it endorses a concept of masculinity and manhood that supports risky sexual behaviors [62]. Within the broader literature on parental influences on Latino adolescent sexual behavior, this double standard has often been examined through the cultural lens of machismo and marianismo [63], Latino cultural constructs that have been associated with gender norms and risky sexual behavior among males (e.g., engaging in casual sex, having multiple sexual partners, and not using condoms during sex) and females (e.g., diminished power and sexual decision-making in sexual relationships and reduced condom use) [64, 65]. The precise basis of this double standard for adolescents needs to be better understood and should be explored in future research with diverse Latino youth. Here, prospective studies will be particularly important to determine if maternal and adolescent endorsement of this double standard changes over time and how such potential changes might be associated with adolescent’s involvement in risky sexual behaviors. As this research grows, programs can sensitize parents and youth to the double standard, the potential health and social consequences of early sexual behavior for Latino adolescent boys, girls and their partners, and appropriate and inappropriate intimate behaviors for early adolescent couples.

A number of studies have found that adolescents tend to underestimate the extent to which parents disapprove of them engaging in sexual behavior [26, 34]. In the current study, we also observed a general lack of correspondence between what mothers approve and disapprove of in adolescent romantic relationships and what their adolescent children think they approve and disapprove of. Overall, adolescents tended to see their mothers as being more approving (or less disapproving) than the mothers actually were. This suggests that parents need to do a better job at making their behavioral expectations explicit to their children, communicating these expectations in clear ways, and encouraging their children to adhere to those expectations. Our data underscore the need for more research on how to help parents accomplish the above.

A striking finding was the rather sizeable percentage of 13 year old Latino youth who did not view their parents as disapproving of them engaging in sexual intercourse in the context of a romantic relationship. The present research found that perceived maternal approval–disapproval of relationship behaviors that are intimate but less sexual in nature (e.g., holding hands, kissing) were not associated with adolescent intentions to have sex over and above maternal approval–disapproval of relationship behaviors focused on sexual intercourse, if the adolescent did not have a prior romantic relationship history. However, if the adolescent had been in a relationship, then perceived maternal approval of even such innocuous behaviors as holding hands and kissing was associated with intentions to have sex. One reason for this may be that adolescents who have made the step into participating in romantic relationships may have trouble interpreting mixed messages from parents about what is appropriate with respect to less intimate and more intimate behaviors. To the extent this is the case, parents need to be especially clear about their specific expectations and intervention programs may need to include content that helps parents to tailor their messages accordingly. For example, parents may need to be explicit about their expectations that they disapprove of more intimate behaviors such as sexual intercourse even though holding hands and kissing may be acceptable. Separate from this, we found that adolescent intentions to have sex were associated with how adolescents perceived their mothers’ feelings about them engaging in sex, again indicating that mothers need to be clear about their disapproval of sexual intercourse in romantic relationships between such young adolescents.

A final result worth highlighting was our finding that adolescent behavior was associated not only with how adolescents perceived the orientations of their mother with respect to relationships, but also with how the mother actually felt about such behaviors (as self-reported by the mother), holding constant adolescent perceptions. These results suggest that even if mothers are not entirely successful at conveying their dating orientations to their children, the mothers probably act on those orientations in ways that are associated with adolescent behavior, such as through monitoring activities and supervision. In the present study, we focused specifically on maternal attitudes that focused on approval and disapproval of dating and intimate behaviors. Previous research has identified other dating-specific practices such as restrictions and limit setting, dating rules, problem solving around dating, and support and encouragement to date [26, 31, 33, 40]. Future research should explore how a more comprehensive range of dating-specific practices are associated with early adolescents’ intentions to engage in sex and with sexual risk behaviors among sexually active youth, as this could have important implications for the development of parent-based interventions.


The results of the study must be interpreted within the methodological limitations under which it was undertaken. First, the results are based on cross-sectional data and do not permit a discussion of causality. Previous longitudinal studies examining parental influences on adolescent sexual behavior have observed bidirectional effects [66]. As such, it is possible that parents’ attitudes toward dating and intimate behaviors in romantic relationships change once their teen begins dating. In addition, it also is feasible that adolescents perceive lower disapproval once they start dating and romantic relationships take on increasing importance in their own lives. Future research should explore how maternal attitudes about adolescent romantic relationships, and adolescent’s perceptions of these attitudes, may change over time, and how such changes are potentially related to sexual risk taking among Latino youth. In addition, the sample in the present study was not nationally representative, though we suspect it characterizes many Latino youth in large, inner city neighborhoods. The data were based on self-reports, which may be subject to bias. Some of the models we tested may have specification error, which can bias parameter estimates and significance tests of them. Given the large sample size, it also is possible that some of the observed significant associations were due to chance. Although we acknowledged the small magnitude of observed findings whenever appropriate, some of the associations should be interpreted with caution. Finally, we interviewed mothers, not fathers, and presented data on youth’s perceptions of maternal approval. Although we collected data on adolescent’s perceptions of paternal disapproval, these analyses were outside the scope of the present study and are thus being reported in a subsequent manuscript. Fathers remain understudied in the research literature and future research should examine paternal influences on Latino youth’s romantic relationships and sexual behavior during early adolescence. Despite these limitations, the results make an important contribution to the small body of research examining the relationship between parent’s dating-specific practices and adolescent’s involvement in romantic and sexual relationships during adolescence. The findings are worthy of additional research, especially prospective studies that can determine causality and document how parents shape normative developmental trajectories of romantic relationships and sexual behavior among diverse groups of Latino adolescents.


This research was conducted as part of the Entre Familias Study and was supported by funding from the Centers for Disease Control and Prevention, Cooperative Agreement #1 U01 DP000175. The findings and conclusions in this paper are those of the authors and do not necessarily represent the official position of the U.S. Centers for Disease Control and Prevention.

Copyright information

© Springer Science+Business Media, LLC 2011

Authors and Affiliations

  • Alida Bouris
    • 1
  • Vincent Guilamo-Ramos
    • 2
    • 3
  • James Jaccard
    • 2
    • 3
  • Michelle Ballan
    • 4
  • Catherine A. Lesesne
    • 5
  • Bernardo Gonzalez
    • 2
  1. 1.School of Social Service AdministrationUniversity of ChicagoChicagoUSA
  2. 2.Center for Latino Adolescent and Family Health (CLAFH)New York UniversityNew YorkUSA
  3. 3.Silver School of Social WorkNew York UniversityNew YorkUSA
  4. 4.School of Social WorkColumbia UniversityNew YorkUSA
  5. 5.Centers for Disease Control and PreventionAtlantaUSA

Personalised recommendations