Relational factors of vulnerability and protection for adolescent pregnancy: a cross-sectional comparative study of Portuguese pregnant and nonpregnant adolescents of low socioeconomic status.

Pereira, Ana I.F. ; Canavarro, Maria C. ; Cardoso, Margarida F. 等

Adolescent pregnancy has been the target of various studies because it elicits innumerable questions. In developed societies it is considered a serious social problem that is associated with low schooling, precarious employment, and poverty. Reproductive health remains a priority issue for European Union (EU) member states even at a time when pregnancy-related mortality rates are at historically low levels. Although the mean age of mothers at the time they give birth, an important determinant of maternity outcome, is increasing throughout the EU, there is a worrisome persistent increase or stability in teenage pregnancy (Ferrinho, Bugalho, & Pereira, 2004).

Portugal has seen a sustained downward trend in teenage pregnancies since the mid-eighties when specific health service measures were introduced for teenagers and improvement in family planning reduced the number of unwanted pregnancies (Mendonca & Calado, 2002). Nevertheless, in 2003 approximately 6,144 Portuguese infants were born to adolescent mothers and 5.5% of all births were from mothers under age 20 (INE, 2003). Portugal has the second highest teenage birth rate in the former 15 European Union member states (UNICEF, 2001).

In the era of information and availability of contraceptive methods, why do adolescents get pregnant? Consistent with pluralist and multi-factorial development theory and models, adolescent pregnancy is seen as a multidetermined phenomenon with different developmental outcomes. Similar to what is postulated in the principles of equifinality and multifinality of general system theory (von Bertalanffy, 1968), different courses can lead to pregnancy during adolescence and different trajectories are drawn from there. Therefore, in lieu of looking for the determinants of adolescent pregnancy, it is preferable to study the different risk and protective factors involved.

Several authors (Corcoran, 1999; Corcoran, Franklin, & Bennett, 2000; Pistole, 1999; Scaramella, Conger, Simons, & Whitbeck, 1998) have claimed the importance of a relational approach to studying risk for adolescent pregnancy. The present study assumes both a relational and ecological perspective. Bronfenbrenner (1979, 1987) believes that comprehension of human development requires "the analysis of multi-personal interractional systems, that are not restricted to only one environment" (p. 40). Although an ecological perspective considers the influence of contexts in which children do not participate, this perspective also predicts that the relationships that form a direct part of the child's and adolescent's life are those that have a most profound influence (Cauce, Reid, Landesman, & Gonzales, 1990). A series of studies identified three main "support systems" in adolescence: family, friends, and school (Cauce, Mason, Hiraga, & Liu, 1994). The following is a brief review of the relational factors of familial and extra-familial contexts that are antecedents to and associated with adolescent pregnancy.

Family Factors

Concerning the familial context, most studies focus on sociodemographic characteristics and family structure. These studies (Coley & Chase-Lansdale, 1998; Hardy & Zabin, 1991; Marlove, 1997; Miller, 2002; Woodward, Fergusson & Horwood, 2001; Yampolskaya, Brown, & Greenbaum, 2002) support the view that there is a large incidence of pregnancy in adolescents raised by single parents, by parents with low educational attainment and low income, and in adolescents from larger families. The existence of adolescent pregnancy/ maternity models in the family is another variable that seems to have influence. The pregnancy in this life cycle period occurs more frequently in daughters of mothers with an adolescent pregnancy history (Manlove, 1997; Woodward et al., 2001) and in sisters of adolescent mothers (Records, 1993). This suggests the existence of intergenerational patterns in teenage motherhood.

With regard to psychological and relational variables of the family context, investigation results are not consistent. Some studies suggest as predictors of early pregnancy a family environment characterized by stress, depression, and conflict (Ravert & Martin, 1997; Woodward, et al., 2001), more dysfunctionality and rigidity (Garrett & Tidwell, 1999), physical, sexual, and emotional abuse (Garrett & Tidwell, 1999; Kellogg, Hoffman, & Taylor, 1999; Woodward et al., 2001), and a lower relationship quality between parents and daughters (Landy, Schubert, Cleland, Clark, & Montgomery, 1983). Nevertheless, other studies (Townsend & Worobey, 1987) revealed the absence of significant differences in family relations when pregnant adolescents are compared with nonpregnant adolescents. Limitations inherent to some of these studies are the cross-sectional design, small selected samples of pregnant/mother adolescents, sample collection methods, and the utilization of instruments with nonevaluated psychometric properties.

Few studies examined the mediating mechanisms that explain how family variables influence adolescent sexual behavior and pregnancy. In an exceptional study using a prospective longitudinal design, Scaramella et al. (1998) followed a cohort of rural adolescents from 7th to 12th grade. Results of this study show that relationships with parents characterized by warmth and involvement during early adolescence are significantly and negatively associated with risk behavior during middle adolescence. This kind of family environment is also negatively correlated with deviant-peer affiliations and positively correlated with academic competence. The effect of parental warmth and involvement during early adolescence in pregnancy status during middle and late adolescence was examined by structural equation analysis controlling for parents' education. Results show that the effects of parenting were indirect and mediated by academic competence, deviant-peer affiliations, and risk-taking behavior. In summary, this study suggests that family relationships indirectly influence the risk for adolescent pregnancy, affecting risk and protective factors within other contexts: peers and schools.

Relations Outside the Family

Relationship with peers is also associated with teenage pregnancy. Some studies (Scaramella et al., 1998; Woodward et al., 2001) found an association between affiliation with deviant peers and teenage pregnancy risk. Also, larger interpersonal difficulties, a restricted number of friends, and a lower quality of friendships characterize the pregnant adolescent's peer relations. The pregnant/mother adolescents, compared to nonpregnant adolescents have fewer friends (Alvarez, Burrows, Zvaighat, & Muzzo, 1987; Passino et al., 1993), more often consider their peer relations as less important than family links, are more submissive, and exhibit poorer interpersonal communication abilities in their relations with other adolescents (Hardy & Zabin, 1991). These findings suggest overall unsatisfactory peer relations. However, other studies (Connelly, 1998; Corcoran et al., 2000; Orshan, 1999) that compare social support perceived by pregnant and nonpregnant teenagers report no differences, which lead us to conjecture that girls who become pregnant may compensate for these unsatisfactory relations with support received from other persons (e.g., boyfriend).

Little is known about the romantic relations of the pregnant teen. Most studies gather information only from the mothers, and studies that include the adolescent's partner consider only the sociodemographic characteristics of the child's father and the objective data of the relationships (e.g., longevity). Some studies (Bull & Hogue, 1998; Coley & Chase-Landsdale, 1998; Males, 1993) show that the partner tends to be 2 to 4 years older and have low educational achievement. Dating at an earlier age is a factor that distinguishes pregnant teenagers from adolescents without a history of early pregnancy (Furstenberg, 1976). Romantic relationships of pregnant teens are relatively steady, although they have a lower mean length of relationship with the partner prior to pregnancy (compared with the relationship preceding marriage for married women without a history of early pregnancy) (Schamess, 1993). Also, the pregnancy tends to occur most often within an exclusive relationship (De Anda, 1983). However, emotional motivations in romantic relationships research tend to be disregarded in studies of teen pregnancy (Pistole, 1999).

School Context

Pregnant adolescents, in comparison to nonpregnant adolescents, have lower levels of education, lower educational aspirations and performance, and higher levels of school dropout (Alvarez et al., 1987; Coley & Chase-Lansdale, 1998; Garrett & Tidwell, 1999; Harris, 1998; Ireson, 1984; Manlove, 1998; Olson & Worobey, 1984; Townsend & Worobey, 1987; Yampolskaya et al., 2002). Nevertheless, this does not mean that those who get pregnant during adolescence have lower cognitive capabilities compared to their nonpregnant peers. Most often, these adolescents attend schools that have fewer resources along with students with lower capabilities, provide less stimulation, and subject them to negative cataloguing (Manlove, 1998).

METHOD

Subjects

Two groups of young women: pregnant (n = 57) and without a pregnancy history (n = 81), aged 14 through 18 years, comprised the convenience sample population. Both groups were from a low socioeconomic status. Data from the pregnant adolescents were collected in three national health services from different geographic areas. For those without a history of pregnancy, data were collected in public schools. A balance between the two groups with regard to socioeconomic status and geographic area guided school selection. All those in the pregnancy group were primiparous: 18% in the first trimester, 35% in the second trimester, and 47% in the third trimester. The majority (85%) of these pregnancies were unplanned.

Procedure

The pregnant adolescents were contacted while they were waiting for an obstetrics consultation in a public service facility. There the study objectives were explained, confidentiality guaranteed, and oral consent obtained. The adolescents were interviewed individually and then completed the study instruments that are described in the next section.

Before the instruments were administered to the nonpregnant adolescents during class time, confidentiality was guaranteed and the voluntary nature of participation was made clear. All adolescents approached in both groups agreed to participate. All adolescents who were not from a low socioeconomic status were excluded. With regard to the nonpregnant adolescent group, those younger than 14 and those who had a previous pregnancy history, were excluded. Adolescents in the pregnant group who had a previous pregnancy were also excluded.

Measures

Sociodemographic, Personal, and Familial Antecedents Questionnaire. This questionnaire was composed of 61 questions for both groups and 16 additional questions for the pregnant adolescents. The questions were organized in five different thematic parts: personal data (sociodemographic and antecedents); family data (sociodemographic and antecedents); school antecedents and professional and school future; sexual life and contraceptive methods; life events and preoccupations. Three additional topics were addressed to the pregnant adolescents: pregnancy, child's father's sociodemographic data and relationship with the child's father, and future projects for the adolescent and her baby.

Perceived Parental Child Rearing (EMBU). Evaluation of the parental child rearing was accomplished through administration of the Portuguese shortened version of EMBU (Perils, Jacobson, Lindstorm, von Knorring, & Perils, 1980), developed by Canavarro (1996). This instrument was designed to assess memories of both maternal and parental rearing behavior, measured separately on three dimensions: Rejection, Emotional Warmth, and Overprotection. The items are scored on a four-point Likert scale for each parent.

Following a personal contact with M. Eisemann (November, 1999), a few changes to the Portuguese version (Canavarro, 1996, 1999) were made to ensure its appropriateness for the adolescent sample of the present study.

EMBU is one of the most widely used retrospective measures of perceived parental child rearing during childhood and adolescence. This instrument was selected among others that evaluate perceived parental rearing based on its psychometric qualities which were supported by various studies in different countries (Arrindell, Perris, Eisemann, van der Ende, Gasgner, Iawawaki, Maj, & Zhang, 1994; Arrindel & van der Ende, 1984; Gerlsma, Emmelkamp, & Arrindel, 1990; Ross, Campbell, & Clayer, 1982; Ruiter, 1994; Vandereycken, 1994; Winefield, Tiggermann, & Winefield, 1994). The Portuguese psychometric studies (Canavarro, 1996) also revealed that this version has good validity and reliability.

Convoy model. Social support was measured by administering a part of the Portuguese version (Araujo, 1997) of the measure developed by Kahn and Antonucci (1980). The Convoy model evaluates the respondent's social networks through a spatial representation technique. Respondents are presented with a set of three concentric circles with the word "you" in the center. Respondents are asked to name the "people who, in your life, are closer, in function of the support (affective, familial, professional ...) received from and given to them," in each of three levels as depicted by the three concentric circles. They are asked to place in the inner circle those persons who are "so close and important" to them that they "can't imagine life without them." Those considered less close but still important are placed in the middle and outer circles. These persons are identified according to relationship (e.g., father, mother, friend, boyfriend, neighbor).

Different support providers are given a score indicating the importance of their position in the three circles, ranging from 0 to 3. The persons placed in the first circle, the circle of the highest emotional proximity, are given three points; those placed in the second circle are given two points; and those placed in the third circle are given one point. A score of 0 indicates that the type of relation is nonexistent or if it did exist the girl did not include it in her social support network.

The respondents then have to evaluate the satisfaction with the support received from others and the satisfaction with the support given to others. The evaluation is done on a five-point Likert scale, from "Not satisfied" to "Very much satisfied." This measure was chosen for the evaluation possibilities it offers; namely, the simultaneous evaluation of the structural aspects and evaluation of the adequacy of the identified social network. Furthermore, it has the advantage of being a less restrictive procedure in the social network identification because respondents are allowed to identify as many people as they wish (Antonucci, 1985).

Statistical Methodology

In order to investigate possible predictors of adolescent pregnancy, the first stage of the analysis was to establish which of the factors were significantly associated with the pregnancy. Comparisons were made between the pregnant and nonpregnant adolescent groups using chi-square tests for association of categorical variables (such as area of residence or emotional proximity to peers) and independent t-tests for continuous variables (such as emotional support or overprotection from father/mother). The second stage of this analysis was to determine which of the antecedent factors to pregnancy were significantly and independently associated with adolescent pregnancy. A multifactorial analysis using logistic regression was employed. Results are reported as adjusted odds ratio of adolescent pregnancy and the corresponding 95% confidence interval. The effect of each factor was obtained controlling for all the others in the model, and nonsignificant factors (p > 0.05) were dropped from the final model, with the exception of age and area of residence. Interactions between variables were also investigated.

RESULTS

Family Factors of Pregnant and Nonpregnant Adolescents

As expected, some associations between family factors and adolescent pregnancy were found. All family factors assessed through the Socio-Demographic and Familial Antecedents Questionnaire were significantly different for the two groups (Table 1). Pregnant adolescents were more likely to have more siblings, to belong to a non-intact family, to have parents with a history of previous psychological/psychiatric intervention, and to have mothers with a history of adolescent pregnancy. Concerning Perceived Parental Child Rearing assessed through EMBU, significant differences between the groups were also found (Table 2). Pregnant young girls perceived less emotional support and less overprotection from both parents. There were no statistical significant differences concerning the rejection dimension.

Relations with Peers and Partners and Social Support in Pregnant and Nonpregnant Adolescents

A considerable percentage of adolescents from the pregnant group reported having no friends (31.6%), while only one adolescent of the nonpregnant group did not include any friend in her social network. Also, teenagers from the pregnant group reported significantly lower levels of emotional proximity to peers (Table 3).

In regard to the romantic partner, there are also significant differences. While the great majority (89.3%) of pregnant teens include romantic partners in their social network, only a minority (38.3%) of the nonpregnant adolescents include the boyfriend. Further, pregnant adolescents report significantly higher levels of emotional proximity to romantic partners (Table 3). It is important to note that for the majority of teens, the child's father is the first sexual partner, and that their relationship has considerable longevity (M = 1.66 years, SD = 1.06). Finally, no significant differences were found concerning the global evaluation of perceived and received support.

School Context in Pregnant and Nonpregnant Adolescents

In the pregnant adolescent group, only a minority attend school (12.3%), and for the majority of them, dropout occurred before the pregnancy (52.6%). All the girls in the nonpregnant sample attend school (a consequence of the sample procedures). However, a Portuguese study (Lourenco, 1996) that collected the sample of both pregnant and nonpregnant adolescents in health centers found similar differences between the two groups concerning school dropout.

Also, there is a significant difference (t = 6.61, p < 0.001) between the two groups concerning grade retention. The pregnant adolescents have a higher number of retentions in grade (M = 1.54, SD = 1.21) than the nonpregnant adolescents (M = 0.38, SD = 0.64). Again, these results should be interpreted with caution. The sample of nonpregnant adolescents was collected in schools and does not include nonpregnant adolescents who dropped out of school. This fact can lead to a greater difference in the number of retentions between pregnant and nonpregnant adolescents than is found in the general population.

Logistic Regression Considering Family, School and Peers Factors

The comparison of pregnant and nonpregnant adolescents was also made through a logistic regression method. Results are shown in Table 4. Variables were selected among those factors found to be significant at the 5% level in the bivariate analyses, with the exception of emotional proximity to boyfriend. Although this variable can contribute to an understanding of the factors that promote teenage pregnancy, we chose not to include it in the analysis for two reasons: this variable is more sensitive to changes associated with the occurrence of pregnancy, and only 38% of the nonpregnant adolescents had a boyfriend. Among the possible predictor variables included in the multifactorial model, five factors remain significantly and independently associated with adolescent pregnancy after adjusting for the effects of age, area of residence, and the remainder in the model. This main effects model fit the data adequately and no interactions were deemed necessary.

After adjusting for the effects of the other variables in the model, lower overprotection from mother and lower emotional support from father were positively associated with teenage pregnancy. Also, pregnancy was more common in girls whose mother had a history of adolescent pregnancy. Concerning school factors, for each year of school failure, the risk of pregnancy increased significantly. Finally, lower emotional proximity to peers was also positively associated with pregnancy. In sum, the most effective predictors of adolescent pregnancy were family, school, and peer level factors.

DISCUSSION

The results of the present study confirm the importance of the simultaneous study of different relational contexts for the understanding of adolescent pregnancy. Factors belonging to different levels of analysis--family, school, peer relations--make independent contributions toward characterization of the two adolescents groups, pregnant and without a pregnancy history.

In spite of the transverse nature of the study, we can infer how the various factors studied, assumed as antecedent variables from a theoretical point of view, can create vulnerabilities for the occurrence of pregnancy in this developmental period. The following describes some hypothetical mechanisms that can explain how these variables could lead the adolescents to desire the pregnancy or, in the case of nonplanned pregnancy, the factors that caused the adolescent to choose the maternity when she became aware that she was pregnant.

Concerning the adolescent's family, the results showed that pregnant adolescents perceive less emotional warmth from father and less overprotection from mother in comparison with the nonpregnant adolescents. The importance of parental child rearing in understanding early pregnancy leads us to search for the different mechanisms that explain how these variables act.

A young woman who perceives less emotional support from her parents may seek this support in romantic relationships (Pereira, 2001; Pereira, Canavarro, Mendonca, & Cardoso, 2002). Some of the data collected during the interviews reveal that many of the adolescents who planned the pregnancy did it to please the boyfriend, who was older and wanted to have a child. Others did it because they wanted to begin a new life with their partners.

Less supervision/overprotection from parents is associated with earlier sexual activity and more risk behaviors, including nonprotected sex (Donoghue, 1992-93; Peterson, Sripada, & Barglow, 1982; Scaramella et al., 1998; Tobas & Ricer, 1998). Frequently these adolescents grow up on their own, under the influence of street culture and without well-defined limits to their behavior. Concerning adolescents who live in poor communities, with little resources and few adult role models, parents' supervision emerges as an important factor of protection from deviant and risky behaviors (Pereira, 2001; Pereira et al., 2002). Poor supervision and involvement by parents also affect the school context and are associated with lack of academic competence and school dropout (Scaramella et al., 1998). In sum, greater emotional support and adequate supervision seem to be protective factors for adolescents of low socioeconomic status. Miller (2002), in a recent literature review, points two family relations variables--parent/child connectedness, closeness and warmth, and parental regulation, monitoring, and supervision, as important predictors of adolescent sexual behavior and pregnancy. Warm and supportive relations with parents promote sensitive environments attuned to adolescents' needs and provide a general sense of competence and responsibility (Scarmella et al., 1998). This environment also has a positive effect on adolescents in other contexts; e.g., adherence to school demands and avoidance of deviant peers.

Consistent with other empirical studies (Manlove, 1997; Woodward et al., 2001) another factor that differentiated adolescent pregnant and nonpregnant groups was the existence of a history of adolescent pregnancy in the adolescent's mother. The mother's early pregnancy is associated with other risk factors for being pregnant during adolescence, among them, single parenthood and larger families. Further, the model provided by the mother can lead to more permissiveness in the adolescent regarding early sexual activity and easy acceptance of adolescent pregnancy.

School is equally a reference context for understanding early pregnancy (Alvarez et al., 1987; Coley & Chase-Lansdale, 1998; Garrett & Tidwell, 1999; Harris, 1998; Ireson, 1984; Manlove, 1998; Olson & Worobey, 1984; Townsend & Worobey, 1987; Yampolskaya et al., 2002). School failure and low success expectancy in association with difficult economic conditions of the families of our sample, can lead the adolescents to drop out and early entrance into the work world. Maternity is one more adult role the adolescent is forced to play (Pereira, 2001). From a different perspective, the teenager with a history of school failure may see maternity as entailing less costs in terms of future projects, leading to easier acceptance of pregnancy (Manlove, 1998). School dropout is also associated with the loss of a potential source of social support and friendship.

In line with findings of other studies (Alvarez et al., 1987; Passino et al., 1993) young pregnant women perceive less emotional proximity to peers and some do not have friends in their social support network. This can be another factor in producing vulnerability for early pregnancy. Isolation from peers can lead the adolescent to seek the support and care of others through becoming pregnant--a guarantee of a "more secure" relation with the boyfriend, or even that her child can fulfil the need for companionship (Canavarro & Pereira, 2001). This kind of compensating mechanism seems to be supported by the results concerning emotional proximity to the partner and social support. A higher emotional proximity to the romantic partner, and a steady relationship may lead to greater social support. Thus, it seems that a multicontextual approach to the study of early pregnancy contributes to a better understanding of the process.

Although the present study contributes to some additional understanding of early pregnancy, there are some study limitations. The cross-sectional design and retrospective assessment of the variables make it difficult to ascribe causality to relationships between pregnancy and some variables. Another limitation concerns the difference between the contexts of data collection for the two groups. The pregnant group was recruited in health centers, while the nonpregnant group was recruited in schools. Given the fact that approximately half the sample of pregnant adolescents was not attending school, the differences found between the two groups may simultaneously reflect school dropout and pregnancy factors.

Nevertheless important contributions can be drawn from this study. The low socioeconomic adolescent sample is one of its main strengths since some previous studies did not take this into account. The focus here on adolescents from low socioeconomic status provided some understanding of the risk and protective factors for adolescent pregnancy in this high-risk group. Finally, the use of psychometrically sound instruments and the simultaneous consideration of different relational contexts through multivariate procedures are other assets of the present work.

REFERENCES

Alvarez, M. L., Burrows, R., Zvaighat, A., & Muzzo, S. (1987). Sociocultural characteristics of pregnant and nonpregnant adolescents of low socioeconomic status: A comparative study. Adolescence, 22, 149-156.

Antonucci, T. (1985). Personal characteristics, social support, and social behavior. In R. H. Binstock & E. Shanas (Eds.), Handbook of aging and social sciences (2nd ed., pp. 94-128). New York: Van Nostrand Reinhold.

Arrindell, W. A., Perris, C., Eisemann, M., van der Ende, J. F., Gasgner, P., Iawawaki, S., Maj, S., & Zhang, J. (1994). Parental rearing behavior from a cross-cultural perspective: A summary of data obtained in 14 nations. In C. Perris, W. Arrindell, & M. Eisemann (Eds.), Parenting and Psychopathology (pp. 145-171). New York: John Wiley & Sons.

Arrindell, W. A., & van der Ende, J. (1984). Replicability and invariance of dimensions of parental rearing behavior: Further Dutch experiences with the EMBU. Person. Individ. Diff., 5, 671-682.

Bronfenbrenner, U. (1987). La ecologia del desarrollo humano. [The ecology of human development]. Barcelona: Ediciones Paidos Iberica. (Original work published in 1979.)

Bull, S., & Hogue, C. J. R. (1998). Exploratory analysis of factors associated with teens' repeated childbearing. Journal of Health Care for the Poor and Undeserved, 9(1), 42-61.

Canavarro, M. C. (1996). Avaliacao das praticas educativas parentais atravees do EMBU: Estudos psicometricos [Evaluation of parental rearing behavior with the EMBU: Psychometric studies]. Psychologica, 16, 5-18.

Canavarro, M. C. (1999). Relacoes afectivas e saude mental. Urea abordagem ao longo da vida [Affective relations and mental health]. Coimbra: Quarteto Editora.

Canavarro, M. C., & Pereira, A. I. (2001). Gravidez e maternidade na adolescencia: Perspectivas teoricas [Pregnancy and maternity in adolescence: Theoretical perspectives]. Coimbra: Quarteto Editora.

Cauce, A. M., Mason, C., Hiraga, Y., & Liu, G. (1994). Social support during adolescence: Methodological and theoretical considerations. In F. Nestmann & K. Hurrelmann (Eds.), Social networks and social support in childhood and adolescence (pp. 89-108). Berlin: Walter de Gruyter.

Cauce, A. M., Reid, M., Landesman, S., & Gonzales, N. (1990). Social support in young children: Measurement, structure and behavioral impact. In B. R. Sarason, I. G. Sarason, & G. R. Pierce (Eds.), Social support: An interactional view (pp. 64-95). New York: Wiley.

Coley, R. L., & Chase-Lansdale, P. L. (1998). Adolescent pregnancy and parenthood: Recent evidence and future directions. American Psychologist, 53, 152-166.

Connelly, C. D. (1998). Hopefulness, self-esteem, and perceived social support among pregnant and nonpregnant adolescents. Western Journal of Nursing Research, 20, 195-209.

Corcoran, J. (1999). Ecological factors associated with adolescent pregnancy: A review of the literature. Adolescence, 34(135), 603-620.

Corcoran, J., Franklin, C., & Bennett, P. (2000). Ecological factors associated with adolescent pregnancy and parenting. Social Work Research, 24(1), 29-40.

De Anda, D. (1983). Pregnancy in early and late adolescence. Journal of Youth and Adolescence, 12(1), 33-42.

Donoghue, E. (1992-93). Sociopsychological correlates of teen-age pregnancy in the United States Virgin Islands. International Journal of Mental Health, 21(4), 39-49.

Ferrinho, P., Bugalho, A., & Pereira M. J. (Eds.). (2004). For better health in Europe. Report with the support of the European Commission. Instituto de Medicina Preventiva da Faculdade de Medicina, Universidade Lisboa, Portugal.

Furstenberg, F. F., Jr. (1976). Unplanned parenthood: The social consequences of teenage childbearing. New York: Free Press.

Garrett, S. C., & Tidwell, R. (1999). Differences between adolescent mothers and nonmothers: An interview study. Adolescence, 34, 91-105.

Gerlsma, C., Emmelkamp, P. M. G., & Arrindell, W. A. (1990). Anxiety, depression, and perception of early parenting: A meta-analysis. Clinical Psychological Review, 10, 251-277.

Hardy, J. B., & Zabin, L. S. (1991). Adolescent pregnancy in an urban environment: Issues, programs, and evaluation. Baltimore: Urban & Schwarzenber.

Harris, J. L. J. (1998). Urban African American adolescent parents: Their perceptions of sex, love, intimacy, pregnancy, and parenting. Adolescence, 33, 833-844.

Instituto Nacional de Estatistica (INE, 2003). Estatisticas Demograficas [Demografic Statistics]. Lisboa: INE.

Ireson, C. J. (1984). Adolescent pregnancy and sex roles. Sex Roles, 11, 189-201.

Kahn, R. L., & Antonucci, T. (1980). Convoys over the life course: Attachment, roles and social support. In P. B. Baltes & O. B. Brim (Eds.), Life span development and behavior: Vol. 3 (pp. 253-368). New York: Academic Press.

Kellogg, N. D., Hoffman, T. J., & Taylor, E. R. (1999). Early sexual experiences among pregnant and parenting adolescents. Adolescence, 34, 293-303.

Landy, S., Schubert, J., Cleland, J. F., Clark, C., & Montgomery, J. S. (1983). Teenage pregnancy: Family syndrome? Adolescence, 18, 781-794.

Lourenco, M. M. (1996). Textos e contextos da gravidez na adolescencia. A adolescente, a familia e a escola [Texts and contexts of adolescent pregnancy: Adolescent, family and school]. Coimbra: Faculdade de Psicologia e de Ciencias da Educacao da Universidade de Coimbra.

Males, M. (1993). School-age pregnancy: Why hasn't prevention worked? Journal of School Health, 63(10), 429432.

Manlove, J. (1997). Early motherhood in an intergenerational perspective: The experiences of a British cohort. Journal of Marriage and the Family, 59, 263-279.

Manlove, J. (1998). The influence of high school dropout and school disengagement on the risk of school age pregnancy. Journal of Research on Adolescence, 8, 187-220.

Mendonca, D., & Calado, B. P. (2002). Situacao da Satide Reprodutiva em Portugal. In I. Craveiro, & P. Ferrinho, Saude reprodutiva nos paises e territorios de lingua Portuguesa ]Reproductive health in countries and territories of Portuguese language] (pp. 153-170). Lisboa: Associacao para o Desenvolvimento e Cooperacao Garcia da horta.

Miller, B. C. (2002). Family influences on adolescent sexual and contraceptive behavior. Journal of Sex Research, 39(1), 22-27.

Olson, C., & Worobey, J. (1984). Perceived mother-daughter relations in a pregnant and nonpregnant adolescent sample. Adolescence, 19, 679-694.

Orshan, S. A. (1999). Acculturation, perceived social support and self-esteem and pregnancy status among Dominican adolescents. Health Care for Women International, 20, 245-238.

Passino, A. W., Whitman, T. L., Borkowski, J. G., Schellenbach, C. J., Maxwell, S. E., Keogh, D., & Rellilnger, E. (1993). Personal adjustment during pregnancy and adolescent parenting. Adolescence, 28, 97-122.

Peterson, C., Sripada, B., & Barglow, P. (1982). Psychiatric aspects of adolescent pregnancy. Psychosomatics, 23, 723-733.

Pereira, A. I. F., Canavarro, M. C., Mendonqa, D. V., & Cardoso, M. F. (2002). O mundo interpessoal da gravida adolesclente [The interpersonal world of the pregnant adolescent]. Psicologia, Teoria, Investigacao e Pratica, 7(2), 19-36.

Pereira, A. I. (2001). Contexctos relacionais de vulnerabilidade e protect(to para a gravidez ns adolescencia [Relational contexts of vulnerability and protection for adolescent pregnancy]. Unpublished master's thesis. University of Coimbra, Coimbra, Portugal.

Perris, C., Jacobsson, L., Lindstrom, H., yon Knorring, L., & Perris, H. (1980). Development of a new inventory assessing memories of parental rearing behavior. Acta Psychiatr. Scand, 61, 265-274.

Pistole, M. C. (1999). Preventing teenage pregnancy: Contributions from attachment theory. Journal of Mental Health Counseling, 21(2), 93-113.

Ravert, A. A., & Martin, J. (1997). Family stress, perception of pregnancy, and age of first menarche among pregnant adolescents. Adolescence, 32, 261-269.

Records, A. A.., & Martin, J. (1997). Family stress, perception of pregnancy, and age of first menarche among pregnant adolescents. Adolescence, 32, 261-269.

Records, K. A. (1993). Life events of pregnant and nonpregnant adolescents. Adolescence, 28(110), 325-338.

Ross, M. W., Campbell, R. L., & Clayer, J. R. (1982). Adolescent attachment and psychopathology. Journal of Consulting and Clinical Psychology, 64, 244-253.

Ruiter, C. (1994). Anxious attachment in agoraphobia and obsessive-compulsive disorder: A literature review and treatment implications. In C. Perris, W. A. Arrindell, & M. Eisemann (Eds.), Parenting and psychopathology (pp. 281-307). New York: John Wiley and Sons.

Scaramella, L. V., Conger, R. D., Simons, R. L., & Whitbeck, L. B. (1998). Predicting risk for pregnancy by late adolescence: A social contextual perspective. Developmental Psychology, 1233-1245.

Schamess, S. (1993). The search for love: Unmarried adolescent mothers' views of and relationship with men. Adolescence, 28(110), 425-438.

Tobias, B. B., & Ricer, R. E. (1998). Counseling adolescents about sexuality. Primary Care, 25(1), 49-70.

Townsend, J. K., & Worobey, J. (1987). Mother and daughter perceptions of their relationship: The influence of adolescent pregnancy status. Adolescence, 12, 487-496.

UNICEF (2001, July). A league table of teenage births in rich nations. (Innocenti Report Card, No. 3). Florence: UNICEF Innocenti Research Centre.

Vandereycken, W. (1994). Parental rearing behavior and eating disorders. In C. Perris, W. A. Arrindell, & M. Eisemann (Eds.), Parenting and psychopathology (pp. 219-234). New York: John Wiley and Sons.

von Bertalanffy, L. (1968). General systems theory: Foundations, developments and applications. New York: Braziller.

Winefield, H. R., Tiggermann, M., & Winefield, A. H. (1994). Parental rearing, behavior, attributional style and mental health. In C. Perris, W. A. Arrindell, & M. Eisemann (Eds.), Parenting and psychopathology (pp. 33-54). New York: John Wiley and Sons.

Woodward, L., Fergusson, D. M., & Horwood, L. J. (2001). Risk factors and life processes associated with teenage pregnancy: Results of a prospective study from birth to 20 years. Journal of Marriage and the Family, 63, 1170-1185.

Yampolskaya, S., Brown, E. C., & Greenbaum, P. E. (2002). Early pregnancy among adolescent females with serious emotional disturbances: Risk factors and outcomes. Journal of Emotional & Behavioral Disorders, 10, 108-126.

This research was supported in part by a grant from the Portuguese Science and Technology Ministry, Technology and Science Foundation, provided to the first author. The authors gratefully acknowledged the participation of the adolescents, schools, and national health services in the project.

Ana I. F. Pereira, School of Education Joao de Deus, Lisbon, Portugal.

Maria C. Canavarro, Faculty of Psychology and Sciences of Education, Coimbra University, Portugal.

Margarida F. Cardoso and Denisa Mendonqa, Institute for Biomedical Sciences Abel Salazar, ICBAS, Oporto University, Portugal.

Requests for repents should be sent to Aria I. F. Pereira, School of Education Joao de Deus, Av. Alvares Cabral, 69, 1269-094 Lisboa, Portugal. E-mail: aifreitaspereira@sapo.pt

Table 1
Family Factors

 Pregnant Nonpregnant
 Adolescents Adolescents
Factor (n = 57) (n = 81)

Number of siblings: M (SD) 3.09 (2.92) 1.44 (1.10)
Intact family: n (%) 30 (55.6) 66 (89.2)
Parent psychological/ 19 (33.3) 10 (12.3)
 psychiatric intervention: n (%)
Mother adolescent 17 (29.8) 5 (6.2)
 pregnancy: n (%)

Factor Statistical Tests

Number of siblings: M (SD) t = 4.06, p < 0.001
Intact family: n (%) [chi square] (1) = 18.84, p < 0.001
Parent psychological/ [chi square] (1) = 9.27, p < 0.01
 psychiatric intervention: n (%)
Mother adolescent [chi square] (1) = 12.60, p < 0.001
 pregnancy: n (%)

Table 2
Perceived Parental Rearing Behavior: Results from EMBU

 Pregnant Nonpregnant
 Adolescents Adolescents
Parental Behavior (n = 57) (n = 81)

Emotional support-father: 18.12 (5.44) 21.20 (4.73)
 M (SD)
Overprotection-father: 12.65 (2.69) 14.97 (3.40)
 M (SD)
Rejection-father: M (SD) 9.65 (2.79) 9.80 (2.23)
Emotional support-mother: 20.83 (4.71) 22.31 (3.77)
 M (SD)
Overprotection-mother: 13.45 (2.79) 15.61 (3.44)
 M (SD)
Rejection-mother: M (SD) 11.14 (6.67) 11.45 (3.18)

Parental Behavior Statistical Tests

Emotional support-father: t = -3.40, p < 0.001
 M (SD)
Overprotection-father: t = -4.09, p < 0.001
 M (SD)
Rejection-father: M (SD) ns
Emotional support-mother: t = -2.01, p < 0.05
 M (SD)
Overprotection-mother: t = -3.98, p < 0.001
 M (SD)
Rejection-mother: M (SD) ns

Table 3
Emotional Proximity to Peers and Boyfriend and Satisfaction
with Received and Given Social Support: Results from Convoy Model

 Pregnant Nonpregnant
Emotional Proximity and Adolescents Adolescents
Satisfaction with Support (n = 57) (n = 81)

High emotional proximity to 23 (40.4) 74 (91.4)
 peers *: n (%)
High emotional proximity to 49 (87.5) 28 (34.6)
 boyfriend/partner*: n (%)
Satisfaction with received 3.81 (1.09) 3.88 (0.71)
 support: M (SD)
Satisfaction with given 3.84 (0.73) 3.94 (0.73)
 support: M (SD)

Emotional Proximity and
Satisfaction with Support Statistical Tests

High emotional proximity to [chi square] (1) = 41.68, p < 0.001
 peers *: n (%)
High emotional proximity to [chi square] (1) = 37.69, p < 0.001
 boyfriend/partner*: n (%)
Satisfaction with received ns
 support: M (SD)
Satisfaction with given ns
 support: M (SD)

* Low emotional proximity (0, 1), high emotional proximity (2, 3).

Table 4
Potential Risk Factors for Adolescent Pregnancy:
Results from Logistic Regression Analysis

 Adjusted 95% Confidence
Predictor Variables Odds Ratio Interval p

Age 1.593 0.72-3.53 0.251
Area of residence 0.190
 Rural 1 (a) --
 Urban 3.026 0.58-15.87
Emotional proximity to peers 0.001
 High 1 (a) --
 Low 23.979 4.26-135.09
Number of school failures 7.401 2.18-25.10 0.001
Overprotection-mother 0.658 0.50-0.88 0.004
Emotional support-father 0.808 0.69-0.95 0.011
Mother adolescent pregnancy 0.037
 No 1 (a) --
 Yes 8.115 1.14-57.80

(a) Reference group.