The Association Between Adverse Childhood Experiences and Adolescent Pregnancy, Long-Term Psychosocial Consequences, and Fetal Death
Objectives. Few reports address the impact of cumulative exposure to childhood abuse and family dysfunction on teen pregnancy and consequences commonly attributed to teen pregnancy. Therefore, we examined whether adolescent pregnancy increased as types of adverse childhood experiences (ACE score) increased and whether ACEs or adolescent pregnancy was the principal source of elevated risk for long-term psychosocial consequences and fetal death.
Design, Setting, and Participants. A retrospective cohort study of 9159 women aged ≥18 years (mean 56 years) who attended a primary care clinic in San Diego, California in 1995–1997.
Main Outcome Measure. Adolescent pregnancy, psychosocial consequences, and fetal death, compared by ACE score (emotional, physical, or sexual abuse; exposure to domestic violence, substance abusing, mentally ill, or criminal household member; or separated/divorced parent).
Results. Sixty-six percent (n = 6015) of women reported ≥1 ACE. Teen pregnancy occurred in 16%, 21%, 26%, 29%, 32%, 40%, 43%, and 53% of those with 0, 1, 2, 3, 4, 5, 6, and 7 to 8 ACEs. As the ACE score rose from zero to 1 to 2, 3 to 4, and ≥5, odds ratios for each adult consequence increased (family problems: 1.0, 1.5, 2.2, 3.3; financial problems: 1.0, 1.6, 2.3, 2.4; job problems: 1.0, 1.4, 2.3, 2.9; high stress: 1.0, 1.4, 1.9, 2.2; and uncontrollable anger: 1.0, 1.6, 2.8, 4.5, respectively). Adolescent pregnancy was not associated with any of these adult outcomes in the absence of childhood adversity (ACEs: 0). The ACE score was associated with increased fetal death after first pregnancy (odds ratios for 0, 1–2, 3–4, and 5–8 ACEs: 1.0, 1.2, 1.4, and 1.8, respectively); teen pregnancy was not related to fetal death.
Conclusions. The relationship between ACEs and adolescent pregnancy is strong and graded. Moreover, the negative psychosocial sequelae and fetal deaths commonly attributed to adolescent pregnancy seem to result from underlying ACEs rather than adolescent pregnancy per se.
Despite reductions in the rate of adolescent pregnancy and birthrates during the 1990s, adolescent pregnancy rates in the United States still exceed those of other developed countries by 2- to 15-fold.1–4 Four of 10 adolescent girls get pregnant before the age of 20 in the United States, leading to ∼900 000 teenage pregnancies each year.3
Consequences frequently attributed to adolescent pregnancy include long-term psychosocial and economic disadvantages for the teen mother, as characterized by lower educational attainment, lower income, and a greater risk of being a single parent.5,6 The impact of adolescent pregnancy on the fetus or infant is multifaceted: many studies report an increased risk of fetal death, and infants born to teen mothers have an increased tendency to be low birth weight or premature and are apt to have poorer cognitive development, lower educational attainment, more frequent criminal activity, and higher risk of abuse, neglect, abandonment, and behavioral problems during childhood.5,7–14 Whether adolescent pregnancy itself leads to adverse reproductive outcomes or whether these outcomes are consequences of the same environment of adversity that led to the teen pregnancy has been the focus of debate.15 The association between adolescent pregnancy and long-term consequences seems to be attenuated when factors that precede and increase the risk of adolescent pregnancy are controlled.15,16
Previous reports have identified a host of community-, family-, school-, and individual-level antecedents of adolescent pregnancy,3,17–30 including restrictive contraceptive marketing laws, higher unemployment rates, higher violent crime rates, higher adolescent suicide rates, higher high school drop-out rates, lower parental education, inadequate health care insurance coverage, having an older sister or friend who was an adolescent parent, minority race/ethnicity, lower family connectedness, physical or sexual abuse, general maltreatment by family, greater number of sex partners, alcohol use, substance use, and attempted suicide.17–30 Despite the plethora of antecedents, none of them have been shown to account independently for a large portion of the teen pregnancy risk.3,17–30 The more effective adolescent pregnancy-prevention programs, which have focused broadly on youth development,3,31–33 may ameliorate several of these risk factors.
Although 2 decades of research have been dedicated to identifying antecedents and consequences of adolescent pregnancy, little attention has been devoted to the role of adverse childhood experiences (ACEs) as either an antecedent or a factor that may influence those consequences commonly attributed to teen pregnancy. Numerous reports have used the ACE Study (conducted in California in 1995–1997) to address the association between major causes of death and disability in the United States and ACEs. Indicators included in the ACE measure, each of which is given a score of 1 (maximum: 8) are self-reported household exposure during childhood to emotional abuse, physical abuse, sexual abuse, intimate partner violence, living with a substance-abusing, mentally ill, or criminal household member, or having a parent who was divorced or separated. Published reports have demonstrated strong, graded associations between the ACE score and many unfavorable outcomes including suicide attempts, smoking, alcohol abuse, high-risk sexual behavior, sexually transmitted diseases, and unintended pregnancy in adult women.34–42
We used the ACE Study to evaluate whether exposure to ACEs increased the risk of adolescent pregnancy. We hypothesized that each category of ACE would independently increase the risk of adolescent pregnancy and that, as the ACE score rose, the risk of adolescent pregnancy would increase. In terms of unfavorable consequences of adolescent pregnancy, we examined whether ACEs or adolescent pregnancy itself was the principal source of elevated risk for long-term psychosocial consequences including serious family problems, job problems, financial problems, high stress, and uncontrollable anger in adulthood. We also examined whether ACEs or adolescent pregnancy per se was associated with negative effects on viability of the infant, as measured by fetal death.
The epidemiologic methods used for the ACE Study have been described elsewhere.34–42 Briefly, a retrospective cohort study was conducted among adult enrollees in a large health maintenance organization (Kaiser Permanente Medical Care Program) in San Diego, California. The protocol was approved by the institutional review boards of Emory University and Kaiser Permanente and by the office of Human Research Protection, Department of Health and Human Services (formerly Office of Protection from Research Risks, National Institutes of Health).
Each year >50 000 adult plan members underwent a standardized health evaluation. In a typical 4-year period in the 1990s, >80% of continuously enrolled members obtained this service, which included a detailed history, psychosocial evaluation, lab studies, and physical examination. Within 2 weeks after this evaluation, participants received a mailed ACE questionnaire that assessed self-reported exposure to childhood abuse or household dysfunction.
Data were collected in 2 survey waves, August to November 1995 and January to March 1996 (wave 1) and June to October 1997 (wave 2). The overall response rate was 68% (wave 1: 70%; wave 2: 65%). Of the 17 421 respondents, 9420 (54.1%) were female. After excluding women who denied having had sexual intercourse (n = 208) and who had missing data on race or education (n = 53), our sample included 9159 women.
Definitions of ACEs, Adolescent Pregnancy, Fetal Death, and Psychosocial Consequences
All measures of ACEs were ascertained by the ACE questionnaire and pertained to experiences during the respondent’s first 18 years of life (Table 1). Each category of abuse and dysfunction has been described in detail.34–42 Questions regarding verbal and physical abuse, intimate partner violence,43 contact sexual abuse during childhood,44 and household substance abuse45 were adapted from previously used scales.
Information about adolescent pregnancy, defined as a pregnancy reported from ages 11 to 19 years, was also obtained through the ACE questionnaire. During wave 2, the question was direct: “How old were you the first time you became pregnant?” During wave 1, the occurrence of a first pregnancy during adolescence was estimated from questions measuring the age when first pregnancy ended. The first question was: “During what month and year did your first pregnancy end?” Age when the first pregnancy ended was computed by subtracting the participant’s date of birth from the date the first pregnancy ended. A combination of age when first pregnancy ended and outcome of first pregnancy (live birth, stillbirth/miscarriage, tubal pregnancy, or elective abortion) was used to impute the age when the participant first became pregnant. For live births, the age when the participant first became pregnant was assumed to have been 8 months before the age when pregnancy ended. For elective abortions or ectopic pregnancies, age at first pregnancy was assumed to have been 2 months before the age when the first pregnancy ended. For reporting fetal death (defined as stillbirth/miscarriage), age at first pregnancy was assumed to be 4 months before the pregnancy ended. Because both direct and estimated variables were available for wave 2, we evaluated the validity of our imputation by using data from that wave. We found that 89.2% of women from wave 2 who would have been classified by the estimated variable as having had an adolescent pregnancy were identified correctly. The assumption of pregnancy beginning 8 months before live birth, 4 months before stillbirth/miscarriage, and 2 months before abortion or ectopic pregnancy, was more accurate than alternate assumptions (such as pregnancy beginning 9 months earlier for births and 3 months earlier for abortions and ectopic pregnancies; data not shown). The frequency of adolescent pregnancy was nearly identical in the 2 waves (23.2% in wave 1 using the estimated variable and 23.8% in wave 2 using the direct variable).
Information regarding psychosocial consequences was obtained from the Kaiser Health Appraisal questionnaire. These questions defined family problems, job problems, and financial problems: “Are you now having serious or disturbing problems with your: family (yes/no), job (yes/no), financial matters?” The question used to define high stress was “Please fill in the circle that best describes your stress level (high/medium/low).” Finally, the question defining fear of uncontrollable anger was “Have you ever had reason to fear your anger getting out of control?”
Persons with incomplete information about an ACE were considered not to have had that experience. Analyses using this approach were almost identical to analyses that excluded participants with incomplete information. The unadjusted association between each of the 8 categories of ACEs and adolescent pregnancy was estimated by using relative risks (RRs) and 95% confidence intervals (CIs).46 The Mantel-Haenszel χ2 test for linear trend in proportions was used to evaluate whether the risk of adolescent pregnancy increased as the number of categories of ACEs (ACE score, classified as 0, 1, 2, 3, 4–5, 6, or 7–8) increased.47 Separate logistic regression models were used to obtain adjusted odds ratios (ORs) (as estimates of RR) and 95% CIs for the association between ACE score and adolescent pregnancy, long-term psychosocial consequences, and fetal death.46 We used the population attributable risk percent to estimate the proportion of cases of adolescent pregnancy attributable to ACEs.48 For the examination of whether adolescent pregnancy was an independent predictor of each of the long-term psychosocial sequelae, maximum-likelihood ratio tests46 demonstrated highly significant differences according to whether participants had been exposed to ACEs; therefore, the associations between adolescent pregnancy and long-term sequelae are reported separately for those with (ACE score: ≥1) and without (ACE score: 0) ACEs. Finally, we examined the association between ACEs and each of our major outcomes (adolescent pregnancy, fetal death, and psychosocial consequences) in analyses restricted to participants in the youngest birth cohort (birth years 1962–1978) and thus aged ≤34 years at interview.
Our study population tended to be ≥50 years of age at interview (62%), to have attended college (72%), and to be white (77%). We found that 66% (6015 of 9159) of participants reported exposure to ≥1 categories of childhood adversity (Table 1). Compared with women who did not report ACEs, those who reported ≥1 ACEs were significantly more likely to be <35 years of age at interview, to be black or Hispanic, and to report certain risk behaviors in adolescence, including street drug use, alcohol use, and ≥1 suicide attempts as well as having had ≥5 sex partners in their lifetimes (Table 2).
Exposure to each of the 8 types of ACEs was associated with an increased risk of adolescent pregnancy (Table 3). Compared with women who did not report the specific ACEs, RRs showed increases in adolescent pregnancy: 90% for those who, as children, had an incarcerated family member; 60% for those who experienced childhood sexual abuse; 60% for those reporting childhood emotional abuse; 50% for those reporting physical abuse; 60% for those reporting household substance abuse; 20% for those who lived with a mentally ill family member during childhood; 60% for those exposed to intimate partner violence; and 60% for those whose parents were divorced or separated.
After adjustment for age at interview, education, and race, ORs for adolescent pregnancy increased significantly and incrementally as exposure (the ACE score) increased (Table 4). The adjusted ORs for adolescent pregnancy were 1.0, 1.4, 1.8, 2.2, 2.4, 3.2, 3.4, and 5.5 among those with 0, 1, 2, 3, 4, 5, 6, and 7 to 8 ACEs, respectively. Whereas 16.0% of those who experienced no ACEs reported their first pregnancy occurring during adolescence, this risk rose to 53.0% among women who experienced 7 to 8 ACEs. Our findings regarding the dose-gradient effect of ACEs on adolescent pregnancy were essentially unchanged by adjusted analyses restricted to wave 2 (wave 2 included expanded behavioral data), which allowed us to control for additional potential confounders, including multiple sex partners, drinking during adolescence, drug use during adolescence, daughter of an adolescent mother, and suicide attempt during adolescence. If one assumes that exposure to ACEs causes adolescent pregnancy, then ACEs lead to a population attributable risk percent of 33%, indicating that one third of adolescent pregnancies could be prevented by eliminating exposure to ACEs.
We observed that women who were exposed during childhood to even modest levels of ACEs (ACE score: 1–2) were significantly more likely to report at interview (mean age at interview: 55.8 years) that they were having current “serious or disturbing” family problems, job problems, financial problems, or high levels of stress or that they had experienced fear of uncontrollable anger (Table 5). As exposure to ACEs rose from zero to 1 to 2, 3 to 4, and ≥5, the adjusted ORs, respectively, for serious family problems (1.0, 1.5, 2.2, and 3.3), job problems (1.0, 1.4, 2.3, and 2.9), financial problems (1.0, 1.6, 2.3, and 2.4), high stress (1.0, 1.4, 1.9, and 2.2), or uncontrollable anger (1.0, 1.6, 2.8, and 4.5) also increased steadily and consistently. In contrast, adolescent pregnancy was not associated with even 1 of these long-term outcomes among women who did not experience ACEs (ACE score: 0). For those exposed to childhood adversity (ACE score: ≥1), adolescent pregnancy was associated with modest increases in family and financial problems and in high stress and fear of uncontrollable anger.
Among women who had ever been pregnant, a strong and independent trend was observed for the association between ACE score and fetal death as the outcome of the first pregnancy (Table 6). This risk increased from 6.9% to 7.6%, 8.7%, and 10.2%, respectively, for those with exposure to 0, 1 to 2, 3 to 4, or ≥5 ACEs. Among women who had a second pregnancy (n = 6789), the risk of fetal death as the outcome of the second pregnancy increased from 7.0% for those with no exposure to ACEs to 8.6%, 9.9%, and 10.3%, respectively, for those with 1 to 2, 3 to 4, and ≥5 types of ACEs. Women whose first pregnancy occurred during adolescence, however, did not have an elevated risk of fetal death (Table 6).
Finally, in analyses restricted to women aged ≤34 at interview (11% of our sample), we found that the significant trend persisted for the association between ACEs and adolescent pregnancy (occurring in 8.7%, 18.7%, 18,2%, 26.1%, 29.3%, 31.8%, 34.9%, and 51.9% of those with an ACE score of 0, 1, 2, 3, 4, 5, 6, and 7–8, respectively; P for trend < .0001). As the ACE score rose in these younger women, the risk of fetal death as the outcome of the first pregnancy increased (occurring in 4.8%, 5.7%, 6.0%, and 11.3% of those with 0, 1–2, 3–4, and ≥5 ACEs, respectively; P for trend < .05). Each of the psychosocial consequences considered in this age group also increased incrementally as the ACE score increased from zero to 1 to 2, 3 to 4, and ≥5 (family problems: 4.3%, 9.3%, 12.4%, and 24.8% [P for trend < .0001]; job problems: 7.2%, 10.0%, 14.5%, and 22.6% [P for trend < .0001]; financial problems: 12.0%, 22.4%, 27.4%, and 27.1% [P for trend < .001]; difficulty controlling anger: 4.8%, 8.7%, 11.5%, and 19.6% [P for trend < .0001]; and current high stress: 12.0%, 17.0%, 27.8%, and 27.1% [P < .0001]).
We found a strong dose-response link between the number of ACEs and teen pregnancy: As the ACE score increased, the risk of adolescent pregnancy increased incrementally. This link is more potent than the relationships found for many other previously identified risk factors and accounts for a sizable fraction of the teen pregnancies, assuming that the relationship is causal. Incremental increases in the ACE score were associated with enduring consequences reported decades after childhood, at a mean age of 56 years. A sizeable proportion of the adult participants who, as children, suffered ACEs, years later described their lives as beset by high stress, uncontrollable anger, and serious or disturbing problems with their families, jobs, and finances. We found that these adverse psychosocial sequelae so commonly thought to be related to the adolescent pregnancy actually seem to be due to being reared in families with difficult ACEs. When the family environment did not have ACEs, becoming pregnant as an adolescent did not raise the likelihood of these long-term, negative, psychosocial consequences.
We also found that as the ACE score rose, the risk of fetal death after both the first and second pregnancies increased. In contrast to previous reports,8–10,13,49 however, adolescent pregnancy itself was not associated with an increased risk of fetal death. The strength, consistency, and gradient character of the associations between ACEs and adolescent pregnancy, long-term psychosocial sequelae, and fetal death suggest that ACEs may be causal risk factors for each of these unfavorable outcomes.
ACEs were shown previously to increase the frequency of high-risk sexual behaviors that are associated with adolescent pregnancy, such as early onset of intercourse and high numbers of sexual partners.34 Engagement in early, unprotected sexual activity that leads to adolescent pregnancy may represent a misguided attempt to achieve the interpersonal connectedness, support, and hope50 that may have been lacking in childhood among those exposed to ACEs.51–56 Although reports have demonstrated that several of the individual ACEs, such as sexual abuse and physical abuse, increase the risk of adolescent pregnancy,26,30,32 few reports have addressed the cumulative impact of exposure to multiple types of adversity during childhood57,58 on adolescent pregnancy. This limitation is salient, because abuse rarely takes place as an isolated event but typically occurs within a broader social context of multiple adversities including family dysfunction and disconnectedness.42
Previous reports have suggested that long-term psychosocial disadvantages are common and costly consequences of teen pregnancy.5,6 However, findings of such studies may be biased by their failure to adequately account for differences in the family environments of those teens who become pregnant and those teens who delay pregnancy and childbearing. In contrast to reports not accounting for familial variation, a study comparing teens who gave birth with their sisters who delayed childbearing found that family background accounted for many of the disadvantages commonly attributed to adolescent pregnancy per se.59,60 Furthermore, a study comparing teens from similar environments who gave birth with those who delayed childbearing found no long-term psychosocial consequences of early childbearing.15 In the present study, we found no evidence to support an association between adolescent pregnancy and long-term psychosocial consequences (high stress, uncontrollable anger, and serious or disturbing problems with their families, jobs, and finances) in women who had not suffered ACEs.
An increase in adverse pregnancy outcomes including fetal death, prematurity, and low birth weight has been observed more frequently in studies comparing younger teen mothers (≤17 years) with mothers aged 20 to 34.6,8–10,14,49,61,62 We found no increased risk of fetal death among adolescent mothers even in analyses comparing young adolescent mothers with mothers aged 20 to 34 years (data not shown). Both behavioral and physiologic mechanisms may be operative in increasing the risk of fetal death among those exposed to ACEs. Those who, as children, experience adversity and dysfunction are known to be at greater risk for alcohol abuse,41 smoking,36 and sexually transmitted diseases,37 all of which may increase the risk of fetal loss.11,49,63–66 Prolonged exposure to deficient or dysfunctional environments during childhood may lead to long-term physiologic alterations including neuroendocrine disruption and abnormalities in immunologic regulation.55,56,67 Our findings suggest that the cumulative and chronic exposure to stress associated with ACEs during childhood led to an increased risk of nonviability for the infant, persisting through both the first and second pregnancies.
We considered limitations that may have biased our findings. Because exposure to ACEs was measured by self-report, it is likely that these childhood exposures were underreported. Any such underreporting, however, would likely have led to underestimation of the strength of association between ACEs and adolescent pregnancy. Similarly, our classifying persons with incomplete information about an ACE as having not had that experience would lead to a conservative estimate of the relationship between ACEs and adolescent pregnancy. Because information was not available regarding the time exposure to ACEs began, we were unable to evaluate the number of participants who had their initial exposure to ACEs during adolescence, possibly at the same time as the teen pregnancy. However, because the focus of most questions regarding ACEs was on events occurring during childhood, it is probable that the majority of cases of reported adolescent pregnancies followed, rather than preceded, the onset of exposure to various categories of ACEs. Future studies would be strengthened by measuring the age during childhood when exposure to each ACE first occurred and by including in their study population women with more variable sociodemographic characteristics. Although the generalizability of our findings cannot be guaranteed, their persisting significance for the youngest birth cohort suggests that the observed impact of ACEs is not a historical phenomenon that disappeared along with known changes in sexual behaviors and values that occurred during the latter part of the 20th century.
In light of the magnitude of the adolescent pregnancy problem and its consequences, the National Campaign to Prevent Teen Pregnancy proposed to reduce adolescent pregnancy rates by 30% by 2005, with programs strategically focused on youth development, changing sexual practices, or a combination of the two.3,33 Similarly, Healthy People 2010 includes as one of its primary national objectives the lowering of pregnancy rates for 15- to 17-year-olds by ∼35% by 2010.68 From our findings, it seems that the proposed reductions in teen pregnancy could be achieved by including programs successful in preventing or ameliorating family dysfunction. Olds et al69,70 demonstrated that interventions directed toward preventing exposure to familial violence and household dysfunction by having public health nurses make home visits to high-risk families during the early years of a child’s life can be successful. Although explanations for the success of youth development programs in preventing adolescent pregnancy have been somewhat elusive,3,33 it is reasonable that successful programs may have components that help ameliorate the detrimental effects of prolonged exposure to family dysfunction. Effective youth development programs seem to build competence and confidence by promoting supportive relationships with peers and mentors; by strengthening education, decision-making, and autonomy; and by providing opportunities for community service.3,31,33 Programs that successfully ameliorate the effects of family dysfunction may have the potential to prevent not only adolescent pregnancy but also fetal death and long-term psychosocial sequelae.
The essential contribution of family connectedness to promoting adolescent health was demonstrated clearly and convincingly at the end of the millennium, largely through the National Longitudinal Study on Adolescent Health.50,71,72 Findings from this national survey of >90 000 students indicated that for adolescents, family connectedness was protective against emotional distress, suicidal thoughts and behaviors, violence, cigarette use, alcohol use, marijuana use, and young age at sexual debut.50 Our analysis supports and extends previous reports by demonstrating that family dysfunction has enduring and unfavorable health consequences for women during the adolescent years, the childbearing years, and beyond. The implications of our findings for public health and societal policies are profound. If we are to reduce both adolescent pregnancy and long-term negative psychosocial outcomes, we must help families avoid these most difficult personal circumstances.
“It is a wonderful idea to record a consultation. Everyone should. How can you really listen, really listen, as you sit in a strange room and dice over your own life with a doctor you have never met before?”
Gearin-Tosh M. Living Proof. Scribner; 2002
Submitted by Student
- Received February 7, 2003.
- Accepted May 14, 2003.
- Address correspondence to Susan D. Hillis, PhD, MS, Mailstop K-34, Centers for Disease Control and Prevention, 4770 Buford Hwy NE, Atlanta, GA 30341-3724. E-mail:
- ↵Ventura SJ, Mathews TJ, Hamilton BE. Births to teenagers in the United States, 1940–2000. Natl Vital Stat Rep.2001;49(10) :1– 23
- ↵Kirby D. Emerging Answers: Research Findings on Programs to Reduce Teen Pregnancy. Washington, DC: National Campaign to Prevent Teen Pregnancy; 2001
- ↵McElroy SW, Moore KA. Trends over time in teenage pregnancy and childbearing: the critical changes. In Maynard RA, Ed. Kids Having Kids: Economic Costs and Social Consequences of Teen Pregnancy. Washington, DC: Urban Institute Press; 1997:23–54
- ↵Maynard RA. The study, the context, and the findings in brief. In Maynard RA, Ed. Kids Having Kids: Economic Costs and Social Consequences of Teen Pregnancy. Washington, DC: Urban Institute Press; 1997:1–22
- ↵Moore DA, Morrison DR, Greene AD. Effects on the children born to adolescent mothers. In Maynard RA, Ed. Kids Having Kids: Economic Costs and Social Consequences of Teen Pregnancy. Washington, DC: Urban Institute Press; 1997:145–180
- Strobino DM, Ensminger ME, Kim YJ, Nanda J. Mechanisms for maternal age differences in birth weight. Am J Epidemiol.1995;142 :504– 514
- Grogger J. Incarceration-related costs of early childbearing. Maynard RA, Ed. Kids Having Kids: Economic Costs and Social Consequences of Teen Pregnancy. Washington, DC: Urban Institute Press; 1997:231–256
- ↵Hotz VJ, McElroy SW, Sanders SG. The impacts of teenage childbearing on the mothers and the consequences of those impacts for government. Maynard RA, Ed. Kids Having Kids: Economic Costs and Social Consequences of Teen Pregnancy. Washington, DC: Urban Institute Press; 1997:55–94
- Ku L, Sonenstein FL, Pleck JH. Neighborhood, family, and work: influences on the premarital behaviors of adolescent males. Soc Forces.1993;72 :479– 503
- Moore KA, Blumenthal C, Sugland BW, Hyatt B, Snyder NO, Morrison DR. State Variation in Rates of Adolescent Pregnancy and Childbearing. Washington, DC: Child Trends, Inc; 1994
- Bearman PS, Bruckner H. Power in Numbers: Peer Effects on Adolescent Girls’ Sexual Debut and Pregnancy. Washington, DC: National Campaign to Prevent Teen Pregnancy; 1999
- Alan Guttmacher Institute. Sex and America’s Teenagers. New York, NY: Alan Guttmacher Institute; 1994
- Smith CA. The link between childhood maltreatment and teenage pregnancy. Soc Work Res.1996;20 :131– 141
- Springarn RW, DuRant RH. Male adolescents involved in pregnancy: associated health risk and problem behaviors. Pediatrics.1996;98 :262– 268
- ↵Hillis SD, Anda RF, Felitti VJ, Nordenberg D, Marchbanks PA. Adverse childhood experiences and sexually transmitted diseases in men and women, a retrospective study. Pediatrics.2000;106(1) . Available at: http://www.pediatrics.org/cgi/content/full/106/1/e11
- Anda RF, Felitti VJ, Chapman DP, et al. Abused boys, battered mothers, and male involvement in teen pregnancy: new insights for pediatricians. Pediatrics.2001:107(2) . Available at: http://pediatrics.aappublications.org/cgi/content/full/107/2/e19
- ↵Straus M, Gelles RJ. Physical Violence in American Families: Risk Factors and Adaptations to Violence in 8,145 Families. New Brunswick, NJ: Transaction Press; 1990
- ↵Kleinbaum DK, Kupper LL, Morgensern H. Epidemiologic Research. New York, NY: Van Nostrand Reinhold Company; 1982:255–257
- ↵Fleiss JL. Statistical Methods for Rates and Proportions. New York, NY: John Wiley & Sons; 1981:143–146
- ↵Hennekens CH, Curing JE. Epidemiology in Medicine. Boston, MA: Little, Brown, and Company; 1987
- ↵Resnick MD, Bearman PS, Blum RW, et al. Protecting adolescents from harm: findings from the National Longitudinal Study on Adolescent Health. JAMA.1994;278 :823– 832
- ↵Strasburger VC, Donnerstein E. Children, adolescents, and the media: issues and solutions. Pediatrics.1999;103 :129– 139
- Seligman ME. Learned Optimism. New York, NY: Knopf; 1991
- Institute of Medicine. Conclusions and recommendations. In: Brown SS, Eisenberg L, Eds. The Best Intentions: Unintended Pregnancy and the Well-Being of Children and Families. Washington, DC: National Academy Press; 1995:250–270
- ↵Cousins N. Head First: The Biology of Hope. New York, NY: E. P. Dutton; 1989
- ↵Diamond M, Hopson J. Magic Trees of the Mind. New York, NY: Plume; 1998
- ↵Geronimus AT, Korenman S. Maternal youth or family background? On the health disadvantages of infants with teenage mothers. Am J Epidemiol.1993;137 :213– 225
- ↵Geronimus A, Korenman S. The socioeconomic consequences of teen childbearing reconsidered. Q J Econ.1992;107 :1187– 1214
- ↵Winbo I. Smoking, alcohol, and reproduction. Am J Public Health.1983;73 :115406
- ↵US Department of Health and Human Services. Healthy People 2010: Understanding and Improving Health. 2nd Ed. Washington, DC: US Government Printing Office; 2000
- Copyright © 2004 by the American Academy of Pediatrics