10 slide power point

EMPIRICAL RESEARCH

Peer and Individual Risk Factors in Adolescence Explaining the Relationship Between Girls’ Pubertal Timing and Teenage Childbearing

C. Emily Hendrick1 • Jessica Duncan Cance1 • Julie Maslowsky1

Received: 17 October 2015 / Accepted: 4 January 2016 / Published online: 14 January 2016

� Springer Science+Business Media New York 2016

Abstract Girls with early pubertal timing are at elevated

risk for teenage childbearing; however, the modifiable

mechanisms driving this relationship are not well under-

stood. The objective of the current study was to determine

whether substance use, perceived peer substance use, and

older first sexual partners mediate the relationships among

girls’ pubertal timing, sexual debut, and teenage child-

bearing. Data are from Waves 1—15 of the female cohort

of the National Longitudinal Surveys of Youth 1997

(NLSY97), a nationwide, ongoing cohort study of U.S.

men and women born between 1980 and 1984. The analytic

sample (n = 2066) was 12–14 years old in 1997 and eth-

nically diverse (51 % white, 27 % black, 22 % Latina).

Using structural equation modeling, we found substance

use in early adolescence and perceived peer substance use

each partially mediated the relationships among girls’

pubertal timing, sexual debut, and teenage childbearing.

Our findings suggest early substance use behavior as one

modifiable mechanism to be targeted by interventions

aimed at preventing teenage childbearing among early

developing girls.

Keywords Pubertal timing � Female adolescents � Teen pregnancy � Teen childbearing � Substance use

Introduction

Teenage childbearing is associated with deleterious out-

comes across the life span for both mother and child. In

addition to increased risk for adverse birth outcomes (Chen

et al. 2007), teen mothers (Patel and Sen 2012) and their

children (Jaffee et al. 2001) are at elevated risk for phys-

ical, emotional, and social adversity throughout their lives.

Despite declines in teenage childbearing in recent decades

(The Office of Adolescent Health 2015), the U.S. still has

one of the highest teen birth rates in the developed world

(United Nations Statistics Division 2014), demonstrating a

critical need to better understand the mechanisms by which

this disparity occurs and the necessary areas for interven-

tion among adolescents.

A number of socio-ecological factors in childhood and

adolescence are associated with girls’ increased risk for

teenage pregnancy and childbearing. Some of the most

salient socio-ecological predictors of teenage pregnancy

and childbearing in the U.S. include minority race or

ethnicity (Meade et al. 2008) and childhood family

characteristics such as lower maternal education, being

born to a teen mother, and father absence from the

household (Driscoll and Abma 2015). It is not assumed

that these socio-ecological factors are directly causal of

teenage childbearing. Rather, they are easily measured

characteristics commonly associated with environmental

conditions that may place girls at increased risk for

teenage pregnancy and childbearing such as low levels of

parental monitoring, lower economic and educational

prospects, and family, peer, and community norms

accepting of early childbearing (Meade et al. 2008). It is

important that research examining the precursors of

teenage childbearing take into account these socio-eco-

logical factors.

& Jessica Duncan Cance jdcance@austin.utexas.edu

1 Health Behavior and Health Education, Department of

Kinesiology and Health Education, The University of Texas

at Austin, 2109 San Jacinto Blvd., UT Mail Code: D3700,

Austin, TX 78712-1415, USA

123

J Youth Adolescence (2016) 45:916–927

DOI 10.1007/s10964-016-0413-6

One notable population at risk for teenage pregnancy and

childbearing consists of girls who experience pubertal

development earlier than their same-age peers (Deardorff

et al. 2005). Two primary psychosocial mechanisms have

been hypothesized to explain why early pubertal timing may

place girls at elevated risk for earlier sexual debut and teenage

childbearing, the developmental readiness and maturational

deviance hypotheses. Per the developmental readiness

hypothesis (Mendle et al. 2007), because early developing

girls physically appear older than same-age peers, they may

be accepted into older, more deviant peer groups. Conse-

quently, they are more likely to be exposed to and initiate risk

behaviors such as substance use (Negriff and Trickett 2012)

and risky sexual behavior (Savolainen et al. 2015) at earlier

agesthan their same-age peers. Althoughthese girlsmay be as

physically developed as girls 1–4 years their seniors, their

cognitive and emotional developmental levels are more

similar to their same-age peers. Hence, they may not have

developed the decision-making capabilities necessary to

successfully navigate the social situations they experience

within older, more deviant peer groups. Relatedly, according

to the maturational deviance hypothesis, off-timed pubertal

development is posited to create additional stress due to the

experience of physical maturational deviance from one’s

peers (Brooks-Gunn et al. 1985). Consequently, early devel-

oping girls’ heightened risk for behaviors leading to teenage

childbearing (such as early substance use and risky sexual

behavior) may reflect maladaptive strategies for coping with

additional stress they experience due to maturational

deviance. The developmental readiness and maturational

deviance hypotheses are not mutually exclusive mechanisms

and both have been shown to explain the link between early

pubertaldevelopment andriskbehavior inadolescence.There

is a dearth of research elucidating the modifiable mechanisms

by which girls’ early pubertal timing leads to teenage child-

bearing, but both the developmental readiness and matura-

tional deviance hypotheses point to the need to explore the

role of substance use and peer influence.

Early developers are more likely to engage in substance

use throughout adolescence compared with their peers

(Cance et al. 2013). In turn, substance use in adolescence is

associated with risky sexual behaviors and teen pregnancy

(Salas-Wright et al. 2015). The relationship between sub-

stance use and riskier sexual behaviors in adolescence is

often attributed to the direct influence of alcohol and illicit

drugs on impairing sexual decision-making (Townshend

et al. 2014). However, tobacco, alcohol, and illicit drug use in

adolescence are each associated with earlier sexual debut

(Kirby 2002). This suggests that substance use in adoles-

cence could also be a proxy for ‘‘risk-prone’’ adolescents

who tend to engage in multiple types of risk behaviors across

adolescence (Crockett et al. 2006). While research examin-

ing the relationship between girls’ pubertal timing and

teenage pregnancy and childbearing to date has not directly

assessed the influence of older and more deviant peers, recent

studies that have assessed the influence of peers on early

developing girls’ risk behaviors in adolescence have pro-

duced mixed findings. Some find older and more deviant

peers draw early developing girls into risk behaviors in that

exposure to these peers leads to earlier sexual debut (Savo-

lainen et al. 2015) and increased use of alcohol and marijuana

(Negriff and Trickett 2012). Others find no mediating peer

effects (Negriff et al. 2011) or mediating effects only among

subsamples of girls (Negriff et al. 2015). However, percep-

tion of peer deviance (e.g., substance use) is associated with a

greater likelihood of personal deviance (Deutsch et al. 2015).

It is, therefore, difficult to disentangle the influence of peers

on early developing girls’ risk behaviors in adolescence that

may lead to teenage pregnancy and childbearing. Presently,

pubertal timing is not a modifiable target for teenage child-

bearing prevention programs, but it is possible to target both

substance use and peer influence and thus there is a need to

explore the potential role of these factors.

A seminal retrospective study conducted by Deardorff et al.

(2005) examined the associations between age at menarche,

sexual debut, alcohol initiation, and first pregnancy within a

sample of ethnically diverse women from Arizona. Earlier age

at initiation of alcohol use and earlier sexual debut together

explainedthe relationshipbetweenearlypubertaldevelopment

and age at first pregnancy. This study marked an essential first

step in identifying potential points of intervention to reduce

early developing girls’ risk for teen pregnancy. However, the

study was retrospective in design, thereby limiting the

researchers’ ability to control for confounding socio-ecologi-

cal factors in childhood other than race and ethnicity. They

were also unable to assesstheimpact of olderandmoredeviant

peers in adolescence who are assumed to influence early

developing girls’ substance using behaviors and initiation of

sex per the developmental readiness hypothesis. Further, this

study assessed women’s age at first pregnancy as the outcome

of interest rather than age at first childbirth. Approximately

40 % of teen pregnancies in the U.S. do notresult in a live birth

(KostandHenshaw 2014) andthe social and health trajectories

of girls who become pregnant in adolescence but do not

become mothers differ from those who become adolescent

mothers (Patel and Sen 2012). Thus, there is a need to extend

the work of Deardorff and her colleagues.

The Current Study

Theaimofthecurrentstudyistoinformtheunderstandingofthe

modifiable mechanisms of the association between girls’

pubertal timing and teenage childbearing and the salient areas

forinterventiontoreduceearlydevelopinggirls’riskforteenage

childbearing. To the authors’ knowledge, this is the first study to

J Youth Adolescence (2016) 45:916–927 917

123

assess the influence of individual and peer risk factors in ado-

lescence on the relationship between girls’ pubertal timing and

teenage childbearing. The current study advances knowledge

regarding girls’ pubertal timing and teen childbearing by

drawing from an ethnically diverse, nationwide sample of U.S.

adolescents from a prospective, longitudinal study design.

Further, we explore both individual and peer mediators of the

relationship between girls’ pubertal timing, age at sexual debut,

and teenage childbearing while controlling for potentially con-

founding socio-ecological factors (see Fig. 1).

Drawing from the developmental readiness hypothesis,

the maturational deviance hypothesis, and the posited

influence of substance use in early adolescence on sexual

risk behaviors, we predicted that early substance use, per-

ceptions of peers’ substance use, and having an older first

sexual partner would each partially mediate the relation-

ship between girls’ pubertal timing and their age at sexual

debut. We further hypothesized that girls’ age at sexual

debut would mediate the relationship between girls’ risk

factors in adolescence and teenage childbearing.

Methods

Participants

Data are from female participants in the National Longitudi-

nal Surveys of Youth 1997 (NLSY97), a nationwide, ongoing

cohort study of U.S. men and women born between 1980 and

1984 (U.S. Bureau of Labor Statistics 2014). At Wave 1

(1997), participants were between the ages of 12 and 18.

Participants turned 20 years of age between 2000 and 2005

and continue to be interviewed annually. The final analytic

sample for the current study included 2066 participants of the

original female cohort (n = 4385). Only participants ages

12–14 at Wave 1 were included in analyses as they provided

data regarding their own and their peers’ substance using

behaviors in early adolescence—prior to reporting on sexual

behaviors and partners, which begins in the NLSY97 when

participants reach age 15 (see Fig. 1). Similarly, participants

reporting their age at first sex to be prior to Wave 1 or prior to

menarche were excluded from analyses. Childhood sexual

abuse may prompt early pubertal development among girls

(Zabin et al. 2005) and can lead to psychopathology in ado-

lescence, including substance abuse (Kilpatrick et al. 2000).

Thus, the temporality of the relationship between sexual

debut, early pubertal timing, and early adolescent substance

use could not be determined for participants reporting an age

of sexual debut prior to menarche or Wave 1 (n = 143). In

order to maintain the hypothesized temporality, these girls

were not included in the present study. Further, those missing

information on the key predictor of interest (menarche,

n = 10) and outcome of interest (teenage childbearing,

n = 91) were excluded from analyses. Those excluded were

older at Wave 1, more likely to have mothers who attended

college, more likely to have experienced father absence in

Birth before age 20Age at first sex

Early pubertal timing

Older first sex partner

Perceived peer substance use

Substance use

Covariates: • Age at Wave 1 • Race/ethnicity • Maternal

educa�onal a�ainment

• Maternal age at first birth

• Father absence in youth

+

-

+ +

+ + -

-

-

-

Peer smoking Peer drinking Peer illicit drug use

Repor�ng at Wave 1 ages 12-14

Repor�ng at Waves 2+ ages 15+

Repor�ng at Waves 1+ Early menarche age <11.08

+

+

+

+

Fig. 1 Analytic model of peer and individual factors in adolescence explaining the relationship between girls’ pubertal timing and teenage childbearing. All paths control for all covariates. ?, - indicate hypothesized positive (?) or negative (-) direct pathway

918 J Youth Adolescence (2016) 45:916–927

123

youth,reportedmorepeerand personalsubstance use atWave

1, were younger at sexual debut, and were more likely to

experience early menarche and teen childbearing. They did

not differ by other background characteristics (race/ethnicity,

maternal high school graduation, and maternal age at first

birth). Due to the analytic methods employed and because the

analytic sample consisted of a sub-sample of the original

female cohort, we did not use NLSY sampling weights per the

NLSY97 sample weighting recommendations (U.S. Bureau

of Labor Statistics 2014). The present study was deemed not

to be human subjects research by the Institutional Review

Board of the sponsoring university.

Measures

Pubertal Timing

Participants were asked the month and year they had

reached menarche beginning at Wave 1. Participants that

did not know their month or year of menarche reported age

at menarche. Participants that had not reached menarche by

Wave 1 were asked their status of menarche (yes/no) and

month/year or age at menarche in subsequent waves until

menarche was reported. Early menarche was treated as a

dichotomous variable based on prior research with the

NLSY (Walker 2001) and the distribution in ages at

menarche among our sample (M = 12.38, SD = 1.30):

early menarche (\1 standard deviation below the mean, \11.08 years; 14.0 %) versus non-early menarche (11.08 years and older).

Teenage Childbearing

Participants’ month and year of birth was subtracted from

the month and year of the birth of their first child to cal-

culate participants’ age at first birth. Age at first birth was

dichotomized for analyses into having had a birth before

age 20 (15.4 %) versus having their first birth at age 20 or

later. As previous research identifying teen mothers to be at

increased risk for adversity has defined teenage child-

bearing in several ways, we conducted sensitivity analyses

with teenage childbearing defined as have a first birth at

\17 years of age,\18 years of age, and\19 years of age. Results for all sensitivity analysis models were similar to

those for teenage childbearing defined as having a first

birth before age 20.

Age at First Sex

Participants began reporting age at first sex at age 15

(Waves 2 and beyond), which was treated as a continuous

variable in analyses (M = 16.25, SD = 1.80).

Substance Use at Ages 12–14

At Wave 1, participants answered yes (1) or no (0) to three

items concerning their lifetime use of cigarettes (‘‘Have you

ever smoked a cigarette?’’), alcohol (‘‘Have you ever had a

drink of an alcoholic beverage? By a drink we mean a can or

bottle of beer, a glassofwine, a mixed drink, or a shot ofliquor.

Do not include childhood sips that you might have had from an

older person’s drink.’’), and marijuana (‘‘Have you ever used

marijuana, for example: grass or pot, in your lifetime?’’)

(Moore et al. 1999). The NLSY97 Substance Use Index is a

sumof the positive responses (M = 0.62, SD = 0.93) andwas

treated as a continuous variable in analyses.

Perceived Peer Substance Use at Ages 12–14

At Wave 1, participants reported the percentage of their

same grade peers that smoke cigarettes, get drunk at least

once a month, and have ever used marijuana, inhalants, or

other drugs. Answer choices were: 1 = almost none,

2 = about 25 %, 3 = about half (50 %), 4 = about 75 %,

and 5 = almost all (more than 90 %). Responses to these

three items were used to create a continuous latent variable.

Older First Sexual Partner

When participants reported age at first sex, they also

reported the age of their first sexual partner. After removing

extreme outliers (more than 3 standard deviations above or

below the mean age difference), we computed the age dif-

ference between the participant and her first sexual partner

(M = 1.73, SD = 2.03). We then dichotomized the variable

so girls with first sexual partners with an age difference over

1 standard deviation above the mean age difference

(3.76 years) were considered to have ‘‘older’’ first sexual

partners (15.2 %). In research exploring the role of the age

difference of romantic partners in adolescence, age differ-

ence is sometimes considered as a continuous variable

(Volpe et al. 2013) and sometimes as categorical—often

with 2 years in age difference signaling an ‘‘older’’ partner

(Halpern et al. 2006). As such, we conducted sensitivity

analyses with the age difference treated continuously and

with 2 years difference signaling an older partner. Neither

variation significantly altered results.

Covariates

Background factors available in the dataset that have

demonstrated to be associated with both girls’ pubertal

timing and their risk for teenage childbearing were chosen

as covariates for multivariate models: race/ethnicity,

mother’s highest level of education, whether the partici-

pant’s mother had her first birth before age 20, and father

J Youth Adolescence (2016) 45:916–927 919

123

absence during youth (prior to or at Wave 1). Participant

age at Wave 1 was also included as a covariate due the

strong association between older age and increased risk for

lifetime substance use in adolescence.

Statistical Analysis

Structural equation modeling was conducted using MPlus

Version 7.31 (Muthen and Muthen 2012) to assess the

hypothesized pathways between early menarche and teen-

age childbearing. The Weighted Least Squares with Means

and Variance Adjusted (WLSMV) estimator was used as it

has been shown to be a robust estimator appropriate for

models including binary dependent variables (Muthen and

Muthen 2012). Unstandardized coefficients (b) and stan-

dard errors (SE) are reported throughout. Using the

WLSMV estimator, the unstandardized coefficients are

modeled as probit regression coefficients. A positive

coefficient indicates that an increase in the predictor is

associated with an increase in the predicted probability of

the outcome. A negative coefficient indicates that an

increase in the predictor is associated with a decrease in the

predicted probability of the outcome.

Per our hypotheses, we modeled a direct path between age

at first sex and teenage childbearing, while the direct path

between early menarche and teenage childbearing was fixed

to zero per our prediction that early menarche would lead to

teenage childbearing through one’s own and peers’ sub-

stance use, having an older first sex partner, and age at sexual

debut. We tested a total of nine mediation pathways. Par-

ticipants’ substance use at ages 12–14, perceived peer sub-

stance use at ages 12–14, and having an older first sex partner

were included in the model as mediators of the relationship

between early menarche and age at first sex (mediation

pathways 1–3). Age at sexual debut was considered in the

model as a mediator of the relationships between partici-

pants’ substance use at ages 12–14, perceived peer substance

use at ages 12–14, having an older first sex partner and

teenage childbearing (mediation pathways 4–6). Finally, we

tested the compound mediation pathways connecting early

menarche to teenage childbearing (e.g., the relationship

between early menarche and teenage childbearing through

participants’ substance use at ages 12–14 and age at sexual

debut; mediationpathways 7–9). Participants’ substance use,

perceived peer substance use, and having an older first sex

partner were all correlated. All pathways controlled for all

covariates (see Fig. 1). The significance of the hypothesized

mediational pathways were computed using RMediation, a

web-based R package which produces confidence intervals

for indirect parameter coefficients and is robust with con-

tinuous, categorical, and dichotomous variables (Tofighi and

MacKinnon 2011).

Results

Descriptive Statistics

Girls with early menarche were, on average, 1.3 years

younger at sexual debut and 6 months younger at their first

birth than girls with on-time or later age at menarche (see

Table 1). As expected, early menarche was also associated

with greater substance use and perceived peer substance

use at ages 12–14. Non-Hispanic black and Hispanic girls

were more likely to experience early menarche compared

with non-Hispanic white girls. Neither participants’

mothers’ age at first birth nor the age difference of their

first sexual partners were associated with early age at

menarche.

Over 15 % of participants had their first birth before age

20. Non-Hispanic black and Hispanic girls, girls whose

mothers attained less education, whose mothers were teen

mothers themselves, and girls who experienced father

absence during youth were all more likely to give birth

before age 20. Girls who gave birth before age 20 were

1.5 years younger at sexual debut, more often had an older

partner at sexual debut, and had greater substance use and

perceived peer substance use at ages 12–14.

Mediation Model

The model explained 42 % of the variance in teenage

childbearing (see Fig. 2). Scholars recommend assessing a

combination of model fit indices to determine model fit (Hu

and Bentler 1999). A root mean square error of approxi-

mation (RMSEA) value of \.06, comparative fit index (CFI) and Tucker Lewis Index (TLI) values of[.95, and a weighted root mean square residual (WRMR) value of

\.90 all suggest good model fit (Schreiber et al. 2006). Per these standard cutoff criteria, our mediation model fit the

data well (RMSEA = .034, 95 % CI = .027, .042;

CFI = .997; TLI = .987; WRMR = .515). At the 95 %

confidence level, five of the nine hypothesized mediation

pathways were statistically significant, and one was mar-

ginally significant. Unstandardized coefficients for direct

effects are presented in Fig. 2, for indirect effects in

Table 2, and each are described below.

Age at First Sex

Early pubertal timing was not significantly directly asso-

ciated with age at first sex (b = -.10, SE = .06) in the

mediation model. Older age at sexual debut was directly

associated with a lower probability of teenage childbearing

(b = -.33, SE = .04, p \ .005).

920 J Youth Adolescence (2016) 45:916–927

123

Substance Use in Early Adolescence

Early menarche was positively associated with substance

use in early adolescence (b = .09, SE = .03, p \ .005),

which, in turn, was associated with a lower age at first sex

(b = -.48, SE = .05, p \ .005). Participants’ substance use in early adolescence significantly mediated the rela-

tionship between early menarche and age at first sex

Table 1 Unweighted descriptive characteristics of analytic sample and bivariate relationships with timing of menarche and teenage childbearing (n = 2066)

On time or later

menarche (B1 SD

below mean)

Early menarche

([1 SD below mean)

v2 First birth C 20 years

old

First birth

\ 20 years old

v2

n (%) n (%) n (%) n (%)

All c

1776 (86.0) 290 (14.0) 1747 (84.6) 319 (15.4)

Age at first birth 3.87 a

C20 years old 1513 (86.6) 234 (13.4)

\20 years old 263 (82.5) 56 (17.6) Race/ethnicity 51.54

b 61.31

b

Non-Hispanic white (1)

918 (91.3) 88 (8.8) 2,3

912 (90.7) 94 (9.3) 2,3

Non-hispanic black (2)

423 (79.1) 112 (21.0) 1

411 (76.8) 124 (23.2) 1

Hispanic (3)

362 (81.4) 83 (18.7) 1

354 (79.6) 91 (20.5) 1

Maternal education 18.04 a

80.12 b

Did not graduate HS (1)

380 (84.1) 72 (15.9) 3

336 (74.3) 116 (25.7) 2,3

HS diploma only (2)

624 (85.0) 110 (15.0) 3

611 (83.2) 123 (16.8) 1,3

CSome college (3)

743 (89.0) 92 (11.0) 1,2

773 (92.6) 62 (7.4) 1,2

Mother’s age at first birth 2.14 76.52 b

C20 years old 1226 (86.9) 185 (13.1) 1263 (89.5) 148 (10.5)

\20 years old 429 (84.3) 80 (15.7) 374 (73.5) 135 (26.5) Father absence in youth 9.93

b 55.18

b

No 1019 (88.2) 137 (11.9) 1038 (89.8) 118 (10.2)

Yes 753 (83.3) 151 (16.7) 704 (77.9) 200 (22.1)

Older first sex partner 0.08 8.57 b

No (B1 SD above mean age difference) 1055 (85.4) 180 (14.6) 1067 (86.4) 168 (13.6)

Yes ([1 SD above mean age difference) 188 (84.7) 34 (15.3) 175 (78.8) 47 (21.2)

M (se) M (se) t M (se) M (se) t

Age at wave 1 interview (years) 13.14 (0.02) 13.13 (0.05) 0.15 13.14 (0.02) 13.15 (0.04) -0.32

Age at menarche (years) 12.73 (0.02) 10.25 (0.05) 40.12 b

12.42 (0.03) 12.19 (0.07) 2.88 b

Age at first sex (years) 16.31 (0.05) 15.02 (0.12) 3.35 b

16.52 (0.05) 15.04 (0.07) 13.67 b

Age at first child (years) 21.36 (0.11) 20.81 (0.25) 1.92 a

22.85 (0.09) 17.32 (0.07) 36.25 b

Wave 1 substance use index (0–3 scale) 0.60 (0.02) 0.73 (0.06) -2.22 a

0.58 (0.02) 0.85 (0.06) -4.71 b

Wave 1 peer substance use (1–5 scales)

Currently smoke 2.22 (0.03) 2.45 (0.08) -2.92 b

2.20 (0.03) 2.54 (0.08) -4.49 b

Get drunk C 1x/month 1.67 (0.02) 1.88 (0.07) -3.03 b

1.66 (0.02) 1.92 (0.07) -4.01 b

Have used illicit drugs 1.88 (0.03) 2.11 (0.07) -3.13 b

1.85 (0.03) 2.24 (0.07) -5.47 b

HS high school a

p \ .05 b

p \ .005 c Not all groups total to 2066 due to missing data

1,2,3 Different numbered superscripts indicate significant (p \ .05) group comparisons based on pairwise v2 tests

J Youth Adolescence (2016) 45:916–927 921

123

(b = -.04, 95 % CI = -.07, -.01). Also, age at first sex

significantly mediated the relationship between substance

use in early adolescence and teenage childbearing

(b = .16, 95 % CI = .12, .21). Accordingly, substance

use in early adolescence and age at first sex together

significantly mediated the relationship between early

menarche and teenage childbearing (b = .014, 95 %

CI = .003, .024). Thus, early developing girls were more

likely to report engaging in substance use in early ado-

lescence, which was associated with a younger age at

sexual debut and thus led to a higher probability of

teenage childbearing.

Perceived Peer Substance Use

Early menarche was positively associated with perceived

peer substance use (b = .11, SE = .04, p \ .005), which, in turn, was associated with a lower age at first sex

(b = -.14, SE = .06, p \ .05). Participant substance use and perceived peer substance use in early adolescence were

positively correlated (b = .31, SE = .03, p \ .005). Per- ceived peer substance use in early adolescence significantly

mediated the relationship between early menarche and age

at first sex (b = -.016, 95 % CI = -.034, -.002) and age

at first sex significantly mediated the relationship between

Birth before age 20

Age at first sex

Early pubertal timing

Older first sex partner

Perceived peer

substance use

Substance use

-.14(.06)a

-.48(.05)b

-.33(.04)b

.11(.04)b

.09(.03)b

.18(.07)a

-.002(.06)

-.10(.06)

-.03(.05)

.02(.06)

.10(.06)

R2 = 42%

Fig. 2 Unstandardized coefficients (b) and standard errors (SE) of direct effects. Participants’ substance use at ages 12–14, perceived

peer substance use at ages 12–14, and having an older first sex partner

were correlated in the model. All paths controlled for age at Wave 1,

race/ethnicity, mother’s education, whether the participant’s mother

had her first birth before age 20, and father absence during youth

(prior to or at Wave 1). Standard errors are within parentheses.

Nonsignificant pathways are indicated with a dashed line. Significant

direct pathways are indicated with a solid line. Control variables and

correlations of mediators are not shown for simplicity. a p \ .05;

b p \ .005

Table 2 Unstandardized coefficients (b) and 95 % confidence intervals (CI) of indirect effects (b1*b2, b1*b2*b3)

Pathway b1*b2 b CI

1 Early menarche? Substance use? Age at first sex -.042 -.073, -.013

2 Early menarche? Perceived peer substance use? Age at first sex -.016 -.034, -.002

3 Early menarche? Older first sexual partner? Age at first sex .000 -.009, .008

4 Substance use? Age at first sex? First birth before age 20 .161 .115, .213

5 Perceived peer substance use? Age at first sex? First birth before age 20 .047 .009, .088

6 Older first sexual partner? Age at first sex? First birth before age 20 -.006 -.046, .034

Pathway b1*b2*b3 b CI

7 Early menarche? Substance use? Age at first sex? First birth before age 20 .014 .003, .024

8 Early menarche? Perceived peer substance use? Age at first sex? First birth before age 20 .005 .000, .011

9 Early menarche? Older first sexual partner? Age at first sex? First birth before age 20 .000 -.003, .003

922 J Youth Adolescence (2016) 45:916–927

123

perceived peer substance use in early adolescence and

teenage childbearing (b = .05, 95 % CI = .01, .09).

However, perceived peer substance use in early adoles-

cence and age at first sex together were marginally sig-

nificant in mediating the relationship between early

menarche and teenage childbearing (b = .005, 95 %

CI = .000, .011). Hence, early developing girls were more

likely to perceive their peers were engaging in substance

use, which was associated with a younger age at sexual

debut and thus led to a higher probability of teenage

childbearing, although this mediating relationship was

marginally significant.

Older First Sexual Partner

Having an older first sexual partner was positively asso-

ciated with teenage childbearing (b = .18, SE = .07,

p \ .05). However, it was not correlated with participant nor perceived substance use in early adolescence (each

b = .05, SE = .04), nor was it associated with early

menarche (b = -0.002, SE = .063) or age at sexual debut

(b = .02, SE = .06). Consequently, having an older part-

ner at sexual debut did not significantly mediate the rela-

tionships among girls’ pubertal timing, sexual debut, and

teenage childbearing (see Table 2).

Discussion

The aim of the current study was to inform the under-

standing of the modifiable mechanisms of the relationship

between girls’ pubertal timing and teenage childbearing

and the salient areas for intervention to reduce early

developing girls’ risk for teenage childbearing. Our study

drew from an ethnically diverse, nationwide sample of U.S.

adolescents from a prospective, longitudinal study design

to test whether early substance use, perceived peer sub-

stance use, or having an older first sexual partner mediate

the relationship among girls’ early pubertal timing, sexual

debut, and teenage childbearing. As such, we add to pre-

vious literature in this field in important ways. Specifically,

we extended our outcome of interest beyond adolescent

risk behaviors and teen pregnancy to teenage childbearing,

which is associated with elevated risk for physical, emo-

tional, and social adversity across the life span for women

and their children. Further, we considered salient, poten-

tially confounding background characteristics in our mul-

tivariate models so as not to overestimate the influence of

pubertal timing and substance use in adolescence on sexual

risk behaviors and risk for teenage childbearing. Finally, in

addition to considering individual risk factors in adoles-

cence, we also considered the influence of peer deviance

and older peers on the relationship between girls’ pubertal

timing and teenage childbearing, thus providing a more

nuanced test of the developmental readiness hypothesis

that informs much of the research in this field.

Age at menarche was associated with age at first

childbirth, such that early developing girls in this study had

a younger age of first childbirth. This finding extends

previous research (Deardorff et al. 2005) that found early

pubertal timing was linked to teenage pregnancy. While, on

average, early developing girls were only 6 months

younger at their first birth than their on-time and later

developing peers (20.8 years of age vs. 21.4 years of age),

the odds of being a teenage mother was 1.38 times greater

for early developing girls. Furthermore, this age difference

comes at a critical time in a woman’s life course where a

difference of 6 months could influence the attainment of

postsecondary education. Thus, future research should

explore whether teenage childbearing mediates the rela-

tionship between early pubertal development and educa-

tional attainment.

Girls with early menarche were more likely to report

substance use at ages 12–14 and lower ages at sexual debut.

And, as predicted, the relationship between early menarche

and age at first sex was explained by substance use and

perceived peer substance use in early adolescence. In other

words, girls with early pubertal timing that were not using

substances in early adolescence and perceived that their

peers were not using substances were not at elevated risk

for earlier sexual debut. Also, as predicted, girls who had

sex at earlier ages were at higher risk for teenage child-

bearing. These findings provide support for both the

developmental readiness hypothesis, which posits that

early developing girls are exposed to and engage in more

risk behaviors due to the disparity between their physical,

cognitive, and emotional development, as well as the

maturational deviance hypothesis, which proposes that the

additional stress experienced by early developing girls due

to their physical maturational deviance from their same-age

peers leads to maladaptive coping behaviors in the form of

substance use or sexual risk-taking. Our finding that sub-

stance use in early adolescence partially explains the

relationship among early menarche, early sexual debut, and

teenage childbearing adds to previous findings of early

initiation of alcohol use explaining the relationship among

early menarche, age at sexual debut, and adolescent preg-

nancy among Arizona women (Deardorff et al. 2005). This

is also consistent with the body of literature demonstrating

early initiation of and more intense substance use predict-

ing earlier sexual debut among girls (Cavazos-Rehg et al.

2011; Kirby 2002) and supports the theory that multiple

domains of risk behaviors, such as substance use and

sexual risk, tend to cluster within individuals (Crockett

et al. 2006). Thus, this result provides additional evidence

of the importance of early developing girls’ own substance

J Youth Adolescence (2016) 45:916–927 923

123

use behaviors in early adolescence as a potential inter-

vention target in preventing sexual risk behaviors in ado-

lescence and teenage motherhood.

With regards to the role of peer deviance, which to date

remains unclear in the literature (Negriff et al. 2015;

Savolainen et al. 2015), we did find evidence of a mar-

ginally significant indirect pathway between early menar-

che, perceived peer substance use in early adolescence, age

at sexual debut, and teenage childbearing. Our results

suggest that perceptions of peers’ substance use may par-

tially explain the relationship between early menarche and

teenage childbearing above and beyond that of girls’ per-

sonal substance use. However, the mediating influence of

perceived peer substance use was only marginally signifi-

cant in comparison to the influence of girls’ own substance

use. This may have been due, in part, to the measurement

of perceived peer substance use. Participants were asked

the substance using behaviors of other students in their

grade, not explicitly their ‘‘friends.’’ Thus, the associations

among peer substance use and variables of interest in the

present study may have been strengthened had participants

reported the substance using behaviors of their friends or if

participants’ friends had reported their own substance using

behaviors (Henry et al. 2011).

Contrary to the study hypothesis, having an older partner

at sexual debut was not associated with early pubertal

timing nor age at first sex, although it was associated with a

greater likelihood of teenage childbearing, as expected.

This is contrary to previous research that has found early

developing girls to be more likely to have older romantic

partners than their on-time and later developing peers

(Halpern et al. 2006), and adolescent girls with older

boyfriends to be more likely to have sexual relationships

(Kaestle et al. 2002). It is possible that the measurement of

early pubertal timing could have impacted the results;

research linking early development with older romantic

partners assessed perceived pubertal timing. The percep-

tion of maturity may be more closely linked to the char-

acteristics of girls’ romantic partners than more objective

measures of maturity, such as age at menarche. This points

to a need to replicate this research with multiple measures

of pubertal timing to better understand the relationships

between pubertal timing and older sexual partners.

Hence, in considering the influence of peers on the rela-

tionship between pubertal timing and teenage childbearing,

our findings were mixed. If perceived peer substance use at

ages 12–14 is an appropriate proxy for peer substance use,

this finding provides support for the developmental readiness

hypothesis in that early developing girls’ acceptance by and

association with substance using peers in early adolescence

exposes them to earlier and riskier sexual behaviors than

their on-time and later developing peers which in turn puts

them at increased risk for teenage childbearing. However,

having an older first sexual partner was not associated with a

younger age at sexual debut, which refutes the hypothesis.

This discrepancy in findings provides further evidence for

the need for more nuanced and better measured examination

of the influence of peers on early developing girls’ risks for

adverse outcomes in adolescence and beyond.

The major strengths of the current study are its large,

nationwide sample of ethnically diverse participants; its

prospective, longitudinal design; and that it tightly con-

trolled for background characteristics influential of both

pubertal timing and adolescent risk behaviors. As we did

not use NLSY97 sample weights in analyses, the sample is

not generalizable to all females in the United States, but

rather the 2066 individuals of the female cohort included in

analyses. Our study was limited by the measurement of

pubertal timing. Age at menarche is only one of many

indicators of girls’ pubertal timing and occurs late in the

pubertal development process (Dorn and Biro 2011).

Nonetheless, it is generally considered an acceptable ap-

proximation of pubertal timing (Brooks-Gunn et al. 1987).

Still, other indicators of pubertal timing (e.g., breast

development, perceived pubertal timing, hormone levels)

were not present in the current dataset and may have helped

to provide a more nuanced exploration of which aspects of

girls’ early pubertal development put them most at risk for

teenage childbearing. In addition, the girls excluded from

the analytic sample differed by several study characteris-

tics. Substance use differences at Wave 1 were primarily

due to the older age of the respondents at Wave 1 that were

excluded. Other differences were largely driven by the 143

girls excluded due to reporting sexual debut prior to

menarche or Wave 1. Because the resulting analytic sample

was overall a lower risk sample, it is likely that the results

represent a conservative estimate of the relationships

between early pubertal timing and teenage childbearing.

The current study findings and limitations suggest

important directions for future research. Given that the

subsample of girls excluded from the study due to sexual

debut prior to menarche or Wave 1 differed significantly

from the final analytic sample, future longitudinal surveys

examining the role of pubertal timing on subsequent risk

behavior in adolescence should begin assessing pubertal

development in childhood, prior to pubertal onset. The

current study was unable to assess the role of childhood

sexual abuse, which has been linked with early pubertal

debut and adolescent substance use, as well as teenage

pregnancy. Future research should examine whether child-

hood sexual abuse is an explanatory variable for the mech-

anisms found in this study, or if the relationships between

pubertal timing, peer and individual substance use, sexual

debut, and teenage childbearing remain after controlling for

childhood sexual trauma. Relatedly, we did not assess

alternative models (e.g., whether having an older sex partner

924 J Youth Adolescence (2016) 45:916–927

123

leads to increased substance use) due to the timing of when

substance use and perceived peer substance use were

assessed in the NLSY97; substance use and perceived peer

substance use were only assessed at Wave 1 whereas infor-

mation about sexual behavior was not assessed until par-

ticipants reached age 15. Future studies should

simultaneously and continuously assess substance use, peer

influences, and sexual risk behaviors across adolescence to

explore alternative models and examine the interplay of

these risk behaviors over time in how they may mediate the

relationship between pubertal timing and risk for teenage

childbearing. The current study could also be extended by

measuring multiple aspects of girls’ pubertal timing, beyond

age at menarche, to determine which are most influential in

predicting sexual health behaviors and outcomes. As previ-

ously mentioned, the assessment of peer substance use was

based on participant report and, as such, future replication

studies should compare the relative influence of perceived

peer substance use and peer-reported substance use. Finally,

future research should examine potential psychological

correlates of the behaviors observed in this study. Under-

standing the role of sensation seeking, psychological dis-

tress, body satisfaction, and self-esteem would help

practitioners and researchers better understand the modifi-

able psychological constructs that could decrease risky

behaviors among early developing girls.

Conclusion

The current study provides further evidence that girls who

develop earlier than their peers are at heightened risk for

deleterious outcomes. Because it is not possible, or advis-

able, to alter pubertal development among children who are

clinically within the ‘‘normal’’ range of development, it is

critical to determine modifiable mediating factors that may

aid in the prevention of these negative outcomes. As with

previous research, we found that an earlier age of sexual

debut was associated with a higher likelihood of teenage

childbearing. There is a continued need for comprehensive

sexual education in the United States in order to delay first

sexual intercourse and provide sexually active girls with

the resources needed to prevent pregnancy. Our findings

also provide evidence of the importance of substance use

behaviors, both perceived peer use and individual use, in

early adolescence in explaining how girls’ early pubertal

timing can lead to higher risk for teenage childbearing. In

particular, substance use in early adolescence was found to

mediate the relationship between early pubertal develop-

ment and age of sexual debut. This provides additional

support for the need for comprehensive substance use

prevention in late childhood/early adolescence. In addition,

providers should be screening for behavioral risk taking

among early developing girls at ages that may be tradi-

tionally considered ‘‘too young’’ for risk taking. Further-

more, prevention programming should be tailored for

parents of early developing girls to help families navigate

the potential psychosocial risks associated with their

pubertal development. These comprehensive prevention

efforts may be beneficial in breaking the link between girls’

early pubertal timing and teenage motherhood.

Acknowledgments Ms. Hendrick received support from the grant, 5 T32 HD007081, Training Program in Population Studies, awarded to

the Population Research Center at The University of Texas at Austin

by the Eunice Kennedy Shriver National Institute of Child Health and

Human Development. Dr. Cance and Dr. Maslowsky are Faculty

Research Associates of the Population Research Center at the

University of Texas at Austin, which is supported by the Eunice

Kennedy Shriver National Institute of Child Health and Human

Development Grant 5 R24 HD042849.

Authors’ Contributions CEH conceived of the study, participated in its design, performed statistical analysis, and coordinated and

drafted the manuscript; JDC supervised the study and statistical

analysis, participated in study design, coordination, interpretation of

the data, and critical revision of the manuscript; JM participated in

statistical analysis, interpretation of the data, and critical revision of

the manuscript. All authors read and approved the final manuscript.

Conflict of interest The authors report no conflict of interest.

Ethical Approval The present study was deemed not to be human subjects research by the Institutional Review Board of the sponsoring

university. This article does not contain any studies with human

participants performed by any of the authors.

Informed Consent For this type of study formal consent is not required.

References

Brooks-Gunn, J., Petersen, A. C., & Eichorn, D. (1985). The study of

maturational timing effects in adolescence. Journal of Youth and

Adolescence, 14(3), 149–161. http://doi.org/10.1007/BF020903

16.

Brooks-Gunn, J., Warren, M. P., Rosso, J., & Gargiulo, J. (1987).

Validity of self-report measures of girls’ pubertal status. Child

Development, 58(3), 829–841. http://doi.org/10.2307/1130220.

Cance, J. D., Ennett, S. T., Morgan-Lopez, A. A., Foshee, V. A., &

Talley, A. E. (2013). Perceived pubertal timing and recent

substance use among adolescents: A longitudinal perspective:

Perceived pubertal timing and substance use. Addiction, 108(10),

1845–1854. http://doi.org/10.1111/add.12214.

Cavazos-Rehg, P. A., Krauss, M. J., Spitznagel, E. L., Schootman, M.,

Cottler, L. B., & Bierut, L. J. (2011). Substance use and the risk

for sexual intercourse with and without a history of teenage

pregnancy among adolescent females. Journal of Studies on

Alcohol and Drugs, 72(2), 194–198. http://doi.org/10.15288/

jsad.2011.72.194.

Chen, X.-K., Wen, S. W., Fleming, N., Demissie, K., Rhoads, G. G.,

& Walker, M. (2007). Teenage pregnancy and adverse birth

outcomes: A large population based retrospective cohort study.

International Journal of Epidemiology, 36(2), 368–373. http://

doi.org/10.1093/ije/dyl284.

J Youth Adolescence (2016) 45:916–927 925

123

Crockett, L. J., Raffaelli, M., & Shen, Y.-L. (2006). Linking self-

regulation and risk proneness to risky sexual behavior: Pathways

through peer pressure and early substance use. Journal of

Research on Adolescence, 16(4), 503–525. http://doi.org/10.

1111/j.1532-7795.2006.00505.x.

Deardorff, J., Gonzales, N. A., Christopher, F. S., Roosa, M. W., &

Millsap, R. E. (2005). Early puberty and adolescent pregnancy:

The influence of alcohol use. Pediatrics, 116(6), 1451–1456.

Deutsch, A. R., Chernyavskiy, P., Steinley, D., & Slutske, W. S.

(2015). Measuring peer socialization for adolescent substance

use: A comparison of perceived and actual friends’ substance use

effects. Journal of Studies on Alcohol and Drugs, 76(2),

267–277. http://doi.org/10.15288/jsad.2015.76.267.

Dorn, L. D., & Biro, F. M. (2011). Puberty and its measurement: A

decade in review. Journal of Research on Adolescence, 21(1),

180–195. http://doi.org/10.1111/j.1532-7795.2010.00722.x.

Driscoll, A. K., & Abma, J. C. (2015). Changing sociodemographic

factors and teen fertility: 1991–2009. Maternal and Child Health

Journal, 19(10), 2159–2167.

Halpern, C. T., Kaestle, C. E., & Hallfors, D. D. (2006). Perceived

physical maturity, age of romantic partner, and adolescent risk

behavior. Prevention Science, 8(1), 1–10. http://doi.org/10.1007/

s11121-006-0046-1.

Henry, D. B., Kobus, K., & Schoeny, M. E. (2011). Accuracy and bias

in adolescents’ perceptions of friends’ substance use. Psychology

of Addictive Behaviors, 25(1), 80.

Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in

covariance structure analysis: Conventional criteria versus new

alternatives. Structural Equation Modeling: A Multidisciplinary

Journal, 6(1), 1–55.

Jaffee, S., Caspi, A., Moffitt, T. E., Belsky, J., & Silva, P. (2001). Why

are children born to teen mothers at risk for adverse outcomes in

young adulthood? Results from a 20-year longitudinal study.

Development and Psychopathology, 13(02), 377–397.

Kaestle, C. E., Morisky, D. E., & Wiley, D. J. (2002). Sexual

intercourse and the age difference between adolescent females

and their romantic partners. Perspectives on Sexual and Repro-

ductive Health, 34(6), 304–309. http://doi.org/10.2307/3097749.

Kilpatrick, D. G., Acierno, R., Saunders, B., Resnick, H. S., Best, C.

L., & Schnurr, P. P. (2000). Risk factors for adolescent substance

abuse and dependence: Data from a national sample. Journal of

Consulting and Clinical Psychology, 68(1), 19–30.

Kirby, D. (2002). Antecedents of adolescent initiation of sex,

contraceptive use, and pregnancy. American Journal of Health

Behavior, 26(6), 473–485.

Kost, K., & Henshaw, S. (2014). US teenage pregnancies, births and

abortions, 2010: National and state trends by age, race and

ethnicity. New York: Guttmacher Institute. Retrieved from

https://www.guttmacher.org/pubs/USTPtrends10.pdf.

Meade, C. S., Kershaw, T. S., & Ickovics, J. R. (2008). The

intergenerational cycle of teenage motherhood: An ecological

approach. Health Psychology, 27(4), 419–429. http://doi.org/10.

1037/0278-6133.27.4.419.

Mendle, J., Turkheimer, E., & Emery, R. E. (2007). Detrimental

psychological outcomes associated with early pubertal timing in

adolescent girls. Developmental Review, 27(2), 151–171. Moore, K. A., McGroder, S., Hair, E. C., & Gunnoe, M. (1999).

NLSY97 codebook supplement main file round 1: Appendix 9:

Family process and adolescent outcome measures. Washington,

DC: Child Trends, Inc.

Muthen, L. K., & Muthen, B. O. (2012). Mplus User’s Guide. Seventh

Edition. Los Angeles, CA. Retrieved from https://www.statmo

del.com/download/usersguide/Mplus%20user%20guide%20Ver_

7_r6_web.pdf.

Negriff, S., Brensilver, M., & Trickett, P. K. (2015). Elucidating the

mechanisms linking early pubertal timing, sexual activity, and

substance use for maltreated versus nonmaltreated adolescents.

The Journal of Adolescent Health: Official Publication of the

Society for Adolescent Medicine, 56(6), 625–631. http://doi.org/

10.1016/j.jadohealth.2015.02.012.

Negriff, S., Susman, E. J., & Trickett, P. K. (2011). The develop-

mental pathway from pubertal timing to delinquency and sexual

activity from early to late adolescence. Journal of Youth and

Adolescence, 40(10), 1343–1356.

Negriff, S., & Trickett, P. K. (2012). Peer substance use as a mediator

between early pubertal timing and adolescent substance use:

Longitudinal associations and moderating effect of maltreat-

ment. Drug and Alcohol Dependence, 126(1), 95–101.

Patel, P. H., & Sen, B. (2012). Teen motherhood and long-term health

consequences. Maternal and Child Health Journal, 16(5),

1063–1071.

Salas-Wright, C. P., Vaughn, M. G., Ugalde, J., & Todic, J. (2015).

Substance use and teen pregnancy in the United States: Evidence

from the NSDUH 2002–2012. Addictive Behaviors, 45, 218–225.

http://doi.org/10.1016/j.addbeh.2015.01.039.

Savolainen, J., Mason, W. A., Hughes, L. A., Ebeling, H., Hurtig, T.

M., & Taanila, A. M. (2015). Pubertal development and sexual

intercourse among adolescent girls: An examination of direct,

mediated, and spurious pathways. Youth and Society, 47(4),

520–538. http://doi.org/10.1177/0044118X12471355.

Schreiber, J. B., Nora, A., Stage, F. K., Barlow, E. A., & King, J.

(2006). Reporting structural equation modeling and confirmatory

factor analysis results: A review. The Journal of Educational

Research, 99(6), 323–338.

The Office of Adolescent Health. (2015). Trends in Teen Pregnancy

and Childbearing. Retrieved August 20, 2015, from http://www.

hhs.gov/ash/oah/adolescent-health-topics/reproductive-health/

teen-pregnancy/trends.html.

Tofighi, D., & MacKinnon, D. P. (2011). RMediation: An R package

for mediation analysis confidence intervals. Behavior Research

Methods, 43(3), 692–700.

Townshend, J. M., Kambouropoulos, N., Griffin, A., Hunt, F. J., &

Milani, R. M. (2014). Binge drinking, reflection impulsivity, and

unplanned sexual behavior: Impaired decision-making in

young social drinkers. Alcoholism: Clinical and Experimental

Research, 38(4), 1143–1150.

United Nations Statistics Division. (2014). Live births by age of

mother and sex of child, general and age-specific fertility

rates, 2004–2013. Retrieved August 20, 2015, from http://

unstats.un.org/unsd/demographic/products/dyb/dyb2013/Table10.

pdf.

U.S. Bureau of Labor Statistics. (2014). National longitudinal

surveys: The NLSY97. Retrieved August 19, 2015, from http://

www.bls.gov/nls/nlsy97.htm.

U.S. Bureau of Labor Statistics. National longitudinal surveys sample

weights and design effects. Retrieved August 20, 2015, from

https://www.nlsinfo.org/content/cohorts/nlsy97/using-and-under

standing-the-data/sample-weights-design-effects

Volpe, E. M., Hardie, T. L., Cerulli, C., Sommers, M. S., & Morrison-

Beedy, D. (2013). What’s age got to do with it? Partner age

difference, power, intimate partner violence, and sexual risk in

urban adolescents. Journal of Interpersonal Violence, 28(10),

2068–2087.

Walker, J. R. (2001). Adolescents’ expectations regarding birth

outcomes: A comparison of the NLSY79 and NLSY97 cohorts.

In R. T. Michael (Ed.), Social awakening: Adolescent behavior

as adulthood approaches (p. 117). New York: Russell Sage

Foundation.

Zabin, L. S., Emerson, M. R., & Rowland, D. L. (2005). Childhood

sexual abuse and early menarche: The direction of their

relationship and its implications. Journal of Adolescent Health,

36(5), 393–400.

926 J Youth Adolescence (2016) 45:916–927

123

C. Emily Hendrick is a doctoral student in the Health Behavior and Health Education program in the Department of Kinesiology and

Health Education at The University of Texas at Austin. She received

her Master’s in Public Health in Maternal and Child Health from the

School of Public Health at the University of California, Berkeley. Her

research is focused on understanding and reducing maternal, child,

and adolescent health disparities through the investigation of the

modifiable influences of women’s health behaviors and outcomes

during the reproductive years.

Jessica Duncan Cance is an Assistant Professor in the Health Behavior and Health Education program in the Department of

Kinesiology and Health Education at The University of Texas at

Austin. Dr. Cance received her Master’s in Public Health in

Behavioral Sciences and Health Education from the Rollins School

of Public Health at Emory University and her Ph.D. in Health

Behavior and Health Education from the Gillings School of Global

Public Health at the University of North Carolina at Chapel Hill. Her

research is focused on how the longitudinal interaction of biological,

psychological, and social factors impacts adolescent and young adult

health risk behaviors.

Julie Maslowsky is an Assistant Professor in the Health Behavior and Health Education program in the Department of Kinesiology and

Health Education at The University of Texas at Austin. Dr.

Maslowsky received a B.S. in Human Development and Psycholog-

ical Services from Northwestern University. She then earned a Ph.D.

in Developmental Psychology at the University of Michigan. She

completed postdoctoral training in Population Health with the Robert

Wood Johnson Foundation Health & Society Scholars program at the

University of Wisconsin, Madison. Dr. Maslowsky’s work combines

Developmental Psychology and Population Health to identify the

origins and consequences of adolescent health risk behavior and

develop prevention and intervention programs to reduce its negative

consequences. Specific areas of focus include substance use and

abuse, mental health, sleep, and risk behavior.

J Youth Adolescence (2016) 45:916–927 927

123

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.