Week 6: Reviewing the Literature Accurately
yazmillie13
Loftus2011incorrectuseofGilliganreference.pdf
E M P I R I C A L R E S E A R C H
Depressive Symptoms Among Adolescent Girls in Relationships with Older Partners: Causes and Lasting Effects?
Jeni Loftus • Brian C. Kelly • Sarah A. Mustillo
Received: 26 April 2010 / Accepted: 27 August 2010 / Published online: 15 September 2010
� Springer Science+Business Media, LLC 2010
Abstract Previous research on adolescent girls in rela-
tionships with older partners suggests a range of negative
outcomes for the adolescent. Using three waves of data
from the National Longitudinal Survey of Adolescent
Health and a life course perspective, we explore the con-
nection between involvement in age discordant relation-
ships (girls dating males three or more years older) and the
course of depressive symptoms among adolescent girls. Our
analyses are conducted on the 1,307 girls with data at all
three waves who had been in a relationship in the last
18 months. The sample is 55% White, 22% Black, 7% other
race and 16% Latina. When comparing early adolescent
girls (13–15) and late adolescent girls (16–18), the younger
girls are no more likely to have depressive symptoms before
entry into the age discordant relationship, yet have greater
depressive symptoms shortly after the relationship onset
and 5 years later. Older adolescent girls in age discordant
relationships, however, have similar levels of depressive
symptoms at any time point relative to their peers. These
data suggest that a girl’s developmental stage influences
whether or not she experiences emotional distress as a result
of being in an age discordant relationship.
Keywords Sexuality � Romantic relationships � Depression
Introduction
There has long been concern in American society about
adolescent girls dating older males. Every state has laws
regarding age of sexual consent intended to protect minors
from sexual relationships with significantly older individ-
uals (Sutherland 2003). The presence of these laws indi-
cates a widespread presumption of harm for adolescents
involved in such relationships. Indeed, much research
suggests that the outcomes are often negative for adoles-
cent girls involved with significantly older males. These
concerns continue to merit further investigation.
Age discordant relationships are not uncommon among
adolescents. Indeed, as much as 30% of older adolescents
have relationships with partners three or more years older
at some point (Darroch et al. 1999; Gowen et al. 2004),
while 10% of younger adolescents have relationships with
someone two or more years older (Marı́n et al. 2000). The
research suggests that there may be many negative out-
comes for girls in relationships with older males. For
instance, adolescent girls involved in age discordant rela-
tionships, most typically defined as three or more years
older, are likely to have an earlier sexual debut (Vanoss
et al. 2000), are more likely to engage in various sexual
activities (Gowen et al. 2004; Kaestle et al. 2002; Marı́n
et al. 2000), are less likely to use condoms (Ford et al.
2001; Miller et al. 1997) and other forms of contraception
(Glie 1999), and have higher odds of contracting sexually
transmitted diseases (Lee et al. 2003). Beyond sexual risk,
there are other problems. Girls in age discordant relation-
ships also have higher odds of engaging in substance use
(Mezzich et al. 1997a). There is also some evidence of
increased suicide risk among girls in age discordant rela-
tionships (Leitenberg and Salizman 2000). Overall, the
existing literature indicates a series of negative impacts of
J. Loftus (&) � B. C. Kelly � S. A. Mustillo Department of Sociology, Purdue University, 700 W State St.,
West Lafayette, IN 47907, USA
e-mail: [email protected]
123
J Youth Adolescence (2011) 40:800–813
DOI 10.1007/s10964-010-9589-3
age discordant relationships on the mental and physical
health among adolescent girls.
Although several studies indicate that age discordant
relationships produce a detrimental effect, such studies are
primarily based on cross-sectional data that do not indicate
a causal direction. Additionally, few have used longitudinal
data to examine the long-term influence of these relation-
ships on the mental health and well-being of adolescent
girls in such relationships. This article addresses both of
these issues. Because mental health is a critical component
of adolescent development, our study investigates trajec-
tories of depressive symptoms among teenage girls in age
discordant relationships. First, we examine the extent to
which depressive symptoms play a role in young girls’
entry into age discordant relationships. Second, we explore
the course of depressive symptoms shortly after entry into
the relationship and into young adulthood. This longitudi-
nal perspective allows us to examine the causes and effects
of age discordant relationships.
Adolescence and Mental Health
Adolescence has long been described as a tumultuous period
of the life course. This rapidly changing and confusing
transitional period has been associated with heightened
mental health problems. Rates of depression rapidly increase
during adolescence (Kessler et al. 1993), with girls experi-
encing greater depressive symptoms than boys (Adkins et al.
2009; Costello et al. 2003; Hankin et al. 1998), before these
rates recede to some degree by young adulthood. In partic-
ular, adolescence is a critical period of the life course
regarding the onset of mental health problems with clinical
significance, particularly mood disorders (Kessler and Wang
2008). Approximately half of individuals with a mental
disorder have an onset that occurs by mid-adolescence
(Kessler et al. 2007). Adolescents who are depressed are at
increased risk for a major depressive disorder during
adulthood (Weissman et al. 1999). While the manifestation
of depression during adolescence may differ somewhat from
that in adulthood, the social factors that influence adolescent
depression have parallels to those found for adults. Ado-
lescents experiencing elevated levels of depression report
greater levels of life stress, lower levels of self-esteem, less
social support, fewer positive coping strategies, and a his-
tory of abuse (Schraedley et al. 1999). Due to its wide
prevalence, co-morbidity with other physical and mental
health problems, and lasting consequences, depression is
among the most concerning mental health problems in
adolescence (MacPhee and Andrews 2006).
As noted above, rates of depression increase during
adolescence across the population, but female adolescents
experience elevated rates compared to their male peers
(Hankin et al. 1998; Schraedley et al. 1999). This mirrors
broader gender disparities in depression across the popu-
lation as women are almost twice as likely to experience
depression (Weissman et al. 1996). Some of these dispar-
ities can be attributed to social and cultural influences.
Beyond physiological factors, existing research suggests a
range of social influences—from family dynamics to
adverse childhood experiences to coping styles to social
support—on the gender disparity in depression (Piccinelli
and Wilkinson 2000). It is noteworthy that social rela-
tionships are more likely to influence the mental health of
teenage girls than their male peers (Rudolph et al. 2000).
Romantic partnerships, in particular, are key social rela-
tionships for adolescent girls.
Relationships and Mental Health
Many relationship factors have been shown to affect
mental health during adolescence and into adulthood. For
example, parental preferential treatment of children is
associated with depression, particularly among adolescent
girls (Shanahan et al. 2008). Additionally, parental divorce
during childhood or adolescence impacts the mental health
of those children well into their young adult years (Cherlin
et al. 1998). Furthermore, relationships with peers,
including peer attachment and victimization, also consid-
erably influence depression among adolescents (Hawker
and Boulton 2000; Laible et al. 2000). This wide body of
research identifies the importance of social relationships as
influences of adolescent mental health, but romantic rela-
tionships relate to depression in specific ways.
Learning how to construct and maintain intimate rela-
tionships is an important developmental task for young
people (Erikson 1950). In fact, intimate relationships have
been identified as one of the most important stressors in the
lives of adolescents (Newcomb et al. 1981). Research
indicates that girls are more likely to be affected by
romantic relationships than are boys (Gilligan 1996; Ro-
senfield 1999), and these influences are not necessarily for
the better. Joyner and Udry (2000) find that romantic
relationships are not positive influences on levels of
depression among adolescents. Specifically, they found that
girls who became involved in any romantic relationship
had higher levels of depression than peers who did not
become involved in a romantic relationship during this
period. Research also suggests that adolescents may
become involved in romantic relationships as ways to cope
with stress and depression (Compas and Vannatta 1986;
Plancherel and Bolognini 1995). These findings suggest
that higher levels of depression may not only result from
involvement in such relationships but also increase one’s
likelihood of becoming involved in a romantic relationship
as well.
J Youth Adolescence (2011) 40:800–813 801
123
Precocious Transitions and Depressive Symptoms
While the above findings collectively indicate that
romantic relationships may impact mental health, they are
of particular concern when an adolescent is involved
with someone at a different socio-developmental stage.
Becoming involved in a romantic relationship is a norma-
tive behavior for adolescents. As adolescents, age they are
more likely to become involved in emotionally close and
long lasting relationships (Zimmer-Gembeck 2002). This is
in contrast to when they are early adolescents when rela-
tionships are more likely to be shorter and less emotionally
close (Shulman et al. 2009). Thus, it is normative and
developmentally anticipated that older adolescents will
have longer and closer romantic relationships, and younger
adolescents will have shorter and less close romantic
relationships. This may become a problem when the two
partners in a romantic relationship are at different devel-
opmental stages. The younger partner, in the case of this
study, girls, may become involved in relationships for
which they are not developmentally ready (i.e. close inti-
macy, sexual activity) or they may feel developmentally
unprepared, which may impact mental health.
Relationships with significantly older males may disrupt
a standard developmental trajectory. Adolescent girls who
are romantically involved with older males may experience
‘‘off-time transitions’’ with regard to typical adolescent
development. Research suggests that off-time life course
transitions adversely influence mental health (Jackson
2004). For instance, individuals who marry earlier or later
than expected report worse mental health than their peers
who experience these life course transitions when expec-
ted. Thus, these relationships raise concern from a devel-
opmental perspective.
When applied to adolescents, early transitions are often
referred to as ‘‘precocious transitions’’ (Newcomb and
Bentler 1986). Entering an age discordant relationship can
be conceived as a precocious transition. Indeed, research
suggests that adolescents who conform to the normative
trajectory have better mental health than adolescents who
do not conform to the normative trajectory (Welsch et al.
2003). Research also suggests that these precocious tran-
sitions can have long term as well as short term implica-
tions for the adolescents. For instance, Krohn et al. (1997)
found that precocious transitions can lead to adverse con-
sequences such as substance use, early parenthood, drop-
ping out of school, and moving out of the parental home
early. All of these problems may have long-term effects.
Newcomb and Bentler (1988) argue that the early adop-
tion of roles and behaviors interferes with the acquisition of
skills necessary to act out these behaviors and roles suc-
cessfully. Girls who become involved with older males at a
different life stage may be developmentally propelled
forward along with them, and may be unprepared develop-
mentally. As noted above, previous research on girls in age
discordant relationships suggests that they are likely to take
on other ‘‘adult’’ behaviors earlier than their peers, such as
engaging in sexual activity or using alcohol and drugs.
Therefore, entry into an age discordant relationship may be
considered a precocious transition – or a precocious claim to
adulthood –that adversely affects mental health, thus
potentially precipitating higher rates of depressive symp-
toms among adolescent girls who enter them.
Hypotheses
We examine the connection between depressive symptoms
and involvement in age discordant relationships for ado-
lescent girls. We ask two key questions in pursuing this line
of inquiry. First, are those who enter age discordant rela-
tionships likely to display greater depressive symptoms
before entry into these relationships? Second, how does
entry into an age discordant relationship influence the
experience of depressive symptoms both in the short- and
long-term? Based upon the existing literature, these ques-
tions set up several hypotheses.
The first hypothesis holds that adolescent girls with
elevated depressive symptoms will be more likely to enter
into age discordant relationships. Previous research indi-
cates that some adolescents may enter into romantic rela-
tionships as a way of coping with depression and stress
(Compas and Vannatta 1986; Plancherel and Bolognini
1995). This may be particularly the case for age discordant
relationships, as depression may lead adolescent girls to
seek positive reinforcement from romantic partners and
peers by dating older males. As such, we can posit that
depressive symptoms may be an influence on likelihood of
dating older males.
The second hypothesis holds that adolescent girls who
enter age discordant relationships will experience elevated
depressive symptoms shortly after the relationship (i.e.
within 1 year after the relationship). The literature suggests
that age discordant relationships place adolescent girls into a
vulnerable position with regard to a range of negative health
outcomes (Glie 1999; Lee et al. 2003; Mezzich et al. 1997b),
with some evidence of an impact on mental health (Leiten-
berg and Salizman 2000). Given the intersection of depres-
sion and adolescent romances (Joyner and Udry 2000) we
can posit that, as an off-time transition, age discordant
relationships may elevate risk for depressive symptoms
among adolescent girls, perhaps especially younger ado-
lescent girls who may be subject to greater off-time
transitions.
The third hypothesis holds that adolescent girls who
enter age discordant relationships will experience elevated
802 J Youth Adolescence (2011) 40:800–813
123
depressive symptoms into early adulthood (i.e. approxi-
mately 5 years after entry into the relationship). Precocious
transitions may have long term impacts by interfering with
the cultivation of interactional skills necessary for devel-
opment on a normative trajectory (Newcomb and Bentler
1988). Evidence suggests that such precocious transitions
have a long term developmental impact into adulthood
(Krohn et al. 1997). Coupled with evidence that depressive
symptoms in adolescence are associated with depression in
adulthood (Weissman et al. 1999), we can posit that age
discordant relationships may have a lasting impact on the
mental health of adolescent girls as they age into young
adulthood.
As noted earlier, the previous research on physical and
mental health outcomes related to age discordant rela-
tionships among adolescents is largely cross-sectional;
either the adolescent is surveyed during the time of the
relationship and reports on behaviors occurring at that time
or the studies are retrospective where adults reflect back on
such relationships and their influence. Because of the
design of these studies, the causal order is often unknown.
For instance, are age discordant relationships causing
adolescents to have elevated levels of depressive symptoms
or are adolescents with higher levels of depressive symp-
toms more likely to become involved in age discordant
relationships? We aim to address this deficit by using
longitudinal data to determine causal direction.
Similarly, the outcomes of these relationships several
years later, into adulthood, remain unclear. While extant
research has demonstrated an array of negative outcomes
associated with age discordant relationships during or
shortly after the relationships, we do not fully understand
the long term effects of these relationships. Does
involvement in age discordant relationships have a lasting
impact on depressive symptoms? In this regard, the data we
present are unique in that they come from a longitudinal
study of adolescents and allow us to assess these girls in
young adulthood. Because of this, we are able to examine
not only the mental health precursors to such relationships
but also the outcomes related to these relationships into
young adulthood as well.
Method
Data
To examine the connections between involvement in age
discordant relationships and levels of depressive symptoms
among adolescent girls, we use data from the National
Longitudinal Study of Adolescent Health (Add Health)
(Udry 2003). Add Health is a school-based, longitudinal
study of health related behaviors of adolescents and their
outcomes into adulthood. Add Health is a nationally rep-
resentative sample of adolescents who were in grades 7–12
during the 1994–1995 school year. We utilize the first three
waves of data from this study in the analyses for this
article, which were collected between 1995 and 2002. We
use the restricted use dataset, with permission, for the
analyses contained in this article. Human Subjects approval
was obtained for this project.
Sample
Our sample includes data spanning 6 years of the subjects’
involvement in the study. Wave 1 was conducted in 1995,
Wave 2 was conducted in 1996, and Wave 3 was conducted
in 2001. The 1-year and 6-year follow-up time frame
allows us to examine the effects of these relationships not
only shortly after adolescent girls have entered them but
also into young adulthood. For this project, as we focus on
adolescent girls, we restrict the analyses to those girls
between the ages of 13 and 18 at Wave 1. This means they
are ages 14–19 at Wave 2 and were into young adulthood at
Wave 3 at ages 19–24. We also considered the develop-
mental differences between older and younger adolescent
girls that may influence how these relationships play out,
particularly with regard to the laws pertaining to the legal
age of consent in the US The age of consent is determined
by state law in the United States and therefore varies across
states within the nation and ranges from 12 to 18 (Suth-
erland 2003). The most common legal age of consent
among US states is 16 (Sutherland 2003). Thus, we con-
sider two age groups: girls 13–15 at Wave I (who are
subject to age of consent laws in almost all states) and girls
16–18 at Wave I (who, by their entries into age discordant
relationships, are of the legal age of consent in most states
in the nation). Parsing out these groups allows us to con-
sider how relationships with significantly older males may
differentially affect adolescent girls depending on their
stage of adolescent development.
Our analytic sample consists of the 1,307 girls between
13 and 18 with data at all three waves who had been in a
romantic relationship in the last 18 months. The Add
Health Study had 7,432 girls retained over the first three
waves of data collection. Of those 7,432, we dropped 289
girls who were under the age of 13 or over the age of 18.
We excluded 3,129 girls who reported no romantic rela-
tionships between Wave 1 and Wave 2. Additionally, we
dropped cases in which a relationship at Wave 2 began
prior to Wave 1 (534) or when girls were in an age dis-
cordant relationship as the older partner (555). We lost 927
girls who had incomplete data for weighting. Missing data
on various factors, including subject or partner age (80),
relationships (17), race (12), parental income (505), sex
(10), alcohol use (10), marijuana use (40), and depression
J Youth Adolescence (2011) 40:800–813 803
123
(12), led to the exclusion of 686 girls. Finally, the dropping
of 5 outliers whose relationship age difference was greater
than 13 years left us with a sample of 1,307 girls. The
analytic sample is 55% white non-Latina, 22% black non-
Latina, 7% other race non-Latina and 16% Latina. The
mean household income for the respondents is $46,600.
Measures
Age Discordant Relationship
We measure involvement in an age discordant relationship
by defining it as having a romantic relationship with
someone three or more years older. While there is some
debate in the literature about how one should define ‘‘age
discordant relationships,’’ 3 years and greater is the method
most commonly used, in large part because it coheres with
the age of sexual consent laws of many US states. This age
gap also reflects developmental stage differences that may
be important.
In the Add Health dataset, respondents are asked about
their past three ‘‘romantic relationships’’ during the previ-
ous eighteen months. In addition, if the respondent reported
no romantic relationship, they were asked if they had held
hands, had kissed, and had told someone they liked or
loved them. If a respondent reported all three of these
interactions with a non-family member they were recorded
as having a ‘‘liked relationship’’ by the Add Health
research team. Thus, by using the Add Health team’s
definition of liked relationships, which is based upon
multiple measures of amorous interaction, we are able to
determine amorous relationships even when adolescents do
not define them as ‘‘romantic,’’ which may have particular
meanings to adolescents. These criteria also ensure that the
analyses are not biased towards relationships in which
sexual activity occurs, as many adolescent relationships do
not inevitably lead to sexual activity, particularly for
younger adolescents. Thus, we use respondents who report
either a ‘‘romantic’’ or ‘‘liked’’ relationship in our consid-
eration of adolescent relationships. Involvement in an age
discordant relationship was coded as ‘‘yes’’ if any of these
relationships were age discordant, ‘‘no’’ if none of them
were. Because we wanted to know their levels of depres-
sive symptoms prior to entering into an age discordant
relationship, and whether that might predict entry, we
dropped any respondents who had been in an age discor-
dant relationship at Wave 2 but whose relationship started
prior to Wave 1. In order to remove any potential con-
founding effects, we dropped adolescents in relationships
with partners younger than them by three or more years and
also those respondents who had not been in any romantic
relationship during the last eighteen months. Thus, we have
two groups: girls in relationships with males three or more
years older and a comparison group, girls in relationships
with males close to their own age. The age difference
within age discordant relationships ranged from 3 to
13 years older (Mean = 5.48 years, SD = 2.78). We
dropped four outliers that had age differences ranging from
15 to 20 years.
Demographics
Three demographic variables related to adolescent depres-
sive symptoms serve as controls: race/ethnicity, parental
income, and age, as previous research has indicated their
association with depressive symptoms (Meadows et al.
2006; Mossakowski 2008; Wickrama et al. 2008).
Race/Ethnicity The girl’s race/ethnicity is measured as a
series of dummy variables: Black non-Latino, Latina, and
other-race non-Latino versus White non-Latino (the refer-
ence category).
Parental Income Parental income is derived from the
parent survey conducted at Wave 1. It was reported as
dollars for total household income. We measure parental
income as the log of their reported income given the
skewness of income.
Age Age is the girl’s age at Wave 1 and ranges from 13 to
18 in the sample utilized.
Behavioral Factors
In addition to demographic factors, we consider whether or
not they have engaged in sexual intercourse prior to entry
into the relationship as well as alcohol and marijuana
usage. We control for whether or not they have had sexual
intercourse, as previous literature suggests that girls in age
discordant relationships are more likely to have had sex
(Gowen et al. 2004; Kaestle et al. 2002), and the literature
has also suggested that adolescents who have had sex are
more likely to display depressive symptoms (Spriggs and
Halpern 2008). Thus, we control for the effects of sex prior
to the relationship. We also control for the use of alcohol
and marijuana at each of the three waves. Alcohol and
marijuana use are significantly associated with being in an
age discordant relationship (Mezzich et al. 1997a), and
with depression (Deykin et al. 1987; Owens and Shippee
2009). We control for their use at each wave to ensure they
do not confound the relationship between age discordant
relationships and depressive symptoms.
Sexual Intercourse Sexual intercourse is measured at
Wave 1 and asks the girl whether or not she has ever had
sexual intercourse, defined as ‘‘when a male inserts his
804 J Youth Adolescence (2011) 40:800–813
123
penis into a female’s vagina.’’ This is a dichotomous var-
iable: 0 = no 1 = yes.
Alcohol Usage Alcohol usage is measured at each wave.
The question asks about alcohol usage in the past year with
seven response options ranging from ‘‘never’’ to ‘‘daily.’’
Marijuana Usage Marijuana usage is measured at each
wave. The question asks how many times the respondent
smoked marijuana in the last 30 days. We recoded the
variable to be dichotomous: 0 = did not smoke marijuana
in the last 30 days, 1 = did smoke marijuana in the last
30 days.
Depressive Symptoms
Depressive Symptoms were measured in all three waves
using a scale of ten items from the Center for Epidemiol-
ogic Studies Depression Scale (CES-D). The questions
concern how often over the past week the respondent was
bothered by things that don’t usually bother her, could not
shake the blues, had trouble keeping her mind on task, felt
depressed, felt too tired to do things, felt sad, felt people
disliked her, enjoyed life (reverse coded), and how often
she felt like she was as good as other people (reverse
coded), and how frequently she cried. These items are
scored on a 0 to 3 scale. As is standard practice, we sum the
scores of the ten variables, creating a depressive symptoms
variable with a possible range from 0 to 30. A higher score
indicates a higher level of depressive symptoms. In our
analyses for hypotheses 2 and 3, we also enter depressive
symptoms at Wave 1 as a covariate in order to determine
whether involvement in an age discordant relationship
predicts depressive symptoms above and beyond the level
of depressive symptoms they had at Wave 1. Cronbach’s
alpha for the depression measure demonstrates consistent
reliability: Wave 1 is 0.82, Wave 2 is 0.81, and Wave 3
is 0.81.
Descriptions of all variables and their means appear in
Table 1.
Analyses
Weighted logistic regression analysis was used to predict
whether those with higher levels of depressive symptoms at
Wave 1 are more likely to have entered into an age dis-
cordant relationship by Wave 2. Weighted Ordinary Least
Squares (OLS) Regression analysis was used to determine
whether entry into an age discordant relationship between
Waves 1 and 2 predicts levels of depressive symptoms at
Wave 2, which facilitates understanding of a short term
impact. Similarly, Weighted OLS Regression analysis was
used to determine whether entry into an age discordant
relationship between Waves 1 and 2 predicts levels of
depressive symptoms at Wave 3, which facilitates under-
standing of a long term impact. Results of these analyses,
which test our three hypotheses, appear in Tables 2, 3,
and 4, respectively. Because the depression variables were
slightly skewed, we used the bootstrap to provide standard
errors robust against deviations from normality and
homoscedasticity (Efron and Tibshirani 1993). Bootstrap
replicate weights were created which accounted for the
survey sampling design (Rao and Wu 1988).
We ran a series of analyses to account for the overall
sample of teenage girls as well as among younger and older
adolescent girls to understand differential impacts at dif-
ferent developmental stages. Initially analyses are run on
all girls together, ages ranging from 13 to 18. These results
appear as Model 1 in Tables 2, 3 and 4. However, because
there are many developmental differences amongst this age
group, we consider both early- and middle-adolescents
distinctly and run analyses separately for the younger girls
(ages 13–15) and the older girls (ages 16–18). The analyses
for the younger girls appear in Model 2 in Tables 2, 3,
and 4 and the analyses for the older girls appear in Model 3
in Tables 2, 3, and 4.
Due to Add Health’s complex sampling design, analyses
are adjusted for sample design effects. Add Health
researchers have computed these weights and we use Stata
to run analyses that account for the complex survey design.
For a more thorough discussion of the Add Health sam-
pling design and how weights are computed to correct for
this design, please see Chantala and Tabor (1999).
Results
Depressive Symptoms and Entry into Age Discordant
Relationships
Table 2 presents the results of the logistic regression
analysis predicting the odds of entry into an age discordant
relationship, which allows us to test Hypothesis 1, which
states that adolescent girls with elevated depressive
symptoms will have higher odds of entry into age discor-
dant relationships. The results in Model 1a indicate that
girls who display higher levels of depressive symptoms at
Wave 1 did not have higher odds of entry into an age
discordant relationship between Waves 1 and 2. Race/
ethnicity, parental income, age, sexual intercourse, and
marijuana and alcohol use are controlled for in Model 1b.
Controlling for these factors does not affect the relationship
between depressive symptoms and entry into an age dis-
cordant relationship. These variables also do not predict
entry into an age discordant relationship, although the
results suggest that age is predictive.
J Youth Adolescence (2011) 40:800–813 805
123
Table 1 Weighted means and standard deviations for the variables used in the analysis
Variable Description Metric Mean All
ages
Mean
13–15 year olds
Mean
16–18 year olds
Wave I (SD) Wave I (SD) Wave I (SD)
Age discordant
relationship
In an age discordant relationship as
younger partner?
0 = no; 1 = yes 0.10 (0.30) 0.15 (0.36) 0.05 (0.22)
Age Age at Wave 1 Range from 13 to 18 15.42 (1.48) 14.20 (0.79) 16.73 (0.76)
Sexual Intercourse
Wave 1
Had sexual intercourse by Wave 1? 0 = no; 1 = yes 0.39 (0.49) 0.22 (0.41) 0.57 (0.50)
Log income Log of parental income Range from 0 to 7 3.50 (0.86) 3.43 (0.93) 3.57 (0.77)
White (omitted) Race/ethnicity is white (non-Hispanic) 0 = no; 1 = yes 0.55 (0.50) 0.54 (0.50) 0.56 (0.50)
Black Race/ethnicity is Black (non-Hispanic) 0 = no; 1 = yes 0.22 (0.41) 0.25 (0.43) 0.18 (0.39)
Other race Race/ethnicity is Other Race (non-
Hispanic)
0 = no; 1 = yes 0.07 (0.26) 0.06 (0.23) 0.09 (0.29)
Latina Race/ethnicity is Latina 0 = no; 1 = yes 0.16 (0.37) 0.15 (0.36) 0.17 (0.37)
Alcohol use Wave 1 Alcohol use in the last year From 0 = never to
6 = daily
1.07 (1.43) 0.84 (1.37) 1.32 (1.45)
Alcohol use Wave 2 Alcohol use in the last year From 0 = never to
6 = daily
1.30 (1.58) 1.26 (1.55) 1.35 (1.61)
Alcohol use Wave 3 Alcohol use in the last year From 0 = never to
6 = daily
2.83 (2.12) 2.84 (2.09) 2.81 (2.15)
Marijuana use Wave 1 Used marijuana in the last 30 days? 0 = no; 1 = yes 0.12 (0.32) 0.10 (0.30) 0.13 (0.34)
Marijuana use Wave 2 Used marijuana in the last 30 days? 0 = no; 1 = yes 0.19 (0.40) 0.19 (0.39) 0.20 (0.40)
Marijuana use Wave 3 Used marijuana in the last 30 days? 0 = no; 1 = yes 0.23 (0.42) 0.22 (0.42) 0.24 (0.43)
Depressive symptoms
Wave 1
Level of depressive symptoms at Wave 1 Range from 0 to 30; 7.24 (4.93) 6.74 (5.00) 7.77 (4.79)
Depressive symptoms
Wave 2
Level of depressive symptoms at Wave 2 Range from 0 to 30; 6.81 (4.33) 6.73 (4.30) 6.89 (4.35)
Depressive symptoms
Wave 3
Level of depressive symptoms at Wave 3 Range from 0 to 30; 5.52 (4.42) 5.48 (4.54) 5.57 (4.91)
N = 1,307 N = 688 N = 619
Table 2 Predicting entry into an age discordant relationship between Wave 1 (1995) and Wave 2 (1996)—(odds ratios and 95% confidence intervals)
Model 1: all ages (at Wave I) Model 2: 13–15 (at Wave I) Model 3: 16–18 (at Wave I)
a b a b a b
Odds
ratio
95% CI Odds
ratio
95% CI Odds
ratio
95% CI Odd
ratios
95% CI Odds
ratio
95% CI Odds
ratios
95% CI
Depressive
symptoms
0.97 0.92–1.02 0.98 0.93–1.03 0.98 0.92–1.05 0.99 0.93–1.06 0.98 0.90–1.06 0.97 0.91–1.04
Black 0.76 0.37–1.59 0.52 0.23–1.19 2.36 0.56–9.90
Other race 1.14 0.43–3.05 0.92 0.31–2.71 1.95 0.40–9.63
Latina 1.50 0.80–2.81 1.12 0.51–2.44 3.03 1.00–9.20
Age 0.59*** 0.48–0.72 0.49** 0.33–0.72 0.68 0.36–1.28
Log of income 0.89 0.63–1.25 0.90 0.60–1.35 0.89 0.43–1.86
Sexual
intercourse
1.07 0.56–2.04 1.34 0.56–3.20 0.77 0.28–2.14
Alcohol use 1.04 0.88–1.22 1.07 0.87–1.30 1.02 0.78–1.35
Marijuana use 1.04 0.45–2.44 0.69 0.23–2.09 1.95 0.60–6.27
N = 1,307 N = 688 N = 619
* p \ 0.05; ** p \ 0.01; *** p \ 0.001
806 J Youth Adolescence (2011) 40:800–813
123
T a
b le
3 C
o e ffi
c ie
n ts
fr o
m th
e w
e ig
h te
d O
L S
re g
re ss
io n
o f
in fl
u e n
c e
o f
a g
e d
is c o
rd a n
t re
la ti
o n
sh ip
o n
d e p
re ss
iv e
sy m
p to
m s
a t
W a v
e II
(1 9
9 6
)— (b
o o
ts tr
a p
p e d
S E
)
M o
d e l
1 :
a ll
a g
e s
(a t
W a v
e I)
M o
d e l
2 :
1 3
– 1
5 (a
t W
a v
e I)
M o
d e l
3 :
1 6
– 1
8 (a
t W
a v
e I)
a b
a b
a b
A g
e d
is c o
rd a n
t re
la ti
o n
sh ip
1 .1
5 *
(0 .5
4 )
1 .4
6 *
* (0
.5 4
) 1
.3 8
* (0
.5 5
) 1
.6 5
* *
(0 .5
5 )
0 .8
0 (1
.2 7
) 1
.1 8
(1 .1
9 )
B la
c k
0 .1
5 (0
.4 2
) 0
.3 0
(0 .4
6 )
0 .0
0 (0
.6 8
)
O th
e r
ra c e
- 0
.3 8
(0 .5
7 )
0 .8
1 (0
.8 0
) -
1 .0
7 (0
.7 2
)
L a ti
n a
- 0
.7 4
(0 .9
7 )
- 0
.4 7
(0 .5
5 )
- 1
.2 2
(0 .6
5 )
A g
e 0
.1 7
(0 .1
2 )
0 .4
0 (0
.2 4
) 0
.4 3
(0 .3
3 )
L o
g o
f in
c o
m e
0 .1
4 (0
.1 6
) 0
.1 4
(0 .2
4 )
0 .1
7 (0
.2 8
)
E a rl
y se
x u
a l
in te
rc o
u rs
e -
0 .1
1 (0
.3 7
) 0
.1 1
(0 .5
4 )
- 0
.2 5
(0 .5
0 )
A lc
o h
o l
u se
- 0
.0 1
(0 .1
0 )
0 .0
0 (0
.1 4
) 0
.0 1
(0 .1
4 )
M a ri
ju a n
a u
se 1
.1 0
* (0
.4 5
) 0
.2 0
(0 .5
5 )
2 .0
2 *
* (0
.6 4
)
W a v
e 1
d e p
re ss
iv e
sy m
p to
m s
0 .1
3 *
* *
(0 .0
3 )
0 .1
7 *
* (0
.0 5
) 0
.0 8
(0 .0
5 )
In te
rc e p
t 6
.6 9
* *
* (0
.1 6
) 2
.4 9
(1 .6
7 )
6 .5
3 *
* *
(0 .2
3 )
- 1
.0 0
(3 .5
5 )
6 .8
5 *
* *
(0 .2
1 )
- 1
.5 5
(5 .3
6 )
R 2
0 .0
1 0
.0 4
0 .0
1 0
.0 7
0 .0
0 0
.0 6
N =
1 ,3
0 7
N =
6 8
8 N
= 6
1 9
* p
\ 0
.0 5
; *
* p \
0 .0
1 ;
* *
* p \
0 .0
0 1
T a
b le
4 C
o e ffi
c ie
n ts
fr o
m th
e w
e ig
h te
d O
L S
re g
re ss
io n
o f
in fl
u e n
c e
o f
a g
e d
is c o
rd a n
t re
la ti
o n
sh ip
o n
d e p
re ss
iv e
sy m
p to
m s
a t
W a v
e II
I (2
0 0
1 –
2 0
0 2
) (S
E )
M o
d e l
1 :
a ll
a g
e s
(a t
W a v
e I)
M o
d e l
2 :
1 3
– 1
5 (a
t W
a v
e I)
M o
d e l
3 :
1 6
– 1
8 (a
t W
a v
e I)
a b
a b
a b
A g
e d
is c o
rd a n
t re
la ti
o n
sh ip
0 .7
1 (0
.5 8
) 0
.7 8
(0 .5
8 )
1 .3
4 (0
.7 0
) 1
.4 7
* (0
.7 0
) -
0 .8
8 (0
.7 7
) -
1 .1
3 (0
.8 1
)
B la
c k
- 0
.0 6
(0 .4
3 )
- 0
.3 1
(0 .5
7 )
0 .2
2 (0
.6 7
)
O th
e r
ra c e
- 0
.4 5
(0 .5
3 )
- 0
.7 5
(0 .6
1 )
- 0
.4 4
(0 .9
0 )
L a ti
n a
0 .4
8 (0
.6 1
) 0
.6 2
(0 .7
4 )
0 .2
6 (0
.8 0
)
A g
e 0
.0 9
(0 .1
4 )
0 .2
9 (0
.2 7
) 0
.2 2
(0 .3
5 )
L o
g o
f in
c o
m e
0 .2
2 (0
.2 0
) 0
.1 7
(0 .2
4 )
0 .3
4 (0
.3 2
)
E a rl
y se
x u
a l
in te
rc o
u rs
e 0
.1 9
(0 .4
8 )
0 .7
1 (0
.6 7
) -
0 .3
2 (0
.5 4
)
A lc
o h
o l
u se
0 .0
2 (0
.0 9
) 0
.1 3
(0 .1
2 )
- 0
.0 6
(0 .1
3 )
M a ri
ju a n
a u
se 0
.6 7
(0 .3
7 )
- 0
.3 3
(0 .4
4 )
1 .9
0 *
* (0
.6 2
)
W a v
e 1
d e p
re ss
io n
- 0
.1 5
(0 .0
4 )
- 0
.0 1
(0 .0
4 )
- 0
.0 3
(0 .0
7 )
In te
rc e p
t 5
.4 5
* *
* (0
.1 7
) 3
.0 5
(2 .1
4 )
5 .2
7 *
* *
(0 .2
3 )
0 .4
0 (4
.1 0
) 5
.6 2
* *
* (0
.2 7
) 0
.8 6
(5 .7
3 )
R 2
0 .0
0 0
.0 1
0 .0
1 0
.0 3
0 .0
0 0
.0 3
N =
1 ,3
0 7
N =
6 8
8 N
= 6
1 9
* p \
0 .0
5 ;
* *
p \
0 .0
1 ;
* *
* p \
0 .0
0 1
J Youth Adolescence (2011) 40:800–813 807
123
We see the same pattern when we look at the two age
groups separately. Model 2a shows the bivariate relation-
ship between depressive symptoms and entry into an age
discordant relationship for the younger girls and shows that
this relationship is not significant, which is mirrored in
Model 3a for the older girls. Models 2b and 3b control for
race/ethnicity, parental income, age, sexual intercourse,
and marijuana and alcohol use, but do not change the
relationship between depressive symptoms and entry into
an age discordant relationship, which remains non-signifi-
cant. These control variables do not predict entry into an
age discordant relationship, with the exception of age,
which remains significant when we look at the younger
girls. For the older girls, age is not significant.
Age Discordant Relationships’ Effects on Levels
of Depressive Symptoms
Table 3 presents the results of regression analyses of the
effects of entry into an age discordant relationship between
Waves 1 and 2 on levels of depressive symptoms at Wave
2, which allows us to test Hypothesis 2, which states that
adolescent girls who enter age discordant relationships will
experience elevated depressive symptoms shortly after
relationship onset. Models 1a and 1b examine this rela-
tionship for all girls. Model 1a shows a bivariate rela-
tionship between entry into an age discordant relationship
and levels of depressive symptoms at Wave 2. This is a
significant relationship, showing that girls who entered into
an age discordant relationship between Waves 1 and 2 are
more likely to have elevated levels of depressive symptoms
Wave 2. Model 1b controls for race/ethnicity, parental
income, age, having had sexual intercourse by Wave 1,
marijuana and alcohol usage at Wave 2 and depressive
symptoms at Wave 1. Those who entered into an age dis-
cordant relationship were more likely to display depressive
symptoms at Wave 2, while controlling for these other
factors. Of these other variables, being Latina and recent
marijuana usage are significantly related to depressive
symptoms. As well, depressive symptoms at Wave 1 are
predictive of depressive symptoms at wave 2.
Models 2a and 2b in Table 3 show these analyses for the
younger girls only. The same pattern is evident: entry into
an age discordant relationship is significantly related to
higher levels of depressive symptoms, as seen in Model 2a.
This relationship remains significant even when race/eth-
nicity, parental income, age, sexual intercourse by Wave 1,
marijuana and alcohol usage at Wave 2, and depressive
symptoms at Wave 1 are taken into account. Of these other
variables, only depressive symptoms at Wave 1 are pre-
dictive of depressive symptoms at Wave 2.
Models 3a and 3b in Table 3 show the analyses for the
older girls. In these models, entry into an age discordant
relationship is not significantly related to depressive
symptoms either at the bivariate level or the multivariate
level when other factors are controlled for. Recent mari-
juana usage at Wave 2 is significant, showing that those
girls who have used marijuana in the last 30 days at
Wave 2 are more likely to display depressive symptoms at
Wave 2.
Table 4 presents the results of regression analyses of the
effects of having entered into an age discordant relation-
ship between Waves 1 and 2 on levels of depressive
symptoms at Wave 3, approximately 5 years later. This
allows us to test Hypothesis 3, which states that adolescent
girls who enter age discordant relationships will experience
elevated depressive symptoms into early adulthood. Here
we are interested in whether entering into an age discordant
relationship between Waves 1 and 2 has a lasting effect on
the mental health of adolescent girls.
In Models 1a and 1b we examine the effect of an age
discordant relationship on depressive symptoms at Wave 3
for all girls. The results suggest that there is no lasting
effect of the depressive symptoms that we saw at Wave 2
(Table 4). At the bivariate level there is no significant
relationship between having entered into an age discordant
relationship between Waves 1 and 2 and experiencing
elevated levels of depressive symptoms at Wave 3. At the
multivariate level, when race/ethnicity, parental income at
Wave 1, age, sexual intercourse by Wave 1, marijuana and
alcohol usage at Wave 3, and depressive symptoms at
Wave 1, are controlled for there continues to be no sig-
nificant relationship between having entered in an age
discordant relationship between Waves 1 and 2 and
depressive symptoms at Wave 3.
However, we do see an effect when we look specifi-
cally at the younger girls. In Models 2a and 2b, we look
at the effect of having entered into an age discordant
relationship between Waves 1 and 2 on depressive
symptoms at Wave 3 for the younger age group, girls
between the ages of 13 and 15 at Wave 1. Here we find
that younger adolescent girls who entered into an age
discordant relationship were more likely to have elevated
levels of depressive symptoms years later, though this is
only true at the multivariate level. None of the control
variables is significant in these models.
In Models 3a and 3b, we examine the effect of having
entered into an age discordant relationship on depressive
symptoms at Wave 3 for the older age group, girls of the
age of sexual consent (ages 17 and 19) by Wave 2. We do
not find the same pattern for the older girls that we found
for the younger girls. Here there is no effect of having
entered into an age discordant relationship on depressive
symptoms at Wave 3. Marijuana use is significantly
related to depressive symptoms at Wave 3 for the older
group.
808 J Youth Adolescence (2011) 40:800–813
123
Discussion
Adolescence is a critical period of the life course that may
set the stage for future health outcomes. The literature has
indicated that adolescent girls who date significantly older
males may be at elevated risk for early sexual debut
(Vanoss et al. 2000), sexual risk taking (Ford et al. 2001;
Miller et al. 1997), contracting sexually transmitted dis-
eases (Lee et al. 2003), and drug use (Mezzich et al.
1997a). Given these concerns, we sought to examine the
impact of these relationships on the mental health of ado-
lescent girls, both in the short term and as they age into
young adulthood.
Our results show some interesting patterns with regard
to the mental health of adolescent girls who have rela-
tionships with significantly older males. The primary
findings indicate that younger adolescent girls who date
males at least 3 years older experience elevated levels of
depressive symptoms shortly after entering into the rela-
tionship as well as into young adulthood. Dating signifi-
cantly older males does not have the same impact on the
mental health of older adolescent girls. Notably, in analy-
ses not presented above, we found similar results regarding
clinical levels of depression as defined by the clinical cut
off score for the CES-D. Overall, these findings suggest
that, for early adolescent girls, relationships with older
males may be precocious transitions, by developmentally
pushing them forward with their boyfriends, and such
precocious transitions may have a lasting impact on mental
health.
Consideration of Hypotheses
We found mixed support for our hypotheses. Our first
hypothesis postulated that girls with higher levels of
depression would be more likely to enter into age discor-
dant relationships. This was not supported by the analyses.
These findings indicate that girls in age discordant rela-
tionships do not have higher levels of depressive symptoms
than their peers prior to entering into their relationships.
This was true when all girls were examined together, and
also when the analyses were run separately on the younger
group and the older group. This is noteworthy as it indi-
cates that negative affect is not ‘‘pushing’’ girls into these
relationships. In other words, on the whole, these girls are
not seeking such relationships because they have higher
levels of depressive symptoms. Thus, there is not a mental
health induced vulnerability to enter into age discordant
relationships among teenage girls, and notably drug and
alcohol use do not predict entry either. Overall, while some
evidence suggests that depression may lead some adoles-
cent girls to enter romantic relationships (Compas and
Vannatta 1986; Plancherel and Bolognini 1995), mental
health remains a poor indicator of ‘‘at-risk’’ status for entry
into age discordant relationships.
After the point of entering into an age discordant rela-
tionship, the results vary with regard to support for our
hypotheses depending on whether the girls are considered
as a whole group, or whether we examine younger girls
(aged 13–15 at Wave 1) separately from older girls (aged
16–18 at Wave 1). As described in the Results section,
when we consider the whole group, we find that although
adolescent girls in age discordant relationships are no more
likely to have elevated levels of depressive symptoms prior
to the relationship than their peers, they are more likely to
experience depressive symptoms shortly after entering into
the relationship than peers in age-similar relationships.
However, examining all girls together obscures the pattern
that emerges if we consider that there are developmental
differences within the period of adolescence. When we
stratify by age, we find that our second hypothesis, which
postulated higher levels of depression for girls in age dis-
cordant relationships in the short term, is upheld only for
girls under the legal age of consent. In other words, only
among early adolescent girls do we find a significant
relationship between involvement in an age discordant
relationship and elevated levels of depressive symptoms
shortly thereafter. For older girls there appears to be no
greater likelihood of experiencing elevated levels of
depressive symptoms shortly thereafter if they have been in
an age discordant relationship. This suggests that being in
an age discordant relationship impacts mental health, but
only for younger adolescent girls.
We continue to see differences between the two groups
at Wave 3. If we consider the girls as a whole, we find that
there is no relationship between having been involved in
an age discordant relationship and depressive symptoms
during young adulthood. However, if we consider younger
girls separately from older girls, we find a different pat-
tern. Our third hypothesis, which postulated higher levels
of depression for adolescent girls in age discordant rela-
tionships in the long term, is upheld for adolescent girls
under the age of legal consent. For younger girls, there
continues to be an impact on depressive symptoms after
involvement in age discordant relationships. Those
younger girls who became involved in an age discordant
relationship between Waves 1 and 2 had significantly
higher levels of depressive symptoms years later at
Wave 3. For older girls, there continues to be no rela-
tionship between involvement and age discordant rela-
tionships and depressive symptoms.
Evaluation of Findings
The level of depressive symptoms is related to involvement
in an age discordant relationship for girls at a stage when
J Youth Adolescence (2011) 40:800–813 809
123
they are protected by age of consent laws, and perhaps at a
more vulnerable stage. By contrast, for the older group of
girls, there is no relationship between involvement in an
age discordant relationship and depressive symptoms at
any wave. As such, while we cannot identify girls who are
‘‘at-risk’’ for age discordant relationships on the basis of
mental health, we can infer that girls in such relationships
during early adolescence are ‘‘at-risk’’ for elevated levels
of depressive symptoms. In this regard, the prospects of
impaired mental health relate, to some degree, to devel-
opmental stage in adolescence, and perhaps the capacity to
consent. This also bears on the concern regarding preco-
cious transitions. Girls who do not have the capacity to
consent are the ‘‘early adopters’’ according to the societal
norms tied to the age of consent laws. According to pre-
cocious transitions theory, it is precisely such early
adopters whom we would anticipate having enduring
problems (Newcomb and Bentler 1988). Older adolescent
girls, while perhaps not pleasing their parents by dating
older males, do not shoulder the burden that off-time
transitions produce in exactly the same way, potentially
because they are developmentally closer to adulthood.
Several factors may be shaping elevated depressive
symptoms among early adolescent girls in age discordant
relationships. First, as mentioned earlier, dating takes on
different forms in early and late adolescence (Zimmer-
Gembeck 2002). Relationships in younger adolescence are
characterized as shorter and less intimate as opposed to
relationships in later adolescence which are characterized
as more long lasting and more intimate. For younger
adolescent girls, being involved in an age discordant rela-
tionship likely means being in a relationship with a partner
at a different developmental stage. This may not be quite so
true for older adolescent girls. The relationships for the
younger girls then are precocious or off-time transitions
that older girls do not experience given their more
advanced developmental stage. These younger girls may be
less prepared to deal with a longer lasting and more inti-
mate relationship than their older peers, which may
account for the higher levels of depression among the
younger group and not among the older group.
Second, considerations of power and gender inequity
must be examined within the context of age discordant
relationships. Research suggests that adolescent relation-
ships marked by inequality in the contribution of emotional
resources and decision making are more likely to lead to
higher levels of depression for girls (Galliher et al. 1999;
Greca and Harrison 2005). While evidence suggests a
gendered power dynamic in many relationships, any
imbalance in age discordant relationships may be exacer-
bated by the age differential. It may be that in age dis-
cordant relationships there is a higher level of inequality
between the partners than within age-similar relationships.
This could be especially the case for the younger girls who
are at a different developmental stage than their partners.
The older male partner may have more power in the rela-
tionship and the girl may not feel that she has a voice in
terms of decision making within the partnership. Given that
the partners in such relationships are at different develop-
mental stages, there may be meaningful differences in the
role expectations of each individual in the dyad. Such role
discontinuity may lead to an underlying, perhaps even
unspoken, tension in the relationship. Thus, power differ-
entials driven by the age difference may elevate the stress
levels of the adolescent girl and adversely impact her
mental health.
Third, becoming involved in an age discordant rela-
tionship may affect the girl’s relationship with her female
friends. Even if the girl feels that she is developmentally
ready for a relationship with an older male, it may still have
a detrimental impact on her relationship with her peers. It is
not uncommon for adolescents to begin to spend more time
with their romantic partners and less time with their same-
gender friends as they age and become more involved in
romantic relationships (Hendry et al. 1993). For girls in age
discordant relationships, this may occur earlier than it does
for girls in age concordant relationships. Thus, it may cause
them to have greater conflict in their same gender rela-
tionships or a shrinking of friendship networks which has
been shown to lead to higher levels of depression (Ueno
2005).
Beyond the mechanisms identified above, it remains
important to delineate the social factors that provide a
context for age discordant relationships. As is the case with
age-similar relationships, age discordant relationships may
vary in nature and quality. For example, a teenage girl who
meets her older boyfriend on the street may have a quali-
tatively different relationship than the girl who meets her
older boyfriend at a community service activity. These
circumstances may also subsequently influence mental
health outcomes. Additionally, peer contexts may influence
the adolescent girl. If the older male is enmeshed in her
peer network, the outcomes may be buffered compared to
the girl who weakens her peer ties for the older boyfriend.
Finally, a critical social component of such relationships is
the extent to which it causes strain in parental relationships
(Dowdy and Kliewer 1999; Larsen 1995). Parental disap-
proval may cause either strain from explicit conflict over
the relationship or underlying strain if the relationship is
hidden. Such strain may directly relate to elevated
depressive symptoms (Sheeber 1997 #379) Future research
must further elucidate the psychosocial and contextual
mechanisms by which these relationships influence ado-
lescent mental health outcomes.
810 J Youth Adolescence (2011) 40:800–813
123
Limitations and Conclusions
While these analyses yield important findings on adolescent
mental health, it is important to note some limitations with
regard to this study. First, social desirability biases may have
come into play when adolescents reported their relation-
ships. Thus, we may have an underreporting of age discor-
dant relationships due to cases where adolescent girls are
unwilling to report a relationship they may know is prob-
lematic or illegal. In addition, we are unable to determine
whether these young women entered into an age discordant
relationship during the 5 year span between Wave 2 and
Wave 3. Thus, it remains unclear whether a later relationship
with a significantly older male may influence the results for
the period of young adulthood for some. Yet, this limitation
may serve to depress our findings and yet we still found
elevated levels of depressive symptoms among younger
adolescent girls in age discordant relationships.
In spite of these limitations, the strength of this dataset
and these analyses provide important information on the
mental health and well-being of adolescent girls. Our
findings highlight the linkages between age discordant
relationships and mental health among a nationally repre-
sentative sample of adolescent girls. The results improve
upon cross-sectional studies by indicating the causal
direction of the links between mental health and age dis-
cordant relationships. We have identified that level of
depressive symptoms does not causally precede involve-
ment with older males. Thus, it does not appear that ele-
vated depressive symptoms are a ‘‘risk factor’’ for entry
into age discordant relationships among adolescent girls.
Yet, it appears that, for younger adolescent girls, dating a
significantly older male may impair mental health, not only
proximally to the relationship but into young adulthood.
This indicates that developmental stage may be important
with regard to the adverse effects of dating an older male
and may continue to have an influence later in life, which
highlights the importance of developmental distinctions
within the period of adolescence.
Acknowledgments We wish to thank Timothy Owens, Viktor Gecas, and Brooke Wells for their comments on previous drafts of this
paper. We also extend our thanks to the University of North Carolina
Population Center for permission to use the National Longitudinal
Study of Adolescent Health dataset. This research uses data from Add
Health, a program project directed by Kathleen Mullan Harris and
designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan
Harris at the University of North Carolina at Chapel Hill, and funded by
grant P01-HD31921 from the Eunice Kennedy Shriver National Insti-
tute of Child Health and Human Development, with cooperative
funding from 23 other federal agencies and foundations. Special
acknowledgment is due Ronald R. Rindfuss and Barbara Entwisle for
assistance in the original design. Information on how to obtain the Add
Health data files is available on the Add Health website (http://www.
cpc.unc.edu/addhealth). No direct support was received from grant
P01-HD31921 for this analysis.
References
Adkins, D. E., Wang, V., Dupre, M. E., Van den Oord, E. J. C. G., &
Elder, G. H., Jr. (2009). Structure and stress: Trajectories of
depresive symptoms across adolescence and young adulthood.
Social Forces, 88(1), 31–61. Chantala, K., & Tabor, J. (1999). Strategies to perform a design-
based analysis using the add-health data. Retrieved March 10, 2010, from http://www.cpc.unc.edu/projects/addhealth/faqs/
aboutdata/weight1.pdf.
Cherlin, A. J., Chase-Lansdale, P. L., & McRae, C. (1998). Effects of
parental divorce on mental health throughout the life course.
American Sociological Review, 63(2), 239–249. Compas, B. E., Wagner, B. M., Salvin, L. A., & Vannatta, K. (1986). A
prospective study of life events, social support, and psychological
symptomology during the transition from high school to college.
American Journal of Community Psychology, 14(3), 241–257. Costello, E. J., Mustillo, S., Erkanli, A., Keeler, G., & Angold, A.
(2003). Prevalence and development of psychiatric disorders in
childhood and adolescence. Archives of General Psychiatry, 60, 837–844.
Darroch, J. E., Landry, D. J., & Oslak, S. (1999). Age differences
between sexual partners in the United States. Family Planning Perspectives, 31, 160–167.
Deykin, E. Y., Levy, J. C., & Wells, V. (1987). Adolescent
depression, alcohol, and drug use. American Journal of Public Health, 77(2), 178–182.
Dowdy, B. B., & Kliewer, W. (1999). Dating, parent-adolescent
conflict, and behavioral autonomy. Journal of Youth and Adolescence, 27, 473–492.
Efron, B., & Tibshirani, R. J. (1993). An Introduction to the bootstrap. New York: Chapman and Hall.
Erikson, E. (1950). Growth and crisis of the healthy personality. In
M. J. E. Seen (Ed.), Symposium on the healthy personality (pp. 91–146). New York: Josiah Macy Jr. Foundation.
Ford, K., Sohn, W., & Lepkowski, J. (2001). Characteristics of
adolescents’ sexual partners and their association with use of
condoms and other methods. Family Planning Perspectives, 33(3), 100–105.
Galliher, R. V., Rostosky, S. S., Welsh, D. P., & Kawaguchi, M. C.
(1999). Power and psychological well-being in late adolescent
romantic relationships. Sex Roles, 40, 689–710. Gilligan, C. (1996). The centrality of relationships in human
development. In G. G. Noam & K. W. Fischer (Eds.),
Development and vulnerability in close relationships (pp. 237–261). Mahwah, NJ: Lawrence Erlbaum Associates.
Glie, D. A. (1999). Measuring contraceptive use patterns among
teenage and adult women. Family Planning Perspectives, 31(2), 73–80.
Gowen, L. K., Feldman, S. S., Diaz, R., & Yisrael, D. S. (2004). A
comparison of the sexual behaviors and attitudes of adolescent
girls with older vs. similar-aged boyfriends. Journal of Youth and Adolescence, 33(2), 167–175.
Greca, A. M. L., & Harrison, H. M. (2005). Adolescent peer relations,
friendships, and romantic relationships: Do they predict social
anxiety and depression? Journal of Clinical Child and Adoles- cent Psychology, 34, 49–61.
Hankin, B. L., Abramson, L. Y., Moffit, T. E., Silva, P. A., McGee,
R., & Angell, K. E. (1998). Development of depression from
preadolescence to young adulthood: Emerging gender differ-
ences in a 10-year longitudinal study. Journal of Abnormal Psychology, 107, 128–140.
Hawker, D. S. J., & Boulton, M. J. (2000). Twenty years’ research on
peer victimization and psychosocial maladjustment. Journal of Child Psychology, 41, 441–455.
J Youth Adolescence (2011) 40:800–813 811
123
Hendry, L. B., Shucksmith, J., Love, J. G., & Glendinning, A. (1993).
Young people’s lesiure and lifestyles. New York: Routledge. Jackson, P. B. (2004). Role sequencing: Does order matter for
mental health? Journal of Health and Social Behavior, 75(2), 132–154.
Joyner, K., & Udry, J. R. (2000). You don’t bring me anything but
down: Adolescent romance and depression. Journal of Health and Social Behavior, 42(4), 369–391.
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.
Kessler, R. C., Amminger, G. P., Aguilar-Gaxiola, S., Alonso, J., Lee, S.,
& Üstün, T. B. (2007). Age of onset of mental disorders: A review
of recent literature. Current Opinion in Psychiatry, 20(4), 359–364. Kessler, R. C., McGonagle, K. A., Swarz, M., Blazer, D. G., &
Nelson, C. B. (1993). Sex and depression in the national
comorbidity survey I: Lifetime prevalence, chronicity and
recurrence. Journal of Affective Disorders, 29, 85–96. Kessler, R. C., & Wang, P. S. (2008). The descriptive epidemiology
of commonly occurring mental disorders in the United States.
Annual Review of Public Health, 29, 115–129. Krohn, M. D., Lizotte, A. J., & Perez, C. M. (1997). The
interrelationship between substance use and precocious transi-
tions to adult statuses. Journal of Health and Social Behavior, 38, 87–103.
Laible, D. J., Carlo, G., & Rafaelli, M. (2000). The differential
relations of parent and peer attachment to adolescent adjustment.
Journal of Youth and Adolescence, 29(1), 45–59. Larsen, B. (1995). Conflict and social interaction in adolescent
relationships. Journal of Research on Adolescence, 5, 55–70. Lee, J. K., Jennings, J. M., & Ellen, J. M. (2003). Discordant sexual
partnering: A study of high-risk adolescents in San Francisco.
Sexually Transmitted Diseases, 30(3), 234–240. Leitenberg, H., & Salizman, H. (2000). A statewide survey of at first
intercourse for adolescent females and age of their male partners:
Relation to other risk behaviors and statutory rape implications.
Archives of Sexual Behavior, 29(3), 203–215. MacPhee, A. R., & Andrews, J. J. W. (2006). Risk factors for
depression in early adolescence. Adolescence, 41, 435–466. Marı́n, B. V., Coyle, K. K., Gómez, C. A., Carvajal, S. C., & Kirby,
D. B. (2000). Older boyfriends and girlfriends increase risk of
sexual initiation in young adolescents. Journal of Adolescent Health, 27(6), 409–418.
Meadows, S. O., Brown, S. J., & Elder, G. H., Jr. (2006). Depressive
symptoms, stress, and support: Gendered trajectories from
adolescence to young adulthood. Journal of Youth and Adoles- cence, 35(1), 93–103.
Mezzich, A. C., Tarter, R. E., Giancola, P. R., Lu, S., Kirisci, L., & Parks,
S. (1997a). Substance use and risky sexual behavior in female
adolescents. Drug and Alcohol Dependence, 44(3), 157–166. Mezzich, A. C., Tarter, R. E., Giancola, P. R., Lu, S., Kirisci, L., & Parks,
S. (1997b). Substance use and risky sexual behavior in female
adolescents. Drug and Alcohol Dependence, 44(2), 157–166. Miller, K. S., Clark, L. F., & Moore, J. S. (1997). Sexual initiation
with older male partners and subsequent HIV risk behavior
among female adolescents. Family Planning Perspectives, 29(5), 212–214.
Mossakowski, K. (2008). Dissecting the influence of race, ethnicity,
and socioeconomic status on mental health in young adulthood.
Research on Aging, 30(6), 649–671. Newcomb, M. D., & Bentler, P. M. (1986). Frequency and sequence
of drug use: A longitudinal study from early adolescence to
young adulthood. Journal of Drug Education, 16(2), 101–120. Newcomb, M. D., & Bentler, P. M. (1988). Consequences of
adolescent drug use. Newbury Park: Sage.
Newcomb, M. D., Huba, G. J., & Bentler, P. M. (1981). A
multidimensional assessment of stressful life events among
adolescents: Derivation and correlates. Journal of Health and Social Behavior, 22(4), 649–671.
Owens, T. J., & Shippee, N. D. (2009). Depressed mood and drinking
occasions across high school: Comparing the reciprocal causal
structures of a panel of boys and girls. Journal of Adolescence, 32(4), 763–780.
Piccinelli, M., & Wilkinson, G. (2000). Gender differences in
depression. British Journal of Psychology, 177, 486–492. Plancherel, B., & Bolognini, M. (1995). Coping and mental health in
early adolescence. Journal of Adolescence, 18(4), 459–474. Rao, H. N. K., & Wu, C. F. J. (1988). Resampling inference with
complex survey data. Journal of American Statistical Associa- tion, 83(401), 231–241.
Rosenfield, S. (1999). Splitting the difference: Gender, self and
mental health. In C. S. Aneshensel & J. C. Phelan (Eds.),
Handbook of sociology of mental health. New York: Kluwer Academic/Plenum.
Rudolph, K. D., Hammen, C., Burge, D., Lindberg, N., Herzberg, D.,
& Daley, S. E. (2000). Toward an interpersonal life-stress model
of depression: The developmental context of stress generation.
Development and Psychopathology, 12(2), 215–234. Schraedley, P. K., Gotlib, I. H., & Hayward, C. (1999). Gender
differences in correlates of depressive symptoms in adolescents.
Journal of Adolescent Health, 28, 98–108. Shanahan, L., McHale, S. M., Crouter, A. C., & Osgood, D. W.
(2008). Linkages between parents’ differential treatment, youth
depressive symptoms, and sibling relationships. Journal of Marriage and Family, 70(2), 480–494.
Sheeber, L., Hops, H., Alpert, A., Davis, B., & Andrews, J. (1997).
Family support and conflict: Prospective relations to adolescent
depression. Journal of Abnormal Child Psychology, 25(4), 333–344.
Shulman, S., Walsh, S. D., Weisman, O., & Schelyer, M. (2009).
Romantic contexts, sexual behavior, and depressive symptoms
among adolescent males and females. Sex Roles, 61, 850–863. Spriggs, A. L., & Halpern, C. T. (2008). Sexual debut timing and
depressive symptoms in emerging adulthood. Journal of Youth and Adolescence, 37(9), 1085–1096.
Sutherland, K. (2003). From Jailbird to Jailbait: Age of consent laws
and the construction of teenage sexualities. William and Mary Journal of Women and the Law, 9, 313–349.
Udry, J. R. (2003). The national longitudinal study of adolescent health (Add Health), Waves I & II, 1994–1996; Wave III, 2001–2002 [machine-readable data file and documentation]. Chapel Hill, NC: Carolina Population Center, University of
North Carolina at Chapel Hill.
Ueno, K. (2005). The effects of friendship networks on adolescent
depressive symptoms. Social Science Research, 34(3), 484–510. Vanoss, B., Coyle, K. K., Gomez, C. A., Carvajal, S. C., & Kirby,
D. B. (2000). Older boyfriends and girlfriends increase risk of
sexual initiation in young adolescents. Journal of Adolescent Health, 27(6), 409-418.
Weissman, M. M., Bland, R. C., Canino, G. J., Faravelli, C.,
Greenwald, S., Hwu, H.-G., et al. (1996). Cross-national
epidemiology of major depression and bipolar disorder. Journal of the American Medical Association, 246(4), 293–299.
Weissman, M. M., Wolk, S., Goldstein, R. B., Moreau, D., Adams, P.,
Greenwald, S., et al. (1999). Depressed adolescents grow up.
Journal of the American Medical Association, 281, 18. Welsch, D. P., Grello, C. M., & Harper, M. S. (2003). When love
hurts: Depression and adolescent romantic relationships. In
P. Florsheim (Ed.), Adolescent romantic relations and sexual behavior: Theory, research and practical implications (pp. 185–212). Mahwah, NJ: Erlbaum.
812 J Youth Adolescence (2011) 40:800–813
123
Wickrama, K. A. S., Conger, R. D., Lorenz, F. O., & Jung, T. (2008).
Family antecedents and consequences of trajectories of depressive
symptoms from adolescence to young adulthood: A life course
investigation. Journal of Health and Social Behavior, 49, 468–483. Zimmer-Gembeck, M. J. (2002). The development of romantic
relationships and adaptations in the system of peer relationships.
Journal of Adolescent Health, 31, 216–225.
Author Biographies
Jeni Loftus is an Assistant Professor at Purdue University. She received her Ph.D. in sociology from Indiana University. Her main
research interests include adolescent sexuality, sexuality across the
life course, and the social psychological effects of health conditions.
Brian C. Kelly is an Associate Professor at Purdue University. He received his Ph.D. in sociomedical sciences from Columbia Univer-
sity. He primarily studies drug use, sexual health and youth cultures,
and his overarching research interests include the political economy
of risk, cross-cultural research on youth, and the link between
network and neighborhood factors and health disparities.
Sarah A. Mustillo is an Associate Professor at Purdue University. She received her Ph.D. in sociology from Duke. Her main research
interests include the links between parent and child mental health,
family-level predictors of developmental issues in at-risk youth,
mental health issues in service members returning from combat
operations, and longitudinal models for categorical data.
J Youth Adolescence (2011) 40:800–813 813
123
Copyright of Journal of Youth & Adolescence is the property of Springer Science & Business Media B.V. and
its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for individual use.
