Clin Child Fam Psychol Rev (2008) 11:30–58

DOI 10.1007/s10567-008-0031-2

Conceptualizing and Re-Evaluating Resilience Across Levels

of Risk, Time, and Domains of Competence
Ella Vanderbilt-Adriance Æ Daniel S. Shaw

Published online: 1 April 2008

Ó Springer Science+Business Media, LLC 2008

Abstract This article examines potential theoretical protective factors that may be associated with positive
constraints on resilience across levels of risk, time, and adjustment despite exposure to risk factors. Resilience has
domain of outcome. Studies of resilience are reviewed as been a hot topic both within the context of developmental
they relate to the prevalence of resilience across levels of research and in the popular media. The allure of resilience
risk (e.g., single life events vs. cumulative risk), time, and arises from the success stories of people who have dealt
domains of adjustment. Based on a thorough review of with seemingly insurmountable odds, inspiring hope, and
pertinent literature, we conclude that resilience, as a global projecting the notion that there is no difficulty that cannot
construct, appears to be rare at the highest levels of risk, and be overcome. The study of resilience has implications for
that resilience may benefit from a narrower conceptualiza- understanding child development in general, but also for
tion focusing on specific outcomes at specific timepoints in prevention and intervention efforts aimed at guiding public
development. The implication of this conclusion for future policy and social programs to improve outcomes for chil-
research and intervention efforts is then discussed. dren at risk (Masten 2001). Thus, researchers must contend
with the dual goals of informing the literature and accu-
Keywords Resilience
Chronic risk
Competence
rately reporting findings to public health institutions and the
Prevalence media in their endeavor to elucidate the factors that are
associated with positive outcomes in the face of adversity
(Luthar and Cicchetti 2000).
Introduction When the concept of resilience was first introduced in
the 1970s, it was conceptualized as a stable personal
Over the past several decades, the concept of resilience has characteristic; at-risk children who appeared to be doing
gained prominence as a way to study the processes and well were thought to be ‘‘invulnerable’’ (Pines 1975). This
mechanisms through which exposure to risk factors may be perspective that certain children, due to some internal
associated with children’s positive and negative outcomes. characteristics (e.g., IQ) or positive features of their envi-
The term resilience has been defined as a positive outcome ronment (e.g., strong relationship with a caregiver), could
in the context of adversity (Luthar et al. 2000a), and centers ‘beat the odds’ and demonstrate positive adjustment in the
on the study of various child, family, and community context of adversity, led to a search for protective factors
that could explain such associations. As research in the
E. Vanderbilt-Adriance (&) area of resilience has developed over time, the conceptu-
Department of Psychology, University of Pittsburgh,
alization of resilience has been refined, such that most
210 S. Bouquet St., 4425 Sennott Square,
Pittsburgh, PA 15260, USA researchers now recognize it as a dynamic process that
e-mail: elv4@pitt.edu results from ongoing transactions between a child and the
environment, rather than an internal characteristic of the
D. S. Shaw
child (Luthar and Zelazo 2003). Few researchers now view
Department of Psychology, University of Pittsburgh,
210 S. Bouquet St., 4101 Sennott Square, children with positive outcomes as ‘‘invulnerable,’’ and
Pittsburgh, PA 15260, USA there is increasing recognition that the effects of risk persist

123
Clin Child Fam Psychol Rev (2008) 11:30–58 31

over time or emerge in unexpected ways (Luthar 2006). overview of select protective factors associated with posi-
However, despite improvements in the conceptualization of tive outcomes is also provided. This is followed by a
resilience, challenges remain in establishing truly ‘high’ discussion of potential theoretical constraints on resilience
risk contexts, interpreting the degree to which positive in the context of severe adversity. The next section criti-
adjustment actually occurs in the context of chronic and cally reviews the extant literature on resilience with a focus
severe risk, and determining the stability of resilience on the degree to which positive adjustment occurs across
across time and domains. different types and levels of risk (e.g., chronic and/or
First, the variability in the establishment of ‘high risk’ severe), and the stability of positive adjustment in different
environments has hindered our ability to determine the contexts of risk across time and domains of adjustment.
prevalence of resilience. For example, some studies of Finally, the article concludes by determining implications
resilience have utilized predominantly European American, of this appraisal for future research on resilience, including
middle-class children who, although experiencing a sig- implications for prevention research and social policy.
nificant life event (e.g., divorce), have been exposed to a
qualitatively lower level of adversity than children growing
up in the context of inner-city poverty. Due to the com- Definition of Resilience and Related Constructs
paratively low level of risk in the former context, such
studies may obtain misleadingly high rates of positive Resilience is currently conceptualized as a dynamic process
adjustment compared to children living in more chronic consisting of a series of ongoing, reciprocal transactions
and severe settings. between the child and the environment (Luthar and Zelazo
There are also data to suggest that when positive 2003; Masten 2001). Importantly, this conceptualization
adjustment is identified among children living in adverse rejects the notion of resilience as a personal or individual
contexts, it may vary across time and domains. For example, trait. In fact, researchers have warned against using such
adjustment may fluctuate over time as children pass through terms as ‘‘resiliency’’ because they connote a stable char-
various milestones and their associated challenges, such that acteristic, and may foster perspectives that blame the
some children experiencing high levels of adversity may be individual for their negative outcomes (Luthar et al. 2000a).
doing well socially and academically at school age, but Although personal traits (e.g., IQ, temperament) can influ-
show deterioration in adjustment during the transition to ence outcomes in the context of adversity, they are also often
adolescence. Such results demonstrate the challenge of strongly affected by both genetic and contextual factors, and
showing persistent positive outcomes in the context of are thus not fully attributable to the child (Luthar and Cic-
chronic adversity. Furthermore, children who may be doing chetti 2000). This distinction is particularly important
well in one area, such as school achievement, may dem- because if resilience is interpreted as a personal trait, policy
onstrate problems in other areas, such as depression (Luthar makers may then use it as justification to withhold important
et al. 1993). Thus, resilience may not be generalized, but services to at-risk children by arguing that resilience comes
rather specific, with children showing strengths and weak- from within the individual (Luthar and Cicchetti 2000).
nesses depending on the domain in question. Resilience has been operationalized in many ways, but it
The primary goal of the current article is to evaluate the is most commonly defined as a positive outcome in the
utility of the term resilience in the context of severe and context of risk, or factors known to be associated with
chronic adversity. To this end, potential constraints or negative outcomes (Luthar et al. 2000a). Explicit within
limitations of resilience in the highest risk contexts (e.g., this definition is the requirement of risk, in addition to a
multiple risks, poverty) will be examined, with particular positive outcome; thus high functioning children in situa-
attention to differences between studies utilizing relatively tions of low adversity would not be considered resilient.
lower risk versus higher risk samples. Specifically, this
review will address the following three issues: (1) the Risk
prevalence of resilience in lower versus higher risk studies;
(2) the stability of resilience across time; and (3) the con- Resilience research has utilized a number of different risk
tinuity of resilience across domains. It is expected that the factors, including parental psychopathology (Conrad and
prevalence of resilience will be lower in the context of Hammen 1993; Luthar and Sexton 2007), socioeconomic
relatively higher risk, and that resilience will be limited disadvantage (Buckner et al. 2003; Kim-Cohen et al.
across time and domains. 2004), urban poverty and community violence (Gorman-
The first section of the article discusses definitions of Smith et al. 2004; Hammack et al. 2004), negative life
concepts related to resilience, including the operational- events (D’Imperio et al. 2000; Masten et al. 1999), child
ization of risk, positive outcome, and protective factor, and maltreatment (Cicchetti and Rogosch 1997; Jaffee et al.
identifies important points of controversy therein. A brief 2007), and cumulative risk indices (Seifer et al. 1992).

123
32 Clin Child Fam Psychol Rev (2008) 11:30–58

While all of these factors are associated with negative 2005). Thus, there are many issues involved in determining
outcomes in children, it is important to note that they are what is truly high-risk.
not necessarily equivalent in severity; rather severity
depends upon both the risk factor and the population in Positive Outcome
question. For example, some researchers have utilized
normative middle class samples exposed to varying levels There are also important differences in the operational-
of negative life events (e.g., Masten et al. 1999), while ization of ‘‘positive outcome,’’ with some studies focusing
others have utilized ethnically diverse samples of children on the absence of psychopathology, while others require
growing up in violent, low-income neighborhoods (e.g., more positive outcomes such as academic achievement,
Gorman-Smith et al. 2004). The inner-city poor contend social competence, or meeting appropriate developmental
with a substantial number of stressors and adversities, milestones. Whether resilience is operationalized as the
including community violence, crowding, poor quality absence of a negative outcome or the presence of a positive
schools, and inadequate housing (McLoyd 1998; Sampson outcome (or the combination of both) is largely a matter of
et al. 1999). Arguably, children growing up in chronic theoretical perspective and the nature of the risk factor in
poverty are exposed to a wide array of risks that are both question. For example, some risk factors are considered to
qualitatively and quantitatively more adverse than those be so powerful that simply the absence of psychopathology
experienced by most children living in middle-class envi- may be quite remarkable, while other more delimited risk
ronments. It is not clear that results from middle-class, factors such as parental divorce may necessitate more
predominantly white samples can be generalized to inner- evidence of a positive outcome (Luthar and Zelazo 2003).
city, minority children; thus results from the former studies Finally, there is also variability in whether positive
may be over-estimating the degree to which resilience adjustment must be demonstrated across several domains,
exists in situations of chronic, severe risk. or whether a positive outcome in one domain is considered
Relatedly, some researchers have stressed the impor- adequate. As resilience is not an ‘‘all-or-nothing’’ phe-
tance of identifying proximal risk factors to ensure that nomenon, Luthar and Zelazo (2003) assert that is must be
children within a particular sample are actually exposed to measured across domains to ensure that an accurate portrait
similar levels of risk (Richters and Weintraub 1990). For of positive adjustment is provided. For example, children
example, it has been argued that the variability in outcomes may be doing well on external measures of functioning
for children of psychiatrically ill parents may be due to the such as school achievement, yet demonstrate high levels of
fact that not all can truly be considered high risk. Some of internal distress (Luthar 1991).
these children may live in middle-class homes with ade- However, there are many studies which utilize single
quate resources, and have an engaged, supportive co-parent. domains of adjustment (e.g., Radke-Yarrow and Brown
Furthermore, their psychiatrically ill parent may be well- 1993; Stouthamer-Loeber et al. 2004; White et al. 1989),
monitored and receiving effective treatment. In contrast, and this can be entirely appropriate if the researcher is
other children may have to contend with a hospitalized, interested in the factors associated with positive outcomes
single mother who is unable to provide consistent, nurturing in a particular domain, such as school grades, rather than
care. Obviously, one would expect very different outcomes overall positive adjustment. The authors must be careful,
for these children because they have experienced very dif- however, to emphasize that adjustment is context-specific
ferent levels of overall adversity. Thus, it is important to and may not generalize to other domains. Furthermore, it
select a risk factor that accurately captures the daily expe- should be noted that the likelihood of finding positive
riences of children at risk for negative outcomes. outcomes in one domain is greatly increased when com-
Finally, Luthar (2003) has also warned against relying pared to definitions requiring positive outcomes across
on stereotypes to determine what constitutes ‘‘high risk.’’ several domains. This point should be kept in mind when
She notes that although affluent children are generally evaluating resilience research, so as to avoid overgeneral-
considered ‘‘low risk,’’ they actually display disturbingly izing from studies with less comprehensive definitions of
high rates of anxiety, depression, and substance use in ‘‘positive outcome.’’
adolescence, often well above national norms. Indeed, their
rates of such negative outcomes are also higher than those Protective Factors
of inner-city adolescents (Luthar 2003). Although one
might argue that such problems have less negative conse- Protective factors are defined as characteristics of the child,
quences for affluent adolescents due to their increased family, and wider environment that reduce the negative
resources, research demonstrates that the negative ramifi- effects of adversity on child outcome (Masten and Reed
cations are similar across socioeconomic classes, at least in 2002). Although some protective factors such as parenting
the domain of academic achievement (Luthar and Ansary appear to be important across different risk factors and

123
Clin Child Fam Psychol Rev (2008) 11:30–58 33

outcomes (Masten 2001), there is some evidence that other associated outcomes that might be otherwise obscured in a
protective factors may be more specific. For example, one between-group design (Seidman and Pedersen 2003). Fur-
study found that while an easy-going temperament and thermore, comparisons of children at differing levels of
stimulating activities in the home were associated with high risk can also lead to fine grain distinctions between
positive cognitive outcomes in the context of low SES, protective factors that operate in the context of high risk,
they had no association with externalizing behavior (Kim- but not extreme risk. For example, there are several studies
Cohen et al. 2004). Rutter (2000) has stressed the impor- of children living in urban poverty, which identified pro-
tance of selecting protective factors that are specific to the tective factors that were only helpful for children who had
risk factor and outcome in question, rather than assuming been exposed to low levels of community violence (e.g.,
that the mechanisms are similar across contexts. Further- Kliewer et al. 2004; Miller et al. 1999). Thus, although all
more, although few studies have explicitly examined the of the children in these studies could be considered high
role of development or gender, it seems likely that pro- risk due to poverty, some were at more extreme risk due to
tective factors may be more or less salient for different ages high levels of violence exposure. If these children had been
or genders. For example, although parental warmth is grouped together and compared to a low risk sample of
important across development, it may be particularly children, the differential benefits of the protective factors
important in early childhood when children are most within this high-risk group would most likely have been
dependent on their parents, rather than in adolescence when missed.
influences outside of the family play a larger role. Simi-
larly, Werner and Smith (1982) noted gender differences
for children with resilient outcomes, with emotional sup- Overview of Protective Factors
port from extended family being particularly important for
girls and family structure more important for boys. Protective factors have been identified in three main areas:
Similar to the issues surrounding the definition of (1) within the child, (2) within the family, and (3) within
resilience, controversy extends to the operationalization of the community. Widely researched protective factors are
protective factors. Some researchers have argued that a briefly reviewed in the following section to familiarize the
protective factor should interact with risk status to predict reader with the area; a full discussion of identified pro-
outcome (Garmezy et al. 1984; Rutter 1987). By this def- tective factors is beyond the scope of this article (for more
inition, only variables that are more strongly (or only) comprehensive reviews, see Luthar 2006; Masten and Reed
associated with positive outcomes in the context of high 2002; Rutter 2000).
risk, as opposed to low risk, are considered to be protective.
In more recent years, however, this term has been used to Child Protective Factors
refer to all factors associated with positive outcomes,
regardless of whether relationships are stronger for children Child attributes that have been found to be associated with
living in high-risk contexts (Luthar and Zelazo 2003). positive outcomes include intelligence, emotion regulation,
Luthar et al. (2000b) argue that while interaction effects temperament, coping strategies, locus of control, attention,
(positive effects only, or to a greater degree, for children at and genetic influences (Masten and Powell 2003). As noted
risk) provide useful knowledge on the processes that above, it is important to keep in mind that although child
function specifically under conditions of risk, main effects attributes can be protective in the context of adversity, they
can also be informative. For example, in designing inter- are also influenced by external factors, such as family
ventions for at-risk children, addressing any and all factors environment and the overall context in which the child
that attenuate the effects of risk are likely to be beneficial. lives. As such, they are not entirely ‘‘personal’’ traits. The
Implicit within this controversy is the issue of what type following brief review presents some representative child
of sample is optimal for studying resilience; for example, if protective factors, and discusses ways in which they may
one is primarily focused on identifying factors that are allow the child to interact differently with the environment,
more helpful in the context of risk, then it would be helpful and thus have more positive outcomes.
to have both low risk and high risk subgroups (Masten and Child IQ has consistently been found to predict a range
Reed 2002). Conversely, if the goal is to simply identify of positive outcomes, including academic achievement,
protective factors that help children at high levels of risk, prosocial behavior, and peer social competence (Masten
regardless of their impact at other levels of risk, the low et al. 1988, 1999), as well as the absence of antisocial
risk subgroup is unnecessary. Examining different patterns behavior (Kandel et al. 1988; Kolvin 1988; White et al.
of adjustment within a high-risk group can also help to 1989), and other types of psychopathology (Radke-Yarrow
elucidate the processes that contribute to positive outcomes and Brown 1993; Tiet et al. 1998, 2001; Werner and Smith
by highlighting the variation in protective factors and 1982, 1992). There are several reasons why IQ may be

123
34 Clin Child Fam Psychol Rev (2008) 11:30–58

important in high-risk contexts. First, children with high Second, children who display high levels of positive affect
IQs may be more likely to possess effective information- and are easy to soothe may evoke more sensitive caregiv-
processing and problem-solving skills, which enable them ing and attention from adults in the environment.
to contend with the stresses and challenges they encounter. Conversely, children who display high levels of negative
Children with higher intellectual skills should also perform affect, adjust poorly to change and are difficult to soothe
better at school; increased academic success is associated may initiate negative patterns of interaction with their
with the adoption of social norms and integration into caregivers, which may place them at increased risk for
prosocial peer groups (Masten and Coatsworth 1998). negative outcomes later in life.
Although some studies have found that IQ was more Research on older children has also focused on internal
important in the context of risk (Kandel et al. 1988; Kolvin attributes such as locus of control, appraisal, and coping
et al. 1988; Masten et al. 1988, 1999; Tiet et al. 2001), one skills, finding associations with a range of positive out-
study of inner-city adolescents found that high intelligence comes, including social competence, school grades, and
was only related to positive outcomes in context of low internalizing and externalizing symptomatology (Cauce
negative life events (Luthar 1991). Thus in this particular et al. 2003; Lin et al. 2004; Luthar 1991; Luthar and Zigler
study, IQ seemed to lose its ability to protect children once 1992). Children and adolescents who have less negative
stress became too high. appraisals of difficult situations, or who see themselves as
Emotion regulation refers to monitoring, evaluating, and having control over situations in their lives may respond
modifying the intensity and duration of emotional reactions less negatively to difficult situations and be better equipped
to accomplish one’s goals (Eisenberg et al. 1997a; to problem-solve. Conversely, children who think they
Thompson and Calkins 1996). Research demonstrates that have no control over external situations may feel helpless
a lack of control over emotion is associated with problem and be less likely to take action. Coping skills are also
behaviors (Calkins and Fox 2002; Eisenberg et al. 1996), important because children’s coping during difficult situa-
while the ability to manage one’s emotional expression tions can moderate the impact of the situation. For
predicts more positive social functioning in middle child- example, ignoring a negative situation maintains the status
hood both contemporaneously and longitudinally (Buckner quo, whereas reaching out for social support can generate
et al. 2003; Eisenberg et al. 1997a, b). Furthermore, studies solutions and decrease a sense of isolation.
of resilience have found that factors associated with emo- Finally, a relatively new line of research has begun
tion regulation (e.g., self-help skills, ego control, and ego examining gene-environment interactions, finding that
resiliency) are related to positive adjustment across risk certain genotypes appear to moderate the effect of envi-
status, and that such factors appear to be especially ronmental risk. For example, a study of child maltreatment
important in the context of adversity (Cicchetti and Rog- found that a functional polymorphism at the promoter of the
osch 1997; Cicchetti et al. 1993; Werner and Smith 1982, monoamine oxidase A (MAOA) gene was related to anti-
1992). Children who are adept at managing their emotions social problems in adolescence and adulthood, such that
may be better able to proactively cope with stressors high MAOA activity was protective in the context of severe
(Buckner et al. 2003) and thereby decrease the associated maltreatment (Caspi et al. 2002). Another study of depres-
negative effects. They may also be less likely to engage in sion found that a functional polymorphism in the promoter
oppositional behavior such as hitting or throwing a tantrum regions of the serotonin transporter (5-HTT) gene moder-
because of their ability to modulate negative emotion. Such ated the effect of life stress (Caspi et al. 2003). Although
children may be less likely to become involved in coercive such research is still in its early stages and requires repli-
cycles with their caregivers, and, therefore, may receive cation, these studies suggest that genetic variation, as well
more support from their social environment. Across con- as environmental variation, can be protective.
texts of risk, such children should function better in school
and in social relationships because they are able to mod- Family Protective Factors
ulate negativity and emotional expression.
Researchers have also examined the role of tempera- Researchers agree that one of the most important resources
ment, particularly in infancy and toddlerhood, finding that for normal development is the presence of a caregiver to
an easy-going temperament is associated with positive provide both material resources, such as nutrition and
outcomes in both childhood and adulthood (Kim-Cohen shelter, and more abstract resources, such as love, nurtur-
et al. 2004; Werner and Smith 1982; Wyman et al. 1999). ance, and a sense of safety and security (Masten 2001).
A child with an easy-going temperament may have positive When this system breaks down, the chances for normal
outcomes later in life for a number of reasons. First, they development are severely limited. In extreme instances,
may respond less negatively to stressful situations and be such as the Romanian orphanages where children were
more flexible in their responses to change or unfamiliarity. denied basic care and nurturance, the developmental

123
Clin Child Fam Psychol Rev (2008) 11:30–58 35

consequences are stark and undeniable (Beckett et al. 2006; activities (Wyman 2003) have all been shown to impact
Fisher et al. 1997; MacLean 2003). Even among materially child functioning. Bronfenbrenner (1979) has written
privileged children, the absence of a close parent–child extensively on the importance of such community-level or
relationship is linked with negative outcomes (Luthar and exosystem factors. The exosystem can affect the child both
Latendresse 2005). Conversely, Masten (2001) argues that directly, through his or her experience of it (e.g., attending
if the caregiving system is functional, this can help children school), or indirectly, through influences on parents and
to overcome considerable adversity. Parents teach their family. For example, a single mother living in poverty who
children the skills they need to succeed in later develop- has to commute 3 h/day to get to her job will be less able
mental tasks, set guidelines for acceptable behavior, and to monitor her child, or even to be physically present to
provide opportunities for cognitive and social stimulation provide the same level of care as a parent who can afford to
(Masten and Coatsworth 1998). In addition to specific work part-time or to pay for high quality after-school care.
parenting practices, having a good relationship with a Community-level influences can also be protective in
parent prepares the child to engage in healthy, productive the context of family and neighborhood risk; for example,
relationships with other people in the social environment. risk for serious chronic delinquency in adolescents from
Resilience research clearly demonstrates the importance inner-city families low on warmth and cohesion was
of the caregiving system. Researchers have examined pro- decreased in the context of high social organization in the
tective factors such as the quality of the parent–child community (Gorman-Smith et al. 2000). The authors sug-
relationship, attachment security in toddlerhood, and the gest that emotional needs for closeness and belonging can
type of parenting strategies employed. Indeed, a high sometimes be addressed at the community level, and rec-
quality relationship with at least one parent, characterized ommend that interventions focus on community-level
by high levels of warmth and openness and low levels of protective factors, as well as improving family functioning
conflict is associated with positive outcomes across levels (Gorman-Smith and Tolan 2003).
of risk and stages of development (Emery and Forehand
1996; Luthar and Latendresse 2005; Owens and Shaw 2003; Summary
Radke-Yarrow and Brown 1993; Stouthamer-Loeber et al.
1993, 2002; Vanderbilt-Adriance and Shaw in press; Wer- A wide variety of protective factors have been identified
ner and Smith 1982). Similarly, warm, responsive parenting that are associated with positive outcomes for children
styles are associated with positive child adjustment across exposed to adversity, including those at the level of the
social, emotional, and academic domains (Kim-Cohen et al. child, family, and community. Notably, child protective
2004; Masten et al. 1999; Werner and Smith 1982, 1992). factors have been most heavily studied, perhaps due to the
Parental monitoring is another protective factor that has earlier conceptualization of resilience as a ‘‘personal’’ trait.
been investigated in older children and adolescents. The majority of protective factors have been found to
Research shows that adolescents whose parents are familiar help across levels of risk, sometimes with an increased
with their friends and know their child’s activities and benefit for children at high levels of risk. Some protective
whereabouts are less likely to engage in deviant behavior factors may help at-risk children more than low risk chil-
(Dishion and McMahon 1998), be diagnosed with a psy- dren because low risk children may not need as many
chiatric disorder (Tiet et al. 1998, 2001), or display resources to have positive outcomes, given that they have
problems across a range of areas (Buckner et al. 2003). fewer stressors to contend with. However, some studies of
Once again, however, monitoring does not always coun- particularly high-risk children and adolescents (e.g., those
teract high levels of risk (Sullivan et al. 2004), suggesting living in the inner-city or low-income households), suggest
that while parental monitoring is important, it may not be that certain factors may not provide protection at the
enough to overcome other prominent risk factors. highest levels of risk (e.g., Luthar 1991; Sullivan et al.
2004). Theoretically, this makes sense because it seems
Community-Level Protective Factors unlikely that a single protective factor would be able to
counteract the impact of so many interrelated risks. The
Although community-level protective factors have been next section discusses issues related to potential constraints
less extensively studied than attributes of the child and upon resilience at high levels of risk in more depth.
family, they are also important for child outcomes.
Neighborhood quality (Barbarin et al. 2006), neighborhood
cohesion (Gorman-Smith et al. 2000; Jaffee et al. 2007; Potential Theoretical Constraints on Resilience
Kliewer et al. 2004; Li et al. 2007), youth community
organizations (Cauce et al. 2003), quality of the school Why might resilience be constrained in the context of extreme
environment (Ozer and Weinstein 2004), and after-school or severe risk? Two potential reasons have to do with

123
36 Clin Child Fam Psychol Rev (2008) 11:30–58

the nature of both risk and protective factors. First, risk tends In addition to the cumulative nature of risk, the high
to be cumulative and stable (Rutter 2000), thereby magni- continuity over time also magnifies its impact. A longitu-
fying the negative consequences associated with it. Second, dinal study of cumulative risk by Sameroff et al. (1993)
protective factors appear to be less frequently identified at found that the stability of risk between ages 4 and 13 was
the highest level of risk (Luthar and Goldstein 2004). .77, rivaling the stability of IQ, which is generally con-
sidered to fluctuate very little. Intuitively, the longer a child
Cumulative Risk is exposed to high levels of risk, the higher the chances that
important developmental processes will be disrupted and
Although the association between individual risk factors behavior will be impaired. Indeed, studies of Romanian
and negative outcomes tends to be relatively small, it is orphans who experienced extreme deprivation prior to
rare for risk factors to exist in isolation (Rutter 2000). For adoption demonstrate that the likelihood of pervasive,
example, living in a low-income neighborhood is associ- negative outcomes across a variety of domains (e.g.,
ated with lower educational attainment, exposure to behavior problems, attachment disorders, cognitive delays,
deviant peers, decreased access to resources, and higher attention problems) increased the longer the children lived
levels of negative life events (Leventhal and Brooks-Gunn in the orphanages (see Maclean 2003, for a review).
2000). Relatedly, environmental risks and genetic risks Obviously, this is an extreme example, but similar dose-
often covary as well. For example, the well-documented response findings have emerged from studies of children
link between maternal depression and negative child out- living in poverty, with those experiencing extreme or
comes is likely due to a combination of genetic and chronic poverty exhibiting worse outcomes than children
environmental factors. Extended twin studies, which exposed to less severe or intermittent poverty (Duncan
include monozygotic and dizygotic twins, as well as their et al. 1994; Korenman et al. 1995).
parents, have demonstrated that there is a genetic compo-
nent to the intergenerational transmission of depression Protective Factors at the Highest Levels of Risk
(Rice et al. 2002, 2005, 2006). However, depressed
mothers are also more likely to display higher rates of Not only do risks covary and generally remain fairly stable
negativity, coercive control, inconsistency, and unrespon- over time, but they also can decrease the likelihood of
sivity than non-depressed mothers when parenting their protective factors. Several studies have shown that children
children (Goodman and Gotlib 1999); these styles are, in at higher levels of risk have significantly lower levels of
turn, associated with negative child outcomes. Similarly, protective factors (Dubow et al. 1997; Farber and Egeland
some researchers have argued that families ‘‘select’’ envi- 1987). In particular, potential child protective factors are
ronments, such that families at high genetic risk for greatly impacted by the environment (Luthar and Cicchetti
externalizing or internalizing symptoms tend to cluster in 2000). For example, a child who does not receive cognitive
poor neighborhoods (Plotnick and Hoffman 1999; Rowe stimulation and appropriate caregiving in the home may be
and Rodgers 1997), which are also associated with negative less likely to demonstrate high intelligence than another
child outcomes (Leventhal and Brooks-Gunn 2000). Thus child without such risks. Similarly, the likelihood of a child
children are often exposed to a ‘‘double whammy’’ of risk retaining an internal locus of control when he or she is
factors, both environmental and genetic. experiencing a high number of uncontrollable, chronic
Furthermore, many studies have demonstrated that stressors is greatly reduced compared to a child who is
cumulative risk is highly associated with negative out- accustomed to life going smoothly. Even potential pro-
comes, and that the probability of a negative outcome tective factors outside of the child can be affected by the
increases as the number of risk factors increases (Fergusson larger environmental context. For example, parenting can
and Lynskey 1996; Kolvin et al. 1988; Rutter 2000). In a be influenced by a number of factors, including work sit-
sample of 4-year-old children, an index of cumulative risk uation, income, social support, and daily stressors (Belsky
explained three times the variation in outcomes compared 1984). A parent who is concerned with having enough
to individual risk factors (Sameroff et al. 1987). In fact, money for food and has little social support may have more
cumulative risk scores predicted outcome even after SES, difficulty providing his or her child with warm, sensitive
minority status, and maternal IQ were partialled out, sug- parenting.
gesting that the type of risk factor matters less than the Even in the context of identified protective factors,
number of risks (Sameroff et al. 1993). Perhaps even more higher risk samples (e.g., low SES, multiple risks) may
startling, another study found that rates of crime recidivism demonstrate lower rates of resilience than would be
increased drastically as the number of risks increased, from expected because protective factors may not equally benefit
11% recidivism with no family risk to 47% with five risks children across various levels of risk. For example, a study of
(Stattin et al. 1997). 97, predominantly ethnic minority, urban boys (ages 6–10)

123
Clin Child Fam Psychol Rev (2008) 11:30–58 37

with adjudicated older brothers examined the effects of Finally, a study of a nationally representative sample of
community violence exposure on antisocial behavior 1,116 twin pairs in the UK examining maltreatment found
(Miller et al. 1999). The authors found that among this that the protective effects of high IQ and positive temper-
sample of high-risk boys, low levels of family conflict were ament disappeared once cumulative family stressors were
only associated with lower levels of antisocial behavior in examined (Jaffee et al. 2007). This suggests that while
the context of low community violence exposure (Miller certain factors may be associated with resilience, they may
et al. 1999). Thus, living in a family low in conflict was not lose their ability to counteract risk once it reaches a certain
a protective factor for antisocial behavior when community level.
violence exposure was high. In line with this result, another In short, there is significant evidence that protective
study of urban African American youths found that family factors do not always generalize across levels of risk.
support was less important in the context of high levels of Importantly, this does not mean that there are no protective
either violence exposure or hassles (Li et al. 2007). factors that benefit children exposed to severe levels of
Using a range of protective factors and outcomes (e.g., risk; indeed, many of the studies above also found evidence
internalizing/externalizing, drug initiation, adaptive func- of main effects (i.e., protective factors beneficial at all
tioning, school achievement), several articles from a levels of risk). However, it is important to note that there
special series on community violence exposure also found do appear to be limits to the effects of some protective
that not all protective factors were beneficial for children factors in the highest level of risk (e.g., low SES, multiple
who had been exposed to high levels of violence (Ham- risks), which suggests that it may be difficult for children
mack et al. 2004; Kliewer et al. 2004; Sullivan et al. 2004). exposed to severe adversity to demonstrate positive
The samples were predominantly low-income, urban, eth- outcomes.
nic minority preadolescents (Hammack et al. 2004; In summary, given that risks tend to covary and that
Kliewer et al. 2004), but one study of children living in cumulative risk is highly predictive of negative outcomes,
rural poverty also found that there were fewer protective as well as the fact that protective factors are less frequent
factors at the highest level of exposure (Sullivan et al. in situations of high risk, children at the highest levels of
2004), suggesting that such findings are not limited to risk appear to have rather low odds for success. Rather, one
urban settings. Although it should be noted that main would expect rates of positive outcomes to be considerably
effects were most common (i.e., protective factors worked lower at the highest levels of risk compared to lower levels
similarly across levels of risk), there was only one study of risk. Furthermore, when positive outcomes are identified
from this special series that found protective factors to be at the highest levels of risk, one would expect them to be
more important at high levels of risk (Ozer and Weinstein qualified across time and domains of adjustment. With
2004). these hypotheses in mind, the next section reviews the
Several studies have found that some protective effects extant literature on resilience, with particular emphasis on
are diminished in the context of neighborhood poverty resilience at the highest levels of risk.
(Silk et al. 2007; Stouthamer-Loeber et al. 2002; Vander-
bilt-Adriance and Shaw in press), with differences
emerging even between low-income urban neighborhoods Literature Review
and inner city neighborhoods or projects (Gorman-Smith
et al. 1999; Shaw et al. 2004). Results from the Pittsburgh Rutter (2000) has commented that although there are many
Youth Study, a longitudinal study of public school boys studies which are relevant to resilience, the number of
oversampled for high levels of antisocial behavior are studies that directly compare resilient and non-resilient
consistent with the above findings (Stouthamer-Loeber groups is fairly limited. This is particularly relevant to the
et al. 2002). The authors examined the overall balance of present review since one of the main goals is to compare
risk and protective factors and found that, at least for older rates of resilience in children exposed to higher versus
adolescents, a score indicating higher levels of protective lower rates of risk. Such percentages can only be deter-
factors and lower levels of risk factors was not entirely mined if researchers utilize person-centered approaches,
protective for those living in disadvantaged neighborhoods. where children are divided into groups based on risk and
Twenty-two percent of these boys were classified as outcome status. In addition, this review also focuses on
‘‘serious, persistent delinquents’’ at age 19, indicating that rates of resilience over time and across domains, further
they had repeatedly engaged in crimes such as robbery, limiting the number of relevant studies. Given the specific
assault, or selling drugs. The authors suggest that risks for nature of the question at hand (i.e., the nature of resilience
adolescent boys from disadvantaged neighborhoods may be at the highest levels of risk), the following review limits
of larger magnitude and, therefore, may be more likely to discussion to studies that fall into three categories: (1)
overwhelm protective factors. studies reporting prevalence rates of positive outcomes; (2)

123
38 Clin Child Fam Psychol Rev (2008) 11:30–58

studies that examine positive outcomes across time; and (3) For example, Masten et al. (1999) followed a commu-
studies that examine positive outcomes across multiple nity sample of children from elementary school through
domains of functioning. With these constraints in mind, a early adulthood and found that 57% of children exposed to
search was conducted in Ovid psychINFO using the key- high levels of negative life events were judged ‘‘resilient’’
words ‘‘resilience’’ and ‘‘protective factors’’; relevant on measures of childhood and adolescent competence. A
studies were also selected from references in review cross-sectional study of 1,285 children from a household
papers. It is important to keep in mind that not all studies of probability sample found that 62% of the girls and 50% of
resilience met the criteria for the current review, and the boys whose mothers displayed significant psychopa-
therefore may have been excluded. thology were judged to be resilient (Tiet et al. 2001). An
earlier study utilizing the same sample found that when
children experienced both maternal psychopathology and
Rates of Resilience in Lower Versus Higher Risk
high levels of negative life events, 40% of them still had
Contexts
positive outcomes (Tiet et al. 1998). It should be noted,
however, that a positive outcome was defined as the
There has been great variability in the operationalization of
absence of a psychiatric disorder and the presence of good
risk, ranging from children who have experienced a neg-
functioning on a psychiatric assessment. It is still possible
ative life event, such as divorce, to children who have
though that these children evidenced significant problems,
experienced chronic poverty, community violence, and any
albeit not a psychiatric diagnosis, so this is likely not the
number of related risk factors. While there are certainly
best measure of their overall functioning. Furthermore, as
negative consequences for some children related to life
this was a cross-sectional study, there is no prospective
events such as parental divorce, the experience of divorce
measurement of their functioning over time.
can vary greatly from child to child, with some being
In another study of parental psychopathology, Radke-
exposed to a high number of related risks and others
Yarrow and Brown (1993) found that 41% of their sample
maintaining a relatively low-risk environment. Arguably,
of middle- to upper-middle-class children of psychiatri-
children living in poverty in the inner city are less able to
cally ill parents displayed positive outcomes (e.g., lack of
avoid risk exposure due to the all-compassing nature of the
psychiatric diagnosis or borderline criteria). Although the
risk. Similarly, cumulative risk indices also ensure signif-
sample size for this study was considerably smaller than
icant exposure to risk. As many studies of resilience have
the previous study, it has several important strengths that
utilized relatively lower risk samples of white, middle class
should be noted. First, in addition to a diagnosis of severe
children (Masten et al. 1999; Radke-Yarrow and Brown
maternal depression (e.g., early onset, multiple severe
1993) or examined single risks (White et al. 1989), it is
episodes), the authors also required a paternal diagnosis of
possible that such studies have overestimated the percent-
depression, anxiety, or substance abuse; the presence of
age of children with resilient outcomes. The following
affective illness in first- or second-degree relatives of one
section compares rates of positive outcomes in the context
or both parents; and high levels of chronic stress or chaos
of lower versus higher risk.
within the family. Thus children were at extremely high
levels of both environmental and genetic risk for psychi-
Lower Risk Contexts: Single Risk Factors and Middle atric diagnoses. Second, children had to demonstrate good
Class Samples functioning across four assessments over a period of
10 years in order to ensure that children defined as
Rates of positive outcomes vary greatly from study to ‘‘resilient’’ were consistently doing well. Despite the
study, depending on sample demographics and the opera- stringent criteria for assessing both risk and positive out-
tionalization of risk. When single measures of risk are come, it is striking that 41% of these children were still
used, or samples consist of predominantly white, middle- without a diagnosis. This is perhaps accounted for by the
class samples, rates of positive outcomes are considerably fact that they were relatively privileged in other ways (e.g.,
higher (see Table 1) than those found in studies of multiple high SES).
risks or in demographically at-risk samples (e.g., ethnic Finally, a retrospective study of a nationally represen-
minority status, low SES). Although studies of predomi- tative UK sample found that 45% of adults who reported
nantly white samples with single risk factors have found childhood maltreatment were ‘‘resilient,’’ based on the
rates of positive outcomes ranging from 25% (Jaffee et al. absence of lifetime psychiatric diagnoses (Collishaw et al.
2007) to 92% (White et al. 1989), the majority of studies 2007). Interestingly, another study of a representative UK
report rates of 40–60% (Collishaw et al. 2007; Kandel et al. sample found only 25% of maltreated children had positive
1988; Lin et al. 2004; Masten et al. 1999; Radke-Yarrow outcomes, defined as at or below the median on teacher-
and Brown 1993; Tiet et al. 1998, 2001). rated behavior problems at ages 5 and 7 (Jaffee et al. 2007).

123
 Table 1 Studies of predominantly white, middle class children or single risk factors
Authors Sample Design Risk Protective factors Outcome Results

Collishaw et al. N = 378 total, N = 44 Longitudinal Retrospective Perceived parental care Absence of lifetime 45% resilient
(2007) maltreated (Ages 14–15 reports of Adolescent peer relationships psychiatric diagnoses Appeared to be doing well across
Ages 42–46 to ages 42– childhood domains of functioning
Quality of adult love relationships
46) maltreatment
Isle of Wight Personality style
epidemiological study
Jaffee et al. N = 1,167 twin pairs Longitudinal Retrospective IQ At or below the median 25% resilient
(2007) Ages 5 and 7 (Ages 5–7) parental report Positive temperament on teacher-reported Doing well across domains
of childhood behavior problems at
UK representative sample Absence of parental psychiatric 1/3 of resilient children at
maltreatment ages 5 and 7
symptoms age 5 not resilient at age 7
Neighborhood safety & cohesion
Clin Child Fam Psychol Rev (2008) 11:30–58

Lin et al. (2004) N = 179 Cross-sectional Parental Parental warmth, discipline, & Below clinical cut-off 44% resilient
Ages 8–16 bereavement mental health on parent-, child-, and Main effects for all parent
Child self-esteem, self-efficacy, teacher-reported measures, child threat
62% white
appraisal of threat, unknown internalizing and appraisal and self-efficacy
Middle class community externalizing
control beliefs, & active
sample recruited for a symptomatology
inhibition of emotional
prevention program
expression
Kandel et al. N = 94 adult males Cross-sectional Father received IQ Absence of jail sentence 60% resilient
(1988) Subset of 1930s Danish at least one or recorded offenses Main effect and interaction
male birth cohort jail sentence of IQ and risk
Highest IQ for resilient group
Masten et al. N = 202 Longitudinal Life events IQ Conduct problems 57% resilient
(1999) Ages 17–23 (ages 7–12 Parenting quality Academic achievement Main effects and interactions
through ages
73% white Social competence Resilient group low on
17–23)
Normative school sample Psychological well-being internalizing
Radke-Yarrow N = 63 Longitudinal Severe familial IQ, favored child status, No psychiatric diagnosis 41% resilient
et al. (1993) Ages 11–13, 15–18 (followed psychopathology, positive self-perception, over course of study But 56% resilient children had
over 10 years) high chronic good relationships with somatic complaints, low self-
Predominantly white,
stress teachers and peers, coping, confidence, poor coping
middle to upper middle
physical health, strategies
class
temperament, social
Subsample of the NIMH Main effects for a number of
support
study of offspring of protective factors, particularly
affectively ill and well those related to social
parents relationships
39

123
 40

Table 1 continued
Authors Sample Design Risk Protective factors Outcome Results

123
Tiet et al. (2001) N = 1,285 Cross-sectional Maternal Child protective factors: gender, No psychiatric disorder 62% high risk girls and 50%
Ages 9–17 psychopathology IQ, educational aspiration, or high risk boys = resilient
physical health, low adverse IQ more important in context
51% white Functional impairment
life events of maternal psychopathology
Household probability
Family protective factors: Main effects for parental
sample (MECA)
SES, lack of paternal monitoring, family
psychopathology, family functioning, educational
structure, marital relationship aspiration, and gender
quality, parental monitoring,
family functioning, # adults
in family
Tiet et al. (1998) N = 1,285 Cross-sectional Negative life Child protective factors: gender, No psychiatric disorder 40% resilient, even when high
Ages 9–17 events IQ, educational aspiration, or negative life events and
physical health maternal psychopathology
51% white Functional impairment
Family protective factors: Main effects for parental
Household probability
SES, lack of maternal monitoring, IQ, family
sample (MECA)
psychopathology, family functioning, # adults in
structure, marital relationship family, educational aspiration.
quality, parental monitoring,
family functioning, # adults
in family
White et al. N = 976 Longitudinal Antisocial behavior IQ Juvenile delinquency 84% high risk boys & 92% high
(1989) Age 15 (ages 5–15) at age 5 at ages 13, 15 risk girls had nondelinquent
outcomes
Predominantly white
Main effect of IQ, no interaction
1970s Dunedin, NZ, birth
cohort
Clin Child Fam Psychol Rev (2008) 11:30–58
Clin Child Fam Psychol Rev (2008) 11:30–58 41

It is likely that method differences account for the dis- low scores on externalizing symptomatology at ages 15 and
crepancy in resilience rates. For example, requiring 16. However, the authors also noted that whereas the
individuals to be low on symptomatology, rather than resilient children had high cumulative risk scores by defi-
merely diagnosis-free is a much more stringent definition nition, their overall levels of risk were significantly lower
of resilience. Furthermore, requiring positive outcomes than children who had less positive outcomes. Furthermore,
across time and informants also decreases the likelihood when they examined outcomes for adolescents from the top
that a child will be considered ‘‘resilient’’ (Jaffee et al. 5% most disadvantaged backgrounds, they found that the
2007). In fact, given that the comprehensive way in which likelihood of being problem-free at age 15 was only 13%
‘‘positive outcome’’ was defined, it is remarkable that 25% (Fergusson et al. 1994). In comparison, 80% of adolescents
of the high-risk subsample still met classification criteria. from the top 50% most advantaged backgrounds were
In sum, studies of single risk factors and studies utilizing problem-free, and the likelihood of an adolescent from an
predominantly white, middle-class samples tend to find advantaged background having multiple problems was 1 in
rates of positive outcomes ranging from approximately every 400–500 (Fergusson et al. 1994).
30–90%, with the majority clustering around 40–60%. This One of the longest running resilience studies followed
wide range of outcomes across studies clearly illustrates the 1955 birth cohort on Kauai from birth to middle
the difficulty inherent in attempting to summarize research adulthood (Werner and Smith 1982, 1992). The authors
on resilience, even among relatively lower-risk children. defined risk status as having four or more risks, covering a
The considerable heterogeneity with which risk and posi- range of domains including demographic factors, child
tive outcome are operationalized contributes to differences physical health and behavior, and family problems;
in results. How does one pull together results from off- approximately half of the families were also living in
spring of psychiatrically ill parent with results from poverty. Twenty-six percent of the high-risk subsample
adolescents at risk due to high rates of childhood antisocial demonstrated positive outcomes on behavioral, mental
behavior? Despite the wide variety of differences, studies health, and learning problem measures at age 18. This rate
of children with single risk factors or from predominantly is higher than most of the other multiple risk studies dis-
white, middle-class backgrounds generally have relatively cussed above, but there are a number of potential
high rates of positive outcomes; as will be demonstrated in explanations for this difference. First, it has been noted that
the next section, these rates are consistently higher than more recent studies of low-income samples have demon-
among children who come from impoverished backgrounds strated considerably worse outcomes than the Kauai study
or experience multiple risks. (Egeland et al. 1993), perhaps because of societal changes
affecting the experience of poverty since the 1950s. Ege-
Higher Risk Contexts: Multiple Risk Factors and land and colleagues suggested that poverty might be
Impoverished Samples associated with a higher number of risks than in the past
due to increases in single parents, divorce, and substance
Indeed, studies examining the impact of multiple risks or use. Furthermore, although approximately half of the
utilizing impoverished samples are much less optimistic in families in the Kauai study were living in poverty, there
their findings (see Table 2). The Rochester Longitudinal was excellent health and prenatal care available when this
Study, which studied children from birth through early study commenced (Werner and Smith 1982), which is
adulthood, utilized a cumulative risk score and found that certainly not the case for most low-income families
only 3 out of 50 high-risk children were above the sample currently.
mean on positive outcomes at age 13 (Seifer et al. 1992). More recent studies of multiple risks in the context of
The authors point out that these three children all experi- poverty have found even more disheartening results. For
enced decreases in their risk scores over time, suggesting example, a study of low birthweight, premature infants
that perhaps their more positive outcomes were actually from predominantly ethnic minority families living in
due to lower levels of risk, rather than protective factors poverty found that only 12% of the children had met
enabling them to ‘‘overcome’’ risk. Regardless of the rea- appropriate developmental milestones in cognitive, health,
son for positive outcomes, however, the small number of and behavioral domains at age 3 (Bradley et al. 1994). In
high-risk children who achieved outcomes at the sample comparison, 40% of low-birth weight, premature infants
mean is striking. not living in poverty had met these same milestones by age
The Christchurch Health and Development Study, a 3. Thus one can see that, as the number of risk factors
16-year longitudinal study of a New Zealand birth cohort, increase, the likelihood of positive outcomes decreased
also utilized a cumulative risk approach (Fergusson and drastically. Furthermore, the authors determined that in
Lynskey 1996). They found that approximately 37% of order for children at the highest level of risk to be
high-risk children had positive outcomes as measured by considered ‘‘resilient,’’ they needed to have three or more

123
 42

Table 2 Studies of children from low SES backgrounds or multiple risks

Authors Sample Design Risk Protective factors Outcome Results

123
Bradley et al. N = 243 Longitudinal Low-income Physical aspects of caregiving Cognitive and behavioral 12% highest risk vs. 40% non-poverty
(1994) Age 3 (birth—age 3) Preterm environment competence; health and children had positive outcomes
Parental acceptance and growth status at age 3 Only 15% have good outcomes
85% ethnic minority Low birthweight
responsivity Normal range on all four with [3 protective factors
Preterm, low
measures
birthweight infants
living in poverty
Buckner et al. N = 155 Cross-sectional Low-income IQ, self-regulatory skills, Composite including behavior 29% resilient
(2003) Ages 8–17 self-esteem, social support, problems, mental health Nonresilient children had more
& parental monitoring, symptoms, adaptive negative events, chronic stressors,
35% white
functioning, & competence and abuse
Low-income families
in Worchester, MA Resilient did not differ from non-
resilient on social support or IQ
Cicchetti et al. N = 206 Cross-sectional Maltreatment Ego control 7 measures of adaptive Resilient = 18%
(1993) Ages 8–13 Low SES Ego resiliency functioning Maltreated more disruptive/
69% ethnic minority, Self-esteem Adaptive composite aggressive, withdrawn,
59% below poverty internalizing, and lower overall
IQ
line, 83% on public competence, lower IQ and ego
assistance resiliency
Number of maltreated in resilient
group = nonmaltreated, but more
maltreated in lower functioning
groups too
Cicchetti et al. N = 213 Longitudinal Maltreatment Ego control 7 measures of adaptive Replicated above results. Resilience
(1997) Ages 6–11 (over 3 years) Low SES Ego resiliency functioning over 3 years = 1.5%. 10%
Adaptive composite maltreated had no indicators of
81% ethnic minority, Self-esteem
competence at any timepoint
87% on public IQ
assistance Relationship factors important for
Relationship quality with nonmaltreated only
mother and camp counselor
Cicchetti et al. N = 677 Cross-sectional Maltreatment Ego control 7 measures of adaptive Maltreated resilience = 6.1%.
(2007) Ages 6–12 Low SES Ego resiliency functioning Nonmaltreated
Adaptive composite resilience = 11.8%
81% ethnic minority, Cortisol
94.8% had All protective factors independently
DHEA
received public predicted resilience
assistance
Fergusson et al. N = 942 Longitudinal Cumulative risk N/A Multiple problem outcomes Only 13% of adolescents from 5%
(1994) Age 15 (birth—age 15) score (early sexual activity, most disadvantaged childhoods
conduct/oppositional were problem-free
Birth cohort from
disorder, police contact for Chances of children from advantaged
Christchurch, NZ
offending, cannabis use, backgrounds having multiple
alcohol use) problems was 1 in every 400–500
Clin Child Fam Psychol Rev (2008) 11:30–58
 Table 2 continued
Authors Sample Design Risk Protective factors Outcome Results

Fergusson et al. N = 940 Longitudinal Cumulative risk Child protective factors: IQ, Substance use 37% resilient
(1996) Age 16 (birth—age 16) score emotional/behavioral problems, Delinquency Resilient children did not differ from
temperament, interests, close maladjusted on internalizing
Birth cohort from School problems
relationships symptoms
Christchurch, NZ
Family protective factors: Lower family adversity
Parenting, parental attachment,
Main effects for IQ, novelty seeking,
home environment
deviant peer affiliation
Luthar et al. N = 360 Cross-sectional Maternal Negative parenting behavior Average social competence 7–21% resilient
(2007) Ages 8–17 diagnosis of Limit setting Internalizing & externalizing 23% resilient if no maternal diagnosis
affective symptomatology
2/3 ethnic minority Closeness Negative parenting most assoc w/ neg
disorder or drug
Clin Child Fam Psychol Rev (2008) 11:30–58

Approximately 50% Low parenting stress child outcomes

use disorder
on welfare
Low SES
Seifer et al. N = 152 Longitudinal (ages Cumulative risk Child and mother personality Changes scores from ages 4–13 Only 3/50 high risk kids above
(1992) Age 13 4–13) score disposition, social support, and on IQ and socio-emotional sample mean at age 13
family cohesion indices Most protective factors supported
*50% low SES,
*60% white across risk status
Subsample of Some interactions suggesting more
Rochester importance in the context of risk
Longitudinal Study
Stouthamer- N = 506 Longitudinal (ages Serious persistent Low physical punishment Absence of serious persistent 40% resilient
Loeber et al. Age 25 13–25) delinquency in Employed or in school delinquency at follow-up However, 56% of those individuals
(2004) adolescence continued to offended at lower
*50% African
American rates
*40% on public Showing difficulties in other domains
assistance too
Werner et al. N = 505 Longitudinal Cumulative risk Child protective factors: e.g., Delinquency 26% resilient
(1982, 1992) Predominantly ethnic (birth to middle score temperament, IQ Mental health problems Many child and family protective
minority adulthood) Family protective factors: e.g., factors
Judgment of ‘‘doing well’’
54% poverty parent–child relationship across domains Rates of somatic & physical
quality, parenting complaints 29 higher for
1955 Kauai birth
cohort ‘‘resilient’’ group
43

123
44 Clin Child Fam Psychol Rev (2008) 11:30–58

protective factors. However, even with three or more pro- Other samples of predominantly low-income, ethnically
tective factors, children at the highest level of risk still had diverse samples also show a high rate of problems among
very low rates of positive outcomes; with fewer than three at-risk children (Buckner et al. 2003; Luthar and Sexton
protective factors, none of the children at the highest level 2007; Stouthamer-Loeber et al. 2004). For example, a
of risk had positive outcomes. This finding further study of low-income children of mothers with psychiatric
emphasizes the difficulty of achieving positive outcomes at diagnoses found that only 7–21% of the children displayed
the highest level of risk. positive outcomes (i.e., average levels of social compe-
A study of maltreatment in low-income children found tence and low externalizing or internalizing symptoms;
that 18% of children in the maltreated group had positive Luthar and Sexton 2007). Sadly, even among a control
outcomes (Cicchetti et al. 1993). Seven domains of adap- group of children whose mothers had no diagnoses, only
tive functioning were measured, and children who were 23% had positive outcomes, demonstrating that although
within the top third of adaptive functioning on four having a mother with a psychiatric illness is detrimental, so
domains were considered competent. Interestingly, while is living in a low-income family. Another study examining
maltreated and nonmaltreated children were equally like to desistance from persistent serious delinquency found that
be in the competent group, maltreated children were 40% of adolescents had desisted by young adulthood
overrepresented in the lowest functioning group that had (Stouthamer-Loeber et al. 2004). However, closer exami-
zero or only one domain of competent functioning. Fur- nation reveals that over half of those individuals had
thermore, maltreated children were higher than committed criminal offenses, albeit at lower levels, sug-
nonmaltreated children on continuous measures of disrup- gesting that in fact positive outcomes were much lower, on
tive/aggressive behavior, withdrawal, internalizing, and par with the previous study.
total competence. Rates of positive outcomes were higher In summary, although rates of positive outcomes in
in this sample than in Bradley et al. (1994) study, which studies of higher-risk children (e.g. multiple risks, low
also utilized a low-income sample. However, it is impor- SES) range widely from 1.5 to 40%, positive outcomes are
tant to note that Cicchetti et al. (1993) based on their generally much less common in these studies than in those
definition of positive outcomes on relative standing with utilizing white, middle-class samples or single risk factors.
other high risk, low- income children in the sample, Only two studies found prevalence rates over 35%, while
whereas Bradley and colleagues used cut-offs from nor- nine studies found that approximately a quarter or less of
mative samples. Given that low-income is a risk factor in the high-risk group was resilient. Importantly, of the two
and of itself, and that 83% of the overall sample was studies finding higher rates of resilience, one found that
receiving public assistance, one would expect that children approximately half of the ‘‘resilient’’ group was still
in the maltreated group would fare even more poorly in exhibiting problems (Stouthamer-Loeber et al. 2004),
comparison with a normative sample. While the authors’ suggesting that in fact resilient outcomes above 25% are
decision to compare low-income groups makes sense in quite rare in higher risk samples.
terms of parsing out the additional risk associated with As mentioned above, resilience rates fluctuate greatly
maltreatment versus low-income alone, it is likely that the depending on the sample demographics (with white, mid-
children in the maltreatment group were exhibiting lower dle class children faring best), number of risks, and the type
rates of competence in comparison with normative and number of outcomes measured. Certainly, the more
samples. outcomes that are measured and the more stringent the
The results of Cicchetti et al. (1993) were replicated in requirements for ‘‘positive outcome,’’ the lower the number
two other studies of similar groups of low-income children of children who can be considered resilient. Although it is
who varied on maltreatment status (Cicchetti and Rogosch not possible to arrive at a normative rate of resilience due
1997, 2007). For example, when the adjustment group to the substantial variability between studies on method-
classification was averaged across three consecutive yearly ology and measurement, it does seem clear that there are
assessments, only 1.5% (n = 2) of the maltreated children significant differences between studies based on their
were classified as ‘‘competent’’ (top third of functioning in degree of risk, with considerably more constraints upon
four or more domains), versus 41% of the nonmaltreated resilience in the context of multiple, high risks.
children. Only 10% of the maltreated children were ever
classified in the competent group at any of the three Resilience Across Time
timepoints. Perhaps even more striking, 10% of the mal-
treated children exhibited no competence in any of the There are a limited number of studies that examine conti-
seven domains at any of the three timepoints. These results nuity and discontinuity in resilience over time, but those
present the stark reality of the detrimental effects of mal- that do generally demonstrate that resilience is not stable
treatment, particularly in the context of low-income. (see Table 3). For example, the Rochester Longitudinal

123
 Table 3 Studies of resilience across time
Authors Sample Design Risk Protective factors Outcome Results

Farrington et al. N = 411 males Longitudinal Cumulative risk N/A Nine criteria of competent 50% of ‘‘resilient’’
(1988a, b) Age 32 (ages 8–32) score functioning (e.g., successful adolescents convicted of a
employment, cohabitation, crime by age 32
Working-class London
absence of deviant Unconvicted high-risk men
families
behavior, etc.) often had the worst
outcomes
Farber et al. N = 44 maltreated Longitudinal Maltreatment Low Infant temperament and behavior Developmental competence Low continuity of resilience
(1987) N = 88 non-maltreated (12–48 mos) SES Parental characteristics, parenting (e.g., attachment, (none consistently
problem-solving, behavior) competent from 12 to 48
Clin Child Fam Psychol Rev (2008) 11:30–58

Age 48 mos knowledge, parent–child

interaction quality, life stress mos)
Low SES
Decrease in competence over
Subsample of children from time
the Minnesota Mother–
Some protective factors
Child Interaction Project
overall, but not for abused
children
Felsman et al. N = 456 non-delinquent, Longitudinal Low SES Childhood strengths: e.g., Global mental health in middle Anecdotal evidence of
(1987) inner city, adolescent (ages 12–16 to relationship quality with family adulthood (e.g., social ‘‘enormous discontinuity’’
males middle members, school/social competence, employment, in competence over the
Recruited 1940–1944 adulthood) adjustment, physical health, IQ, happy marriage, income) lifespan
etc.
Low SES
Adulthood strengths: e.g., object
Matched w/ reform school
relations, Erickson’s life stage,
boys on IQ, age,
maturity of defenses, SES, etc.
neighborhood crime rate,
ethnicity
Jaffee et al. N = 1,167 twin pairs Longitudinal Retrospective IQ At or below the median on 25% resilient
(2007) Ages 5 and 7 (ages 5–7) parental report Positive temperament teacher-reported behavior Doing well across domains
of childhood problems at ages 5 and 7
UK representative sample Absence of parental psychiatric 1/3 of resilient children at age
maltreatment
symptoms 5 not resilient at age 7
Neighborhood safety & cohesion
Masten et al. N = 173 Longitudinal Life events IQ Age-appropriate competence Continuity in resilience
(2004) Ages 28–36 (ages 8–12 to Parenting quality (e.g., academic achievement, Resilient children had more
28–36) social competence, conduct) protective factors, higher
73% white Adapative resources (e.g., coping,
motivation, support) SES
Normative school sample
Childhood/adol adversity
only modestly associated
with young adult success
45

123
46 Clin Child Fam Psychol Rev (2008) 11:30–58

Study found that competence (e.g., IQ, mental health) at

High stability of risk over

one time point was not related to competence at a later

‘‘recoveries’’ exhibited

Resilient group high on

internalizing, social
age (Sameroff 1998; Sameroff et al. 1987). Rather, the

Child ability severely

environmental risk
level of risk was the most significant predictor of positive
25% of adolescent

illegal behavior

undermined by
isolation, etc.
outcomes over time. As reported by Sameroff (Sameroff
1998, 2005), children were divided into groups based on
their cumulative risk score (e.g., number of family and
Results

time
sociodemographic risk factors) and various competence
measures, such as cognitive ability (e.g., IQ scores) and
development, behavior, etc.) mental health (e.g., social-emotional functioning, psychi-
Measures of child functioning

atric symptomatology), and then tracked on their outcomes

over time. Although children high on social-emotional
functioning or IQ at age 4 tended to do better on similar
(e.g., IQ, language
Criminal offending

measures at the age 18 follow-up, these associations were

Psychopathology

much smaller in magnitude than those explained by risk

Economic life
Personal life

level at age 4. In fact, children in the high competence,

Personality
Outcome

high-risk group at age 4 had worse outcomes at age 18 than

children in the low competence, low risk group. This
procedure was then repeated at age 13, predicting to age 18
outcomes, with the idea that perhaps age 4 competence was
not stable enough to predict positive outcomes in adoles-
cence. However, similar results were found, regardless of
which time point was used. In sum, the early competence
of high-risk children did not seem to predict to later
Protective factors

competence, suggesting little continuity in the positive

outcomes of high-risk children.
Similarly, a longitudinal study of a New Zealand birth
cohort found questionable continuity over time in positive
N/A

N/A

outcomes (Moffitt et al. 2002). A group of boys who had

evidenced high levels of aggression as children, yet dis-
childhood and/
or adolescence

Cumulative risk

played low levels in adolescence, were termed

behavior in

‘‘recoveries’’ for their apparent desistance. Yet, at age 26,

Antisocial

score

the authors noted that a full 25% of these ‘‘recoveries’’ had

Risk

demonstrated illegal behavior, and were in fact more

appropriately termed low-level offenders. Thus, although
their outcomes looked promising in adolescence, these
(ages 5–26)

(ages 4–13)
Longitudinal

Longitudinal

gains deteriorated over time. Another follow-up of ado-

lescent males at high risk for antisocial behavior found
Design

even more disheartening results, in that almost half of the

resilient adolescents (14 of 31) had been convicted of a
Birth cohort from Dunedin,

crime by age 32 (Farrington et al. 1988a). Other longitu-

*50% low SES, *60%

Subsample of Rochester

dinal studies following individuals from adolescence into

Longitudinal Study
Predominantly white

middle adulthood have also anecdotally reported disconti-

nuity in outcomes over time (Felsman and Vaillant 1987).
N = 477 males

However, perhaps it is expecting too much for high-risk

children to maintain positive outcomes over such long
N = 152

white
Sample

Age 26

Age 13

periods of time. What about shorter follow-ups? A study of

NZ

childhood maltreatment in a representative UK sample

Table 3 continued

found that one-third of the children who were classified as

(1987, 1993)
Sameroff et al.
Moffitt et al.

resilient at age 5 fell into the non-resilient group by age 7

(Jaffee et al. 2007). Another study of a low-income, mal-
(2002)
Authors

treated subsample of children from the Minnesota Mother–

Child Project found even more substantial variability in

123
Clin Child Fam Psychol Rev (2008) 11:30–58 47

positive outcomes over time, such that no high-risk chil- other domains (see Table 4). Luthar and colleagues (Luthar
dren were consistently rated as competent from 12 to 1991, 1995; Luthar et al. 1993) have conducted a number
42 months (Farber and Egeland 1987). In fact, by pre- of studies of ethnically diverse, inner city adolescents and
school, none of the maltreated children displayed found discontinuities across domains of functioning. In a
competent outcomes. Furthermore, there appeared to be widely cited study of 9th graders exposed to high levels of
decreases in competence over time among all low-income negative life events, ‘‘resilient’’ children who were doing
children in the sample, regardless of maltreatment status. well in terms of school-based social competence were also
While the sample size for the maltreated group was rela- found to have high rates of internal distress (Luthar 1991).
tively small (N = 44), this is still an impressive finding, These results were replicated and expanded upon in a 6-
suggesting that continuity in competence in the context of month prospective study of positive adjustment across
such risk is unlikely. Taken together, these two studies domains in another sample of inner-city adolescents (Lu-
suggest that there is substantial variability in resilience, thar et al. 1993). The authors reported that 60% of
even over shorter time periods. adolescents who fell within the upper 1/3 on one measure
There is one exception to this trend of low stability of of competence were in the lowest 1/3 of another measure of
resilience and that is Project Competence, a community social competence. Interestingly, while these measures
sample of predominantly white middle class children in were different aspects of social competence, they were all
Minneapolis. Masten et al. (2004) followed this group of still within the overall domain of school-based social
children from elementary school into adulthood, using a competence, and thus one might expect a greater degree of
measure of negative life events (e.g., death or sickness in continuity between them. Furthermore, when the absence
the family, birth of sibling) to determine risk status, and of emotional distress was included as a necessary compo-
found continuity in resilience over time. There are several nent of a positive outcome, only 15% of the original
explanations, which may account for the fact that this study ‘‘resilient’’ group retained that classification. A third study
found continuity in resilience over time, while other studies of a similar sample found that peer-rated sociability pro-
have not. First, this sample was substantially different from spectively predicted lower indices of school functioning,
the children discussed in the other studies above, which and that low anxiety in girls was related to decreased
were generally low-income and were exposed to arguably performance in school over a 6-month period (Luthar
more severe levels of risk. Second, negative life events had 1995). Thus, the author concluded that although there was
only modest associations with outcomes in young adult- some continuity across domains for academic achievement
hood, and then only in two domains (academic and teacher-rated classroom behavior, it was also true that
achievement and conduct problems), suggesting that per- adolescents with the best interpersonal or emotional
haps the risk associated with negative life events was not as adjustment may also be those who are not doing well in
high in magnitude as other risks. Thus, it is not surprising other aspects of functioning.
that Masten et al. (2004) should find continuity in resil- Another study of inner city middle school students who
ience among children who most likely have more resources had experienced differing levels of negative life events also
to begin with and have experienced an overall lower level found increased rates of internalizing in ‘‘resilient’’ chil-
of risk. In contrast, results from the other six samples dren compared to their lower-risk peers (D’Imperio et al.
reviewed demonstrate discontinuity and even decreases in 2000). In fact, rates of internalizing symptomatology were
competence over time. Interestingly, two studies that found related to risk exposure, rather than competence level, such
discontinuity over time could be considered relatively that differences in the rates of distress for resilient and
lower risk, as they focused on single risk factors and stress-affected children were not statistically significant,
consisted of predominantly white children (Jaffee et al. although ‘‘resilient’’ children actually had higher distress
2007; Moffitt et al. 2002), suggesting that this pattern is not (32% vs. 20%).
always limited to children experiencing the highest level of Longitudinal community studies of antisocial behavior
adversity. Overall, this points to the decreased likelihood in boys have also demonstrated discontinuity across
of sustained resilience over time, particularly in the context domains of functioning. The Dunedin Study found that
of higher risk. although there was a group of boys termed ‘‘recoveries’’
because they ceased to exhibit antisocial behavior in ado-
Resilience Across Domains of Competence lescence, this term may have been overly optimistic
because these boys exhibited problems in adulthood
In addition to findings of discontinuity over time, studies (Moffitt et al. 2002). They were characterized by higher
have also examined positive outcomes across domains and rates of internalizing disorders, with 1/3 formally diag-
found that high-risk children who have positive outcomes nosed with depressive or anxiety disorders. They tended to
in one domain do not necessarily have positive outcomes in be neurotic and socially isolated, and had obtained lower

123
 48

Table 4 Studies of resilience across domains

Authors Sample Design Risk Protective factors Outcome Results

123
Anthony St. Louis Risk Research Longitudinal Parent with Child personality traits Psychological functioning ‘‘Psychological cost’’ to resilience
(1987) Project (15 years) schizophrenia or Ability to distance from parent (diagnosis, severity, Difficulties in intimate
Offspring of parents with manic-depression symptomatology) relationships
schizophrenia or manic- Felt ‘‘strangely unsatisfied’’
depression
D’Imperio N = 185 Cross-sectional Stressful events or Coping, self-perception Above median on 2/3 Protective factors didn’t
et al. (2000) 7–8th graders neighborhood Family cohesion, competence factors differentiate btwn outcomes
disadvantage expressiveness, and conflict (antisocial behavior; school High stress assoc with lower
from disadvantaged, urban
grades, behavior, and protective factors
areas Extrafamilial support
attendance)
18% white Similar rates of internalizing for
Internalizing symptomatology resilient and maladjusted
Farrington N = 411 males Longitudinal Cumulative risk N/A Nine criteria of competent 50% of ‘‘resilient’’ adolescents
et al. (1988a, Age 32 (ages 8–32) score functioning (e.g., successful convicted of a crime by age 32
b) employment, cohabitation, Unconvicted high-risk men often
Working-class London
absence of deviant behavior, had the worst outcomes
families
etc.)
Loeber et al. N = 503 Longitudinal Serious persistent Cognitive abilities Absence of serious persistent Desisters had difficulty with
(2007) Age 20 (ages 7–20) delinquency in Skin conductance delinquency at follow-up anxiety, employment,
adolescence educational attainment
56.4% African American Heart rate
Community, family, and peer
protective factors
Luthar (1991) N = 144 Cross-sectional Stressful life events IQ School-based social Resilient children had high rates of
9th graders Social skills competence (e.g., teacher internal distress
and peer ratings, school Main effects for several protective
77% minority, inner city, low Locus of control
grades) factors
SES Ego development
Internalizing symptomatology IQ not protective at highest levels
Positive life events
of risk
Luthar et al. N = 138 Longitudinal Stressful life events N/A School-based social Children not resilient across
(1993) 9th graders (6 mos) competence (e.g., teacher domains. Resilient children had
and peer ratings, school high internal distress.
85% minority, inner city,
grades) Internalizing/
externalizing
low SES
symptomatology
Luthar (1995) N = 138 Longitudinal Inner city poverty N/A School-based social Not as much cross-domain
9th graders (6 mos) competence (e.g., teacher continuity for inner city kids as
and peer ratings, school found in lower risk samples
84% minority, inner city, low
grades) Peer-rated sociability assoc. with
SES
Internalizing/externalizing lower school functioning,
symptomatology anxiety w/ girls assoc. with
academic achievement
Clin Child Fam Psychol Rev (2008) 11:30–58
 Table 4 continued
Authors Sample Design Risk Protective factors Outcome Results

Masten et al. N = 202 Longitudinal Life events IQ Conduct problems 57% resilient
(1999) Ages 17–23 (ages 7–12 Parenting quality Academic achievement Main effects and interactions
through ages
73% white Social competence Resilient group low on
17–23)
Normative school sample Psychological well-being internalizing
Moffitt et al. N = 477 males Longitudinal Antisocial behavior N/A Criminal offending 25% of adolescent ‘‘recoveries’’
(2002) Age 26 (age 5–26) in childhood and/ Personality exhibited illegal behavior
or adolescence Resilient group high on
Predominantly white Psychopathology
internalizing, social isolation,
Birth cohort from Dunedin, Personal life
etc.
NZ Economic life
Radke-Yarrow N = 63 Longitudinal Severe familial IQ, favored child status, No psychiatric diagnosis over 41% resilient
Clin Child Fam Psychol Rev (2008) 11:30–58

et al. (1993) Ages 11–13, 15–18 (followed psychopathology, positive self-perception, course of study But 56% resilient children had
over 10 high chronic stress good relationships with somatic complaints, low
Predominantly white, middle
years) teachers and peers, coping, self-confidence, poor coping
to upper middle class
physical health, strategies
Subsample of the NIMH study temperament, social support
of offspring of affectively ill Main effects for a number of
and well parents protective factors, particularly
those related to social
relationships
Stouthamer- N = 506 Longitudinal Serious persistent Low physical punishment Absence of serious persistent 40% resilient
Loeber et al. Age 25 (ages 13 to delinquency in Employed or in school delinquency at follow-up However, 56% of those individuals
(2004) 25) adolescence continued to offended at lower
*50% African American
rates
*40% on public assistance
Showing difficulties in other
domains too
Werner et al. N = 505 Longitudinal Cumulative risk Child protective factors: e.g., Delinquency 26% resilient
(1982, 1992) Predominantly ethnic minority (birth to score temperament, IQ Mental health problems Many child and family protective
middle factors
54% poverty Family protective factors: e.g., Judgment of ‘‘doing well’’
adulthood) parent–child relationship
1955 Kauai birth cohort across domains Rates of somatic & physical
quality, parenting complaints 29 higher for
‘‘resilient’’ group
49

123
50 Clin Child Fam Psychol Rev (2008) 11:30–58

rates of education and lower-status occupations. These men children were not without ‘‘covert troubles,’’ including low
were also more likely to engage in drug or alcohol use. self-confidence and the employment of escape and denial
Similarly, the Pittsburgh Youth Study found that even coping strategies. Taken together, these studies illustrate
among those who desisted from serious crime in early the difficulty of maintaining positive outcomes across
adulthood, there still appeared to be detrimental effects in domains.
the realms of educational attainment, cigarette and mari- Once again, however, studies of predominantly white
juana use, unemployment, and anxiety (Loeber et al. 2007; community samples have found more evidence of cross-
Stouthamer-Loeber et al. 2004). Furthermore, although domain competence. For example, a retrospective study of
these individuals had better outcomes across domains in childhood maltreatment in a representative UK sample
general than persisters, they still had more negative out- found that adults classified as ‘‘resilient’’ due to the
comes than less serious or non-delinquents. This suggests absence of psychiatric diagnoses also were functioning
that they were experiencing difficulties in other areas of fairly well in the areas of personal difficulties, criminality,
their lives, despite demonstrating improvement in antiso- poor health, and relationship instability; in fact, they
cial behavior. showed more positive outcomes in these areas than the
Farrington et al. (1988a, b) found that non-delinquent, non-maltreated comparison group (Collishaw et al. 2007).
high-risk adolescents were often the least successful on a Another study of childhood maltreatment that examined
variety of later outcomes, and that conviction status in childhood behavioral outcomes in a representative UK
general had little relation to success in adulthood. High-risk sample found that resilient and non-maltreated children did
men who remained unconvicted at age 32 often had the not vary on measures of mental health, social competence,
worst outcomes on other measures of functioning, includ- or academic achievement (Jaffee et al. 2007). Masten et al.
ing poor home conditions, low paid jobs, and poor family (1999) also found that resilient adolescents seemed to be
relationships (Farrington et al. 1988b). They also tended to doing quite well across domains.
have been socially isolated as children; in fact, having few The overall sample demographics of these studies sug-
or no friends at age 8 was the best predictor of remaining gest that, in general, the samples may have experience
unconvicted (Farrington et al. 1988b). Furthermore, the qualitatively different types of risk than some of the other
men who were rated as successful at age 32 tended to be samples from studies discussed above. Specifically, the
neurotic and of low intelligence in childhood (Farrington Masten et al. (1999) sample was predominantly European-
et al. 1988a), suggesting that there was little relation American, middle class children representative of the
between success in one domain and another, particularly Minneapolis area, while the other two comprised of rep-
over time. resentative samples from the UK (Collishaw et al. 2007;
Similarly, the Kauai Longitudinal Study found that Jaffee et al. 2007). As previously noted, compared with
although participants in the ‘‘resilient’’ group in general did samples of low-income, minority children living in violent
have positive outcomes across domains, they had consider- neighborhoods, it is likely that these children did not have
ably higher rates of physical problems and somatic the same overall level of stress to deal with in their lives,
complaints than their low-risk counterparts and even their regardless of the negative life events they may have
high-risk, maladjusted counterparts (Werner and Smith experienced. This may explain the fact that these children
1992). They also tended to report themselves as disconnected were more likely to evidence competence across domains
from their families, and were less likely to rely on their than children in the previous studies.
friends for support. The authors described them as ‘‘inter- In sum, the bulk of studies (10/13) examining resilience
personally aloof.’’ In particular, the men had fewer long-term across domains suggest that while children exposed to high
committed relationships, while the women expressed more levels of risk may show positive outcomes in one domain,
tension between career and family commitments. this does not necessarily generalize to other domains. The
A study of the offspring of individuals with schizo- three exceptions to this pattern comprised of lower risk
phrenic, bipolar, and depressive disorders similarly found samples (Collishaw et al. 2007; Jaffee et al. 2007; Masten
that adults who were classified as ‘‘resilient’’ due to the et al. 1999); however, two other lower risk samples also
absence of a psychiatric illness displayed difficulties in found evidence of discontinuity across domains (Moffitt
intimate relationships and employed less healthy coping et al. 2002; Radke-Yarrow and Brown 1993), suggesting
strategies (Anthony 1987). Another study of offspring of that this finding is not necessarily limited to the highest
psychiatrically ill parents found that of children who were level of adversity.
consistently diagnosis-free across time, 56% had somatic As these studies show, ‘‘resilient’’ individuals may
complaints (Radke-Yarrow and Brown 1993). This was in exhibit high rates of internal distress, physical or somatic
comparison to 21% of the control children from well complaints, or difficulties in intimate relationships. As
families. The authors also pointed out that the resilient Luthar et al. (1993) demonstrated, there may even be

123
Clin Child Fam Psychol Rev (2008) 11:30–58 51

discontinuity within a general domain such as school-based across so many areas of research. Thus, while such breadth
social competence, illustrating the difficulty of achieving can be frustrating, there is also the potential for the concept
positive outcomes in the context of risk. Rutter (2000) has of resilience to inform any number of research areas.
pointed out that a certain amount of discontinuity across Specific to this review, integration efforts are also
domains is to be expected, given that risks and protective qualified by the fact that although there are many studies of
factors may be specific to particular outcomes; for exam- resilience, the majority of them look at continuous mea-
ple, we would not expect that because someone has sures of positive outcomes and do not create and compare
avoided cancer they would be protected against coronary groups based on risk status and outcome. While these
artery disease. However, while this is an important point, it studies provide valuable information on protective factors
is also true that the likelihood of discontinuity across that are associated with positive outcomes, they do not
domains appear to increase as the level of risk increases, allow for the examinations of the prevalence of resilience,
such that the middle class children experiencing negative group differences, change in adjustment status over time,
life events are more likely to have positive outcomes across or fluctuations in outcome across domains. Consequently,
domains than low-income minority children. Once again, the number of studies that were relevant to this particular
this illustrates the importance of not generalizing across review was limited, qualifying the strength of any con-
levels of risk, as well as the importance of looking at clusions that can be drawn. Relatedly, there were a number
outcomes across domains, or at the very least taking a more of studies whose results were reported in books or book
specific approach, such as talking about domain-specific chapters, as opposed to peer-reviewed journals (Anthony
resilience (Luthar 1993). 1987; Farber and Egeland 1987; Farrington et al. 1988b;
Felsman and Vaillant 1987; Werner and Smith 1982,
1992). As such, they were not subject to the same level of
Summary and Integration of Findings rigorous review of methodology. Other limitations include
retrospective reports of risk (Collishaw et al. 2007; Jaffee
The aim of this review was to examine potential constraints et al. 2007; Masten et al. 1999), relatively small samples of
or limitations of resilience in the highest-risk contexts (e.g., high-risk children (Collishaw et al. 2007; Farber and E-
multiple risks, low SES), with particular attention to dif- geland 1987; Radke-Yarrow and Brown 1993), and cross-
ferences between studies utilizing relatively lower risk sectional methodologies (Buckner et al. 2003; D’Imperio
versus high-risk samples. To this end, the article reviewed et al. 2000; Lin et al. 2004; Luthar 1991; Luthar and Sexton
studies that examined rates of resilience across levels of 2007; Tiet et al. 1998, 2001), all of which constrain the
risk, as well as studies looking at resilience across time and strength of the conclusions that can be drawn.
domains of competence. As noted from the outset, inte- In spite of these limitations, however, some interesting
grating findings from the literature on resilience has trends emerged that are worth considering. First, rates of
inherent difficulties due to the variability with which risk, positive outcomes differed widely depending on sample
protection, and positive outcome have been operationalized demographics, number of risks, and the number and type of
(Rutter 2000). It is challenging to determine criteria for outcomes. Although there was some overlap between sets
meaningfully grouping studies together, and questions arise of studies, in general, studies utilizing predominantly
regarding interpreting differences in results across studies. white, middle class samples and single risk factors found
For example, it is unclear whether disparate results are due higher rates of positive outcomes than studies utilizing
to differences in sample demographics, risk factors, pro- ethnically diverse, low-income samples and multiple risk
tective factors, and/or outcomes measured. Such problems factors. While it is hardly surprising that higher risk levels
are inevitable given the many permutations that arise from are associated with higher rates of negative outcome, it is
different combinations of risks, protective factors, and nonetheless an important finding, and suggests that great
outcomes that can be investigated. In order to truly arrive at care should be executed in how results from one study are
consensus about a particular risk or protective factor, each generalized to other samples, so that resilience rates are not
must be thoroughly researched on its own. At present, overestimated. In addition, other related findings support a
while some broad generalizations can be made, we are still cautionary approach to generalizing across levels of risk.
limited about specific conclusions about any particular risk For example, children at the highest level of risk are less
or protective factor and their association with specific likely to have protective factors (e.g., Dubow et al. 1997),
outcomes, and, consequently, much future research is or to benefit from them if they do exist (Luthar and
needed before we can draw firm conclusions about specific Goldstein 2004). In sum, these findings illustrate the sad
associations in specific contexts. However, there is a reality of the negative effects of high risk, and the great
positive side to this heterogeneity in that one could also difficulty in promoting positive outcomes at the highest
argue that part of the appeal of resilience is that it does cut level of risk.

123
52 Clin Child Fam Psychol Rev (2008) 11:30–58

Second, the majority of studies examining positive important to ensure that the protective factors targeted for
outcomes in the context of adversity across time and across promotion will also benefit those at the highest level of
domains of functioning support the idea of resilience as a risk. In support of this goal, it would be helpful for
dynamic process that fluctuates within and across devel- researchers to focus more attention on studying within-
opment. Certainly it does not support the initial perspective group differences among high-risk children, so that pat-
of resilience as a static outcome or a stable characteristic. It terns of adjustment and maladjustment can be better
also points to the fact that while not all children exposed to understood (Seidman and Pedersen 2003).
high levels of risk have disastrous outcomes, it is also rare Second, it has also been noted that even given empirical
for them to completely ‘‘escape’’ the negative effects of support for a specific protective factor in the context of a
risk altogether, particularly in the context of chronic or specific risk factor, prevention efforts are still in no way
cumulative risk. While there is cause to promote and cel- guaranteed. Luthar and Cicchetti (2000) point out that the
ebrate the positive outcomes of children at risk, the deep overall context needs to be taken into consideration, and
negative impact of risk also needs to be recognized and that an understanding of how protective factors emerge,
addressed. develop, and interact with risk is essential. For example,
From a more conceptual standpoint, the lack of consis- they point out that although an internal locus of control has
tency in positive outcomes across time and domains been identified as a protective factor, targeting this in an
suggests that ‘‘global resilience’’ is at best quite rare, if not intervention for low-income, inner-city children may not be
nonexistent. Thus, resilience might be better conceptual- very effective. Such children have no doubt developed
ized in terms of specific outcomes at specific time points. external loci of control because this is the reality of their
Researchers should exhibit caution in discussing resilience lives, dealing with many uncontrollable, negative events.
in a general or global way, and instead focus on circum- The development of this perspective may even be adaptive
scribed outcomes, such as ‘‘resilience in externalizing in some situations. Consequently, attempting to alter this
behaviors’’ or ‘‘resilience in school achievement.’’ Given perspective would most likely prove quite difficult because
this narrower conceptualization of resilience, some might the overall context is working against it. This also speaks to
wonder about its continued utility. What is to be gained the influence of the overall context on protective factors.
from research on resilience if it needs to be defined in such Luthar and Cicchetti (2000) also note that attempting to
constrained ways? While this is certainly a reasonable and change individual protective factors will most likely be of
thought-provoking question, completely dismissing the little benefit because the overall context will remain the
construct of resilience may be excessive. In fact, one could same. Indeed, this is one of the problems that child thera-
argue that a narrower definition of resilience may well pists struggle with: they may work effectively with an
contribute positively to the literature and our understanding individual child, but if the child then returns to the same
of risk and protective processes because it is a more environment, any benefits are likely to be short-lived.
accurate representation of children’s development in gen- Prevention efforts are therefore, better focused on pro-
eral and risk of continuity in psychopathology in particular. moting multiple protective factors across domains,
Furthermore, the study of resilience offers a way to including the child, family, and larger community.
understand the mechanisms through which some children Similarly, researchers have pointed out that in addition to
demonstrate positive outcomes in particular domains, even increasing the number and quality of protective resources
in the context of risk, and has important implications for available to children at risk, we also need to focus on
theory, prevention, and intervention. Thus there is much to decreasing overall exposure to risk because there are limits
be gained from retaining resilience as a construct, albeit in to the amount of risk that can be overcome (Cauce et al.
a more constrained version. 2003; Sameroff 1998). Furthermore, because the likelihood
The findings regarding positive outcomes at the highest of resilience decreases with the number of risks experi-
level of risk and the discontinuity in outcomes across both enced, this also suggests that intervention efforts should
time and domain also have important implications for focus on contexts where children are exposed to multiple
prevention and intervention efforts with children at risk. risks (Rutter 2000). Decreasing the level of risk becomes
Related to the issue of generalizability, prevention particularly important when considered in the context of
researchers and designers of public policy must be careful several studies demonstrating that not all protective factors
to select protective factors that have been shown to be are beneficial at the highest levels of risk (e.g., Miller et al.
beneficial for the targeted population in regards to the 1999; Vanderbilt-Adriance and Shaw in press). Conse-
outcome of interest. First, studies show that among high- quently, even if protective factors are increased for children
risk children, protective factors may not always be bene- at the highest levels of risk, one would still expect a high
ficial at the highest levels of risk (e.g., Miller et al. 1999; percentage of negative outcomes. Furthermore, as many
Vanderbilt-Adriance and Shaw in press). Therefore, it is researchers have noted, prevention is often more effective

123
Clin Child Fam Psychol Rev (2008) 11:30–58 53

and economical than intervention, and in this case, elimi- from such person-centered approaches is provided by
nating or decreasing risk would be the most desirable focus, researchers from Ann Masten’s lab (Obradovic et al. 2006),
because it goes to the root cause of the problem. who examined patterns of competence over time, assigning
individuals to competence trajectories based on their actual
data at each time point. Five patterns of competence over
Future Directions time were identified (low-declining, low-improving, mid-
dle-improving, middle-declining, and consistently high),
There are many exciting new directions for future research with important differences emerging between groups on
on resilience to explore. In general, there is a need for more both levels of risk and protective factors. Supporting earlier
studies examining within-group differences in high-risk theory and research conceptualizing developmental transi-
children to replicate and expand upon findings from pre- tions as a time of both vulnerability and opportunity, the
vious such studies (Cicchetti and Rogosch 1997; Gorman- authors determined that the most dramatic changes in
Smith et al. 2004; Luthar and Sexton 2007; Vanderbilt- competence occur during the period of emerging adulthood
Adriance and Shaw in press). Relating back to a narrower (ages 17–23). This study was conducted utilizing data from
conceptualization of resilience, researchers should clearly Project Competence, which consists of predominantly
specify the particular outcome in question, and discuss white, middle-class participants, the majority of whom were
their results as specific to that outcome, instead of referring considered low risk by the researchers. Thus, it would be
to a general, overall ‘‘resilient’’ outcome. Furthermore, informative to employ similar methods in samples of higher
researchers need to investigate and be cognizant of dif- risk children to examine differences and similarities in
ferent patterns of association between risks, protective patterns of competence over time.
factors, and outcomes, with particular attention to reporting Third, while developmental considerations are often
their findings as specific to their selected factors and implicit within studies (e.g., no one investigates school
sample, rather than generalizing to children ‘‘at risk.’’ This grades as an outcome in toddlerhood), there is little explicit
is particularly important given that resilience among higher attention to this issue. Many studies group diverse ages of
risk children is likely to be less common and to display children together, with minimal regard to potential devel-
more discontinuity over time and domains than among opmental differences in the effects of risks and protective
lower risk children. In the light of the fact that children at factors, or their relation to adjustment (Buckner et al. 2003;
the highest level of risk tend to have lower rates of positive Cicchetti and Rogosch 1997; Cicchetti et al. 1993; Lin
outcomes, it may be important to examine cumulative et al. 2004; Luthar and Sexton 2007; Masten et al. 1999;
protective factors. Researchers have pointed out that while Tiet et al. 1998, 2001). For example, it is possible that
an individual protective factor may not be powerful enough certain protective factors may be more or less helpful at
to counteract high levels of risk, an accumulation of pro- particular stages of development. Indeed, one study of
tective factors may improve outcomes (Bradley et al. 1994; elementary school children found that father involvement
Luthar and Zigler 1992). For example, a study of premature was most important in infancy (Wyman et al. 1991).
infants with multiple risks found that at least three pro- Although this study was cross-sectional and necessitated
tective factors were required in order to predict positive retrospective reporting of early protective factors, it still
outcomes (Bradley et al. 1994). Unfortunately, very few provides some support for the notion that developmental
studies have examined the effects of cumulative protective stage is important to consider.
factors. Future studies addressing these issues will help to Examining periods of developmental transition may be
further delineate the specifics of which protective factors another fruitful area for future research. Developmental
are beneficial in which contexts and for whom, a necessary transitions, such as the emergence of independent mobility
step towards creating more sophisticated conceptualiza- in toddlerhood, beginning formal schooling, or entering
tions of resilience and also for designing empirically adolescence and adulthood, may prove to be key points for
informed prevention and intervention efforts. both increased vulnerability or positive change. To illus-
Second, while the extant literature examining resilience trate, a child may be functioning well in preschool, but
across time and domains is intriguing, there is still a need decline significantly upon reaching elementary school due
for further investigation to make sure that the current to increased demands on attention, impulse control, and
findings are robust. For instance, while there are statistical behavior. At the other end of the spectrum, researchers
and theoretical reasons for using continuous measures of have discussed the importance of turning points, such as
positive adjustment, it would be beneficial if future studies marriage or entering the armed forces, in positively
would also report rates of resilience, so that comparisons changing the life trajectories of individuals at risk (Laub
can be made between groups and across time and domains. et al. 1998; Rutter 2000). Pointing to the dynamic nature of
An innovative example of the knowledge that can be gained resilience, Masten et al. (2004) also noted that although

123
54 Clin Child Fam Psychol Rev (2008) 11:30–58

childhood adversity and protective factors remained affect many aspects of behavior, emotion, and cognition
important in young adulthood, protective factors in ado- and likely mediate and/or moderate the associations
lescence predicted positive adjustment even after between risk, protection, and outcome (Greenberg 2006).
controlling for childhood circumstances. These results Furthermore, there is likely to be both mediation and
point to the dynamic nature of adjustment, for while past moderation across social and neurobiological contexts
circumstances continue to have weight, subsequent cir- (Silk et al. 2007). Silk et al. (2007) present a model for
cumstances are also of importance. cross-contextual mediation and moderation, in which they
Fourth, the extant literature has very little to say about discuss how each context can affect the other. They pro-
the process through which protective factors have their vide examples of cross-contextual mediation, such as how
influence. The vast majority of studies examine factors biological characteristics (e.g., emotional reactivity) may
associated with positive outcomes, but they generally do affect aspects of the environment (e.g., responses from
not attempt to unpack how these protective factors actually caregivers), and vice versa; and cross-contextual modera-
moderate risk, even though there have been calls for more tion, such as how biological or genetic risk and
process research in resilience (Gore and Eckenrode 1996). environmental risk may interact, as in the case of skin
This is an important area for future researchers to investi- cancer. Such models have the potential to greatly increase
gate. For example, why might parenting be associated with our understanding of both risk and resilience processes.
higher academic achievement? There may be cascading Although such integrative efforts are still in the begin-
effects that emerge as outcomes in one domain influence ning stages, a recent New York Academies of Sciences
other domains (Masten et al. 2005), or connections among conference on resilience in children aptly demonstrated the
protective factors, with certain protective factors increasing many ways in which genetics, biology, and neuroscience
the likelihood of others emerging (e.g., high parenting can inform resilience research (Lester et al. 2006). Future
warmth contributing to high self esteem). For instance, one research examining processes that have already been
study of school performance and adjustment in three demonstrated to be important, such as those involving
independent samples of urban, African American preado- human relationships, attention-regulation and stress-regu-
lescents and adolescents determined that associations lation systems are suggested as good places to start
between parent and child protective factors were often integrating neuroscience and biobehavioral research with
bidirectional (Connell et al. 1994). In sum, it is time for resilience research (Masten and Obradovic 2006).
research to move beyond establishing if factors are asso-
ciated with positive outcomes to beginning to examine how
they may play a role. Conclusion
Finally, studies are only just beginning to examine the
role of biology and genetics in resilience (Haglund et al. In conclusion, the present review supports the conceptu-
2007; Nigg et al. 2007). Several studies have investigated alization of resilience as a dynamic process that varies
gene by environment interactions, noting that environ- within and across time, rather than a stable, static trait. It
mental risk was only associated with negative outcomes in also points to the value of reconceptualizing resilience in
the context of genetic risk (Caspi et al. 2002, 2003; Jaffee narrower, specific terms to more accurately represent
et al. 2005). Another ground-breaking study utilized a twin resilience as it is observed in the real world. Relatedly,
design to examine the heritability of positive adaptation in differences in the prevalence of resilience across the
the context of risk, finding that it is both genetically and highest levels of risk, as well as discontinuity across time
environmentally determined (Kim-Cohen et al. 2004). and domains, emphasize the difficulty of ‘‘escaping’’ risk,
There is certainly a need for more studies examining and illustrate the need for both researchers and policy
genetic factors as both risk and protective factors because makers to target established protective factors that have
such lines of research present an exciting new framework been reliably shown to be associated with positive out-
for conceptualizing and investigating resilience. In partic- comes in similar samples. Furthermore, we must be
ular, genetics studies that employ reasonable measures of realistic in our expectations for positive outcomes at the
environmental factors are necessary because they will highest level of risk, and turn towards reducing risk as well
allow us to fully capture the role of both environmental and as increasing protective factors. As Luthar and Goldstein
genetic processes, as well as their interaction. (2004) noted, ‘‘if children are faced with continuing and
Researchers have also pointed out the gains that could severe assaults from the external environment, then they
be made by integrating neuroscience findings with resil- simply cannot sustain resilience adaptation over time—
ience research, in particular through informing models of regardless of how much they are helped to believe in
plasticity and/or constraints (Curtis and Cicchetti 2003; themselves, how intelligent they are, or how well they learn
Greenberg 2006; Luthar et al. 2006). Biological processes to regulate their emotions’’ (pp. 503).

123
Clin Child Fam Psychol Rev (2008) 11:30–58 55

Acknowledgments The authors would like to thank Susan B. Conrad, M., & Hammen, C. (1993). Protective and resource factors in
Campbell, Ronald E. Dahl, and Jennifer S. Silk for their comments on high- and low-risk children: A comparison of children with
earlier versions of this article. unipolar, bipolar, medically ill, and normal mothers. Develop-
ment and Psychopathology, 5, 593–607.
Curtis, W., & Cicchetti, D. (2003). Moving research on resilience into
the 21st century: Theoretical and methodological considerations
References in examining the biological contributors to resilience. Develop-
ment and Psychopathology, 15, 773–810.
Anthony, E. J. (1987). Children at high risk for psychosis growing up D’Imperio, R. L., Dubow, E. F., & Ippolito, M. F. (2000). Resilient
successfully. In E. J. Anthony & B. J. Cohler (Eds.), The and stress-affected adolescents in an urban setting. Journal of
invulnerable child (pp. 147–184). New York, NY: Guilford. Clinical Child Psychology, 29, 129–142.
Barbarin, O., Bryant, D., McCandies, T., Burchinal, M., Early, D., Dishion, T. J., & McMahon, R. J. (1998). Parental monitoring and the
Clifford, R., et al. (2006). Children enrolled in public pre-K: The prevention of child and adolescent problem behavior: A
relation of family life, neighborhood quality, and socioeconomic conceptual and empirical formulation. Clinical Child and Family
resources to early competence. American Journal of Orthopsy- Psychology Review, 1, 61–75.
chiatry, 76, 265–276. Dubow, E. F., Edwards, S., & Ippolito, M. F. (1997). Life stressors,
Beckett, C., Maughan, B., Rutter, M., Castle, J., Colvert, E., neighborhood disadvantage, and resources: A focus on inner-city
Groothues, C., et al. (2006). Do the effects of early severe children’s adjustment. Journal of Clinical Child Psychology, 26,
deprivation on cognition persist into early adolescence? Findings 130–144.
from the English and Romanian Adoptees Study. Child Devel- Duncan, G. J., Brooks-Gunn, J., & Klebanov, P. K. (1994). Economic
opment, 77, 696–711. deprivation and early childhood development. Child Develop-
Belsky, J. (1984). The determinants of parenting: A process model. ment, 65, 296–318.
Child Development, 55, 83–96. Egeland, B. R., Carlson, E., & Sroufe, L. (1993). Resilience as
Bradley, R. H., Whiteside, L., Mundfrom, D. J., Casey, P. H., et al. (1994). process. Development and Psychopathology, 5, 517–528.
Early indications of resilience and their relation to experiences in the Eisenberg, N., Fabes, R. A., Guthrie, I. K., Murphy, B. C., et al.
home environments of low birthweight, premature children living in (1996). The relations of regulation and emotionality to problem
poverty. Child Development, 65, 346–360. behavior in elementary school children. Development and
Bronfenbrenner, U. (1979). The ecology of human development: Psychopathology, 8, 141–162.
Experiments by nature and design. Cambridge, MA: Harvard Eisenberg, N., Fabes, R. A., Shepard, S. A., Murphy, B. C., Guthrie, I.
University Press. K., Jones, S., et al. (1997a). Contemporaneous and longitudinal
Buckner, J. C., Mezzacappa, E., & Beardslee, W. R. (2003). prediction of children’s social functioning from regulation and
Characteristics of resilient youths living in poverty: The role emotionality. Child Development, 68, 642–664.
of self-regulatory processes. Development and Psychopathology, Eisenberg, N., Guthrie, I. K., Fabes, R. A., Reiser, M., Murphy, B. C.,
15, 139–162. Holgren, R., et al. (1997b). The relations of regulation and
Calkins, S. D., & Fox, N. A. (2002). Self-regulatory processes in early emotionality to resiliency and competent social functioning in
personality development: A multilevel approach to the study of elementary school children. Child Development, 68, 295–311.
childhood social withdrawal and aggression. Development and Emery, R. E., & Forehand, R. (1996). Parental divorce and children’s
Psychopathology, 14, 477–498. well-being: A focus on resilience. In R. J. Haggerty, L. R.
Caspi, A., McClay, J., Moffitt, T., Mill, J., Martin, J., Craig, I. W., Sherrod, N. Garmezy, & M. Rutter (Eds.), Stress, risk, and
et al. (2002). Role of genotype in the cycle of violence in resilience in children and adolescents: Processes, mechanisms,
maltreated children. Science, 297, 851–854. and interventions (pp. 64–99). New York, NY: Cambridge
Caspi, A., Sugden, K., Moffitt, T. E., Taylor, A., Craig, I. W., University Press.
Harrington, H., et al. (2003). Influence of life stress on Farber, E. A., & Egeland, B. (1987). Invulnerability among abused
depression: Moderation by a polymorphism in the 5-HTT gene. and neglected children. In E. J. Anthony, & B. J. Cohler (Eds.),
Science, 301, 386–389. The invulnerable child (pp. 253–288). New York, NY: Guilford.
Cauce, A. M., Stewart, A., Rodriguez, M. D., Cochran, B., & Ginzler, Farrington, D. P., Gallagher, B., Morley, L., et al. (1988a). Are there
J. (2003). Overcoming the odds? Adolescent development in the any successful men from criminogenic backgrounds? Psychia-
context of urban poverty. In S. S. Luthar (Ed.), Resilience and try: Journal for the Study of Interpersonal Processes, 51, 116–
vulnerability: Adaptation in the context of childhood adversities 130.
(pp. 343–363). New York, NY: Cambridge University Press. Farrington, D. P., Gallagher, B., Morley, L., St. Ledger, R. J., & West,
Cicchetti, D., & Rogosch, F. A. (1997). The role of self-organization D. J. (1988b). A twenty-four year follow-up of men from
in the promotion of resilience in maltreated children. Develop- vulnerable backgrounds. In R. L. Jenkins (Ed.), The abandon-
ment and Psychopathology, 9, 797–815. ment of delinquent behavior (pp. 156–173). New York, NY:
Cicchetti, D., & Rogosch, F. A. (2007). Personality, adrenal steroid Praeger.
hormones, and resilience in maltreated children: A multilevel Felsman, J. K., & Vaillant, G. E. (1987). Resilient children as adults:
perspective. Development and Psychopathology, 19, 787–809. A 40-year study. In E. J. Anthony, & B. J. Cohler (Eds.), The
Cicchetti, D., Rogosch, F. A., Lynch, M., & Holt, K. D. (1993). invulnerable child (pp. 289–314). New York, NY: Guilford.
Resilience in maltreated children: Processes leading to adaptive Fergusson, D. M., Horwood, L., & Lynskey, M. T. (1994). The
outcome. Development and Psychopathology, 5, 629–647. childhoods of multiple problem adolescents: A 15-year longitu-
Collishaw, S., Pickles, A., Messer, J., Rutter, M., Shearer, C., & dinal study. Journal of Child Psychology and Psychiatry, 35,
Maughan, B. (2007). Resilience to adult psychopathology 1123–1140.
following childhood maltreatment: Evidence from a community Fergusson, D. M., & Lynskey, M. T. (1996). Adolescent resiliency to
sample. Child Abuse & Neglect, 31, 211–229. family adversity. Journal of Child Psychology and Psychiatry,
Connell, J. P., Spencer, M. B., & Aber, J. (1994). Educational risk and 37, 281–292.
resilience in African-American youth: Context, self, action, and Fisher, L., Ames, E. W., Chisholm, K., & Savoie, L. (1997). Problems
outcomes in school. Child Development, 65, 493–506. reported by parents of Romanian orphans adopted to British

123
56 Clin Child Fam Psychol Rev (2008) 11:30–58

Columbia. International Journal of Behavioral Development, 20, Kolvin, I., Miller, J. W., Fleeting, M., & Kolvin, P. A. (1988). Risk/
67–82. protective factors for offending with particular reference to
Garmezy, N., Masten, A. S., & Tellegen, A. (1984). The study of stress deprivation. In M. Rutter (Ed.), Studies of psychosocial risk: The
and competence in children: A building block for developmental power of longitudinal data (pp. 77–95). Cambridge: Cambridge
psychopathology. Child Development, 55, 97–111. University Press.
Goodman, S. H., & Gotlib, I. H. (1999). Risk for psychopathology in Korenman, S., Miller, J. E., & Sjaastad, J. E. (1995). Long-term poverty
the children of depressed mothers: A developmental model for and child development in the United States: Results from the
understanding mechanisms of transmission. Psychological NLSY. Children and Youth Services Review, 17, 127–155.
Review, 106, 458–490. Laub, J. H., Nagin, D. S., & Sampson, R. J. (1998). Trajectories of
Gore, S., & Eckenrode, J. (1996). Context and process in research on change in criminal offending: Good marriages and the desistance
risk, resilience. In R. J. Haggerty, L. R. Sherrod, N. Garmezy, & process. American Sociological Review, 63, 225–238.
M. Rutter (Eds.), Stress, risk, and resilience in children and Lester, B. M., Masten, A. S., & McEwen, B. (Eds.). (2006).
adolescents: Processes, mechanisms, and interventions (pp. 19– Resilience in children (Vol. 1094). New York, NY: New York
63). New York, NY: Cambridge University Press. Academy of Sciences.
Gorman-Smith, D., Henry, D. B., & Tolan, P. H. (2004). Exposure to Leventhal, T., & Brooks-Gunn, J. (2000). The neighborhoods they
community violence and violence perpetration: The protective live in: The effects of neighborhood residence on child and
effects of family functioning. Journal of Clinical Child and adolescent outcomes. Psychological Bulletin, 126, 309–337.
Adolescent Psychology, 33, 439–449. Li, S. T., Nussbaum, K. M., & Richards, M. H. (2007). Risk and
Gorman-Smith, D., & Tolan, P. H. (2003). Positive adaptation among protective factors for urban African American youth. American
youth exposed to community violence. In S. S. Luthar (Ed.), Journal of Community Psychology, 39, 21–35.
Resilience and vulnerability: Adaptation in the context of Lin, K. K., Sandler, I. N., Ayers, T. S., Wolchik, S. A., & Luecken, L.
childhood adversities (pp. 392–413). New York, NY: Cambridge J. (2004). Resilience in parentally bereaved children and
University Press. adolescents seeking preventive services. Journal of Clinical
Gorman-Smith, D., Tolan, P. H., & Henry, D. (1999). The relation of Child and Adolescent Psychology, 33, 673–683.
community and family to risk among urban-poor adolescents. In Loeber, R., Pardini, D. A., Stouthamer-Loeber, M., & Raine, A.
P. Cohen, C. Slomkowski, & L. N. Robins (Eds.), Historical and (2007). Do cognitive, physiological, and psychosocial risk and
geographical influences on psychopathology (pp. 349–367). promotive factors predict desistance from delinquency in males?
Mahwah, NJ: Lawrence Erlbaum. Development and Psychopathology, 19, 867–887.
Gorman-Smith, D., Tolan, P. H., & Henry, D. B. (2000). A Luthar, S. S. (1991). Vulnerability and resilience: A study of high-risk
developmental-ecological model of the relation of family adolescents. Child Development, 62, 600–616.
functioning to patterns of delinquency. Journal of Quantitative Luthar, S. S. (1993). Annotation: Methodological and conceptual
Criminology, 16, 169–198. issues in research on childhood resilience. Journal of Child
Greenberg, M. T. (2006). Promoting resilience in children and youth: Psychology and Psychiatry, 34, 441–453.
Preventive interventions and their interface with neuroscience. Luthar, S. S. (1995). Social competence in the school setting:
Annals of the New York Academy of Sciences, 1094, 139–150. Prospective cross-domain associations among inner-city teens.
Haglund, M., Nestadt, P., Cooper, N., Southwick, S., & Charney, D. Child Development, 66, 416–429.
(2007). Psychobiological mechanisms of resilience: Relevance Luthar, S. S. (2003). The culture of affluence: Psychological costs of
to prevention and treatment of stress-related psychopathology. material wealth. Child Development, 74, 1581–1593.
Development and Psychopathology, 19, 889–920. Luthar, S. S. (2006). Resilience in development: A synthesis of
Hammack, P. L., Richards, M. H., Luo, Z., Edlynn, E. S., & Roy, K. research across five decades. In D. Cicchetti, & D. Cohen (Eds.),
(2004). Social support factors as moderators of community Developmental psychopathology: Risk, disorder, and adaptation
violence exposure among inner-city African American young (2nd ed., Vol. 3, pp. 739–795). Hoboken, NJ: John Wiley &
adolescents. Journal of Clinical Child and Adolescent Psychol- Sons, Inc.
ogy, 33, 450–462. Luthar, S. S., & Ansary, N. S. (2005). Dimensions of adolescent
Jaffee, S. R., Caspi, A., Moffitt, T. E., Dodge, K. A., Rutter, M., rebellion: Risks for academic failure among high- and low-
Taylor, A., et al. (2005). Nature x nurture: Genetic vulnerabil- income youth. Development and Psychopathology, 17, 231–250.
ities interact with physical maltreatment to promote conduct Luthar, S. S., & Cicchetti, D. (2000). The construct of resilience:
problems. Development and Psychopathology, 17, 67–84. Implications for interventions and social policies. Development
Jaffee, S. R., Caspi, A., Moffitt, T. E., Polo-Tomas, M., & Taylor, A. and Psychopathology, 12, 857–885.
(2007). Individual, family, and neighborhood factors distinguish Luthar, S. S., Cicchetti, D., & Becker, B. (2000a). The construct of
resilient from non-resilient maltreated children: A cumulative resilience: A critical evaluation and guidelines for future work.
stressors model. Child Abuse & Neglect, 31, 231–253. Child Development, 71, 543–562.
Kandel, E., Mednick, S. A., Kirkegaard-Sorensen, L., Hutchings, B., Luthar, S. S., Cicchetti, D., & Becker, B. (2000b). Research on
et al. (1988). IQ as a protective factor for subjects at high risk for resilience: Response to commentaries. Child Development, 71,
antisocial behavior. Journal of Consulting and Clinical Psy- 573–575.
chology, 56, 224–226. Luthar, S. S., Doernberger, C. H., & Zigler, E. (1993). Resilience is
Kim-Cohen, J., Moffitt, T. E., Caspi, A., & Taylor, A. (2004). Genetic not a unidimensional construct: Insights from a prospective study
and environmental processes in young children’s resilience and of inner-city adolescents. Development and Psychopathology, 5,
vulnerability to socioeconomic deprivation. Child Development, 703–717.
75, 651–668. Luthar, S. S., & Goldstein, A. (2004). Children’s exposure to
Kliewer, W., Cunningham, J. N., Diehl, R., Parrish, K. A., Walker, J. community violence: Implications for understanding risk and
M., Atiyeh, C., et al. (2004). Violence exposure and adjustment resilience. Journal of Clinical Child and Adolescent Psychology,
in inner-city youth: Child and caregiver emotion regulation skill 33, 499–505.
caregiver-child relationship quality, and neighborhood cohesion Luthar, S. S., & Latendresse, S. J. (2005). Comparable ‘‘risks’’ at the
as protective factors. Journal of Clinical Child and Adolescent socioeconomic status extremes: Preadolescents’ perceptions of
Psychology, 33, 477–487. parenting. Development and Psychopathology, 17, 207–230.

123
Clin Child Fam Psychol Rev (2008) 11:30–58 57

Luthar, S. S., Sawyer, J. A., & Brown, P. J. (2006). Conceptual issues and correlates. Annals of the New York Academy of Sciences,
in studies of resilience: Past, present, and future research. Annals 1094, 340–344.
of the New York Academy of Sciences, 1094, 105–115. Owens, E. B., & Shaw, D. S. (2003). Poverty and early childhood
Luthar, S. S., & Sexton, C. C. (2007). Maternal drug abuse versus adjustment. In S. S. Luthar (Ed.), Resilience and vulnerability:
maternal depression: Vulnerability and resilience among school- Adaptation in the context of childhood adversities (pp. 267–292).
age and adolescent offspring. Development and Psychopathol- New York, NY: Cambridge University Press.
ogy, 19, 205–225. Ozer, E. J., & Weinstein, R. S. (2004). Urban adolescents’ exposure to
Luthar, S. S., & Zelazo, L. B. (2003). Research on resilience: An community violence: The role of support, school safety, and
integrative review. In S. S. Luthar (Ed.), Resilience and social constraints in a school-based sample of boys and girls.
vulnerability: Adaptation in the context of childhood adversities Journal of Clinical Child and Adolescent Psychology, 33,
(pp. 510–549). New York, NY: Cambridge University Press. 463–476.
Luthar, S. S., & Zigler, E. (1992). Intelligence and social competence Pines, M. (1975, Dec). In praise of ‘‘invulnerables’’. APA Monitor, 7.
among high-risk adolescents. Development and Psychopathol- Plotnick, R. D., & Hoffman, S. D. (1999). The effect of neighborhood
ogy, 4, 287–299. characteristics on young adult outcomes: Alternative estimates.
MacLean, K. (2003). The impact of institutionalization on child Social Science Quarterly, 80, 1–18.
development. Development and Psychopathology, 15, 853–884. Radke-Yarrow, M., & Brown, E. (1993). Resilience and vulnerability
Masten, A. S. (2001). Ordinary magic: Resilience processes in in children of multiple-risk families. Development and Psycho-
development. American Psychologist, 56, 227–238. pathology, 5, 581–592.
Masten, A. S., Burt, K. B., Roisman, G. I., Obradovic, J., Long, J. D., Rice, F., Harold, G. T., Shelton, K. H., & Thapar, A. (2006). Family
& Tellegen, A. (2004). Resources and resilience in the transition conflict interacts with genetic liability in predicting childhood
to adulthood: Continuity and change. Development and Psycho- and adolescent depression. Journal of the American Academy of
pathology, 16, 1071–1094. Child & Adolescent Psychiatry, 45, 841–848.
Masten, A. S., & Coatsworth, J. D. (1998). The development of Rice, F., Harold, G. T., & Thapar, A. (2002). Assessing the effects of
competence in favorable and unfavorable environments: Lessons age, sex and shared environment on the genetic aetiology of
from successful children. American Psychologist, 53, 205–220. depression in childhood and adolescence. Journal of Child
Masten, A. S., Garmezy, N., Tellegen, A., Pellegrini, D. S., et al. Psychology and Psychiatry, 43, 1039–1051.
(1988). Competence and stress in school children: The moder- Rice, F., Harold, G. T., & Thapar, A. (2005). The link between
ating effects of individual and family qualities. Journal of Child depression in mothers and offspring: An extended twin analysis.
Psychology and Psychiatry, 29, 745–764. Behavior Genetics, 35, 565–577.
Masten, A. S., Hubbard, J. J., Gest, S. D., Tellegen, A., Garmezy, N., Richters, J. E., & Weintraub, S. (1990). Beyond diathesis: Toward an
& Ramirez, M. (1999). Competence in the context of adversity: understanding of high-risk environments. In J. E. Rolf, A. S.
Pathways to resilience and maladaptation from childhood to late Masten, D. Cicchetti, K. H. Nuechterlein, & S. Weintraub (Eds.),
adolescence. Development and Psychopathology, 11, 143–169. Risk and protective factors in the development of psychopathol-
Masten, A. S., & Obradovic, J. (2006). Competence and resilience in ogy (pp. 67–96). Cambridge: Cambridge University Press.
development.[see comment]. Annals of the New York Academy Rowe, D. C., & Rodgers, J. L. (1997). Poverty and behavior: Are
of Sciences, 1094, 13–27. environmental measures nature and nurture? Developmental
Masten, A. S., & Powell, J. L. (2003). A resilience framework for Review, 17, 358–375.
research, policy, and practice. In S. S. Luthar (Ed.), Resilience and Rutter, M. (1987). Psychosocial resilience and protective mecha-
vulnerability: Adaptation in the context of childhood adversities nisms. American Journal of Orthopsychiatry, 57, 316–331.
(pp. 1–25). New York, NY: Cambridge University Press. Rutter, M. (2000). Resilience reconsidered: Conceptual consider-
Masten, A. S., & Reed, M. J. (2002). Resilience in development. In C. ations, empirical findings, and policy implications. In J. P.
R. Snyder & S. J. Lopez (Eds.), Handbook of positive psychology Shonkoff & S. J. Meisels (Eds.), Handbook of early childhood
(pp. 74–88). New York, NY: Oxford University Press. intervention (2nd ed., pp. 651–682). New York, NY: Cambridge
Masten, A. S., Roisman, G. I., Long, J. D., Burt, K. B., Obradovic, J., University Press.
Riley, J. R., et al. (2005). Developmental Cascades: Linking Sameroff, A. (1998). Environmental risk factors in infancy. Pediat-
Academic Achievement and Externalizing and Internalizing rics, 102, 1287–1292.
Symptoms Over 20 Years. Developmental Psychology, 41, Sameroff, A. (2005). The science of infancy: Academic, social, and
733–746. political agendas. Infancy, 7, 219–242.
McLoyd, V. C. (1998). Socioeconomic disadvantage and child Sameroff, A., Seifer, R., Baldwin, A., & Baldwin, C. (1993). Stability
development. American Psychologist, 53, 185–204. of intelligence from preschool to adolescence: The influence of
Miller, L. S., Wasserman, G. A., Neugebauer, R., Gorman-Smith, D., social and family risk factors. Child Development, 64, 80–97.
& Kamboukos, D. (1999). Witnessed community violence and Sameroff, A., Seifer, R., Zax, M., & Barocas, R. (1987). Early
antisocial behavior in high-risk, urban boys. Journal of Clinical indicators of developmental risk: Rochester longitudinal study.
Child Psychology, 28, 2–11. Schizophrenia Bulletin, 13, 383–394.
Moffitt, T. E., Caspi, A., Harrington, H., & Milne, B. J. (2002). Males Sampson, R. J., Morenoff, J. D., & Earls, F. (1999). Beyond social
on the life-course-persistent and adolescence-limited antisocial capital: Spatial dynamics of collective efficacy for children.
pathways: Follow-up at age 26 years. Development and Psy- American Sociological Review, 64, 633–660.
chopathology, 14, 179–207. Seidman, E., & Pedersen, S. (2003). Holistic contextual perspectives
Nigg, J., Nikolas, M., Friderici, K., Park, L., & Zucker, R. A. (2007). in risk, protection, competence among low-income urban
Genotype and neuropsychological response inhibition as resil- adolescents. In S. S. Luthar (Ed.), Resilience and vulnerability:
ience promoters for attention-deficit/hyperactivity disorder, Adaptation in the context of childhood adversities (pp. 318–342).
oppositional defiant disorder, and conduct disorder under New York, NY: Cambridge University Press.
conditions of psychosocial adversity. Development and Psycho- Seifer, R., Sameroff, A., Baldwin, C. P., & Baldwin, A. L. (1992).
pathology, 19, 767–786. Child and family factors that ameliorate risk between 4 and
Obradovic, J., Burt, K. B., & Masten, A. S. (2006). Pathways of 13 years of age. Journal of the American Academy of Child &
adaptation from adolescence to young adulthood: Antecedents Adolescent Psychiatry, 31, 893–903.

123
58 Clin Child Fam Psychol Rev (2008) 11:30–58

Shaw, D. S., Criss, M. M., Schonberg, M. A., & Beck, J. E. (2004). the American Academy of Child & Adolescent Psychiatry, 37,
The development of family hierarchies and their relation to 1191–1200.
children’s conduct problems. Development and Psychopathol- Tiet, Q. Q., Bird, H. R., Hoven, C. W., Wu, P., Moore, R., & Davies,
ogy, 16, 483–500. M. (2001). Resilience in the face of maternal psychopathology
Silk, J. S., Vanderbilt-Adriance, E., Shaw, D. S., Forbes, E. E., and adverse life events. Journal of Child and Family Studies, 10,
Whalen, D. J., Ryan, N. D., et al. (2007). Resilience among 347–365.
children and adolescents at risk for depression: Mediation and Vanderbilt-Adriance, E., & Shaw, D. S. (in press). Protective factors
moderation across social and neurobiological context. Develop- and the development of resilience in the context of neighborhood
ment and Psychopathology, 19, 841–865. disadvantage. Journal of Abnormal Child Psychology.
Stattin, H., Romelsjo, A., & Stenbacka, M. (1997). Personal resources Werner, E. E., & Smith, R. S. (1982). Vulnerable, but invincible: A
as modifiers of the risk for future criminality: An analysis of longitudinal study of resilient children and youth. New York,
protective factors in relation to 18-year-old boys. British Journal NY: McGraw-Hill.
of Criminology, 37, 198–223. Werner, E. E., & Smith, R. S. (1992). Overcoming the odds: High risk
Stouthamer-Loeber, M., Loeber, R., Farrington, D. P., Zhang, Q., children from birth to adulthood. Ithaca, NY: Cornell University
et al. (1993). The double edge of protective and risk factors for Press.
delinquency: Interrelations and developmental patterns. Devel- White, J. L., Moffitt, T. E., & Silva, P. A. (1989). A prospective
opment and Psychopathology, 5, 683–701. replication of the protective effects of IQ in subjects at high risk
Stouthamer-Loeber, M., Loeber, R., Wei, E., Farrington, D. P., & for juvenile delinquency. Journal of Consulting and Clinical
Wikstrom, P.-O. H. (2002). Risk and promotive effects in the Psychology, 57, 719–724.
explanation of persistent serious delinquency in boys. Journal of Wyman, P. A. (2003). Emerging perspectives on context specificity of
Consulting and Clinical Psychology, 70, 111–123. children’s adaptation and resilience: Evidence from a decade of
Stouthamer-Loeber, M., Wei, E., Loeber, R., & Masten, A. S. (2004). research with urban children in adversity. In S. S. Luthar (Ed.),
Desistance from persistent serious delinquency in the transition Resilience and vulnerability: Adaptation in the context of
to adulthood. Development and Psychopathology, 16, 897–918. childhood adversities (pp. 293–317). New York, NY: Cambridge
Sullivan, T. N., Kung, E. M., & Farrell, A. D. (2004). Relation University Press.
between witnessing violence and drug use initiation among rural Wyman, P. A., Cowen, E. L., Work, W. C., Hoyt-Meyers, L.,
adolescents: Parental monitoring and family support as protec- Magnus, K. B., & Fagen, D. B. (1999). Caregiving and
tive factors. Journal of Clinical Child and Adolescent developmental factors differentiating young at-risk urban chil-
Psychology, 33, 488–498. dren showing resilient versus stress-affected outcomes: A
Thompson, R. A., & Calkins, S. D. (1996). The double-edged sword: replication and extension. Child Development, 70, 645–659.
Emotional regulation for children at risk. Development and Wyman, P. A., Cowen, E. L., Work, W. C., & Parker, G. R. (1991).
Psychopathology, 8, 163–182. Developmental and family milieu correlates of resilience in
Tiet, Q. Q., Bird, H. R., Davies, M., Hoven, C., Cohen, P., Jensen, P. urban children who have experienced major life stress. American
S., et al. (1998). Adverse life events and resilience. Journal of Journal of Community Psychology, 19, 405–426.

123