Development and Psychopathology 24 (2012), 335344

# Cambridge University Press 2012

SPECIAL SECTION ARTICLE

Resilience as a dynamic concept

MICHAEL RUTTER
Kings College London

Abstract
The concept of resilience has as its starting point the recognition that there is huge heterogeneity in peoples responses to all manner of environmental
adversities. Resilience is an inference based on evidence that some individuals have a better outcome than others who have experienced a comparable
level of adversity; moreover, the negative experience may have either a sensitizing effect or a strengthening steeling effect in relation to the response to later
stress or adversity. After noting the crucial importance of first testing for the environmental mediation of risk through natural experiments, findings are
reviewed on steeling effects in animal models and humans. Geneenvironment interaction findings are considered, and it is noted that there is some evidence
that the genetic influences concerns responsivity to all environments and not just bad ones. Life course effects are reviewed in relation to evidence on turning
point effects associated with experiences that increase opportunities and enhance coping. Attention is drawn to both research implications and substantive
findings as features that foster resilience.

Norm Garmezy was one of the most important pioneers in ial causal pathways; and all need to examine geneenviron-
the conceptualization and study of resilience from the early ment interdependence.
1970s onward (Garmezy, 1974, 1985). Several features Third, in his own research, Garmezy had been motivated
made his approach distinctive. First, in keeping with Eisen- by Bleulers (1978) study of the children of mothers with
berg (1977), he viewed development as the unifying concept schizophrenia, which showed that even in this high-risk
in the study of psychopathology. This was the central element group there were numerous examples of individuals who
in the field of developmental psychopathology that he did so showed adaptive patterns of social behavior and work
much to advance (Rutter, 2008; Rutter & Garmezy, 1983). achievement. Garmezy appreciated that the high risk involved
Two key elements defined developmental psychopathology: a genetic liability but, equally, he realized that being raised by
the focus on continuities and discontinuities over time, and a schizophrenic mother involved environmental as well as ge-
continuities and discontinuities between normality and men- netic risks (see Rutter, 1989, for a fuller discussion on this
tal disorder (Rutter, 1986). It was notable that this involved point). He decided that there was need to study stress resis-
no presumption that either continuities or discontinuities tance in high-risk groups but chose to focus on psychosocial
would predominate. Rather, testing constituted an essential disadvantage in community samples. In that connection, he
part of the research endeavor. was clear that risk and protective influences should not be de-
Second, Garmezy was forthright in requiring a methodolo- fined on the basis of theoretical or ideological presumptions.
gically rigorous approach to data analysis (Garmezy, Masten, Rather, the influences needed to be investigated systemati-
& Tellegen, 1984). Resilience should not constitute a theory, cally in order to understand how they actually operated in
nor should it be seen as equivalent to positive psychology or the samples under study. Moreover, it should not necessarily
competence. Both of the latter are valid and useful concepts be expected that a universal answer would be found; effects
(see Masten & Tellegen, 2012 [this issue]) but they differ were likely to be shaped by social context (Rutter, 1999).
from resilience. However, all require longitudinal study for Fourth, Garmezy appreciated the need for resilience re-
their rigorous investigation; all need to consider multifactor- search to include positive personality dispositions, a nurtur-
ant family milieu, and external societal support systems.
This broad-based conceptualization meant that he was resis-
tant to notions of inherent invulnerability that were being
Many thanks to Tytti Solantaus for helpful comments on an earlier draft of put forward by others in the 1970s and 1980s (Anthony,
this article. 1974; Anthony & Cohler, 1987). Resilience had to be viewed
Address correspondence and reprint requests to: Michael Rutter, PO80,
MRC Social Genetic & Developmental Psychiatry Centre, Institute of Psy- as a process and not as a fixed attribute of an individual. Of
chiatry, Kings College London, Denmark Hill, London SE5 8AF, UK; E- course, it was likely that some individuals would show resil-
mail: [email protected]. ience across a range of circumstances and across a range of
335
336 M. Rutter

outcomes, but it could not be assumed that the same features has two great strengths. First, it is obviously a desirable out-
would be protective in relation to all risks. Moreover, resili- come, and second, it is quantifiable. Nevertheless, it has three
ence was an interactive concept and had to be inferred; it important limitations (see Rutter, 2011). First, it implies that,
could not be measured directly as if it was a characterlogical usually, the causal influences will be much the same in non-
trait. stressed groups as in those suffering from extreme adversity.
Fifth and finally, although very much a user of quantitative Nonlinear interactive effects are also systematically consid-
methods, Garmezy was basically interested in what experi- ered, but they have to be derived from mathematical models
ences meant for individuals and how research findings might with all the uncertainties that these require about the assump-
be used to develop better means of helping children who expe- tions (see also Parker & Maestripieri, 2011; Seery, 2011), ra-
rience serious stress and adversity. Accordingly, he paid atten- ther than measured directly. Nevertheless, they can be exam-
tion to qualitative, as well as quantitative, research strategies ined and, when they are, promotive factors that apply outside
(see Rutter, in press-a, for a discussion of the value of both). of resilience also contribute to resilience in the face of adver-
It should be obvious from what I have written how much I sity (see Bowes, Maughan, Caspi, Moffitt, & Arseneault,
am indebted to Norm. My own research has been greatly influ- 2010; Masten et al., 2004). Second, it implies that outcomes
enced by his input as experienced through joint collaborations generally will be explicable on the balance between risk and
and discussions (he was a great talker, full of wit as well as protective factors. That suggests that protective factors can be
wisdom). What made Norm so different from others was the identified on the basis of their nature, rather than their effects.
integrated breadth of his conceptualizations across diverse In many circumstances that works but, as discussed below,
approaches, his rejection of theories that claimed to explain ev- protection may come from risk experiences that lead to suc-
erything, his concern and compassion for those who were dis- cessful coping. Third, and most crucially, it assumes that
advantaged, and his commitment to a positive problem-solving most individuals will respond to stress and adversity in
approach. The whole field of resilience research, and of devel- much the same way and to the same degree or that, at the
opmental psychopathology more generally, was shaped by very least, prevention may best be achieved by acting on
Norms vision; scientific papers today continue to show his im- that assumption.
print (albeit not always explicitly acknowledged). Masten and Powell (2003) have argued that promotive fac-
tors tend to operate in the same way in all populations, and
hence, that resilience can best be promoted by focusing on
Conceptualization of Resilience
competence. Such promotive factors include cognitive abil-
Particularly during the last two decades, there has been a ities, temperament, parenting quality, and good schools.
marked tendency for researchers, clinicians, and policy ma- Their arguments are correct but, nevertheless, do not focus
kers to shift their focus from risk to resilience (e.g., Mohaupt, on the influences that do work differently in the presence of
2008). The aim was to emphasize the positive rather than the adversity; that is what defines resilience.
maladaptive. This was seen in the emergence of positive The concept of resilience has a quite different starting
psychology, as a major movement (Seligman & Csikszent- point. It has its origins in the universal finding from all re-
mihalyi, 2000) and Layards (2005) happiness agenda. search, naturalistic and experimental, human and other ani-
The valuable aspect of this movement was the recognition mals, that there is huge heterogeneity in response to all man-
that eudaimonic socioemotional well-being (including a ners of environmental hazards: physical and psychosocial
sense of purpose and direction) was as important as economic (Rutter, 2006). It is argued that the systematic investigations
success (Keyes, 2007). The less helpful aspect was the trivi- of the causes of this heterogeneity should not just throw light
ality of relabeling family conflict as a risk and family har- on the specifics of different responses to a particular hazard
mony as a protective factor. The most crucial point is that but, in addition, might throw light on a broader range of
there was the downgrading of the seriousness of mental dis- causal processes.
order in order to concentrate on variations in degree of happi- Accordingly, resilience can be defined as reduced vulner-
ness in the general population, and hence the downgrading of ability to environmental risk experiences, the overcoming of
resilience in the face of severe stress and adversity. In addi- a stress or adversity, or a relatively good outcome despite risk
tion, insofar as resilience is concerned, there is the misleading experiences (Rutter, 2006). Thus, it is an interactive concept
implication that it requires generally superior functioning, ra- in which the presence of resilience has to be inferred from in-
ther than relatively better functioning compared with that dividual variations in outcome among individuals who have ex-
shown by others experiencing the same level of stress or ad- perienced significant major stress or adversity (Rutter, 1987).
versity. There are also methodological problems that are in-
herent in the concept of positive mental health (Jahoda,
Testing for Environmental Mediation of Risks
1959) and difficulties in differentiating between hedonic
pleasure and excitement and the quiet satisfaction of a job This interactive concept of resilience necessitates testing the
well done (see Rutter, 2011). postulate that the stress or adversity does entail an envi-
The concept of psychological and social competence ronmentally mediated risk. Of course, it has long been appre-
(Masten et al., 1999) raises a rather different set of issues. It ciated that a statistical correlation or association does not nec-
Resilience as a dynamic concept 337

essarily mean causation. What is new is the range of quasi-ex- on antisocial behavior reflected a genetic liability and not a
perimental research strategies, employing a range of natural causal pathway (Poulton & Moffitt, 2008). Resilience re-
experiments that facilitate causal inferences (Rutter, 2007, search cannot sensibly be based on statistical risk effects
2009, in press-b). These all start with an appreciation of the that have not undergone rigorous testing of the hypothesis
various reasons why a correlation might not imply causation. of environmentally mediated causation.
Thus, the risk might be genetically, rather than envi- Finally, it is crucial that the study of resilience be preceded
ronmentally, mediated; it might represent reverse causation by careful analysis of the elements in the environmental risk
(i.e., the disorder led to the supposed risk factor, rather than variable that actually involve causal influences. History
the other way round); it might reflect social selection (i.e., shows how easy it is to misidentify these. Thus, for many
the process by which individuals select or shape their envi- years it was supposed that broken homes or family breakup
ronments); or it might reflect the origins, rather than the were involved in the causation of both antisocial behavior and
risk actions, of the supposed risk factor. depression (see British Academy Working Group, 2009).
More than a dozen different forms of natural experi- Quantitative analyses of longitudinal data have shown that
ments have been devised and found to be useful. Here, there the risks for antisocial behavior following family separation
is space only to illustrate the strategies through a few exam- are minor compared to the risks from family discord, after
ples. The possibility of genetic mediation has been examined controlling for the other in each case (see Fergusson, Hor-
by discordant twin designs in which one twin experienced the wod, & Lynskey, 1992). Similarly, the proximal risks for de-
risk factor and the other did not (Kendler & Prescott, 2006); pression were found to stem from poor parenting (brought
by assisted reproductive technologies comparing offspring about by family breakup) and not from the breakup as such
born by methods in which the child and mother share genes (Harris, Brown, & Bifulco, 1986).
(e.g., sperm donation) with those such as egg donation in
which that was not the case (Rice et al., 2009; Thapar et al.,
Steeling or Strengthening Effects
2009); and by children of twins designs capitalizing on the
fact that the offspring of monozygotic females are genetically One of the features that particularly characterizes resilience
half-siblings but socially cousins (DOnofrio et al., 2003, research is the recognition of the importance of possible
2008; Silberg & Eaves, 2004; Silberg, Maes, & Eaves, steeling effects. That is exposure to stresses or adversities
2010). The findings have been informative in showing that may either increase vulnerabilities through a sensitization ef-
some risks for psychopathology (such as parental negativity fect or decrease vulnerabilities through a steeling effect. A
and prenatal smoking exposure) are partially genetically key question concerns the circumstances that lead to the
mediated but others (such as physical and sexual abuse) are one rather than the other, plus the equally important question
largely environmentally mediated. The findings have also of the mechanisms that mediate those effects. This research
shown that parental mental disorder that involves a substan- process is most easily illustrated through reference to research
tial genetic liability may nevertheless have environmentally undertaken by David Lyons research group (Lyons et al.,
mediated effects on mental disorder in the children. 2010; Lyons & Parker, 2007; Lyons, Parker, Katz, & Schatz-
Among the designs that can obviate the possibility of so- berg, 2009; Parker, Buckmaster, Schatzberg, & Lyons, 2004)
cial selection are those that examine risks that operate on using squirrel monkeys. This built on the early studies by Gig
the whole population. This is exemplified by the study of Levine in the early 1960s and continued up to the time of his
the effects of prenatal starvation on the risk of developing death (Levine & Mody, 2003).
schizophrenia brought about by the Dutch famine in World In brief, they used a strategy that mimicked the normal ten-
War II (Hoek, Brown, & Susser, 1998; Stein, Susser, Saen- dency in nature for the occurrence of brief motherinfant sep-
ger, & Marolla, 1975); the study of the effects of stopping arations brought about by the mothers going off to forage for
the use of measles, mumps, and rubella vaccines to test the food when the newly weaned offspring reached 36 months
hypothesis that measles, mumps, and rubella had led to an of age. Socially housed squirrel monkeys were randomized
epidemic of autism, finding that it had not done so (Honda, at 17 weeks of age to either brief intermittent separations or
Shimizu, & Rutter, 2005); and the investigation into the ef- a nonseparated control condition. The separated individuals
fects of poverty on psychopathology by examining the bene- were removed from the rearing group for a 2-hr period each
fits for children that followed the relief of poverty as a result week for a total of 10 weeks. After 27 weeks of age, both
of the opening of a casino on a native Indian reservation groups were reared in identical conditions. Behavioral, hor-
(Costello, Compton, Keeler. & Angold, 2003). monal, and brain imaging data were obtained at specified
The possibility for reverse causation could be examined by ages up to adulthood. At 9 months in a novel environment
means of instrumental variable approaches, of which Mende- test, the two groups were initially similar but differences
lian randomization is a specific example (Davey-Smith & Eb- emerged over repeated sessions. Cognitive control was as-
rahim, 2003, 2005). The findings have shown that the sup- sessed at 1.5 years and curiosity in a stress-free situation
posed causal effect of early puberty on the liability to was measured at 2.5 years. On all these measures, the sepa-
alcoholism was largely an artifact (Prescott & Kendler, rated group performed better. In addition, cortisol measures
1999) and the postulated causal effect of early use of alcohol showed decreased reactivity to stress. Neuroimaging showed
338 M. Rutter

that the separated monkeys had a larger ventromedial corticol tended to cope better with the stresses of hospital admission.
volume. Of course, admission to hospital involves multiple stressful
In order to test whether the benefits of intermittent brief elements other than separation. Nevertheless, the implication
separation were mediated by changes in maternal behavior, is that successful happy separations seem to foster resilience
monkeys were randomized to three postnatal conditions, in dealing with unhappy separations.
one of which involved separations of mother and offspring to- The animal model findings have often been interpreted as
gether as a pair. It was found that the changes in arousal reg- meaning that the mechanism involves some sort of inocula-
ulation more closely corresponded to stress exposure than to tion effect; in other words, exposure to a small dose of some
separation-induced changes in maternal care. hazard serves to build up resistance to a major dose through
Other findings showed that similar beneficial effects of the body having had the opportunity to acquire effective de-
brief stress exposure were found in rats. However, this was fenses. That may well be involved, but the human studies
not found with prolonged separation experiences that instead suggest that that is likely to be too narrow a perspective.
led to increased sensitization to later stress experiences rather Thus, the adolescents in Elders study who appeared to be
than steeling. Other rat studies have shown that the adverse strengthened by having to take on new responsibilities as a
effects are a function of uncontrollable stressors (Maier, result of the great economic depression seemed to acquire a
Amat, Baratta, Paul, & Watkins, 2006), a finding that pro- sense of self-efficacy and mastery. It did not appear that
vides a link to human evidence on the benefits of coping the new responsibilities inoculated them against greater re-
(see below). A different randomization study with adult squir- sponsibilities, rather the strengthening effect seemed to oper-
rel monkeys showed again that brief intermittent separations ate more broadly.
(in this case, from a familiar adult male companion) led to in- Similarly, Hauser, Allen, and Goldens (2006) qualitative
creased hippocampal neurogenesis and altered gene expres- study of young adults who had been institutionalized for a se-
sion. rious mental disorder in adolescence but who nevertheless
The overall body of evidence from animal models pro- ended up successful and optimistic was not distinctive in
vides strong evidence of the reality of steeling effects from re- terms of resilience-building inoculation experiences. Rather,
peated brief stress experiences that are not accompanied by what stood out as different from those who were less resilient
overall adversity or deprivation. The next question concerns was a personal agency involving a concern to act to overcome
the applicability of this effect to humans. Clearly, the best ex- adversity, a self-reflective style that meant that they tried to
ample is provided by the resistance to infections that comes assess what was and what was not working for them, and a
about either through the acquisition of natural immunity commitment to relationships.
through exposure to the infectious agents or through immuni- These findings resonate with those from Quinton and Rut-
zation in which induced immunity is brought about by admin- ters (1988) study of girls reared in residential group homes.
istering a controlled dose of a modified version of the patho- As compared with girls from the same geographical area liv-
gen. This is undoubtedly a steeling effect. However, does it ing with their families, many left their institutions in late ado-
apply to psychosocial stressors and psychopathological out- lescence with a feeling that there was nothing they could do to
comes? affect what happened to them. They lacked what was termed
The human evidence on this point is much weaker, largely planning in relation to marriage or work. It was not that
because there have been so few attempts to investigate the they planned badly, but rather that they did not feel able to
matter. However, two examples of possible steeling effects plan at all. However, this did not apply to all the girls.
warrant mentioning. First, there are Elders (1974) longitu- Some did show planning, and it was found that this was as-
dinal analyses of the Californian cohorts going through the sociated with earlier successes in other areas of their lives,
economic depression of the 1920s and 1930s, in which the rarely academic but in some activity important to them,
children had to take on new responsibilities. The follow-up such as sport, music, and positions of responsibility. Those
showed that whereas younger children tended to fair poorly, who did show planning had better outcomes overall. As in
adolescents were sometimes strengthened by the experience the Hauser et al. (2006) study, their resilience seemed to
of having to take on adult roles and doing so successfully. El- stem from successes leading to self-efficacy rather than
ders proposed explanation was that the adolescents with from minor stresses overcome. The findings are far too sparse
greater maturity and experience were better able to take on re- for firm conclusions but they all point to the important quality
sponsibilities and succeed; finding that they could succeed being a state of mind rather than a high IQ or some particular
made them more resilient. By contrast, the younger children temperamental feature.
could not cope so well (and perhaps had a less obvious rele-
vant role) and were therefore sensitized rather than strength-
GeneEnvironment Interactions (G 3 E)
ened.
Second, an example is provided by Stacey, Dearden, Pill, Resilience encompasses resistance to adverse environmental
and Robinsons (1970) finding that children who had experi- influences, as well as steeling following brief intermittent
enced happy separations from their parents (such as staying stressors (Rutter, 2009). This is best illustrated by the findings
with their grandparents or having sleepovers with friends) on G  E. Quantitative genetic studies pointed to the likeli-
Resilience as a dynamic concept 339

hood of G  E operating, perhaps especially with respect to episode of depression. Second, a finding from Uher et al.
depression and antisocial behavior (Rutter & Silberg, (2011) showed that the G  E mainly applied to recurrent or
2002). However, the situation was transformed by the avail- chronic depression rather than a single episode depressive
ability of molecular genetic methods to identify individual disorder. This again suggests a biological effect on liability
susceptibility genes. G  E could then be tested for by using rather than a provoking effect on onset.
those genes in relation to measured environments that have Third, the next finding to emphasize is that the serotonin
been shown to involve environmentally mediated risks for transporter promoter gene G  E applied to depression as an
psychopathology (Moffitt, Caspi, & Rutter, 2005). Moffitt, outcome, but not to antisocial behavior. Conversely, the mon-
Caspi, and colleagues led the way through a series of papers oamine oxidase A GE applied to antisocial behavior but not
based on the Dunedin Longitudinal Study (see, e.g., Caspi to depression. That has two important implications. It means
et al., 2002, 2003, 2005). In each case, there was no signifi- that resilience cannot be viewed as a general trait; the interac-
cant main effect of genes, a marginally significant main effect tion found in relation to one outcome does not necessarily ap-
of environment, but a clearly significant effect of the G  ply to others. In addition, just because there was no measur-
E. The GE held up after rigorous control for possible alter- able effect of maltreatment on depression in the individuals
native explanations (such as the scaling effect, an effect of a with the short allele polymorphism, that does not mean that
geneenvironment correlation [rGE], and GG interactions; these individuals are invulnerable to the effects of maltreat-
see Rutter, Thapar, & Pickles, 2009). ment, because they may develop other adverse outcomes.
The issues are most easily considered in relation to the Du- Fourth and finally, it is necessary to ask whether the ge-
nedin finding of GE between the short allele polymorphism netic effect operates as an influence on vulnerability to ad-
of the serotonin transporter promoter gene and both life verse experiences or rather an influence on susceptibility to
events and maltreatment (Caspi et al., 2003). The finding all experiences, good or bad. Pluess and Belsky (2009) and
has been replicated or partially replicated many times (Uher Boyce and Ellis (2005) have argued that, from an evolution-
& McGuffin, 2008), but Risch et al. (2009) sought to claim ary perspective, an effect on response to all experiences is
that this was likely to be an artifact. The claim was based more likely. In other words the relevant genetic polymor-
on a flawed meta analysis of a biased subset of studies using phism that is associated with vulnerability to bad experiences
life events (see Caspi, Hariri, Holmes, Uher, & Moffitt, 2010; may also be associated with a better response to good experi-
Uher & McGuffin, 2010), which ignored findings based on ences in the absence of environmental adversity. There are
maltreatment and ignored all experimental biological studies findings that are consistent with this proposition (see Ellis,
in humans and in nonhuman animals (see Rutter, 2010; Rut- Boyce, Belsky, Bakermans-Kranenburg, & van IJzendoorn,
ter et al., 2009). Two examples of experimental studies may 2011) but the findings in humans so far leave some important
be given: one in rhesus monkeys focusing on central seroto- questions unanswered. Thus, the Belsky and Beaver (2011)
nin functioning (Bennett et al., 2002) and one in humans, study found the differential susceptibility only in males;
using brain imaging to examine the neural effects of G  E Pluess and Belsky (2011) had findings that suggested possi-
(Hariri et al., 2002; see also Hariri, 2011). Both were strongly ble prenatal programming of postnatal plasticity; and Obrado-
supportive of the reality of biological GE. The human study vic, Bush, and Boyce (2011) found that the effects varied
was additionally important because the findings derived from according to whether the challenges were interpersonal or
a sample of individuals deliberately chosen to be free of psy- cognitive. Nevertheless, the balance of findings indicates
chopathology. The implication is that the G  E biological the likelihood that the genetic polymorphisms that are associ-
pathway, although relevant for depression, is one present in ated with increased vulnerability to bad experiences may also
the general population who do not have depression. The find- be associated with greater sensitivity to good experiences.
ing reinforces the epidemiological claim that the G  E does The finding from the Suomi group studies (Suomi, per-
not reflect a genetic main effect on depression. sonal communication, 2011) has been striking in finding
Further epidemiological studies have been highly informa- that all the G  E interactions they have found show this
tive in focusing down on the specifics of GE with the sero- crossover effect. Moreover, it is even more striking that
tonin transporter promoter gene. First, Karg, Burmeister, the less efficient transcription allele is the one associated
Shedden, and Sen (2011) brought together a diverse range with the greater vulnerability to adversities, whereas the
of studies focusing on different types of stressor. The findings more efficient transcription allele is the one with the vulner-
showed that, although there was a marginally significant G ability effect in advantageous circumstances. The mecha-
E with life events, there was a much larger highly significant nisms involved have yet to be determined but the phenom-
GE with maltreatment. That is important both because it fo- enon certainly seems to have a biological plausibility.
cuses on a more serious environmental risk factor and be-
cause the E applied to experiences in childhood long before
Life Course Effects
the onset of a depressive episode in late adolescence or early
adult life. The implication is that a causal biological pathway The research discussed so far has mainly concerned circum-
brought about earlier changes associated with the liability to stances in early life, and it is necessary to ask whether the
depression, rather than a provoking effect of the onset of an overcoming of adversity can be influenced by experiences
340 M. Rutter

in adult life. There is clear evidence from both animal models sort of career. It may be concluded that positive experiences
and human studies that adult experiences have both behav- in adult life can do much to counter the effects of early adver-
ioral and neural effects (Keating, 2011). However, the issue sities provided that they serve to both cut off the past and
here is the rather different one of whether such experiences provide new opportunities.
can counter or alleviate the effects of earlier adversities. Find- Although not planned as a study of turning points, the
ings from Lyons et al. (2010), studying brief separation ex- adult follow-up of Masten et al.s (2004) Project Competence
periences in adult squirrel monkeys, has already been noted. produced strikingly comparable findings. The seven indi-
Such experiences not only provided a behavioral and neu- viduals who made a dramatic change from maladaptive to re-
roendocrine steeling effect with respect to later stresses, but silient over the transition to adulthood differed from their
also there were measurable neural and epigenetic effects. Al- peers in terms of planfulness, future motivation, autonomy,
though the findings provide a convincing demonstration of and adult support outside the family. Thus, both mental qual-
steeling effects in adult life in relation to later stresses, they ities and new opportunities seemed important. Individual
do not address the question of effects that serve to counter cases suggested that the latter happened through moving to
earlier adversities. There is other animal research that sug- a good job, marrying into better functioning families, experi-
gests that this does occur but it may be more fruitful to turn encing religious conversion, and/or the pursuit of higher edu-
to human studies on life span effects. cation. Bowes et al.s (2010) separate study also showed the
Probably the best research is that undertaken by Sampson role of positive family relationships in fostering resilience.
and Laub (1993; Laub & Sampson, 2003). Their starting
point was boys in residential institutions for delinquents, first
Resilience Versus the Summative Effect of Risk and
studied by the Gluecks (Glueck & Glueck, 1950). Laub and
Protective Factors
Sampson followed the sample up to 70 years of age, and
Laub personally interviewed a purposively chosen subsample It is necessary now to return to the starting point of the differ-
of 52, chosen to represent those who had an outcome much ences between the interactive concept of resilience and the
better than expected on the basis of their previous behavior summative vulnerability and competence approach. The con-
and psychological background, as well as those who had ceptual differences were outlined in Rutter (2006), but here
done averagely well or as expected. One of the adult experi- we focus on the rather different issues of the research impli-
ences they investigated was marriage, which had been postu- cations and substantive findings.
lated to have a protective effect (Laub, Nagin, & Sampson
1998). There first approach was a multivariate analysis that
Research considerations
did indeed show that marriage had a protective effect. They
then argued that if this was a true effect, it should follow Four key research implications stand out. First, the focus on in-
that crime rates should vary over time according to whether dividual differences in outcome requires a specific identifica-
the individuals were currently married or not married. A pro- tion of the key risk factors and, second, it demands rigorous
pensity score approach was used to create groups equivalent testing of the hypothesis that the risks are truly environmentally
with respect to a propensity to marriage. The findings showed mediated. Neither of these needs tend to be met in even the best
that marriage was associated with a reduction in crime of ap- of the vulnerability/competence studies (Fergusson & Hor-
proximately 36% to 43% (Sampson, Laub, & Wimer, 2006). wood, 2003; Masten & Powell, 2003; Sameroff, Gutman, &
The final step lay in Laubs qualitative interviews. These Peck, 2003), although its importance has been recognized (Lu-
showed that marriage constituted far more than the provision thar & Brown, 2007). Third, the examination of individual dif-
of a stable attachment relationship. It brought about a new ferences calls for experimental testing (preferably combined
kinship group and a new peer group, as well as a strong guid- with biological measures). Fourth, it also calls for the use of an-
ing influence from the wife. Marriage was a life-changing ex- imal models. This was evident in the research into steeling ef-
perience and not just an event. fects and also that into geneenvironment interactions. Of
A second adult life experience examined was military ser- course, this is not to say that none of these needs are ever
vice. Perhaps unexpectedly, this was associated with a sub- met in vulnerability/competency research; but it is to say that
stantially better outcome in the socially disadvantaged youths they are to the forefront as a priority in resilience research,
studied (Sampson & Laub, 1996). The meaning of this find- whereas that is less commonly the case in other approaches.
ing has to take account of both the social context (many of the The need for a perspective that spans biology and psychosocial
youths had dropped out of schooling and viewed their future influences has become accepted (see Cicchetti, 2010), but the
in very negative terms) and the benefits (both intended and crucial differences between resilience research and risk and pro-
incidental) that came with military service. The intended ben- tective factors research are less well appreciated.
efit came from the educational opportunities provided by the
GI bill and the unintended from the postponement of mar-
Substantive findings
riage that resulted in having a much wider choice of partners,
many of whom did not have the same disadvantaged back- The resilience research has given rise to several substantive
ground, and marriage at a time of having established some findings that would have been unlikely to arise if the focus
Resilience as a dynamic concept 341

had not been on the individual differences in response to the Seventh and finally, although recognizing the huge strides
same level and type of adversities. First, there are the steeling in understanding resilience that have been made since Garme-
effects finding that intermittent exposure to brief periods of zys pioneering concepts and findings, it has to be accepted
stress, far from being damaging, increases resistance to later that many key questions have still to be addressed. It is cru-
stresses. This has been shown by both behavioral and neu- cially important that, in pointing to new research successes,
roendocrine effects. The conclusion is in keeping with the bi- the results are not oversold. All research builds on a broad
ological understanding that developmental benefits from base of investigations by other research groups, and it is
meeting, and successfully coping with, challenges (Rutter rare indeed for a single finding to constitute a true break-
& Rutter, 1992). It is also in keeping with the consistent find- through.
ing that resistance to infectious agents comes through expo-
sure to, and not avoidance of, contact with those agents. In ad-
Conclusions
dition, it is consonant with the clinical evidence that the
treatment of phobias benefits from exposure to and is ham- Resilience research has as its starting point the universal find-
pered by avoidance of, the feared object. It may be concluded ing of huge heterogeneity in outcomes after all types of envi-
that these steeling effects needed a resilience approach for ronmental adversity, together with the evidence that, in some
their identification. circumstances, exposure to stress may be followed by an in-
Second, the genetic findings on G  E could only have creased resistance to later stress (a steeling effect), rather
come from a focus on interactive effects. In all the replicated than a sensitization or increased vulnerability. In other words,
findings, no main genetic effect has been found. The genetic the focus is on individual differences in response to adversity
influence was not on a liability for a particular mental disor- rather than an assumption that outcomes can be accounted for
der, but rather on a susceptibility to environmental influences. in terms of the balance between positive and negative influ-
The findings were also important in showing that the main ences, with the assumption that they will affect most people
risk effect came from serious, chronic adversities (such as in the same way and to the same degree.
maltreatment) rather than acute stresses. However, this is There are some nine features that serve to characterize re-
even more important in showing that the GE operated in re- silience research as distinctive from the overall field of risk
lation to environmental experiences long before the onset of and protective factors. First, there is a direct analysis of the
disorder. Some sort of biological pathway (reflecting G  E) features associated with heterogeneity in response to adver-
that predisposed to a psychopathological liability, rather than sity, rather than a reliance on statistical approaches to detect
the provoking of an onset, seemed to be operative. nonlinear interactive effects. The statistical power to detect
Third, the life course findings indicate that appropriate ex- interactions is inevitably less than the power to determine
periences in adult life can do much to counter the effects of the associations with heterogeneity of outcomes. In addition,
earlier adversities. It needs to be noted, however, that such ex- there is the requirement to test for environmental mediation
periences are not simply pleasurable happenings but, rather, of risk effects, rather than relying on quantifying a hetero-
experiences that create a helpful discontinuity with the past, geneous mixture of risks that may be either genetically medi-
and increase opportunities and enhance coping. ated or environmentally mediated or both.
Fourth, resilience is accompanied by important biological Second, there is an interest in variables that are without ef-
changes, neuroendocrine and neural. Resilience is a dynamic fect in the general population of lower risk individuals but
concept in which successful coping may involve a compli- which have substantial effects in the presence of adversity.
cated mixture of psychological habituation, changes in mental Adoption is the obvious example of this kind. Of course, it
set, alterations in perceived and actual self-efficacy, hormonal could be identified in risk and protective factor studies but
changes (especially in the hypothalamicpituitaryadrenal it has not been so identified, probably because of its infre-
axis) and neural alterations. quency in the total population. Planning constitutes a further
Fifth, the findings on ideas, attributions, self-reflection, example in which its origins lay in good experiences outside
and planning emphasize the importance of mental phenom- the family. That would be unlikely to have been picked up in
ena in the response to stress and adversity (i.e., in the pro- the usual type of risk and protective factor study. However, it
cesses of coping and not just in terms of enduring tempera- is both relevant and noteworthy that the importance of plan-
mental features). ning was detected in Masten et al.s (2004) competence study
Sixth, the interactive resilience approach emphasizes that once there was a focus on individual differences.
resilience needs to be judged, not in terms of superior overall Third, there is an interest in the steeling effects of success-
functioning as judged in relation to the population as a whole, fully coping with stress or challenge. That could have arisen as
but rather in terms of functioning that is relatively better than a result of risk/protective studies but it has not been a promi-
that shown by others experiencing the same level of adversity. nent feature, probably because what was needed was a focused
As the study of Romanian adoptees indicated (Rutter & So- hypothesis-testing approach, a feature of resilience research but
nuga-Barke, 2010), that may well mean a mixture of impor- not other approaches, at least not to the same extent.
tant real-life successes in the context of some continuing dif- Fourth, as a specific example of hypothesis-driven strate-
ficulties. gies, there is the group of studies of GE interactions. Once
342 M. Rutter

more, the need (that was met) was for hypotheses driven by Eighth, there are the basic science findings on brain plas-
biological findings (see Rutter et al., 2009). In addition, how- ticity (see Rutter, in press-c), which underline the dynamic
ever, there was the explicit acceptance that epidemiological nature of plasticity in terms of its temporal limits and its open-
findings had to be put to the test through human experimental ness to external influences.
studies, animal models, and basic science. Ninth and finally, resilience is defined in terms of a better
Fifth, that brings in the central importance in resilience re- outcome than that seen in other individuals from a similarly
search of animal models, of which the squirrel monkey stud- adverse background. In short, there is no requirement of su-
ies of possible stress inoculation, represent a good example. perior functioning in relation to the nondeprived population
The focus is explicitly on possible steeling effects and an ex- as a whole. The study of Romanian adoptees who experi-
perimental approach is used. enced profoundly depriving care (Rutter & Sonuga-Barke,
Sixth, there is the study of possible turning point effects, as 2010) constitutes a good illustration. Although, in the group
illustrated by the study of the beneficial effects of marriage as a whole, deficits were apparent, some individuals fared sur-
and of early service in the Armed Forces for individuals prisingly well. There was marked relative success of a mean-
from a disadvantaged background living through the Great ingful kind. Once again, the risks were shown to be envi-
Depression of the 1930s. General population longitudinal ronmentally mediated and a hypothesis-testing approach
studies provided the data but it was the focus on individual was followed.
differences that brought this research into the resilience arena. As indicated in the introductory section, resilience concepts
Seventh, a key feature of resilience research has been the accept, and build on, the importance of risk and protective fac-
use of qualitative data to determine the meaning of experi- tors research (and require its operation), but they add to it in cru-
ences. The research into marriage constitutes one example cially important ways that would not have emerged at all
of this and the study of positive outcomes following inpatient readily out of other approaches. The fields of competence, pos-
psychiatric care in adolescence constitutes another rather dif- itive psychology, risk and protection and resilience all have im-
ferent example. portance, but it is a mistake to want to group them together.

References
Anthony, E. J. (1974). The syndrome of the psychologically invulnerable Costello, E. J., Compton, S. N., Keeler, S. N., & Angold, A. (2003). Relation-
child. In E. J. Anthony & C. Koupernik (Eds.), The child in his family: ships between poverty and psychopathology: A natural experiment. Jour-
Children at psychiatric risk (pp. 529545). New York: Wiley. nal of the American Medical Association, 290, 20232029.
Anthony, E. J., & Cohler, B. J. (Eds.). (1987). The invulnerable child. Davey Smith, G., & Ebrahim, S. (2003). Mendelian randomization: Can
New York: Guilford Press. genetic epidemiology contribute to understanding environmental deter-
Bennett, A. J., Lesch, K. P., Heils, A., Long, J. C., Lorenz, J. G., Shoaf, S. E., minants of disease? International Journal of Epidemiology, 32, 122.
et al. (2002). Early experience and serotonin transporter gene variation inter- Davey Smith, G., & Ebrahim, S. (2005). What can Mendelian randomization
act to influence primate CNS function. Molecular Psychiatry, 7, 118122. tell us about modifiable behavioural and environmental exposures. Brit-
Belsky, J., & Beaver, K. M. (2011). Cumulativegenetic plasticity, parenting ish Medical Journal, 330, 10761079.
and adolescent self-regulation. Journal of Child Psychology and Psychia- DOnofrio, B. M., Turkheimer, E., Eaves, L. J., Corey, L. A., Berg, K., So-
try, 52, 619626. laas, M. H., et al. (2003). The role of the children of Ttwins design in elu-
Bleuler, M. (1978). The schizophrenic disorders: Long-term patient and cidating causal relations between parent characteristics and child out-
family studies. New Haven, CT: Yale University Press. comes. Journal of Child Psychology and Psychiatry, 44, 11301144.
Bowes, L., Maughan, B., Caspi, A., Moffitt, T. E., & Arseneault, L. (2010). DOnofrio, B. M., Van Hulle, C. A., Waldman, I. D., Rodgers, J. L., Harden,
Families promote emotional and behavioural resilience to bullying: K. P., Rathouz, P. J., et al. (2008). Smoking during pregnancy and off-
Evidence of an environmental effect. Journal of Child Psychology and spring externalizing problems: An exploration of genetic and environ-
Psychiatry, 51, 809817. mental confounds. Development and Psychopathology, 20, 139164.
Boyce, W. T., & Ellis, B. J. (2005). Biological sensitivity to context: I. An Eisenberg, L. (1977). Development as a unifying concept in psychiatry. Brit-
evolutionarydevelopmental theory of the origins and functions of stress ish Journal of Psychiatry, 131, 225237.
reactivity. Development and Psychopathology, 17, 271301. Elder, G. H. (1974). Children of the Great Depression. Chicago: University
British Academy Working Group Report. (2009). Social science and family of Chicago Press.
policies. London: British Academy Policy Centre. Ellis, B. J., Boyce, W. T., Belsky, J., Bakermans-Kranenburg, M. J., & van
Caspi, A., Hariri, A. R., Holmes, A., Uher, R., & Moffitt, T. E. (2010). Ge- IJzendoorn, M. H. (2011). Differential susceptibility to the environment:
netic sensitivity to the environment: The case of the serotonin transporter An evolutionaryneurodevelopmental theory. Development and Psycho-
gene and its implications for studying complex diseases and traits. Amer- pathology, 23, 728.
ican Journal of Psychiatry, 167, 509527. Fergusson, D. M., & Horwood, L. J. (2003). Resilience to childhood adver-
Caspi, A., McClay, J., Moffitt, T. E., Mill, J., Martin, J., Craig, I. W., et al. sity: Results of a 21-year study. In S. S. Luthar (Ed.), Resilience and vul-
(2002). Role of genotype in the cycle of violence in maltreated children. nerability: Adaptation in the context of childhood adversities (pp. 130
Science, 297, 851854. 155). Cambridge: Cambridge University Press.
Caspi, A., Moffitt, T. E., Cannon, M., McClay, J., Murray, R., Harrington, H., Fergusson, D. M., Horwood, L. J., & Lynskey, M. T. (1992). Family change,
et al. (2005). Moderation of the effect of adolescent-onset cannabis use parental discord and early offending. Journal of Child Psychology and
on adult psychosis by a functional polymorphism in the catechol-o-me- Psychiatry, 33, 10591075.
thyltransferase gene: Longitudinal evidence of a gene environment inter- Garmezy, N. (1974). The study of competence in children at risk for severe
action. Biological Psychiatry, 57, 11171127. psychopathology. In E. J. Anthony & C. Koupernik (Eds.), The child in
Caspi, A., Sugden, K., Moffitt, T. E., Taylor, A., Craig, I. W., Harrington, H., his family: Children at psychiatric risk (Vol. 3, pp. 7797). New York:
et al. (2003). Influence of life stress on depression: Moderation by a poly- Wiley.
morphism in the 5-HTT gene. Science, 301, 386389. Garmezy, N. (1985). Stress-resistant children: The search for protective fac-
Cicchetti, D. (2010). Resilience under conditions of extreme stress: A multi- tors. In A. Davids (Ed.), Recent research in developmental psychopathol-
level perspective. World Psychiatry, 9, 145154. ogy (pp. 213233). Elmsford, NY: Pergamon Press.
Resilience as a dynamic concept 343

Garmezy, N., Masten, A. S., & Tellegen, A. (1984). The study of stress and Masten, A. S., & Tellegen, A. (2012). Resilience in developmental psycho-
competence in children: A building block for developmental psychopa- pathology: Contributions of the Project Competence Longitudinal Study.
thology. Child Development, 55, 97111. Development and Psychopathology, 24, 345361.
Glueck, S., & Glueck, E. (1950). Unraveling juvenile delinquency. Moffitt, T. E., Caspi, A., & Rutter, M. (2005). Strategy for investigating in-
New York: Commonwealth Fund. teractions between measured genes and measured environments. Ar-
Harris, T., Brown, G. W., & Bifulco, A. (1986). Loss of parent in childhood chives of General Psychiatry, 62, 473481.
and adult psychiatric disorder: The role of lack of adequate parental care. Mohaupt, S. (2008). Review article: Resilience and social exclusion. Social
Psychological Medicine, 16, 641659. Policy & Society, 8, 6371.
Hauser, S., Allen, J., & Golden, E. (2006). Out of the woods: Tales of resilient Obradovic, J., Bush, N. R., & Boyce, T. (2011). The interactive effect of mar-
teens. Cambridge, MA: Harvard University Press. ital conflict and stress reactivity on externalizing and internalizing symp-
Hariri, A. R. (2011). Neurobiological mechanisms supporting geneenviron- toms: The role of laboratory stressors. Development and Psychopathol-
ment interaction effects. In K. A. Dodge & M. Rutter (Eds.), Geneenvi- ogy, 23, 101114.
ronment interactions in developmental psychopathology (pp. 5970). Parker, K. J., Buckmaster, C. L., Schatzberg, A. F., & Lyons, D. M. (2004).
New York: Guilford Press. Prospective investigation of stress inoculation in young monkeys. Ar-
Hariri, A. R., Mattay, V. S., Tessitore, A., Kolachana, B., Fera, F., Goldman, chives of General Psychiatry, 61, 933941.
D., et al. (2002). Serotonin transporter genetic variation and the response Parker, K. J., & Maestripieri, D. (2011). Identifying the key features of early
of the human amygdala. Science, 297, 400403. stressful experiences that produce stress vulnerability and resilience in
Hoek, H. W., Brown, A. S., & Susser, E. (1998). The Dutch famine and primates. Neuroscience & Biobehavioral Reviews, 35, 14661483.
schizophrenia spectrum disorders. Social Psychiatry and Psychiatric Epi- Pluess, M., & Belsky, J. (2009). Differential susceptibility to rearing experi-
demiology, 33, 373379. ence: The case of childcare. Journal of Child Psychology and Psychiatry,
Honda, H., Shimizu, Y., & Rutter, M. (2005). No effect of MMR withdrawal 50, 396404.
on the incidence of autism: A total population study. Journal of Child Pluess, M., & Belsky, J. (2011). Prenatal programming of postnatal plastic-
Psychology and Psychiatry, 46, 572579. ity? Development and Psychopathology, 23, 2938.
Jahoda, M. (1959). Current concepts of positive mental health. New York: Poulton, R., & Moffitt, T. E. (2008). Is it important to prevent early exposure
Basic Books. to drugs and alcohol among adolescents? Psychological Science, 19,
Karg, K., Burmeister, M., Shedden, K., & Sen, S. (2011). The serotonin 10371044.
transporter promoter variant (5-HTTLPR), stress, and depression meta- Prescott, C. A., & Kendler, K. S. (1999). Age at first drink and risk for alco-
analysis revisited: Evidence of genetic moderation. Archives of General holism: A noncausal association. Alcoholism: Clinical and Experimental
Psychiatry, 68, 444454. Research, 23, 101107.
Keating D. P. (Ed.). (2011). Nature and nurture in early child development. Quinton, D., & Rutter, M. (1988). Parenting breakdown: The making and
New York: Cambridge University Press. breaking of intergenerational links. Aldershot: Avebury.
Kendler, K. S., & Prescott, C. A. (2006). Genes, environment, and psychopa- Rice, F., Harold, G., Boivin, J., Hay, D., van den Bree, M., & Thapar, A.
thology: Understanding the causes of psychiatric and substance use dis- (2009). Disentangling prenatal and inherited influences in humans with
orders. New York: Guilford Press. an experimental design. Proceedings of the National Academy of Sci-
Keyes, C. L. M. (2007). Promoting and protecting mental health as flourish- ences of the United States of America, 106, 24642467.
ing: A complementary strategy for improving national mental health. Risch, N., Herrell, R., Lehner, T., Liang, K. Y., Eaves, L., Hoh, J., et al.
American Psychologist, 62, 95108. (2009). Interaction between the serotonin transporter gene (5-HTTLPR),
Laub, J. H., Nagin, D. S., & Sampson, R. J. (1998). Trajectories of change in stressful life events, and risk of depression: A meta-analysis. Journal of
criminal offending: Good marriages and the desistance process. Ameri- the American Medical Association, 301, 24622471.
can Sociological Review, 63, 225238. Rutter, M. (1986). Child psychiatry: The interface between clinical and de-
Laub, J. H., & Sampson, R. J. (2003). Shared beginnings, divergent lives: velopmental research. Psychological Medicine, 16, 151169.
Delinquent boys to age. 70 Cambridge, MA: Harvard University Press. Rutter, M. (1987). Psychosocial resilience and protective mechanisms. Amer-
Layard, R. (2005). Happiness. New York: Penguin Press. ican Journal of Orthopsychiatry, 57, 316331.
Levine, S., & Mody, T. (2003) The long-term psychobiological conse- Rutter, M. (1989). Psychiatric disorder in parents as a risk factor in children.
quences of intermittent postnatal separation in the squirrel monkey. Neu- In D. Shaffer, I. Philips, N. Enver, M. Silverman, & V. Anthony (Eds.),
roscience & Biobehavioral Reviews, 27, 8389. Prevention of psychiatric disorders in child and adolescent: The
Luthar, S. S., & Brown, P. J. (2007). Maximizing resilience through diverse project of the American Academy of Child & Adolescent Psychiatry.
levels of inquiry: Prevailing paradigms, possibilities, and priorities for the OSAP Prevention Monograph 2 (pp. 157189). Rockville, MD: US
future. Development and Psychopathology, 19, 931955. Department of Health and Human Services, Office of Substance Abuse
Lyons, D. M., Buckmaster, P. S., Lee, A. G., Wu, C., Mitra, R., Duffey, L. Prevention.
M., et al. (2010). Stress coping stimulates hippocampal neurogenesis in Rutter, M. (1999). Social context: Meanings, measures and mechanisms. Eu-
adult monkeys. Proceedings of the National Academy of Sciences of ropean Review, 7, 139149.
the United States of America, 107, 1482314827. Rutter, M. (2006). Implications of resilience concepts for scientific under-
Lyons, D. M., & Parker, K. J. (2007). Stress inoculation-induced indi- standing. Annals of the New York Academy of Sciences, 1094, 112.
cations of resilience in monkeys. Journal of Trauma Stress, 20, Rutter, M. (2007). Proceeding from observed correlation to causal inference:
423433. The use of natural experiments. Perspectives on Psychological Science,
Lyons, D. M., Parker, K. J., Katz, M., & Schatzberg, A. F. (2009). Develop- 2, 377395.
mental cascades linking stress inoculation, arousal regulation, and resili- Rutter, M. (2008). Developing concepts in developmental psychopathology.
ence. Frontiers in Behavioral Neuroscience, 3, 32. In J. J. Hudziak (Ed.), Developmental psychopathology and wellness:
Maier, S. F., Amat, J., Baratta, M. V., Paul, E., & Watkins, L. R. (2006). Be- Genetic and environmental influences (pp. 322). New York: American
havioral control, the medial prefrontal cortex, and resilience. Dialogues in Psychiatric Publications.
Clinical Neuroscience, 8, 397406. Rutter, M. (2009). Understanding and testing risk mechanisms for mental
Masten, A. S., Burt, K. B., Roisman, G. I., Obradovic, J., Long, J. D., & disorders. Journal of Child Psychology and Psychiatry, 50, 4452.
Tellegen, A. (2004). Resources and resilience in the transition to adult- Rutter, M. (2010). Geneenvironment interplay. Depression & Anxiety, 27,
hood: Continuity and change. Developmental Psychopathology, 6, 14.
10711094. Rutter, M. (2011). Resilience: Causal pathways and social ecology. In M. Un-
Masten, A. S., Hubbard, J. J., Gest, S. D., Tellegen, A., Garmezy, N., & Rai- gar (Ed.), The social ecology of resilience. New York: Springer.
merz, M. (1999). Competence in the context of adversity: Pathways to re- Rutter, M. (in press-a). The role of science in understanding family troubles.
silience and maladaptation from childhood to late adolescence. Develop- In J. McCarthy (Ed.), Family troubles? London: Policy Press.
ment and Psychopathology, 11, 143169. Rutter, M. (in press-b). Natural experiments as a means of testing causal
Masten, A. S., & Powell, J. L. (2003). A resilience framework for research, inferences. In C. Barzini, P. Dawid, & L. Bernardinelli (Eds.), Statistical
policy and practice. In S. S. Luthar (Ed.), Resilience and vulnerability: methods in causal inference. New York: Guilford Press.
Adaptation in the context of childhood adversities (pp. 128). Cam- Rutter, M. (in press-c). Resilience: Clinical implications. Journal of Child
bridge: Cambridge University Press. Psychology and Psychiatry.
344 M. Rutter

Rutter, M., & Garmezy, N. (1983). Developmental psychopathology. In E. Silberg, J. L., & Eaves, L. J. (2004). Analysing the contributions of genes and
M. Hetherington (Ed.), Mussens handbook of child psychology: parentchild interaction to childhood behavioural and emotional prob-
Vol. 4. Socialization, personality, and social development (4th ed., pp. lems: A model for the children of twins. Psychological Medicine, 34,
775911). New York: Wiley. 347356.
Rutter, M., & Rutter, M. (1992). Developing minds: Challenges and continu- Silberg, J. L., Maes, H., & Eaves, L. J. (2010). Genetic and environmental
ity across the life span. New York: Basic Books. influences on the transmission of parental depression to childrens de-
Rutter, M., & Silberg, J. (2002). Geneenvironment interplay in relation to pression and conduct disturbance: An extended children of twins study.
emotional and behavioral disturbance. Annual Review of Psychology, Journal of Child Psychology and Psychiatry, 51, 734744.
53, 463490. Stacey, M., Dearden, R., Pill, R., & Robinson, D. (1970). Hospitals, children
Rutter, M., & Sonuga-Barke, E. J. (Eds). (2010). Deprivation-specific psy- and their families: The report of a pilot study. London: Routledge & Ke-
chological patterns: Effects of institutional deprivation. Monographs of gan Paul.
the Society for Research in Child Development, 75, 1252. Stein, Z. A., Susser, M., Saenger, G., & Marolla, F. (1975). Famine and hu-
Rutter, M., Thapar, A., & Pickles, A. (2009). From JAMA: Commentary on pa- man development: The Dutch hunger winter of 19441945. New York:
per by Risch et al. (2009). Geneenvironment interactions: Biologically va- Oxford University Press.
lid pathway or artefact? Archives of General Psychiatry, 66, 12871289. Thapar, A., Rice, F., Hay, D., Bolvin, J., Langley, K., Van den Bree, M., et al.
Sameroff, A., Gutman, L. M., & Peck, S. C. (2003). Adaptation among youth (2009). Prenatal smoking may not cause ADHD: Evidence from a novel
facing multiple risks: Prospective research findings. In S. S. Luthar (Ed.), design. Biological Psychiatry, 66, 722727.
Resilience and vulnerability: Adaptation in the context of childhood ad- Uher, R., Caspi, A., Houts, R., Sugden, K., Williams, B., Poulton, R., et al.
versities (pp. 364391). Cambridge: Cambridge University Press. (2011). Serotonin transporter gene moderates childhood maltreatments
Sampson, R. J., & Laub, J. H. (1993). Crime in the making: Pathways and effects on persistent but not single-episode depression: Replications
turning points through life. Cambridge, MA: Harvard University Press. and implications for resolving inconsistent results. Journal of Affective
Sampson, R. J., & Laub, J. H. (1996). Socioeconomic achievement in the life Disorders. Advance on-line publication.
course of disadvantaged men: Military service as a turning point, circa Uher, R., & McGuffin, P. (2008). The moderation by the serotonin trans-
19401965. American Sociological Review, 61, 347367. porter gene of environmental adversity in the aetiology of mental illness:
Sampson, R. J., Laub, J. H., & Wimer, C. (2006). Does marriage reduce Review and methodological analysis. Molecular Psychiatry, 13, 131
crime? A counterfactual approach to within-individual causal effects. 146.
Criminology, 44, 465508. Uher, R., & McGuffin, P. (2010). The moderation by the serotonin trans-
Seery, M. D. (2011). Resilience: A silver lining to experiencing adverse life porter gene of environmental adversity in the etiology of depression:
events? Current Directions in Psychological Science, 20, 390394. 2009 update. Molecular Psychiatry, 15, 1822.
Seligman, M. E. P., & Csikszentmihalyi, M. (2000). Positive psychology: An
introduction. American Psychology, 55, 514.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.