PEDIATRIC DENTISTRY V 36 / NO 2 MAR / APR 14

Conference Paper
The Lifelong Effects of Early Childhood Adversity and Toxic Stress
W. Thomas Boyce, MD

Abstract: A rapidly expanding body of research indicates that early social environments characterized by adversity, subordination and stress, along
with individual differences in susceptibility to such environments, create risks for lifelong chronic diseases, including declines in oral health.
Emerging findings suggest that gene-environment interplay, resulting in epigenetically regulated differences in gene expression, underlie many
such declines in health. The origins of these processes in early life reveal how many of the chronic morbidities of adulthood should be viewed
as developmental disorders, with etiologic roots in childhood. (Pediatr Dent 2014;36:102-8)
KEYWORDS: DEVELOPMENTAL ORIGINS, CHILDHOOD STRESS, GENE-ENVIRONMENT INTERPLAY

The convergence of 'nature' and 'nurture'. Within the pro- Twin studies and other genetically informed research, how-
fessional lifetimes of a single generation, dentists, physicians ever, repeatedly found that almost no disorders were completely
and other health care providers have witnessed wholesale heritable; that is, virtually none showed the 100 percent con-
shifts in accepted theories of disease causation, shifts as dra- cordance among monozygotic twins that would be expected if
matic as those attending the historical emergence of 'germ a disease were exclusively genetic in origin. Rather, the picture
theory' or the confirmation of smoking and air pollution as that began to emerge, within fields such as "behavior genetics,"
etiologic agents in respiratory disease. During the 1960s and was one of separable, shared variance in health outcomes, attri-
1970s, the training decades for many health professionals in butable to the combined, additive influences of environmental
practice today, a strong, pervasive environmental determinism and genetic factors. Behavior genetics thus utilized twin studies
held that most diseases of modernity would be eventually attri- and other genetically informed research designs to estimate the
butable to the pathogenic effects of external agents, such as "heritable" portion of complex behavioral disorders. None-
toxins, diets, pollutants and social conditions. In the theless, new work has increasingly revealed that this effort to
"epidemiologic transition" from acute, often lethal diseases of disaggregate outcome variance into that attributable to genes
the past to the chronic, disabling diseases of the late 20th versus environments was deeply flawed by its inattention to a
century, attention shifted from air- and water- borne micro- far more fundamental and consequential process: the interplay
bial pathogens to dietary fat, lead in paint, and hydrocarbons in between genetic predispositions and contextual exposures in
cigarette smoke, but the locus of pathogenesis remained square- the causal events leading to disorders of health and devel-
ly assignable to environmental agents. Within the realm of opment.
psychiatric medicine, causal theories similarly held that severe This narrative—of alternating, wholesale genetic and en-
developmental disorders such as schizophrenia and autism were vironmental accounts for disease, followed by a new synthesis
the consequences of social environmental causes: "cold revealing the conjoint, cooperative influences of both genes and
mothering" and "disturbed family relationships"—attributions environments—has been instantiated, for this author, within
that unfairly and mistakenly assigned culpability to the already two "stories" that have together defined a career in pediatric
distressed parents of children with such disorders. research. These two stories are those describing:
With the advent of the Human Genome Project and mole- 1. How neurogenomic differences in susceptibility to
cular level understanding of heritability and genomic processes environmental adversities amplify and diminish risk
in disease, however, a sea change occurred in theories of cause, for disorders of mental and physical health; and
and the health sciences entered a period of biological or even 2. How early exposures to social stratification and sub-
genetic determinism. As the promised sequencing of the entire ordination may have lifelong influences on human
human genome became a nearing reality, sanguine predictions morbidities.
heralded the discoveries, surely not far away, of the genes for
heart disease, diabetes, schizophrenia and depression. The focus Each of these accounts will be examined, in turn.
of thinking about pathogenesis in disorders of human health
moved, abruptly and convincingly (and ultimately disappoint Story 1: Neurogenomic differences in environmental sus-
ingly), from external causes to internal, heritable etiologies. ceptibility. The single most well replicated finding in all of
Theories of causation switched from a focus on "nurture" to an child health services research is that portrayed in Figure 1.
even stronger, alternate focus on "nature." Within any given childhood population, 15-20 percent of
children—approximately one in five—will sustain over half
of the biomedical and psychiatric (and likely dental) morbidities
Dr. Boyce is a professor of Pediatrics and Psychiatry and heads the Division of and will be responsible for the majority of health/dental care
Developmental-Behavioral Pediatrics, University of California, San Francisco, utilization within the population at large. Thus, the distribu-
Calif., USA. tion of disease within childhood populations is highly non-
Correspond with Dr. Boyce at [Link]@[Link] random in character, with a small subset of children bearing

102 EFFECTS OF EARLY CHILDHOOD ADVERSITY AND TOXIC STRESS

 PEDIATRIC DENTISTRY V 36 / NO 2 MAR / APR 14

the SES spectrum, not just in conditions of poverty, with the

children occupying each step of the scale having slightly worse
health than those in step just above, even at the high SES end
where the differences are between the most affluent families
and those only slightly less affluent. We easily recognize and
acknowledge the many ways in which true poverty influences
health—through exposures to toxins, inadequate diets, poor
medical care, and experiences of adversity, to name just a few.
The character of the linear SES-health association implies,
however, a deeper, less well understood causal process, in-
volving differences in life experience at every level of the social
class spectrum.
Further, childhood SES is associated not only with con-
current disease burden but with longitudinal, adult morbid-
Figure 1. The nonrandom distribution of childhood morbidities. ities, as well. The work of a variety of investigators3,4,5 shows
how growing up in disadvantaged communities and homes
leaves a biological "residue" of chronic physiological changes
most of the burdens of disease, disorder and illness. The public and susceptibilities, resulting in higher rates of adult morbidity,
health implications of this observation are clear: if we could chronic disability and premature mortality.
find the predisposing factors responsible for this highly uneven Although the mediators of these linkages between child-
distribution of childhood morbidity, we could potentially ad- hood social status and both child and adult health are incom-
dress over half the illnesses sustained by the population at large. pletely understood, such associations are almost certainly
Although the forces and factors that create this non- driven, in part, by differences in parenting and exposures to
random distribution of illness are multiple and complex, a single, adversity and stress. The well known, ethnographic work of
most salient predictor stands out: the socioeconomic status of Hart and Risley6 in 42 welfare, working class and professional
a child's family and community. Socioeconomic status (SES) is families, for example, showed that, by kindergarten, children
a broad, umbrella-like construct—usually indexed, for children, from welfare families had heard 32 million fewer words than
as a combination of family income, education, and parental those from professional class families. Both the volume and
job prestige —that constitutes the single most powerful epide- complexity of parent-to-child communication was vastly dif-
miologic predictor of health and disease we know. Indeed, ferent in households of differing SES, and these differential
SES is so uniquely and potently predictive of disease burden exposures were associated with disparate levels of school readi-
that we only provisionally believe any other association that ness and academic performance, as early as five years of age.
has not first controlled, in some manner, for the potentially We also know that family SES is associated with drama-
confounding influence of SES. Despite this widely recog- tically different rates of children's exposures to a wide variety
nized, predictive strength of SES, it is only within the past two of adversities and stressors. The work of Evans and collea-
decades of biomedical research that it has been studied as an gues7,8 (shown in Figure 3), for example, documents the sys-
important causal factor in its own right. tematic differences in levels of physical housing problems,
Among the most intriguing aspects of the SES-health as- ambient noise, family chaos, and witnessing of violence, which
sociation, shown, for example, in Figure 2, is its graded linearity are found among children from impoverished versus middle
and its implications for both concurrent, childhood health class families and communities. Further, the many, more ad-
status, as well as for future, adult health1,2. The graded, linear verse childhood events sustained by children lower in the
association suggests that the effects of SES are exerted across SES continuum are now known to predict striking differences

Figure 2. Associations between family SES and common pediatric Figure 3. The socioeconomic partitioning of stressors and adversities
disorders (abstracted from the work of Chen et al.38) (abstracted from thework of Evans et al.7)

EFFECTS OF EARLY CHILDHOOD ADVERSITY AND TOXIC STRESS 103

PEDIATRIC DENTISTRY V 36 / NO 2 MAR / APR 14

in the rates of virtually all adult disease states linked to earlier

mortality 9,10. Together, these accumulating data suggest the
existence of pervasive, profound differences in the early life
experiences and exposures of children reared in lower versus
higher SES circumstances.
Despite the power and pervasiveness of these SES dis-
parities in exposures and health, there is also remarkable varia-
bility in the consequences of such exposures among individual
children. Figure 4, for example, shows the relations between
SES and reading and literacy levels in a large population of
Canadian children11. Although the data show the previously
noted, strong, linear association between SES and literacy,
they also reveal a striking degree of individual variability. There
are, for example, many children from high SES families with
unusually low reading skills and likewise many low SES chil-
dren with anomalously high reading levels. The same varia-
bility can be found in the relations between childhood adversity
exposure and health outcomes: striking differences in the health
consequences of such exposures among individual children.
A number of years ago, my UCSF colleague Abbey Alkon
and I began considering whether such variability in the link-
ages between stress and child health was simply "noise" caused
by imperfect measures or was instead the "music" to which we
should be attending to garner a more complete and nuanced
understanding of the SES-health association. If such associa-
tions were intrinsically powerful but highly variable, perhaps
individual variation in children's neurobiological responses to
stressors could be a salient additional factor that could further
illuminate the connections among SES, stress and health. We
began studying diverse groups of young children, usually
between the ages of three and eight years, in laboratory set-
tings, where they were asked to complete a series of ecologically Figure 4. SES differences in literacy and reading levels in Canadian
valid, mildly challenging stressors, such as an interview with a children11.
previously unknown examiner, completing difficult cognitive
tasks like digit span recitation, or watching emotion-evoking
video clips. Figure 5 is an illustration, from a single study 12 , of
During these highly scripted and controlled laboratory what we have found in now multiple studies, over many years
stressors, we measured peripheral indicators of the activity with- of time, and in diverse samples of children. When the
in the brain's two principal stress response circuits: the cortico- samples of children have been divided into those with
tropin releasing hormone (CRH) system, which activates the low and high neurobiological reactivity (usually at about
hypothalamic-pituitary-adrenocortical (HPA) axis; and the locus the 80-85th percentile), we find that the vast majority of
coeruleus-norepinephrine (LC-NE) system that initiates the children, low or moderate in reactivity in both the CRH
fight-or-flight responses of the autonomic nervous system and LC-NE systems, show little or no relation between
(ANS). Responsivity in the former system was
measured with salivary levels of cortisol, the hu-
man adrenocortical hormone influencing a broad
variety of physiologic processes, including
blood pressure control, glucose homeostasis, and
immune system activity. Reactivity in the LC-
NE system was indexed using impedance car-
diography, with separate measures of the ANS
sympathetic and parasympathetic branch
responses. We found broad individual differ-
ences in children's stress-associated reactivity
in both systems, with some children showing
almost no response, some with quite vigorous
physiological changes, and many in mid-point
between the extremes. We then employed these
measures of stress reactivity within epidemio-
logic study designs, examining the role of
laboratory-based reactivity in the relation
between naturally occurring social conditions in
the child's real world and levels of biomedical
and psychiatric morbidities. Figure 5. Biological sensitivity to social context (from Obradović et al.12)

104 EFFECTS OF EARLY CHILDHOOD ADVERSITY AND TOXIC STRESS

 PEDIATRIC DENTISTRY V 36 / NO 2 MAR / APR 14

typical levels of social environmental stressors (in this case, childhood populations results in a greater variability in de-
marital conflict) and health outcomes (here, externalizing be- velopmental and health outcomes than would be seen in their
havior problems). By contrast, however, the subgroup of highly absence, especially within disadvantaged environments of
reactive children, usually constituting 15-20 percent of our adversity and stress.
samples, shows a remarkable pattern of outcomes: they have
the worst health outcomes in conditions of high adversity but Story 2: The health effects of social stratification and sub-
the best outcomes within supportive, low adversity contexts. ordination. Environments of poverty and disadvantage can
In this example, the high reactivity subset of children showed thus influence liability to disease and appear to upregulate
the highest rates of behavior problems when exposed to neurobiological reactivity within aversive social contexts, as well.
high levels of marital conflict at home but the lowest rates One illustrative example is children's dental caries, among the
when marital conflict was low or absent—rates even lower most common, visible, and socioeconomically stratified chronic
than those of their low or moderate reactivity counterparts. diseases of childhood, well known to the readership of this
Our interpretation of this recurring pattern of results is journal. It is a disease with striking SES and racial disparities
that the smaller, highly reactive subgroup is revealing a neuro- in incidence, with a variety of established etiologic factors
biological susceptibility to social contexts, with exaggerated (including diet, lead, tobacco smoke, and lack of water fluori-
responses to both highly negative and highly positive environ- dation), and with a known causal link to the presence of the
ments. Thus, the differentially susceptible children show either oral cariogenic bacteria, Streptococcus mutans and Lactobacillus
the most or least healthy outcomes, depending critically upon acidophilus. The conventional wisdom regarding the dispro-
the character of their social worlds. In a kind of shorthand, we portionate rates of caries found in children from impoverished
have referred to the low reactive children, with approximately communities, however, is that the principal origin of these
equivalent health and morbidity under a broad array of con- socioeconomic disparities is low SES parents' inattention to
ditions, with the Swedish idiomatic expression, maskrosbarn or the proper routines of oral hygiene. In a 2010 paper in Social
"dandelion child" (by which the Swedes mean a child that can Science and Medicine, we tested an alternate hypothesis:
grow strongly and well under most any circumstance, like the i.e., that the inordinate psychosocial stressors sustained by low
dandelion that thrives in a diversity of settings). In contrast to SES families and the neuroendocrine changes that accompany
these, we have designated their highly reactive, highly suscep- such stress could place children from disadvantaged commu-
tible counterparts as orkidebarn ("orchid child"), who like the nities in jeopardy for more prevalent and severe disorders of
orchid flower can reveal exceptional beauty and hardiness in oral health.21
carefully maintained and protected environments but may In two studies of kindergarten children from varying
falter and fail when such conditions are unavailable. socioeconomic backgrounds in the San Francisco Bay Area of
This phenomenon of individual differences in neurobio- California (ns = 94 and 38), we performed detailed dental
logical susceptibility to social contexts, first noted in 199513, examinations to count decayed, missing or filled dental surfaces
was explicated and tested within an evolutionary biology and microtomography to assess the thickness and density of
framework in 2005 14,15, has been confirmed in the parallel microanatomic compartments in exfoliated, deciduous teeth.
results of other investigators16,17,18,19, and has now been docu- Cross-sectional, multivariate associations were examined be-
mented at multiple levels of biological complexity and scale, tween these measures and SES-related risk factors, including
ranging from behavior and temperament to allelic and epi- household education, financial stressors, basal and reactive
genetic variation 20. In 2011, a special issue of the journal salivary cortisol secretion, and the number of oral cariogenic
Development and Psychopathology summarized emerging bacteria. Results of these multivariate analyses are illustrated
evidence and assembled a series of empirical papers revealing in Figure 6. Low SES, higher basal salivary cortisol secretion,
differential susceptibility to both naturally occurring positive and larger numbers of cariogenic bacteria were each signifi-
and negative social environments, as well as psychosocial/ cantly and independently associated with caries, and higher
developmental interventions. Further, heightened sensitivity to salivary cortisol reactivity was associated with thinner, softer
the social world appears more prevalent in both extremely pro- enamel surfaces in exfoliated teeth. The highest rates of dental
tected and highly adverse social context. In sum, there now pathology were found among children with the combination
exists strong evidence for a heightened neurobiological of elevated salivary cortisol expression and high counts of
sensitivity to social contexts—both positive and negative— cariogenic bacteria. The socioeconomic partitioning of child-
within a small subgroup of children. Representation of such hood dental caries may thus involve social and psychobiological
differential sensitivity has been established at the levels of pathways through which lower SES is associated with higher
behavior, autonomic/adrenocortical reactivity, brain circuitry numbers of cariogenic bacteria and higher levels of stress-
activation, allelic variation, and cellular aging. The presence of associated salivary cortisol. This convergence of psychosocial,
these more contextually sensitive individuals within broader infectious, and stress-related biological processes appears to

Figure 6. Interactive influences of family stressors, stress reactivity and cariogenic bacteria on dental microanatomy
and care (from Boyce et al.39)

EFFECTS OF EARLY CHILDHOOD ADVERSITY AND TOXIC STRESS 105

PEDIATRIC DENTISTRY V 36 / NO 2 MAR / APR 14

Convergence: The biological embedding of early social

adversity. These two "stories"—of how neurogenomic dif-
ferences in susceptibility to social contexts can jeopardize
health and development and how a social stratification of
health occurs at the levels of both societies and classrooms—
converge in an new, emerging story of how early, adverse
childhood exposures become biologically embedded within
the epigenetic and cellular processes guiding neurode-
velopment.23 We know from a now growing literature that
the diversely challenging exposures that accompany rearing
in impoverished or disadvantaged settings can result in both
structural and functional variations in brain development,
leading to premature foreclosure on academic achievement,
risk for developmental psychopathology, and jeopardy in
the attainment of individual potential24,25,26. Areas of the
prefrontal cortex, as well as the limbic-prefrontal circuitry
involved in emotion regulation appear particularly affected
by the exposures inherent within impoverished early envi-
ronments, producing deficits in executive functions, atten-
tional abilities, and behavioral regulation.27,28 Increasingly,
Figure 7. Interactive influences of social dominance position and teachers’ these systematic variances in brain development have been
learner-centered practices (LCPs) on teacher-reported child depression related to socially partitioned differences in experiences of
(from Boyce et al.22) early stress and adversity.
Observations such as these lead logically to questions
of how differences in early experiences, along with differ-
be implicated in the production of greater cariogenic bacterial ences in inherited DNA sequences within neuroregulatory
growth and in the conferral of an increased physical vulnera- genes, lead to systematic variation in brain circuitry, cognitive
bility of the developing dentition. ability, and regulatory functions. How does allelic variation co-
While low SES and its concomitants clearly exert power- operate with social contextual variation to produce such sub-
ful influences on health in childhood, there appear to be stantive differentiation in phenotypic, developmental outcomes?
deeper, even more fundamental risks associated with the gen- What are the mechanisms by which nature and nurture
eral experiences of social subordination, per se. We know that converge? Answers to such questions have begun to unfold in
animal social groups are arrayed in ordered, linearly transitive the field of social epigenetics, which examines how early social
dominance hierarchies, from round worms and fruit flies to conditions and gene variants together produce chromatin mo-
fish and nonhuman primates. Even within our own species, difications controlling the expression of key developmental
children as young as two years of age assemble visible pecking genes.29
orders within weeks of entering new social groups, such as pre- "Epigenetics" refers to heritable changes in gene expression
school classrooms or neighborhood play groups. Guided by that are not encoded in the DNA sequence itself. Although
animal models of hierarchical organization and the known the first insights into epigenetic control of the human genome
health correlates of subordination, we conducted a prospective were derived from studies of cancer biology, pioneering work
study examining the partitioning of children's adaptive be- in rodent models began to show how naturally occurring vari-
havioral development by their positions within kindergarten ation in early social environments could also effect enduring
classroom hierarchies.22 A sample of 338 five-year-old children differences in adult social behavior and psychobiologic suscep-
was recruited from Berkeley, California public school class- tibility to adversity and challenge. Work by Meaney, Szyf and
rooms. An observational measure of social rank, parent-reported colleagues, for example, 30,31 showed how epigenetic changes
family SES, and child-reported classroom climate were used in expression of the gene encoding the glucocorticoid recep-
in estimating multilevel, random-effects models of children's tor (GR) are mediated, during an early sensitive period, by
adaptive behavior at the end of the kindergarten year. Children typical differences in maternal behavior, specifically the inten-
occupying subordinate positions had significantly more teacher- sity and frequency of licking, grooming and arched-back
reported depressive symptoms than their dominant peers. nursing (LG-ABN) of the pups. Naturally or induced low levels
Further, as shown in Figure 7, interaction terms revealed that of maternal LG-ABN behavior were associated, among off-
teachers' child-centered pedagogical practices (a measure of spring, with down-regulation of GR expression, resultant
classroom climate) diminished the adverse influences of social increases in HPA reactivity, and temperamental characteristics
subordination. These results suggest that, even within early suggesting more fearful and anxious behavior. High LG-ABN,
childhood groups, social stratification is associated with a parti- by contrast, results in down-regulation of HPA reactivity and
tioning of adaptive behavioral outcomes and that the character bolder, less fearful behavior, characteristics that are subse-
of larger societal and school structures in which such groups quently passed on to a subsequent generation through epige-
are nested can moderate rank-behavior associations. netic or behavioral means.
In sum, there appears to be a conserved predisposition, Taking these groundbreaking experimental studies as a
across species, to hierarchical social organization. Experiences foundation, a new generation of human research has begun to
of social subordination are associated with diminished health illuminate how early social experience can similarly result in
and development, irrespective of the scale or complexity of epigenetic modifications with influences on brain development
the social context, but the consequences of inequality can and health outcomes32,33,34,35,36. One such study 37, using data
be tempered by the social climate and culture shaped by from the longitudinal Wisconsin Study of Families and Work,
those who teach and lead. examined differences in DNA methylation in 15-year-old

106 EFFECTS OF EARLY CHILDHOOD ADVERSITY AND TOXIC STRESS

 PEDIATRIC DENTISTRY V 36 / NO 2 MAR / APR 14

adolescents in relation to parent reports of adversity during 6. Hart T and Risley TR. Meaningful Differences in the
the adolescents' infancy and preschool periods. Microarray Everyday Experience of Young American Children. 1995,
technology applied to 28,000 methylation sites within DNA Baltimore, MD: Paul H. Brookes.
derived from buccal epithelial cells showed differential methy- 7. Evans GW and Schamberg MA. Childhood poverty, chro-
lation among adolescents whose parents had reported high levels nic stress, and adult working memory. Proc Natl Acad
of stress during their children's early lives. Maternal stressors Sci USA 2009;106:6545-9.
in infancy and paternal stressors in the preschool years were 8. Evans GW et al. How poverty gets under the skin: A life-
most strongly predictive of differential methylation, and the course perspective. In: Maholmes V. and King RB, eds. The
pattern of such epigenetic marks varied by children's gender. Oxford Handbook of Poverty and Child Development.
This first report of prospective associations between adversi- Oxford: Oxford University Press; 2012:13-36.
ties in early childhood and the epigenetic conformation of 9. Brown DW, Anda RF, Tiemeier H, et al. Adverse childhood
adolescents' genomic DNA has anticipated a new generation experiences and the risk of premature mortality. Am J Prev
of studies linking developmental changes in epigenetic marks Med 2009;37(5):389-96.
with earlier experiences of adversity and stress. 10. Felitti VJ, Anda RF, Nordenberg D, Williamson DF et
Taken together, these emerging themes in the early psy- al. Relationship of childhood abuse and household dys-
chobiology of adversity and inequality trace an increasingly function to many of the leading causes of death in adults.
coherent picture of how social environmental conditions in The Adverse Childhood Experiences (ACE) Study. Am
childhood and genetically derived differences in vulnerability to J Prev Med 1998;14(4):245-58.
such conditions operate together, at the level of epigenetic 11. Willms JD. Variation in Literacy Skills among Canadian
change, to bias development toward outcomes of dispropor- Provinces: Findings from the OECD PISA. 2004, Culture,
tionate morbidity, suboptimal achievement, and premature Tourism and the Centre for Education Statistics Division:
mortality. It is a picture that requires and deserves the attention Ottawa, Canada.
of physicians, dentists and other professionals charged with 12. Obradovic J, Bush NR, and Boyce WT. The interactive
the protection and care of children's health. The pervasive effect of marital conflict and stress reactivity on external-
insufficiencies, perils, and inequalities that characterize life in izing and internalizing symptoms: the role of laboratory
disadvantaged early settings become, in the end, biologically stressors. Dev Psychopathol 2011;23(1):101-14.
embedded within epigenetic and neural processes that simply 13. Boyce WT, Chesney M, Alkon A, Tschann JM et al. Psy-
disallow biomedical or dental health, success or well being, chobiologic reactivity to stress and childhood respiratory
especially among those whose genomes encode special sensi- illnesses: Results of two prospective studies. Psychosom
tivities to social environmental influence. The result is an Med 1995;57:411-22.
enduring "cycle of poverty," first recognized decades ago as a 14. Boyce WT and Ellis BJ. Biological sensitivity to context:
kind of cross-generational, sociological entrenchment and now [Link] evolutionary-developmental theory of the origins
increasingly revealed as a deep and persistent biology of mis- and functions of stress reactivity. Development & Psy-
fortune. chopathology 2005;17(2):271-301.
15. Ellis BJ, Essex MJ and Boyce WT. Biological sensitivity
Acknowledgments to context: [Link] explorations of an evolutionary-
The research upon which this report is based was supported developmental hypothesis. Development & Psychopath-
by the National Institute of Mental Health (grant awards R01 ology 2005;17(2):303-28.
MH62320 and R01 MH62320- S1), the MacArthur Foun- 16. Belsky J. Variation in susceptibility to environmental in-
dation Research Network on Psychopathology and Develop- fluence: An evolutionary argument. Psych Inquiry 1997;8
ment, the British Columbia Leadership Chair in Child (3):182-6.
Development at the University of British Columbia, and the 17. Belsky J. Differential susceptibility to rearing influence:
Canadian Institute for Advanced Research. An evolutionary hypothesis and some evidence, in Origins
of the Social Mind: Evolutionary Psychology and Child
References Development. Ellis BJ and Bjorklund DF, Editors.
1. Keating DP and Hertzman C. Developmental Health and Guilford: New York; 2005:139-63.
the Wealth of Nations: Social, Biological, and Educational 18. Van Ijzendoorn MH and Bakermans-Kranenburg MJ.
Dynamics. 1999, New York: The Guilford Press. DRD4 7-repeat polymorphism moderates the association
2. Hertzman C and Boyce WT. How experience gets under between maternal unresolved loss or trauma and infant
the skin to create gradients in developmental health. disorganization. Attach Hum Dev 2006;8(4):291-307.
Annu Rev Public Health 2010;31:329-47, 3p following 19. Manuck SB, Craig AE, Flory JD, Halder I, and Ferrell
347. RE. Reported early family environment covaries with me-
3. Kittleson MM, Meoni LA, Wang NY, Chu AY and narcheal age as a function of polymorphic variation in
Ford DE. Association of childhood socioeconomic status estrogen receptor-alpha. Dev Psychopathol 2011;23(1):
with subsequent coronary heart disease in physicians. 69-83.
Arch Intern Med 2006;166(21):2356-61. 20. Ellis BJ, Boyce WT, Belsky J, et al. Differential suscepti-
4. Miller GE, Chen E, Fok AK, Walker H, Lim A, Nicholls bility to the environment: an evolutionary-- neuro
EF, Cole S, Kobor MS. Low early-life social class leaves a developmental theory. Dev Psychopathol 2011;23(1):
biological residue manifested by decreased glucocorticoid 7-28.
and increased proinflammatory signaling. Proc Natl Acad 21. Boyce WT, Den Besten PK, Stamperdahl J, et al. Social
Sci U S A 2009;106(34):14716-21. inequalities in childhood dental caries: the convergent
5. Cohen S, Cohen S, Janicki-Deverts D, Chen E, Matthews roles of stress, bacteria and disadvantage. Soc Sci Med
KA. Childhood socioeconomic status and adult health. 2010;71(9):1644-52.
Ann N Y Acad Sci 2010;1186:37-55.

EFFECTS OF EARLY CHILDHOOD ADVERSITY AND TOXIC STRESS 107

PEDIATRIC DENTISTRY V 36 / NO 2 MAR / APR 14

22. Boyce WT, Obradović J, Bush NR, Stamperdahl J et al. 30. Szyf M, Weaver IC, Champagne FA, et al. Maternal pro-
Social stratification, classroom 'climate' and the behavioral gramming of steroid receptor expression and phenotype
adaptation of kindergarten children. Proc Natl Acad Sci through DNA methylation in the rat. Front Neuroendo-
USA 2012;109(Suppl 2):17168-73. crinol 2005;26(3-4):139-62.
23. Boyce WT, Sokolowski MB, and Robinson GE. Toward a 31. Gluckman PD, Hanson MA and Beedle AS. Non-genomic
new biology of social adversity. Proc Natl Acad Sci USA transgenerational inheritance of disease risk. Bioessays
2012;109(Suppl 2):17143-8. 2007;29(2):145-54.
24. Stevens C, Lauinger B, and Neville H. Differences in the 32. Rutter M, Moffitt TE, and Caspi A. Gene-environment
neural mechanisms of selective attention in children from interplay and psychopathology: multiple varieties but real
different socioeconomic backgrounds: an event-related effects. J Child Psychol Psychiatry 2006;47(3-4):226-61.
brain potential study. Dev Sci 2009;12(4):634-46. 33. Szyf M, McGowan P, and Meaney MJ. The social environ-
25. McEwen BS, and Gianaros PJ. Central role of the brain ment and the epigenome. Environ Mol Mutagen 2008;
in stress and adaptation: links to socioeconomic status, 49(1):46-60.
health, and disease. Ann N Y Acad Sci 2010;1186: 34. Meaney MJ. Epigenetics and the biological definition of
190-222. gene x environment interactions. Child Dev 2010;81
26. Kishiyama MM, Boyce WT, Jimenez AM, Perry LM (1): 41-79.
and Knighy RT. Socioeconomic disparities affect pre- 35. Petronis A. Epigenetics as a unifying principle in the
frontal function in children. J Cogn Neurosci 2009;21 aetiology of complex traits and diseases. Nature 2010;
(6):1106-15. 465(7299):721-7.
27. Noble KG, McCandliss BD, and Farah MJ. Socioecono- 36. Essex MJ, Boyce WT, Hertzman C, Lam LL, et al. Epige-
mic gradients predict individual differences in neurocog- netic vestiges of early developmental adversity: childhood
nitive abilities. Dev Sci 2007;10(4):464-80. stress exposure and DNA methylation in adolescence.
28. Sweatt JD, Meaney M, Nestler E, and Akbarian S. Epige- Child Dev 2013;84(1):58-75.
netic Regulation in the Nervous System: Basic Mecha- 37. Chen E, Matthews KA, and Boyce WT. Socioeconomic
nisms and Clinical Impact. London: Elsevier; 2013. differences in children's health: how and why do these
29. Meaney MJ, Szyf M, and Seckl JR. Epigenetic mechanisms relationships change with age? Psychol Bull 2002;128(2):
of perinatal programming of hypothalamic-pituitary- 295-329.
adrenal function and health. Trends Mol Med 2007;13
(7):269-77.

108 EFFECTS OF EARLY CHILDHOOD ADVERSITY AND TOXIC STRESS