Table 1 Childhood adversity and candidate gene methylation in samples of children.

Cicchetti & Handley,²⁶

N = 534; 53% exposed to CM

Characterization of sample: Low-income

School age (M = 9.4 years)

49% female

61.2% black, 9.9% white, 8.2% biracial or other; 20.6%

Latino

NR3C1

Saliva

Physical, sexual, emotional abuse, supervisory and physical neglect via child protection case files coded with the MCS

Maltreated children demonstrated hypermethylation compared to non-maltreated children; CM during infancy and/or toddlerhood associated with greater methylation than children with no CM; no difference between children who experienced CM in the preschool period or later and those with no CM; greater chronicity of CM associated with greater methylation; exposure to more CM subtypes associated with greater methylation

Hecker et al.,²⁹

N = 60; 58% with high exposure to CM; 42% with low exposure to CM

Characterization of sample: Participants resided in Tanzania

9–15 years

(M = 11.3 years in high-exposure group; M = 11.8 years in low-exposure group)

60% female in high-exposure group; 56% female in low-exposure group

100% Tanzanian

NR3C1

POMC

CRH

AVP

Saliva; blood lymphocytes

Physical and emotional abuse measured via the Maltreatment and Abuse Chronology of Exposure – Pediatric Version, structured interview

Children with high CM exposure had higher methylation of POMC and CRH, at one CpG site in each gene that survived multiple comparisons correction; differential methylation was also present for several sites in NR3C1, POMC, and AVP but these did not survive multiple comparisons correction

Parade et al.,¹²

N = 171; 42% exposed to CM

Characterization of sample: Low-income; enrolled in Kids Markers Study;

Follow-up of Tyrka, Parade et al. 2015, to examine association of methylation with symptoms/behavior

3–5 years

(M = 4.2 years)

52% female

23% white non-Hispanic, 48% Hispanic, 15% black, 14% other

NR3C1

Saliva

Physical, sexual, emotional abuse, supervisory and physical neglect assessed via child protective services records; death or separation of a caregiver, frequent change of residence or homelessness, inadequate food or clothing, witnessing violence via parent contextual stress interview and Diagnostic Infant and Preschool Assessment

As expected based on prior work with the Kids Markers sample, adversity composite was associated with greater methylation, and CM, past month stress, and lifetime stress were each individually associated with greater methylation whereas traumatic life events were not associated with methylation; new to this analysis, methylation mediated the effect of adversity exposure on internalizing, but not externalizing symptoms

Parent et al.,¹³

N = 260; 53% exposed to CM

Characterization of sample: Low-income; enrolled in Kids Markers Study;

Six-month follow-up of baseline results reported in Tyrka, Parade et al., 2015

3–5 years

(M = 4.2 years)

52% female

27.7% white non-Hispanic, 45.6% Hispanic, 16.3% black, 21.9% biracial, 2.7% other races

NR3C1

Saliva

Physical, sexual, emotional abuse, supervisory and physical neglect assessed via child protective services records; death or separation of a caregiver, frequent change of residence or homelessness, inadequate food or clothing, witnessing violence via parent contextual stress interview and Diagnostic Infant and Preschool Assessment

As expected based on prior work with the baseline data in the Kids Markers sample, CM was associated with higher levels of methylation within 6 months of CM; CM was negatively associated with change in methylation one year after CM, at which point maltreated children demonstrated lower levels of methylation relative to non-maltreated children

Radtke et al.,¹⁴²

N = 46

Characterization of sample: Participants resided in Germany

11–21 years

(Median = 15 years)

61% female

Not reported

NR3C1

Blood lymphocytes

Physical, emotional, and sexual abuse; witnessed violence toward parents, witnessed violence toward siblings, peer physical violence, physical and emotional neglect via a German version of the pediatric Maltreatment and Abuse Chronology of Exposure interview

CM not associated with average methylation; CM was positively associated with methylation at one of 41 examined CpG sites, after adjusting for multiple comparisons using a false discovery rate

Romens et al.,¹⁴³

N = 56; 32% with substantiated physical abuse

11–14 years

(M = 12.1 years)

46% female

66% white, non-Hispanic, 30% black, 4% white, Hispanic

NR3C1

Whole blood

Physical abuse via child protective services records

Children with CM had higher levels of methylation at 3 out of 13 CpG sites than children without CM (using a mixed ANOVA method to first test overall interaction of CM and CpG sites [within subjects] followed by planned examination of individual CpG sites when overall interaction was significant)

Tyrka, Parade, et al.,²⁵

N = 184; 40% exposed to CM

Characterization of sample: Low-income; enrolled in Kids Markers Study

3–5 years

(M = 4.2 years)

51% female

22% white non-Hispanic, 47% Hispanic, 16% black, 15% other races

NR3C1

Saliva

Physical, sexual, emotional abuse, supervisory and physical neglect assessed via child protective services records; death or separation of a caregiver, frequent change of residence or homelessness, inadequate food or clothing, witnessing violence via parent contextual stress interview and Diagnostic Infant and Preschool Assessment

An adversity composite, including CM, contextual stress, and traumatic life events, was positively associated with mean methylation; CM and contextual stress were each individually positively associated with methylation

van der Knaap et al.,¹⁴⁴

N = 468

Characterization of sample: Enrolled in the Tracking Adolescents’ Individual Lives Survey

14–18 years

(M = 16.1 years)

50% female

100% Dutch

NR3C1

Whole blood

Perinatal stress assessed using parent interview and record review; traumatic youth experiences assessed using adolescent retrospective self-report; stressful life events assessed using parent interview (childhood events), child self-report (early adolescence events), Event History calendar (middle adolescence events)

Stressful life events and traumatic experiences associated with greater NR3C1 methylation; Stressful life events in adolescence associated with methylation independent of stressful life events in childhood; Perinatal stress not associated with methylation

Non et al.,²⁷

N = 117; n = 82 institutionalized children, n = 35 non-institutionalized children

Characterization of sample: Enrolled in Bucharest Early Intervention Project

M = 12.5 years

49% female

63% Romanian, 27% Rroma, 9% other

FKBP5;

SLC6A4

Buccal cells

Percentage of time spent in institutional care

Greater time spent in institutional care was associated with lower methylation at one of two CpG sites for FKBP5 (effect survived multiple comparison correction for analysis within institutionalized group but not in whole sample);

Greater time spent in institutional care was associated with lower methylation at two of six CpG sites for SLC6A4, significant after multiple comparisons correction;

FKBP5 genotype by adversity interaction was not significant

Parade, Parent, et al.,¹⁰⁷

N = 231; 53% exposed to CM

Characterization of sample: Low-income; enrolled in Kids Markers Study;

Six-month follow-up of baseline results reported in Tyrka, Ridout, et al., 2015

3–5 years

(M = 4.3 years)

52% female

40% white, 16% black, 21% biracial, 23% other races

FKBP5

Saliva

Physical, sexual, emotional abuse, supervisory and physical neglect assessed via child protective services records; death or separation of a caregiver, frequent change of residence or homelessness, inadequate food or clothing, witnessing violence via parent contextual stress interview and Diagnostic Infant and Preschool Assessment

Maltreated children had lower levels of baseline methylation compared to non-maltreated children, but maltreatment did not predict change in methylation over time; when contextual stress was high, maltreated children had consistently low methylation over time whereas non-maltreated children demonstrated a decline in methylation from baseline to follow-up; FKBP5 genotype did not moderate associations of CM and methylation

Tyrka, Ridout, et al.,²⁵

N = 174; 40% exposed to CM

Characterization of sample: Low-income; enrolled in Kids Markers Study

3–5 years (M = 4.2 years)

52% female

22% white non-Hispanic, 47% Hispanic, 17% black, 15% other races

FKBP5

Saliva

Physical, sexual, emotional abuse, supervisory and physical neglect assessed via child protective services records; death or separation of a caregiver, frequent change of residence or homelessness, inadequate food or clothing, witnessing violence via parent contextual stress interview and Diagnostic Infant and Preschool Assessment

Children exposed to CM had lower levels of methylation at two out of two examined CpG sites; there was a trend-level association of lifetime contextual stress and methylation at one site; an adversity composite, including CM, contextual stress, and traumatic life events was negatively associated with methylation at one site; FKBP5 genotype did not moderate links between adversity exposure and methylation

Timothy et al.,³¹

N = 100; n = 50 children of alcoholics; n = 50 controls, significantly higher adversity in COA group compared to controls Characterization of sample: Resided in urban India

8–16 years

(M = 11.3 years in COA group; M = 11.2 years in control group)

0% female

100% Indian

SLC6A4

Saliva

Physical, emotional and sexual abuse, neglect; peer and collective violence; stress of family and friends; pregnancy; problems with alcohol/drugs measured using the WHO Adverse Childhood Experiences Scale

Greater adversity was associated with hypermethylation, particularly in the COA group

van der Knaap et al.,³²

N = 939

Characterization of sample: Enrolled in the Tracking Adolescents’ Individual Lives Survey

14–18 years

(M = 16.2 years)

51% female

100% Dutch

SLC6A4

Whole blood

Perinatal stress assessed using parent interview and record review; traumatic youth experiences assessed using adolescent retrospective self-report; stressful life events assessed using parent interview (childhood events), child self-report (early adolescence events), Event History calendar (middle adolescence events)

More stressful life events associated with higher methylation; effect of stressful life events in adolescence stronger than stressful life events in childhood; 5HTTLPR genotype moderated effect of stressful life events such that this association was only observed in l-allele homozygotes; perinatal adversity and traumatic youth experiences not associated with methylation

Parade, Novick, et al.,³⁰

N = 228; 52% exposed to CM

Characterization of sample: Low-income; enrolled in Kids Markers Study

3–5 years

(M = 4.2 years)

53% female

39% white, 17% black, 22% biracial, 22% other races, 45% Hispanic

HTR2A

Saliva

Physical, sexual, emotional abuse, supervisory and physical neglect assessed via child protective services records; death or separation of a caregiver, frequent change of residence or homelessness, inadequate food or clothing, witnessing violence via parent contextual stress interview and Diagnostic Infant and Preschool Assessment

CM, traumatic life events, and contextual stress were not associated with methylation; genotype moderated contextual stress to predict methylation, such that contextual stress was positively associated with methylation among A homozygotes, negatively associated among G homozygotes, and not associated among heterozygotes

Barker et al.,³³

N = 785

Characterization of sample: Enrolled in ARIES study nested within ALSPAC

7 years

50% female

100% white

Inflammation-related epigenetic polygenic

risk scores (i-ePGS)

Whole blood

Life events (e.g., death in family, accident, and illness), contextual risks (e.g., poor housing conditions and financial problems), parental risks (e.g., parental psychopathology, criminal involvement, and substance use), interpersonal risks (e.g., intimate partner violence and family conflict), and direct victimization (e.g., child bullied by peers or physically hurt) via maternal reports

Postnatal adversity was associated with higher i-ePGS methylation at age 7, which was associated with internalizing symptoms from ages 7–15

Fujisawa et al.,³⁴

N = 85; 52% exposed to CM

Characterization of sample: Those with CM resided in a child welfare facility; controls were recruited from the community.

6–20 years

(M = 12.9 years)

35% female

100% Japanese

OXTR

Saliva

Physical, emotional, sexual abuse, and/or neglect, as determined by the local child welfare facilities

Children who were exposed to CM had higher methylation than children not exposed to CM; methylation was negatively correlated with gray matter volume in the left orbitofrontal cortex; children exposed to CM showed lower gray matter volume compared to the non-CM children