Prevalence and Symptomatic Sequelae of Self-Report Childhood Physical and Sexual Abuse in a General Population Sample of Men and Women

Abstract

This study examined the prevalence and psychological sequelae of childhood sexual and physical abuse in adults from the general population.
A national sampling service generated a geographically stratified, random sample of 1,442 subjects from the United States. Subjects were mailed a questionnaire that included the Traumatic Events Survey (TES) [Traumatic Events Survey, Unpublished Psychological Test, Harbor-UCLA Medical Center, Los Angeles] and the Trauma Symptom Inventory (TSI) [Trauma Symptom Inventory Professional Manual, Psychological Assessment Resources, Odessa, FL]. Of all potential subjects, 935 (64.8%) returned substantially completed surveys.
Sixty-six men and 152 women (14.2% and 32.3%, respectively) reported childhood experiences that satisfied criteria for sexual abuse, and 103 males and 92 females (22.2% and 19.5%, respectively) met criteria for physical abuse. Twenty-one percent of subjects with one type of abuse also had experienced the other type, and both types were associated with subsequent adult victimization. After controlling for demographics, adult history of interpersonal violence, and other child abuse, childhood sexual abuse was associated with all 10 scales of the TSI, and physical abuse was related to all TSI scales except those tapping sexual issues. Sexual abuse predicted more symptom variance than did physical abuse or adult interpersonal victimization. Various aspects of both physical and sexual abuse experiences were predictive of TSI scores. Abuser sex, however, both alone and in interaction with victim sex, was not associated with additional TSI symptomatology.
Childhood sexual and physical abuse is relatively common in the general population, and is associated with a wide variety of psychological symptoms. These relationships remain even after controlling for relevant background variables.

Full text

Child Abuse & Neglect 27 (2003) 1205–1222
Prevalence and psychological sequelae of self-reported
childhood physical and sexual abuse in a general
population sample of men and women
John Briere∗, Diana M. Elliott
Department of Psychiatry and the Behavioral Sciences, Keck School of Medicine,
University of Southern California, Los Angeles, CA, USA
Received 13 April 2001; received in revised form 22 February 2002; accepted 2 March 2002
Abstract
Objective: This study examined the prevalence and psychological sequelae of childhood sexual and
physical abuse in adults from the general population.
Method: A national sampling service generated a geographically stratified, random sample of 1,442
subjects from the United States. Subjects were mailed a questionnaire that included the Traumatic
Events Survey (TES) [Traumatic Events Survey, Unpublished Psychological Test, Harbor-UCLA Med-
ical Center, Los Angeles] and the Trauma Symptom Inventory (TSI) [Trauma Symptom Inventory
Professional Manual, Psychological Assessment Resources, Odessa, FL]. Of all potential subjects, 935
(64.8%) returned substantially completed surveys.
Results:Sixty-sixmenand152women(14.2%and32.3%,respectively)reportedchildhoodexperiences
thatsatisfiedcriteriaforsexualabuse,and103malesand92females (22.2% and 19.5%, respectively)met
criteria for physical abuse. Twenty-one percent of subjects with one type of abuse also had experienced
the other type, and both types were associated with subsequent adult victimization. After controlling for
demographics, adult history of interpersonal violence, and other child abuse, childhood sexual abuse
was associated with all 10 scales of the TSI, and physical abuse was related to all TSI scales except
those tapping sexual issues. Sexual abuse predicted more symptom variance than did physical abuse or
adult interpersonal victimization. Various aspects of both physical and sexual abuse experiences were
predictive of TSI scores. Abuser sex, however, both alone and in interaction with victim sex, was not
associated with additional TSI symptomatology.
∗Corresponding author address: USC Psychiatry, Psychological Trauma Program, 2020 Zonal Avenue, IRD
Building, Los Angeles, CA 90033, USA.
0145-2134/$ – see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.chiabu.2003.09.008

1206 J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222
Conclusions: Childhood sexual and physical abuse is relatively common in the general population, and
is associated with a wide variety of psychological symptoms. These relationships remain even after
controlling for relevant background variables.
© 2003 Elsevier Ltd. All rights reserved.
Keywords: Physical abuse; Sexual abuse; Symptoms; Prevalence
Introduction
Although the incidence and long-term effects of childhood sexual and physical abuse are
thought by most clinicians and researchers to be significant, some have suggested that the
methodology and samples used in such studies may obscure or confound the actual relation-
ship between child maltreatment and its potential effects. Among criticisms offered are (a)
the fact that most studies examine clinical samples (Rind, Tromovitch, & Bauserman, 1998)
or, alternatively, focus on university students (Wyatt & Peters, 1986), neither of which may
provide a clear picture of the rates and impacts of child abuse in the general population, (b)
the generic (i.e., not trauma-specific) measures used in many studies may underevaluate the
actual symptomatology arising from childhood trauma (Carlson, 1997; Elliott, 1994), (c) the
failure of many studies to examine both physical and sexual abuse as separate variables may
overestimate the effects of sexual abuse where physical abuse is also present (Briere, 1992),
and(d) backgroundfactors suchas demographicand socioeconomicvariables,family environ-
ment, and non-abuse-related traumas may confound what otherwise appear to be abuse-related
effects (Fromuth, 1986; Rind et al., 1998).
Some controversy has arisen regarding the last of these points, that is, the possible con-
founding of abuse effects with the potential impacts of other variables that correlate with
abuse. In this regard, there are significant statistical issues associated with controlling for
abuse-correlated variables when abuse is antecedent to such variables (Davis, 1985; Pedhazur,
1982) or when the abuse variable is, itself, logically inseparable from the controlled variable
(Briere & Elliott, 1993). For example, in the case of family environment, child abuse may
further disrupt an already dysfunctional family, and a dysfunctional family may be an im-
portant aspect of child abuse (especially intrafamilial sexual and physical abuse). As a result,
controlling for family environment when examining the relationship between abuse and later
psychological symptoms may be a highly conservative, or even nonsensical procedure (e.g.,
examining the effects of incest after removing variance associated with living in a disturbed
or dysfunctional family environment) (Briere & Elliott, 1993).
Despite these concerns, statistical control procedures can be helpful in assessing potential
child abuse effects, especially when the control variables are less confounded with child mal-
treatment than is the case for concurrent family environment. In this regard, researchers have
found it helpful to control for race or family income when examining potential risk factors
for adult psychopathology, although this has been done rarely in child abuse research. Simi-
larly, although other forms of child maltreatment and later (adult) victimization experiences
may also covary with a given form of child abuse, controlling for such events can provide a
clearer—albeit more conservative—view of the unique mental health sequelae of child abuse.

J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222 1207
For this reason, the current authors have advocated for the application of multivariate research
strategies in the study of potential child abuse effects, while, at the same time, warning of
the potential problems associated with partialing out “family dysfunction” in such research
(Briere, 1988, 1992; Briere & Elliott, 1993).
Despite these methodological concerns, over two decades of research suggest with relative
unanimity that childhood physical and sexual abuse have a wide number of psychological
sequelae. Among these are low self-esteem, anxiety, depression, anger and aggression, post-
traumatic stress, dissociation, substance abuse, sexual difficulties, somatic preoccupation and
disorder, self-injurious or self-destructive behavior, and most of the various symptoms and be-
haviorsseenin those diagnosedwith borderline personalitydisorder (see reviewsbyBerliner&
Elliott, 2002; Briere & Runtz, 1993; Finkelhor, 1990; Kolko, 2002;Neumann, Houskamp,
Pollock, & Briere, 1996;Polusny & Follette, 1995). Given this wide range of findings, an
important goal of current research in this area is to determine the mediators and potential
confounds associated with these abuse-symptom relationships. Equally important is the deter-
mination of the extent to which such symptom correlates occur in the general population, as
opposed to solely in the university and clinical contexts where most studies in this area have
been conducted.
Beyond the issue of possible abuse effects, also unresolved is the actual prevalence of
childhood sexual and physical abuse in the general population. Reported sexual abuse rates
for women, for example, have ranged from 8% to 32% in retrospective general population
samples, whereas the rates for men range from 1% to 16% (Finkelhor, 1994). Among the
possible reasons for such variability include differences in abuse definitions, survey methods,
and representativeness of the samples collected. Controversy has occurred especially for male
abuse rates, with several writers suggesting that the reported rates for males who were sex-
ually abused as children are serious underestimates of their actual prevalence in the general
population (e.g., Mendel, 1995).
Inresponseto these continuingissues,thepresent paper reportsonthe prevalenceandmental
health correlates of child abuse in a random sample of the general population. In order to boost
the external validity/generalizability of this study, special efforts were made to insure a rela-
tivelyhighresponserate, sothatconcerns aboutrepresentativenesscouldbe addressed. Inaddi-
tion, in response to concerns about measurement sensitivity, the present study used the Trauma
Symptom Inventory (TSI; Briere, 1995), a standardized test that may be more likely than some
other measures to be responsive to the specific impacts of childhood abuse. Finally, the present
study controlled for important demographic variables, and accounted for non-abuse-related
traumas that otherwise might confound abuse-symptom analyses, while at the same time con-
sidering the statistical and interpretive implications of such a conservative procedure.
Methods
Procedure
A national sampling service generated a geographically stratified, random sample of 1,442
subjects with deliverable addresses, based on records of registered owners of automobiles

1208 J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222
and/or individuals with listed telephones. According to the 1990 U.S. Census, over 95% of
all households have telephones, allowing this sample to tap the majority of individuals in the
United States. These subjects were mailed a questionnaire that included, among other mea-
sures, the Traumatic Events Survey (TES; Elliott, 1992) and the Trauma Symptom Inventory.
Informed consent and a guarantee of confidentiality was included on the coversheet of the
questionnaire. Research approval for this study was obtained from Biola University when the
second author was affiliated there.
Three follow-up mailings were sent to nonrespondents at approximately one month inter-
vals. Of all potential subjects, 935 (64.8%) returned surveys with substantially complete data
for the measures used in this study. Although response rates varied according to three types of
solicitation approaches (some received no money, some received $5 with the questionnaire,
and some received $5 if they mailed back the completed questionnaire), no demographic vari-
able, self-reported trauma experience, or TSI scale score differed according to method. Given
this equivalence, the different solicitation samples were combined into a single group.
Measures
Traumatic Events Survey (TES). The TES evaluates a wide range of childhood and adult
traumas. This instrument appears to be a valid measure of exposure to potentially traumatic
events, and has been used in several published studies of trauma impacts (see Briere, 1997 for
a more detailed review of this measure). Of the 30 interpersonal and environmental traumas
examined by the TES, 20 address adult events and 10 are devoted to childhood events.
Trauma Symptom Inventory (TSI). The TSI is a 100-item test of posttraumatic stress and other
psychologicalsequelaeoftraumaticevents.Eachitemasksaboutthefrequencyatwhicha given
symptom has occurred in the last 6 months, rated on a 0 (“never”) to 3 (“often”) Likert-like
scale. Typical items are “Feeling tense or ‘on edge’,” “Nightmares or bad dreams,” “Bad
thoughts or feelings during sex,” “Feeling helpless,” and “Trouble controlling your temper.”
The TSI has three validity scales and 10 clinical scales, although only the clinical scale results
are reported for this study. The latter are Anxious Arousal,Depression,Anger-Irritability,
Intrusive Experiences,Defensive Avoidance,Dissociation,Sexual Concerns,Dysfunctional
Sexual Behavior,Impaired Self-Reference, and Tension Reduction Behavior. TSI scales are
internally consistent (mean alphas of .86, .87, .84, and .84 in general population, clinical,
university,andmilitarysamples,respectively;Briere,1995),andexhibitconvergent,predictive,
and incremental validity in a variety of studies (e.g., Briere & Elliott, 1998; Runtz & Roche,
1999; Shapiro & Schwartz, 1997).
Abuse specification
Child physical and sexual abuse histories were determined by subjects’ responses to the
relevant sections of the TES. The two TES physical abuse items ask “Before the age of 18, did
your parents or caretaker ever do the following:” (1) “hit you with a fist, kick you, or throw you
down on the floor, into a wall, or down stairs,” or (2) “do something to you on purpose that left
marks, bruised, burned, or caused you to bleed, lose teeth, or have broken bones.” Additional

J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222 1209
items ask how often the events had occurred, the subject’s age at the first and last times it ever
happened, and how upsetting the subject found the event to be at the time it occurred.
The two sexual abuse items are: (1) “Before the age of 18, did anyone 5 or more years older
than you ever kiss or touch you in a sexual way or have you touch them in a sexual way,” and
(2) “Before the age of 18, did anyone less than 5 years older than you use physical force to kiss
or touch you in a sexual way, or force you to touch them in a sexual way?” Additional items
assess how often these acts occurred, the subject’s age at first and last time it occurred, whether
any event included oral/anal/vaginal penetration, total number of individuals who perpetrated
these acts, whether any of these individuals were family members, and how upsetting the
subject found the event to be at the time it occurred.
Childhood physical abuse was considered present in this study if either of the two TES
physicalabuseitemswere endorsed. Similarly,childhoodsexual abuse was defined aspresentif
eitherTES sexualabuse itemwas endorsed.For bothphysical and sexual abuse,characteristics
were summarized as follows: age at first abuse was considered to be the earliest endorsement
of either item, age at last abuse the latest endorsement of either item, frequency of abuse was
the highest number for either item, and how “upset” the subject felt was defined as the highest
rating for either item. For sexual abuse, penetration and incest were considered present if
endorsed on either TES sexual abuse item.
Adult trauma specification
The variable used in the present study to index adult interpersonal violence was created by
determiningfor eachsubject anyexposure to sexualor physicalviolence (withinor outsideof a
relationship) at age 18 or older, as self-reported on the TES. If any such violence was reported,
this variable was scored as positive, whereas no reports of adult interpersonal violence of any
type meant that this variable was scored as negative.
This variable was created to serve as a control variable, since child abuse history has been
related to subsequent adult trauma exposure in several studies (Neumann et al., 1996). In this
regard, it is possible that some of what have been considered child abuse effects may be due,
in fact, to adult traumas that are statistically associated with child abuse. For example, an adult
rape experience might produce symptoms that would correlate with child maltreatment to the
extent that rape and child maltreatment were correlated. In such an instance, were the rape not
taken into account, child abuse would account for symptom variance that was potentially due
to the more proximal rape experience (Briere, 1992). To address this possible confound, adult
trauma—as operationalized above—was entered in regression analyses before child abuse was
considered, thereby controlling for its potential impact on symptomatology. Because the focus
of the present paper is child abuse, as opposed to adult victimization, adult victimization was
treated solely as a summary variable. Additional papers are in progress regarding specific adult
victimization effects.
Subjects
Ofthe935subjects in thissample,464(49.6%) were male and471(50.4%)were female. The
mean subject age was 46 years, with a range of 18 to 90. The modal marital status was married/

1210 J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222
cohabiting(n=521,55.4%),followedbynevermarried(n=171,18.2%),separated/divorced
(n=158, 16.8%), and widowed (n=88, 9.4%). Racial breakdown was Caucasian (n=698,
74.7%), African American (n=106, 11.3%), Hispanic (n=68, 7.3%), Asian (n=28,
3.0%), Native American (n=19, 2.0%) and “other” (n=16, 1.7%). The modal education
level was some college or trade school (n=306, 32.5%) and the modal family income was
$10,000–$19,999 (n=182, 19.3%), followed by $20,000–$29,999 (n=174, 18.5%).
Statistical issues
Because of the relatively large number of statistical tests performed in this study, and
thus the likelihood of experiment-wise error rate inflation, the minimal alpha for statistical
significance was set at p<.01. In addition, post hoc univariate tests were only performed
when their associated multivariate tests were significant, per Cohen and Cohen’s approach
to experiment-wise error rate protection (Cohen & Cohen, 1983). Finally, all 2 ×2χ2were
corrected for continuity.
Results
Prevalence and characteristics of abuse histories
Sexual abuse. A total of 66 men and 152 women (14.2% and 32.3%, respectively) reported
childhood experiences that satisfied the current criteria for sexual abuse. This sex difference
was statistically significant, χ2(1)=41.6, p<.001. Of these cases, a number of abuse
characteristics did not differ according to sex of subject. These were incest (abuse within the
immediate or extended family: 46.8%), oral, anal, or vaginal penetration by a penis or other
object (52.8%), number of incidents (mode =2–5, 32.2%), number of perpetrators (M=1.7,
SD =1.2), age at first abuse (M=9.7, SD =3.8), and age at last abuse (M=12.1,
SD =4.0). However, males were more likely than females to report at least one female sexual
abuse perpetrator (39.4% vs. 8.5%, χ2(1)=13.5, p<.001), and females were more likely
than males to report at least one male perpetrator (92.7% vs. 69.7%, χ2(1)=8.6, p<.003).
There also were sex differences in reports of how upsetting the sexual abuse was at the time it
happened.On ascale of 0 (“not at all”) to 3 (“very”),males had a mean score of 1.5 (SD =1.4),
whereas females have a mean score of 2.3 (SD =1.4), t(215)=5.4, p<.001.
Physical abuse. Childhood experiences satisfying criteria for physical abuse were reported
by 103 males and 92 females (22.2% and 19.5%, respectively), without a significant sex
difference. Mean ages at first and last physical abuse incidents were 9.0 (SD =4.1) and 14.1
(SD =3.5) years, respectively, with no sex difference for age at first abuse, but with females
reporting abuse that ended at a later age than was the case for males (M=14.8, SD =3.6
vs. M=13.4, SD =13.4, respectively), t(173)=−2.2, p=.007. The modal number of
physical abuse incidents for this group was 2–5, and was equivalent for males and females. As
was true for sexual abuse, males rated their physical abuse as less upsetting at the time than did
females (M=2.4, SD =.8, vs. M=2.8, SD =.5, respectively), t(189)=−3.8, p<.001.

J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222 1211
Overlapping abuse histories. Cross-tabulation of subjects’ physical and sexual abuse histo-
ries indicated that individuals with one form of abuse were statistically more likely to have
experienced the other as well (χ2(1)=23.2, p<.001). In this regard, of all subjects with
some form of child abuse history (n=345), 72 (20.9%) had experienced both physical and
sexual abuse.
Prevalence of adult interpersonal violence
Based on their self-reports on the TES, a total of 337 subjects (36.0%) reported at least one
instance of sexual or physical interpersonal victimization at age 18 or later. This prevalence
varied by sex: 149 males (32.0%) reported victimization, as opposed to 188 females (39.9%),
χ2(1)=5.96, p<.015. Adult victimization histories were more common among both those
who reported childhood physical abuse (χ2(1)=45.2, p<.001) and childhood sexual abuse
(χ2(1)=39.8, p<.001).
Prediction of symptomatology
Of the current sample, 921 (all but 14) subjects had complete data on background vari-
ables, abuse status, and TSI raw scale scores. Hierarchical multiple regression analyses were
performed first, regressing (in order of entry) sociodemographic variables, adult victimization
history, and childhood abuse history (sexual and physical) on each of the 10 TSI scale scores.
At Step 1, demographic variables were significantly associated with psychological symp-
tomatology (see Table 1). Step 1 βweights (evaluating the unique effects of each demographic
variable, controlling for all other demographic variables) indicated that raw TSI scores were
(a) negatively related to age (younger subjects scored higher on all TSI scales), (b) positively
related to female sex (women scored higher than men on Anxious Arousal, Depression, Dis-
sociation, and Impaired Self-Reference), (c) positively or negatively related to racial minority
status, according to TSI scale (Caucasians [coded as “0”] scored higher on Anxious Arousal,
whereas non-Caucasians [coded as “1”] scored higher on Tension Reduction Behavior), and
(d) negatively related to income (those with less family income scored higher on all TSI scales
except Anger-Irritability, Dissociation, Sexual Concerns, and Tension Reduction Behavior).
At Step 2, multiple regression analysis revealed significant R2changes for all 10 TSI scores,
indicating that the adult interpersonal trauma variable predicted additional TSI score variance
beyond that predicted by Step 1 demographic variables.
AtStep 3, physicalandsexual abuse were enteredsimultaneously and werefoundto produce
significant R2changes for all 10 TSI scores, indicating that childhood abuse was associated
with additional TSI variance after controlling for demographic variables, adult trauma, and
each other. As indicated in Table 1, Step 3 βweights revealed that sexual abuse was uniquely
related to all 10 TSI scales, and physical abuse was associated with all scales except Sexual
Concerns and Dysfunctional Sexual Behavior.
Finally, at Step 4, all 2-way interactions between sex, sexual abuse, and physical abuse
were entered. Only in one instance was the R2change associated with this step statistically
significant: subjects with a sexual abuse history but no physical abuse scored higher than other
subjects on Dysfunctional Sexual Behavior (β=−.17, p<.001).

1212 J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222
Table 1
Multiple regression of demographic variables, adult victimization, and child abuse history on Trauma Symptom
Inventory (TSI) scales
TSI scale Step 1 R2F
(4, 917) p
Age
(β)Sexa
(β)Raceb
(β)Income
(β)
Anxious Arousal −.17∗.11∗−.14∗−.09∗.06 14.9 <.001
Depression −.16∗.16∗−.06 −.14∗.08 19.2 <.001
Anger-Irritability −.27∗.03 −.08 −.04 .07 17.8 <.001
Intrusive Experiences −.15∗.12 .05 −.11∗.06 14.9 <.001
Defensive Avoidance −.17∗.07 .02 −.18∗.07 17.6 <.001
Dissociation −.16∗.09∗−.08 −.08 .04 10.2 <.001
Sexual Concerns −.20∗−.07 −.03 −.04 .04 10.2 <.001
Dysfunctional Sexual
Behavior −.22∗−.06 .06 −.09∗.06 14.9 <.001
Impaired
Self-Reference −.26∗.10∗.06 −.12∗.09 23.5 <.001
Tension Reduction
Behavior −.17∗.05 .09∗.00 .07 18.4 <.001
Step 2 Step 3 R2change F
(2, 914) p
Adult
IVc(β)R2
change F
(1, 916) pPhysical
abuse (β)Sexual
abuse (β)
Anxious Arousal .21∗.04 43.7 <.001 .12∗.23∗.08 36.8 <.001
Depression .21∗.04 43.3 <.001 .13∗.25∗.08 44.5 <.001
Anger-Irritability .20∗.04 40.1 <.001 .19∗.23∗.09 53.7 <.001
Intrusive Experiences .26∗.07 72.1 <.001 .15∗.21∗.07 41.2 <.001
Defensive Avoidance .26∗.07 73.4 <.001 .10∗.25∗.09 43.8 <.001
Dissociation .21∗.05 45.0 <.001 .13∗.27∗.09 48.0 <.001
Sexual Concerns .16∗.03 25.2 <.001 .08 .32∗.10 56.9 <.001
Dysfunctional Sexual
Behavior .12∗.01 13.0 <.001 .04 .25∗.06 32.7 <.001
Impaired
Self-Reference .19∗.04 37.9 <.001 .12∗.24∗.07 40.8 <.001
Tension Reduction
Behavior .14∗.02 19.6 <.001 .12∗.30∗.10 56.3 <.001
Step 4 (interactions) Final statistics
R2
change F
(4, 911) pR
2F
(10, 911) p
Anxious Arousal .01 2.9 ns .18 19.8 .001
Depression .00 .4 ns .20 22.5 .001
Anger-Irritability .00 1.3 ns .21 23.9 .001
Intrusive Experiences .00 .6 ns .20 23.1 .001
Defensive Avoidance .00 .4 ns .22 25.3 .001
Dissociation .00 .5 ns .18 18.5 .001
Sexual Concerns .01 2.6 ns .18 19.6 .001
Dysfunctional Sexual
Behavior .01 5.0 <.002 .15 16.1 .001
Impaired
Self-Reference .01 3.3 ns .21 24.1 .001
Tension Reduction
Behavior .00 1.4 ns .20 22.3 .001
aMale (0) versus female (1).
bCaucasian (0) versus non-Caucasian (1).
cAdult interpersonal violence.
∗p≤.01.

J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222 1213
Table 2
Sexual abuse and physical abuse effect sizes for males and females
TSI scale Males Females
Sexual abuse Physical abuse Sexual abuse Physical abuse
F(1, 461) η2F(1, 461) η2F(1, 458) η2F(1, 458) η2
AA 36.2a.07 4.4b.01 12.2c.03 119d.03
D 37.0 .07 11.6 .02 24.1 .05 10.2 .02
AI 34.4 .07 29.1 .06 20.8 .04 17.7 .04
IE 20.4 .04 13.0 .03 24.4 .05 26.1 .05
DA 24.5 .05 3.3 .01 38.5 .08 13.8 .03
DIS 36.8 .07 12.7 .03 23.2 .05 10.0 .02
SC 37.2 .08 5.4 .01 39.5 .08 1.2 .00
DSB 19.8 .04 0 .00 20.6 .04 1.1 .00
TSR 31.4 .06 7.5 .02 15.0 .03 8.7 .02
TRB 42.9 .09 9.3 .02 31.5 .06 7.0 .02
Note. AA: Anxious Arousal, D: Depression, AI: Anger/irritability, IE: Intrusive Experiences, DA: Defen-
sive Avoidance, DIS: Dissociation, SC: Sexual Concerns, DSB: Dysfiinctional Sexual Behavior, TSR: Impaired
Self-Reference, TRB: Tension Reduction Behavior.
aAll male sexual abuse effects significant at p≤.01.
bAll male physical abuse effects significant at p≤.01 except AA (p=.037), DA (ns), SC (p=.020), and
DSB (ns).
cAll female sexual abuse effects significant at p≤.01.
dAll female physical abuse effects significant at p≤.01 except SC (ns), and DSB (ns).
Although no sex ×abuse interactions were significant, some meta-analytic studies of child
abuse effects calculate effect sizes separately for males and females (e.g., Rind et al., 1998).
For this reason, univariate child abuse effect sizes (η2, representing the amount of TSI scale
variance associated with a given type of abuse) are presented for each sex in Table 2. However,
given the absence of significant sex×abuse interactions in this study, apparent sex differences
in these effects should not be considered statistically meaningful.
In order to determine the clinical significance of the sexual and physical abuse effects found
in this study, a 2(sex)×2(sexualabuse)×2(physical abuse)ANOVA was conducted, where
the dependent variable consisted of the number of TSI scales (ranging from 0 to 10) each
subject had that were in the clinical range (i.e., a T-score of 65 or higher, per Briere, 1995).
As shown in Table 3, the number of clinically elevated scales varied as a function of sex,
childhood sexual abuse, and childhood physical abuse, although inspection of the associated
η2suggests that the sexual abuse effect was the most meaningful finding. In that instance,
those with a positive history had a mean of two clinically elevated TSI scales, whereas those
without a sexual abuse history had a mean of less than one elevation.
Because the distribution of scores on the “elevated TSI scales” measure was not normally
distributed (skew =2.5), nonparametric analyses were also performed on the relationship
between sexual abuse history and TSI elevations. The TSI elevation variable was reformed
into a five-point scale: 0 (no clinical elevations), 1 (one clinical elevation), 2 (two to four
elevations), 3 (five to seven elevations), and 4 (eight to ten elevations). As presented in Table 4,

1214 J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222
Table 3
ANOVAs of sex and abuse history on number of Trauma Symptom Inventory (TSI) scale elevations above T65
Independent variable Number of elevated TSI scores
MSDF(1, 933) pη2
Sex 5.6 <.018 .006
Male (n=469) 1.0 2.2
Female (n=472) .9 2.0
Childhood sexual abuse 60.36 <.001 .061
No (n=721) .7 1.7
Yes (n=220) 2.1 3.0
Childhood physical abuse 11.2 <.001 .012
No (n=744) .8 1.9
Yes (n=197) 1.7 2.8
the greater the number of elevated TSI scores, the greater the likelihood of the subject having
a sexual abuse history [χ2(4)=75.5, p<.001]; Somer’s d=.28, p<.001). In the most
extreme case, of the 34 subjects with 8–10 elevated TSI scales, 22 (64.7%) had a self-reported
history of childhood sexual abuse. A total of 46.4% of sexual abuse survivors had at least one
elevatedTSIscore,as opposedto only 20.1%of those who did notreport a sexualabusehistory.
Abuse characteristics effects
As presented in Table 5, canonical correlation analysis indicated that five sexual abuse
characteristics were associated with all TSI scores [Rc=.50, F(70,1073)=1.50, p<
.006]. Examination of the canonical structure coefficients (c), which represent the redundant
contribution of each abuse characteristic to the overall multivariate relationship between abuse
characteristics and TSI scale scores, revealed that older age at last abuse, a greater number of
abuse incidents, a greater number of abuse perpetrators, the presence of oral, anal, or vaginal
penetration, and how upset subjects reported being at the time of the abuse were associated
with a general increase in TSI scores. Not associated with TSI scores were age at first sexual
abuse incident or whether any abuse involved incest.
Table 4
Cross-tabulation of number of elevated Trauma Symptom Inventory (TSI) scales and history of childhood sexual
abuse
Number of elevated
TSI scales Negative sexual abuse
history (n=721) Positive sexual abuse
history (n=220)
0 569 (78.9%) 118 (53.6%)
1 56 (7.8%) 22 (10.0%)
2–4 61 (8.5%) 35 (15.9%)
5–7 23 (3.2%) 23 (10.5%)
8–10 12 (1.7%) 22 (10.0%)

J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222 1215
Table 5
Canonical correlation of sexual abuse characteristics with Trauma Symptom Inventory (TSI) scales
ca
Abuse characteristic
Age at first abuse −.26
Age at last abuse −.78
Number of incidents −.53
How upset at time of abuse −.41
Number of perpetrators −.70
Penetration −.50
Incest −.14
TSI scale
Anxious Arousal −.50
Depression −.42
Anger-Irritability −.62
Intrusive Experiences −.53
Defensive Avoidance −.71
Dissociation −.73
Sexual Concerns −.82
Dysfunctional Sexual Behavior −.75
Impaired Self-Reference −.57
Tension Reduction Behavior −.74
aCanonical structure coefficient, considered meaningful at |c|>.35.
Table 6
Canonical correlation of physical abuse characteristics with Trauma Symptom Inventory (TSI) scales
ca
Abuse characteristic
Age at first abuse −.16
Age at last abuse .84
Number of incidents .50
How upset at time of abuse .62
TSI scale
Anxious Arousal .61
Depression .74
Anger-Irritability .35
Intrusive Experiences .71
Defensive Avoidance .67
Dissociation .64
Sexual Concerns .18
Dysfunctional Sexual Behavior .10
Impaired Self-Reference .66
Tension Reduction Behavior .17
aCanonical structure coefficient, considered meaningful at |c|>.35.

1216 J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222
Effects of sexual abuse offender and victim sex. Because the TES also assesses the gender
of the abuser in instances of self-reported sexual abuse, a final abuse characteristic analysis
was performed to determine if the presence of at least one male or at least one female abuser
determined symptomatic outcome, both as a main effect and as it interacted with victim sex.
Multivariate analyses of covariance (controlling for total number of perpetrators) indicated
that—beyond being sexually abused, per se—being abused by a male had no specific impact
on TSI scores [F(10,97)=.6, ns] for either victim sex [F(10,97)=1.5, ns], nor did abuse
by a female [F(10,97)=1.0, ns] for either victim sex [F(10,97)=1.2, ns].
Characteristicsof physical abusewere also related to TSIscales bycanonical analysis [Rc=
.49, F(40,597.2)=2.02, p<.001]. All TSI scales but Sexual Concerns, Dysfunctional
Sexual Behavior, and Tension Reduction Behavior were associated with older age at last
abuse, a greater number of physical abuse incidents, and how upset the victim was at the time
of the abuse (see Table 6). Age at first physical abuse incident was not related to TSI scores.
Discussion
The results of this study are discussed below in terms of prevalence, associations with
current symptomatology, and methodological issues.
Prevalence
As per other general population studies, the current results indicates that a self-reported
history of sexual or physical abuse is relatively common. Approximately 32% of general
population females and 14% of males reported childhood sexual abuse, and 22% and 20%
of males and females, respectively, reported childhood physical abuse. Overall, a total of 345
subjects (37% of all those studied) described either sexual or physical maltreatment as a child.
Of those subjects with child abuse histories, approximately 21% reported having been exposed
to both physical and sexual maltreatment.
Thesexualabuseprevalencefiguresdescribedaboveare somewhathigher than somegeneral
population studies (e.g., Bagley & Ramsay, 1986; Finkelhor, 1984) but in the same range as
others (e.g., Bagley, 1991;Finkelhor, Hotaling, Lewis, & Smith, 1990;Wyatt, 1985). Such
numbers reinforce the notion that child maltreatment is a significant phenomenon in North
Americanculture.Itshouldberecalled, however,thatthesedatareflectthestateofvictimization
incidence approximately 30 years ago, as opposed to whatever its current rate may be. At
minimum, however, these data suggest that a significant proportion of the current cohort of
American men and women in America have been exposed to childhood sexual or physical
abuse, as indexed by their self-report.
In addition to the prevalence of sexual and physical abuse, the present study reports on the
sex of sexual abuse offenders for male and female victims. The gender breakdown for female
victims is generally in line with the sexual abuse literature (93% had been abused by at least
one male, and 9% had been abused by at least one female), whereas the data on perpetrator
sex for male victims represents a newer finding. In this regard, 39% of males reported having
been sexually abused by at least one female, and 70% described sexual abuse by at least

J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222 1217
one male. Although there are few community studies with which to compare these latter
numbers, Finkelhor (1994) estimates that, extrapolating from available retrospective report
studies, approximately 20% of prepubescent boys have been sexually victimized by a female.
At the other end of the report continuum, Mendel (1995) found that 60% of his clinical sample
of males reported childhood sexual contact with an older female. Given that boys are probably
at least as likely be abused by females during puberty as they would be prepubertally, it is not
surprising that the current sample, which considered sexual abuse experiences up through age
17, would report a 39% female abuse rate. Such relatively high rates of female offending are
noteworthy, since the common assumption is that the vast majority of sexual child abusers are
male. Although nothing in the present study contradicts the notion that most sexual abusers
are men, the current data suggest that the child abuse field may have underestimated the rate
of women’s sexual offending, primarily against boys.
Association with psychological symptoms
Sexual abuse. The current data suggest that, as has been found in clinical and university stu-
dent studies, childhood sexual abuse is a significant risk factor for a range of psychological
symptoms in the general population. Specifically, reports of sexual abuse were associated with
elevations on all 10 scales of the TSI, even after controlling for a variety of sociodemographic
variables, including sex, age, race, and family income, as well as subsequent interpersonal
victimization as an adult and physical abuse in childhood. These data support not only the
majority of the literature on mental health sequelae of childhood sexual abuse, but also the
findings of one of the only general population studies in this area (Saunders, Villeponteaux,
Lipovsky, Kilpatrick, & Veronen, 1992). Saunders et al. found that in a 391-person random
sample of Charleston County, South Carolina, self-reported childhood sexual abuse was asso-
ciated with a wide range of psychiatric disorders and problems, including depression, phobias,
obsessive-compulsive disorder, panic disorder, posttraumatic stress disorder, sexual disorders,
and both suicidal ideation and suicide attempts.
In the current study, certain characteristics of the sexual abuse experience were specifically
associated with psychological symptomatology. Predictive of TSI scores were sexual abuse at
a later age, a greater number of abuse incidents, multiple abusers, victimization that involved
oral,anal, or vaginal penetration,and a greater levelof emotional upset atthe time of the abuse.
As has been shown in some other studies (see reviews by Briere, 2000; Berliner & Elliott,
2002),whether theabusewas intrafamilial orextrafamilial did notseem toaffectpsychological
outcome, at least as measured by TSI scores. Finally, contrary to expectation, sex of the abuser
was not a significant predictor of symptomatology among those who reported having been
abused, even under those conditions sometimes thought to be most detrimental (e.g., male
abuser-male victim or female abuser-female victim).
Physical abuse. Also associated with most TSI scores was self-reported childhood physical
abuse, although to a lesser extent than sexual abuse. Physical abuse was associated with all
TSI scales except for those tapping sexual symptoms (Sexual Concerns and Dysfunctional
Sexual Behavior) and Tension Reduction Behavior. The tendency for sexual, but not physical
abuse to predict sexual symptoms has been found in other studies (e.g., Briere & Runtz, 1990),

1218 J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222
reinforcing the notion that sexual distress or dysfunction may be a specific impact of sexual
trauma for some individuals. Similar to sexual abuse, physical abuse characteristics associated
withmoreseveresymptomatologywereabuseat a later age,agreaternumber of abuseincidents
and abuse perpetrators, and greater distress at the time of the abuse.
Effects of controlling for related variables. It should be noted that the multiple regression ap-
proachusedinthisstudyisintrinsicallyconservativewithrespecttouncoveringabuse-symptom
relationships. By entering interpersonal violence in adulthood in the step prior to simultane-
ously entering sexual and physical abuse, each type of child abuse was considered only after
all variance associated with adult victimization and the other form of childhood abuse was
removed. Because childhood physical and sexual abuse were found in the present study to be
risk factors for later adult victimization, and sexual abuse was significantly associated with
physical abuse, controlling for adult victimization and the other form of child abuse (i.e.,
controlling for sexual abuse while examining physical abuse effects, and the reverse) is likely
to have removed significant abuse-related variance that otherwise might have correlated with
symptomatology. For this reason, the continuing relationship between self-reported childhood
maltreatment and adult symptomatology is all the more noteworthy.
Effect size and clinical significance issues. Although the relationship between child abuse
history and both (a) the likelihood of adult victimization, and (b) a variety of types of psy-
chological symptomatology proved to be statistically robust in this study, it should be noted
that the size of the abuse-symptom relationships are relatively small. Once demographics and
adult interpersonal violence exposure were taken into account (admittedly a conservative pro-
cedure), the additional variance in any given TSI scale accounted for by physical and sexual
abuse ranged from 6% (for Dysfunctional Sexual Behavior) to 10% (for Sexual Concerns
and Tension Reduction Behavior). Effect sizes of this magnitude suggest that the impact of
child abuse may vary from individual to individual, probably as a function of a range of
other variables, abuse-specific and otherwise. For example, the canonical correlation results
of the current study join findings from other studies (see Berliner & Elliott, 2002 for a re-
view) in suggesting that more severe outcomes are likely to arise when sexual abuse occurs
later in childhood (perhaps as a function of closer proximity to the point of [adult] symptom
evaluation), involves a greater number of separate abuse incidents and multiple abuse perpe-
trators, includes oral/anal/vaginal penetration, and is perceived as especially upsetting at the
time.
As a result, it is important to qualify the results of this or other studies by emphasizing
that abuse need not always produce long-term effects, nor need all such effects necessarily be
major. It is likely that any given child’s short- and long-term response to maltreatment will be
a function of a variety of variables, including his or her temperament, general pre-abuse devel-
opmental history and psychological functioning, the characteristics of the abuse experience(s),
and how others responded to any subsequent disclosure. On the other hand, effect sizes of the
magnitude found in the present study can have substantial clinical implications. Ondersman,
Chaffin, and Berliner (1999), for example, performed a meta-analysis and estimate that the re-
lationship between smoking and lung cancer is approximately r=.12; a considerably smaller
effect size (1% of variance accounted for) than what is demonstrated here (6–10% of symptom

J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222 1219
variance)—especially given that the latter refers to variance still available after partialing-out
related socioeconomic and victimization variables.
Another way to consider the meaningfulness of the effects found here is to examine their
psychometric significance. In the present study, sexual abuse victims were more than twice
as likely to have at least one TSI scale elevated into the clinical range, and those with a
greater number of clinical TSI elevations were considerably more likely to be sexual abuse
survivors. Such data suggest that, at minimum, sexual abuse survivors are at increased risk for
the subsequent development of clinically significant psychological disturbance, as measured
by psychological tests.
Limitations of this study. Like most other studies on adult abuse sequelae, the cross-sectional
self-report methodology of the current research limits the interpretations that can be made
from its results. Participants’ descriptions of their maltreatment history may have been af-
fected by memory distortion, and their symptom endorsements may have been biased by
under- or over-reporting of their actual level of distress. Further, although a 65% return rate
is considered well within the range of acceptability for mail-out research (Dillman, 1978),
it is possible that those who did not return questionnaires would have responded in ways
that could have changed the findings of this study. For example, the methodology of the cur-
rent research precluded participation by those unable to read English, as well as those who
were not surveyed by the mailing list, such as the homeless, incarcerated, or psychiatrically
hospitalized.
In addition, the cross-sectional nature of this study precludes definitive causal interpreta-
tions regarding the long-term impacts of child abuse (Briere, 1992). The relationship between
self-reported childhood maltreatment and adult symptomatology may have been due to other
intervening variables. Although we took care to control for demographics, other child abuse,
and adult trauma exposure, we did not examine childhood neglect, nor did we assess for in-
tervening life stressors (other than adult interpersonal violence) that might have covaried with
childhood maltreatment and adult symptomatology. It should be noted, however, that the cur-
rent data generally agree with the findings of most other studies in the area (Berliner & Elliott,
2002; Kolko, 2002; Neumann et al., 1996; Polusny & Follette, 1995) and are in concordance
with longitudinal studies that directly implicate childhood abuse in the development of lasting
psychological symptoms (Egeland, 1997; Starr & Wolfe, 1991).
Conclusion
Thecurrent report on the prevalenceandsymptomatic correlates ofself-reported childabuse
in the general population suggests that not only is child maltreatment relatively common, it
also is associated with a variety of types of psychological dysfunction years later. Because this
study was based on self-report, retrospective data, the relationship between childhood abuse
and adult symptoms cannot be assumed to be causal. Nevertheless, in combination with a
variety of other studies, the current findings suggest that a reliable relationship exists, even
after controlling for background variables such as socioeconomic status and other forms of
trauma.

1220 J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222
Together, these various studies and the current data reinforce the proposition that childhood
sexual and physical abuse is endemic in our culture, and suggest that, in contrast to the conclu-
sions of Rind et al. (1998), sexual abuse is likely to have significant long-term effects. These
findings not only support of child abuse prevention activities as a way to potentially decrease
the rate of psychological disorder in American society, but also serve as a continuing reminder
to clinicians that at least some of the basis for their clients’ psychological disturbance may
involve childhood maltreatment experiences.
Acknowledgments
This research was conducted when the second author was affiliated with Biola University,
La Mirada, California.
References
Bagley, C. (1991). The prevalence and mental health sequels of child sexual abuse in a community sample of
women aged 18 to 27. Canadian Journal of Community Mental Health,10, 103–116.
Bagley, C., & Ramsay, R. (1986). Sexual abuse in childhood: Psychosocial outcomes and implications for social
work practice. Journal of Social Work and Human Sexuality,4, 33–47.
Berliner,L.,&Elliott,D.M.(2002). Sexualabuseofchildren. In J. E. B. Myers, L. Berliner, J. Briere, C. T.Hendrix,
T. Reid, & C. Jenny (Eds.), The APSAC handbook on child maltreatment (2nd ed., pp. 55–78). Newbury Park,
CA: Sage Publications.
Briere, J. (1988). Controlling for family variables in abuse effects research: A critique of the “partialing” approach.
Journal of Interpersonal Violence,3, 80–89.
Briere, J. (1992). Methodological issues in the study of sexual abuse effects. Journal of Consulting and Clinical
Psychology,60, 196–203.
Briere, J. (1995). Trauma Symptom Inventory professional manual. Odessa, FL: Psychological Assessment Re-
sources.
Briere,J.(1997).Psychologicalassessmentof adult posttraumaticstates.Washington,DC: American Psychological
Association.
Briere, J. (2000). Incest. In A. E. Kazdin (Ed.), Encyclopedia of psychology. Washington, DC: American Psycho-
logical Association and Oxford Press.
Briere, J., & Elliott, D. M. (1993). Sexual abuse, family environment, and psychological symptoms: On the validity
of statistical control. Journal of Consulting and Clinical Psychology,61, 284–288.
Briere, J., & Elliott, D. M. (1998). Clinical utility of the Impact of Event Scale: Psychometrics in the general
population. Assessment,5, 135–144.
Briere, J., & Runtz, M. (1990). Differential adult symptomatology associated with three types of child abuse
histories. Child Abuse & Neglect,14, 357–364.
Briere, J., & Runtz, M. (1993). Child sexual abuse: Long-term sequelae and implications for psychological assess-
ment. Journal of Interpersonal Violence,8, 312–330.
Carlson, E. B. (1997). Trauma assessments: A clinician’s guide. New York: Guilford.
Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation analysis for the behavioral sciences (2nd
ed.). Hillsdale, NJ: Lawrence Erlbaum.
Davis, J. A. (1985). The logic of causal order. Sage University Paper Series on Quantitative Applications in the
Social Sciences, number 07-055. Beverly Hills, CA: Sage.
Dillman, D. A. (1978). Mail and telephone surveys: The Total Design method. New York: Wiley.

J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222 1221
Egeland, B. (1997). Mediators of the effects of child maltreatment on developmental adaptation in adolescence. In
D. Cicchetti & S. L. Toth (Eds.), Developmental perspectives on trauma: Theory, research, and intervention
(pp. 403–434). Rochester, NY: University of Rochester Press.
Elliott, D. M. (1992). Traumatic Events Survey. Unpublished psychological test. Los Angeles: Harbor-UCLA
Medical Center.
Elliott, D. M. (1994). Assessing adult victims of interpersonal violence. In J. Briere (Ed.), Assessing and treating
victims of violence (pp. 4–16). San Francisco: Jossey-Bass.
Finkelhor, D. (1984). Child sexual abuse: New theory and research. New York: Free Press.
Finkelhor, D. (1990). Early and long-term effects of child sexual abuse: An update. Professional Psychology:
Research and Practice,21, 325–330.
Finkelhor, D. (1994). Current information on the scope and nature of child sexual abuse. The Future of Children,
4, 31–53.
Finkelhor, D., Hotaling, G., Lewis, I. A., & Smith, C. (1990). Sexual abuse in a national survey of adult men and
women: Prevalence, characteristics, and risk factors. Child Abuse & Neglect,14, 19–28.
Fromuth, M. E. (1986). The relationship of childhood sexual abuse with later psychological and sexual adjustment
in a sample of college women. Child Abuse & Neglect,10, 5–16.
Kolko, D. J. (2002). Child physical abuse. In J. E. B. Myers, L. Berliner, J. Briere, C. T. Hendrix, T. Reid, &
C. Jenny (Eds.), The APSAC handbook on child maltreatment (2nd ed., pp. 21–54). Newbury Park, CA: Sage
Publications.
Mendel, M. P. (1995). The male survivor: The impact of sexual abuse. Thousand Oaks, CA: Sage.
Neumann, D. A., Houskamp, B. M., Pollock, V. E., & Briere, J. (1996). The long-term sequelae of childhood sexual
abuse in women: A meta-analytic review. Child Maltreatment,1, 6–16.
Ondersman,S. J., Chaffin,M., & Berliner,L.(1999).Commentary on the Rind et al. meta-analysis. APSAC Advisor,
12, 1–5.
Pedhazur, E. J. (1982). Multiple regression in behavioral research: Explanation and prediction (2nd ed.). New
York: Holt, Rinehart and Winston.
Polusny, M. A., & Follette, V. M. (1995). Long term correlates of child sexual abuse: Theory and review of the
empirical literature. Applied & Preventive Psychology,4, 143–166.
Rind, B., Tromovitch, P., & Bauserman, R. (1998). A meta-analytic examination of assumed properties of child
sexual abuse using college samples. Psychological Bulletin,124, 22–53.
Runtz, M. G., & Roche, D. N. (1999). Validation of the Trauma Symptom Inventory (TSI) in a Canadian university
sample. Child Maltreatment,4, 69–80.
Saunders, B. E., Villeponteaux, L. A., Lipovsky, J. A., Kilpatrick, D. G., & Veronen, L. J. (1992). Child sexual
assault as a risk factor for mental disorders among women: A community survey. Journal of Interpersonal
Violence,7, 189–204.
Shapiro, B. L., & Schwartz, J. C. (1997). Date rape: Its relationship to trauma symptoms and sexual self-esteem.
Journal of Interpersonal Violence,12, 407–419.
Starr, R. H., & Wolfe, D. A. (Eds.). (1991). The effects of child abuse and neglect: Issues and research. New York:
Guilford.
Wyatt, G. E. (1985). The sexual abuse of Afro-American and White American women in childhood. Child Abuse
& Neglect,9, 507–519.
Wyatt, G. E., & Peters, S. D. (1986). Methodological considerations in research on the prevalence of child sexual
abuse. Child Abuse & Neglect,10, 241–251.
Résumé
Objectif: Cette étude a voulu examiner la prévalence et les séquelles psychologiques des agressions
sexuelles et des mauvais traitements physiques dans la population en général.
Méthode: Un échantillon aléatoire et géographiquement stratifié comprenait 1.442 sujets vivant aux
ÉtatsUnis. Ils ontreçu un questionnairepar la poste contenant le Traumatic Events Survey (TES; Elliott,

1222 J. Briere, D.M. Elliott / Child Abuse & Neglect 27 (2003) 1205–1222
1992) et le Trauma Symptom Inventory (TSI; Briere, 1995); 935 individus (64.8%) l’ont retourné,
généralement bien rempli.
Résultats: Soixante-six hommes et 152 femmes (14.2% et 32.3% respectivement) ont relaté des ex-
périences qui rencontraient les critères pour agressions sexuelles; et 103 hommes et 92 femmes (22.2%
et 19.5% respectivement) ont satisfait les critères pour les mauvais traitements physiques. Vingt-et-un
pour cent des individus qui avaient connu au moins un type de mauvais traitements avaient aussi connu
l’autre type, et pour les deux catégories de mauvais traitements, on note que les individus deviennent
victimes d’agressions en ˆ
age adulte. Ayant contrˆ
olé les facteurs démographiques ainsi que la présence
de violence durant la vie adulte et autres types de mauvais traitements en enfance, les auteurs ont noté
que les abus sexuels s’associent aux dix échelles du TSI tandis que les mauvais traitements physiques
s’associent à toutes les échelles du TSI sauf celles qui mesurent les agressions sexuelles. La présence
d’expériences d’agressions sexuelles prédit une plus grande variance de symptˆ
omes que la présence de
mauvais traitements physiques ou d’agressions subies durant la vie adulte. Divers aspects des expéri-
ences sexuelles et physiques prédisent des scores du TSI. Cependant, on n’a remarqué aucun lien entre
des symptˆ
omes additionnels du TSI et la présence d’actes sexuels des agresseurs, en soi, ou combiné
avec la présence d’expériences sexuelles de la victime.
Conclusions: On retrouve communément la présence d’agressions sexuelles et physiques en enfance
dansla population engénéral, et celles-ci s’accompagnentd’une gamme de symptˆ
omespsychologiques.
Ces liens persistent mˆ
eme après avoir contrˆ
olé diverses variables.
Resumen
Objetivo: Este estudio examina la prevalencia y las secuelas psicológicas del maltrato y el abuso sexual
en la infancia en adultos de la población general.
Método: Un servicio estad´
ıstico nacional generó una muestrea aleatoria geográficamente estratificada
de 1.442 sujetos de USA. A todos los sujetos de les envió por correo un cuestionario que incluyó la
Encuesta de Acontecimientos Traumáticos (TES; Elliot, 1992) y el Inventario de S´
ıntomas Traumáticos
(TSI; Briere, 1995). De todos los potenciales sujetos, un total de 935 (64.8%) devolvieron las entrevistas
cumplimentadas.
Resultados: Sesenta y seis varones y 152 mujeres (14.2% y 32.3%, respectivamente) notificaron ex-
periencias infantiles que satisfacen los criterios para abuso sexual. Además, 103 varones y 92 mujeres
(22.2% y 19.5%, respectivamente) notificaron experiencias infantiles que cumplieron los criterios de
maltrato f´
ısico. Un 21% de los sujetos con un tipo de maltrato también hab´
ıa experimentado el otro tipo
y ambos tipos de maltrato estaban asociados con posterior victimización como adultos. Después de con-
trolar el efecto de las variables demográficas, la historia adulta de violencia interpersonal y otros tipos
de maltrato, se observó una asociación del abuso sexual infantil con el total de las 10 escalas del TSI y
una asociación del maltrato f´
ısico con todas las escalas del TSI, excepto con aquellas relacionadas con
cuestiones sexuales. El abuso sexual tiene un mayor valor predictivo de la varianza de los s´
ıntomas que
el maltrato f´
ısico o la victimización adulta interpersonal. Varios aspectos de las experiencias de maltrato
f´
ısico y abuso sexual fueron predictivos de las puntuaciones en el TSI. El género del abusador sexual,
ni de manera aislada ni en interacción con el género de la v´
ıctima, estuvo asociada con sintomatolog´
ıa
adicional en el TSI.
Conclusiones: El maltrato f´
ısico y el abuso sexual infantil es relativamente frecuente en la población
general y están asociados con una amplia variedad de s´
ıntomas psicológicos. Estas relaciones per-
manecen incluso después de haber controlado variables históricas relevantes.