Child Abuse & Neglect, Vol. 22, No. 11, pp.

1079 –1091, 1998

Copyright © 1998 Elsevier Science Ltd
Pergamon Printed in the USA. All rights reserved
0145-2134/98 $19.00 1 .00

PII S0145-2134(98)00089-1

ALCOHOL DEPENDENCE AND DOMESTIC VIOLENCE

AS SEQUELAE OF ABUSE AND CONDUCT DISORDER
IN CHILDHOOD
STEPHEN J. KUNITZ
Department of Community and Preventive Medicine, University of Rochester School
of Medicine and Dentistry, Rochester, NY, USA

JERROLD E. LEVY
Department of Anthropology, University of Arizona, Tucson, AZ, USA

JOANNE MCCLOSKEY
Center on Alcoholism, Substance Abuse, and Addictions, University of New Mexico, Albuquerque, NM, USA

K. RUBEN GABRIEL
Department of Statistics, University of Rochester, Rochester, NY, USA

ABSTRACT

Objectives: To examine in the Navajo population: (1) the importance of childhood abuse as a risk factor for conduct
disorder; (2) the importance of each form of abuse and conduct disorder as risk factors for alcohol dependence; and (3) the
relative importance of each form of abuse, conduct disorder, and alcohol dependence as risk factors for being a perpetrator
and/or victim of domestic violence.
Method: The study is based on a case-control design. Cases (204 men and 148 women) between the ages of 21 and 65 were
interviewed in alcohol treatment program and matched to community controls. There were two groups of controls: alcohol
dependent (374 men, 60 women) and nonalcohol dependent (157 men, 143 women). When adjusted for stratification by age,
community of residence, and sex, the combined control groups comprise a representative sample of the Navajo male and
female population 21– 65 years of age.
Results: The prevalence of physical and sexual abuse before age 15 is within limits observed in other populations. Each
form of abuse is a risk factor for conduct disorder. Along with conduct disorder, physical abuse is a risk factor for alcohol
dependence. Physical abuse and alcohol dependence are independent risk factors for being involved in domestic violence
as both perpetrator and victim. There appears to have been no secular trend in the incidence of childhood abuse over the
past several generations, but there is suggestive evidence that domestic violence has become more common.
Conclusions: Physical abuse is a significant risk factor for alcohol dependence as well as for domestic violence independent
of the effects of alcohol abuse. The effects of sexual abuse with regard to both domestic violence and alcohol dependence
do not appear to be significant. © 1998 Elsevier Science Ltd

Key Words—North American Indians, Alcohol dependence, Abuse, Conduct disorder.

The research on which this paper is based was supported by grant #R01 AA09420 from the National Institute on Alcohol
Abuse and Alcoholism.
Submitted for publication May 9, 1997; final revision received March 2, 1998; accepted March 5, 1998.
Requests for reprints should be sent to Dr. S. J. Kunitz, Department of Community and Preventive Medicine, Box 644,
University of Rochester Medical Center, 601 Elmwood Avenue, Rochester, NY 14642.
1079
1080 S. J. Kunitz, J. E. Levy, J. McCloskey, and K. R. Gabriel

INTRODUCTION

The causes, consequences, and secular trends of abuse of children have been of increasing interest
in the past several decades. Much of that concern has manifested itself in American Indian
communities as well. In this paper we use data from a study of alcohol use and abuse by Navajo
Indians to examine one small piece of the problem: the degree to which abuse in childhood is a risk
factor for alcohol dependence and domestic violence.
The evidence from Native American communities, limited though it is, indicates that child abuse
is not unknown (US Congress Office of Technology Assessment, 1990). For Navajo children,
White and Cornely (1981) and Hauswald (1987) report a rate of 13.5 per 1,000. Drawing on a
medical chart review and staff survey at the San Carlos (Apache) Indian Health Service Hospital,
Fischler (1985) found a rate of 5.7 per 1,000. For Cheyenne River Sioux Reservation children,
Wichlacz, Lane, and Kempe (1978) reported a rate of 26 per 1,000, derived from a register of
suspected cases. In a study of an Alaskan village, one third (28 of 84) of the native children were
considered to have severe problems of abuse, neglect, and homelessness related to poverty and
demoralization in the village (Jones, 1969); and a study in an unidentified southwestern Indian
community reported a history of childhood sexual abuse of 49% among women and 14% among
men (Robin, Chester, Rasmussen, Jaranson, & Goldman, 1997). The reported age-specific inci-
dence rate of abuse of children less than 18 years of age on the Navajo Reservation in 1992-95
varied between 3.3 and 4.3 per 1,000 (Northern Navajo Medical Center, 1996). Because these data
come from the Navajo Criminal Justice System and represent reported cases, they are very likely
much lower than the true incidence.
Human service providers who work with Indian children and adolescents in the Albuquerque and
Phoenix Indian Health Service (IHS) attributed abuse and neglect “to chaotic family situations and
to other mental health problems such as alcoholism and depression” (Piasecki, Manson, Biernoff,
Hiat, Taylor, & Bechtold, 1989; see also Lujan, DeBruyn, May, & Bird, 1989). Indeed, there are
data to support the association of alcohol abuse by parents and other caretakers and the abuse of
Indian children. Based on a case-control study in several Southwestern Indian communities
DeBruyn, Lujan, and May (1992) claim that alcohol abuse is a necessary but not sufficient cause
of abuse and neglect. They also observed that in addition to alcohol abuse, abuse of other
substances by parents, “histories of divorce, death in the immediate family, single-parent house-
holds, alcohol abuse by grandparents, and deaths in the family associated with alcoholism” were
all risk factors for abuse of children. And an early case-control study of neglected infants in one
Navajo community indicated that neglected infants were especially likely to have mothers who
were single, widowed, or divorced and who came from smaller families than did controls (Oakland
& Kane, 1973). Although none of these is inconsistent with what is known from studies of
nonIndian populations (e.g., Mullen, Martin, Anderson, Romans, & Herbison, 1996), virtually all
studies in Indian communities have been of cases known to official agencies. Thus their reported
severity as well as the significance of multi-problem families may be greater than would be found
in community surveys.
What is less clear than the risk factors for abuse are the long term sequelae. Studies in other
populations indicate that there are a number of untoward consequences, including depression
(Mullen, Martin, Anderson, Romans, & Herbison, 1996), a wide variety of other psychiatric as
well as somatic complaints (McCauley et al., 1997), and violence (Widom, 1989). There is
some disagreement about alcohol abuse, however. In a recent review of the literature and
report of a retrospective cohort study of abused and nonabused children, Widom, Ireland, and
Glynn (1995) have suggested that studies of clinical populations tend to show an association
between the experience of childhood abuse and subsequent alcohol dependence, but that in
 Childhood abuse and subsequent alcohol dependence 1081

community studies, including their own, no such association was observed. On the other hand,
Mullen, Martin, Anderson, Romans, and Herbison (1996) found that sexual abuse, but not
physical and emotional abuse, was a risk factor for heavy drinking among severely abused
women. Holmes and Robins (1988) have found that severe parental discipline in childhood was
a risk factor for subsequent alcohol abuse. McCauley and colleagues (1997) have shown that
among women childhood physical and sexual abuse are risk factors for substance (including
alcohol) misuse. The first two were community-based studies. The third was of women
enrolled in practices of general internal medicine. And in a study of a southwestern Indian
community, a history of childhood sexual abuse was found to be a risk factor for multiple
psychiatric problems in adulthood, including alcohol dependence and abuse (Robin, Chester,
Rasmussen, Jaranson, & Goldman, 1997).
Finally, it is commonly agreed that there is a cycle of violence, that violence begets violence,
and that people who were abused as children are more likely to become abusive as adults, but
Straus (1995) has observed that, “For the most part, quantitative data demonstrating such an
association are lacking.” His own study did not show an impressive association between the
degree to which parents reported having experienced physical punishment as children and the
degree to which the were violent towards their own children. According to these data,
therefore, the link between having been abused as a child and being abusive as an adult has not
been clearly demonstrated.
Regarding being the recipient of violence, however, calculations from data published by
McCauley and colleagues (1997) from women patients in general internal medicine practices
indicate that of the 400 who had experienced some form of childhood abuse, 196 (49%) also
experienced abuse in adulthood, whereas of 1,460 who did not experience childhood abuse, 203
(13.9%) experienced abuse in adulthood. This more than 3-fold difference suggests that abuse in
childhood is a risk factor for being the victim of abuse in adulthood.
The associations between abuse in childhood and being the giver or recipient of abuse in
adulthood are complicated by the association between alcohol misuse and violence (Martin, 1993).
Kantor and Straus (1995) found that 76% of cases of family violence did not involve the use of
alcohol. They did find, however, that for people who drank a great deal or drank in binges, a much
higher proportion of cases of family violence involved alcohol than among more moderate drinkers.
Using a subset of data from Indian informants from the same national survey used by Kantor and
Straus (1995), Bachman (1992) found a similar relationship.
Because there is evidence that abuse in childhood is associated with subsequent alcohol
misuse, and because both abuse in childhood and alcohol misuse are thought to be associated
with domestic violence, it is worth exploring further the associations among them. Moreover
because many of the risk factors for conduct disorder are the same as those for childhood
abuse, most notably unstable families, low socioeconomic status, and alcohol abuse by parents
and other caretakers (Offord, Alder, & Boyle, 1986; McGaha & Leoni, 1995; Salzinger,
Feldman, Hammer, & Rosario, 1991; Velleman, 1992a, 1992b); and because the adult sequelae
of childhood abuse and conduct disorder are also similar and in some studies include alcohol
abuse and domestic violence, it is important to examine and if possible disentangle the
associations between abuse and conduct disorder and their relative contributions to subsequent
alcohol dependence and domestic violence.
Elsewhere we have shown that conduct disorder before age 15 is a risk factor for alcohol
dependence among Navajo Indian men and women (Kunitz et al., in press). In this paper we extend
that analysis to consider: (1) the degree to which both physical and sexual abuse are risk factors for
conduct disorder; and (2) the joint effects of abuse and conduct disorder before age 15 on
subsequent alcohol dependence and domestic violence.
1082 S. J. Kunitz, J. E. Levy, J. McCloskey, and K. R. Gabriel

METHODS

The study uses a case control design, the details of which have been described elsewhere (Kunitz
et al., in press). It suffices to say here that the study was designed to examine the association of
conduct disorder before age 15 with subsequent alcohol dependence. The field work was done in
1993–1995 on the Navajo Reservation in northwestern New Mexico and northern Arizona. Alcohol
dependent cases resident is either one of two of the eight Indian Health Service (IHS) catchment
areas were identified in both in-patient and out-patient treatment programs. There were 204 male
and 148 female cases (CAS) between the ages of 21 and 65.
Controls were matched to the cases by age, sex, and community of residence. They were
randomly selected from lists of the service unit populations known to the IHS because they had
been seen at least once in an IHS facility within the previous 10 years (1982–1992). Because the
IHS is the major provider of health services to Navajos living on and adjacent to the reservation,
virtually all individuals are included in the data system, whether in-patient or out-patient. Thus
there is unlikely to be any major bias introduced by this method of sampling. For example, in a
previous study using this same technique, we identified more individuals 65 and above than had
been enumerated in the decennial census (Kunitz & Levy, 1991).
For each case, potential controls were interviewed until a nonalcohol dependent informant was
identified. It was not always possible to identify such an individual. In the end, therefore, there were
two groups of male and two groups of female controls. There were 374 alcohol dependent (DEP)
and 157 nonalcohol dependent (NADC) male controls, and 60 alcohol dependent (DEP) and 143
nonalcohol dependent (NADC) female controls.
Cases, and therefore their matched controls, came from different types of communities and
ranged in age from 21– 65. The communities were classified as Agency towns (service centers on
the reservation), nonAgency town reservation communities, and border towns. Drinking patterns in
these various types of communities, as well as by age and sex, had been shown in previous research
to differ considerably (Levy & Kunitz, 1974). To adjust for any possible biases introduced by these
sampling strata, a 12-fold stratification variable was created combining age (,50, 501), gender,
and community of residence (border town, agency town, and other reservation community), which
is used in the regression analyses. On the other hand, as shown elsewhere, within sampling strata
the controls are unbiased with regard to alcohol dependence or any variable associated with it, and
they represent a reasonable sample of the adult Navajo population when suitable adjustments are
made for stratification (Kunitz et al., in press).
The interview instrument was the Diagnostic Interview Schedule (DIS) modified for the Revised
Third Edition of the Diagnostic and Statistical Manual of the American Psychiatric Association
(DSM-III-R) (American Psychiatric Association, 1987). The DIS was developed for the Epidemi-
ological Catchment Area Study (Regier & Robins, 1991) and is designed so that lay interviewers
can collect data suitable for making psychiatric diagnoses. The series of questions we used were
those for the diagnosis of conduct disorder and alcohol dependence. The conduct disorder questions
are 15 in number and are scored Yes or No. According to DSM-III-R, more positive responses
indicate greater severity of conduct disorder. A log transformation of the total number of affirma-
tive replies (logASYES) is used as a measure of the degree of severity of conduct disorder
problems. When treated as a dichotomous variable, a minimum of three criteria are required to
make the diagnosis of conduct disorder. Only behaviors reported as occurring before age 15 are
included in this study because the purpose was to examine the degree to which such early behavior
was predictive of subsequent alcohol dependence.
The DIS also includes a series of questions to elicit information diagnostic of alcohol depen-
dence, again to match the criteria of DSM-III-R. As in the case of conduct disorder, more
affirmative responses to the relevant DIS questions indicate greater severity. The scale is called
ALCUMSAB, and a square root transformation is used. The diagnosis of alcohol dependence was
 Childhood abuse and subsequent alcohol dependence 1083

confirmed in all the cases, but the greatest utility of the scale was to distinguish between alcohol
dependent (DEP) and nonalcohol dependent (NADC) controls.
The DIS has not previously been used in the Navajo population and questions of validity may
thus arise. We have discussed this issue in more detail elsewhere (Kunitz et al., in press). It suffices
here to say that the purpose of using both interval and nominal variables from the DIS (i.e.,
logASYES and conduct disorder yes/no, and ALCSUMAB and alcohol dependence yes/no) was to
avoid assuming that the cut-off points for diagnoses of conduct disorder and alcohol dependence
are the same among Navajos and nonNavajos. The interval variables are simply the total number
of affirmative responses to questions about particular behaviors and thus avoid the problem of
labeling those behaviors.
The questions concerning abuse asked whether the individual had experienced sexual and
physical abuse, and if so, at what age? These questions were not part of a standardized instrument.
Informants were asked if they had ever been abused sexually and/or physically and then were asked
to describe what had occurred. Examples were often given if the informant was uncertain what was
meant. Unlike the questions regarding alcohol use and childhood misbehaviors consistent with
conduct disorder, abuse has not been a subject of widespread and open discussion among Navajos
any more than it has been among other societies until recently. Because of the sensitivity of the
topic and the exploratory nature of the investigation, we believed that an open-ended approach was
most appropriate and would provide more valid data than a standardized instrument.
Our respondents generally described physical abuse as a severe thrashing including among other
things parental beating with fists, whipping with a belt or bailing wire, and twisting of ears or arms,
but it also included being forced by parents or others to fight other children of the same age with
the threat of a beating if the informant refused, being locked up, and being burned with cigarette
butts. Sexual abuse was defined as inappropriate touching and fondling as well as actual physical
penetration. The reported episodes ranged from fondling by an older relative or acquaintance to
homosexual gang rape.
With regard to domestic violence, respondents were asked if they had ever struck or been struck
by their partners, if they had struck first, if it had occurred more than once, if alcohol was involved
in any of the episodes, and if it was, who was drinking.
Interviewing occurred between May 1993 and September 1995. Interviews ranged in length from
2 to 4 hours. Interviewees were requested to sign a consent form which had been approved by both
the University of Rochester’s IRB and the IRB comprised of representatives of both the Navajo
Tribe and the IHS. A Certificate of Confidentiality had been obtained to protect informants should
they have reported any illegal activities. At the end of the interview, each informant was paid $30.

RESULTS

In the first part of the analysis, we consider only the controls (NADC1DEP). We describe the
prevalence of both sexual and physical abuse before age 15 in each of the 12 strata. Then we
examine both forms of abuse as risk factors for conduct disorder; then abuse and conduct disorder
as risk factors for alcohol dependence; and then alcohol dependence, abuse, and conduct disorder
as risk factors for domestic violence. In the second part of the analysis we use all three samples
(CAS, DEP, and NADC) to consider the importance of abuse and conduct disorder in family
violence given different levels of alcohol use.

The Prevalence of Abuse

Table 1 displays the proportions of controls within each sampling stratum who experienced
physical and/or sexual abuse before age 15. The differences in proportions who were physically
1084 S. J. Kunitz, J. E. Levy, J. McCloskey, and K. R. Gabriel

Table 1. Proportions Who Experienced Abuse Before 15 Within Strata (%)

Physical Abuse ,15 Sexual Abuse ,15

Strata Prop. S.E. Prop. S.E.

Female $50, Agency Town 14.3 too few 0 -

Female $50, Border Town 0 - 0 -
Female $50, Reservation 16.7 11.2 0 -
Female ,50, Agency Town 13.9 3.7 10.7 3.4
Female ,50, Border Town 17.9 7.3 25.9 8.6
Female ,50, Reservation 13.9 4.3 13.8 4.3
Male $50, Agency Town 0 - 5.8 5.8
Male $50, Border Town 0 - 0 -
Male $50, Reservation 11.5 4.1 1.6 1.6
Male ,50, Agency Town 10.6 2.6 4.5 1.5
Male ,50, Border Town 21.1 4.9 1.4 1.4
Male ,50, Reservation 11.4 2.1 2.2 1.0

abused are not significant among strata, so the overall prevalence reflects that of the population.
Overall, 12.7% of the sample (NADC1DEP) experienced physical abuse. There are, however,
significant differences among the strata with regard to sexual abuse, accounted for by the difference
between men and women. Among men, 2.4% and among 12.7% experienced sexual abuse before
age 15.

Abuse as a Risk Factor for Conduct Disorder

In Table 2 the results of a regression, given stratification, of logASYES onto physical and sexual
abuse are displayed. (The results of logist regressions of conduct disorder/no conduct disorder are
virtually identical to those using logASYES.) In analyses which have not been displayed sex was
included as an independent variable because conduct disorder is less frequent among females than
males. The interaction terms were not significant and are omitted. The reduction in logASYES (and
in conduct disorder) from males to females is constant, so the effects of physical and sexual abuse
can be studied by themselves. The effects of the two types of abuse are much the same.

Abuse and Conduct Disorder as Risk Factors for Alcohol Dependence

Table 3 displays two regressions of the square root of ALCUMSAB onto both forms of abuse
and logASYES, given stratification. (The results are the same in a logistic regression using DEP vs.
NADC as the dependent variable.) The effect of sex is the same as in the previous analysis and is
therefore included in the stratification variable (the same is true in subsequent analyses).
Physical abuse has a significant effect but sexual abuse does not and can be omitted from the
model. In the final model, neither physical abuse nor logASYES can substitute for the other as a
predictor of the square root of ALCUMSAB (or the probability of being DEP). In other words,
physical abuse has an effect beyond what is measured by logASYES.

Table 2. Regression of logASYES Onto Physical Abuse <15 and Sexual Abuse <15, Given Stratification,
Controls Only
Risk factors Estimate S.E. p-value

Physical Abuse ,15 0.141 0.033 ,0.0001

Sexual Abuse ,15 0.159 0.074 0.0015
 Childhood abuse and subsequent alcohol dependence 1085

Table 3. Regression of Square Root of ALCUMSAB Onto Physical Abuse <15, Sexual Abuse <15, and logASYES,
Given Stratification, Controls Only
Model 1 Model 2 Model 3

Risk Factors Est. S.E. P Est. S.E. P Est. S.E. P

Physical Abuse ,15 0.238 0.080 0.0021 0.145 0.078 0.0640 0.156 0.075 0.0385
Sexual Abuse ,15 0.206 0.121 0.1409 0.100 0.118 0.3941 - - -
logASYES - - - 0.661 0.088 ,0.0001 0.655 0.037 ,0.0001

The Prevalence of Domestic Violence

Table 4 displays the result of a logistic regression of answers to the question whether the
informant ever struck his/her partner, onto the three stratification variables. Only age is significant.
Among informants 50 and above, 18.6 percent answered affirmatively. Among those below 50,
43.2% answered affirmatively (see Table 5).
Table 4 also displays the results of a similar regression of answers to the question whether the
informant had ever been struck by his/her partner. Both age and sex are significant. Women are
more likely than men, and people less than 50 more likely than people 50 and above, to answer
affirmatively. The proportions answering affirmatively are displayed in Table 5.
The differences between the proportions of men less than 50 and 50 and above who admit
striking their partners (43.4 and 21.0%, respectively) and women in the same age groups who say
they were struck by their partners (52.7 and 28.6%, respectively) is about 7–10%, which may be
a rough approximation of the amount of under reporting of the commission of family violence by
the men in this population.

Abuse, Conduct Disorder, and Alcohol Dependence as Risk Factors for Family Violence.
Table 6 displays the results of two logistic regressions of whether or not the informant has struck
his/her partner. Again sex has the same effect as in previous analyses and is only included in the
stratification variable. Sexual abuse and logASYES have no effects additional to those of physical
abuse and the square root of ALCUMSAB. Physical abuse affects the probability of hitting beyond
the effect of ALCUMSAB, such as for fixed ALCUMSAB, individuals who were physically
abused have a higher probability of hitting than individuals without such abuse.
Table 7 displays similar analyses of whether informants were struck by their partners. The results
are essentially the same as those in the previous analysis. That is, sex is treated as before, sexual
abuse and logASYES are not significant, and both physical abuse and severity of alcohol use
(ALCUMSAB) are significant.
We turn next to a more detailed consideration of the associations between abuse and conduct

Table 4. Logistic Regression of Measures of Family Violence Onto Stratification Variables, Controls Only
Struck Partner Struck by Partner
Stratification
Variables Est. S.E. P Est. S.E. P

Community type:
a. Agency–Res. 20.048 0.121 0.6876 0.208 0.122 0.0887
b. Border–Res. 0.235 0.148 0.1130 0.013 0.152 0.9296
Age ($50–,50) 20.582 0.136 ,0.0001 20.491 0.134 0.0003
Sex (female–male) 20.054 0.092 0.5552 0.283 0.091 0.0018

Interaction terms are not significant.

1086 S. J. Kunitz, J. E. Levy, J. McCloskey, and K. R. Gabriel

Table 5. Proportions of Men and Women Who Struck, and Who Had Been Struck By, Their Partners, Controls Only
Struck Partner (%) Been Struck by Partner

Age Men Women Men Women

Below 50 43.4 42.5 37.5 52.7

50 and Above 21.0 9.5 17.3 28.6

disorder on the one hand and family violence on the other, given level of alcohol abuse. For the
following analyses all three samples, not only the controls are used. Table 8 displays the
proportions who have been involved in family violence by sex and sample. It is clear that there is
a progression of prevalence from NADC to DEP to CAS, and that at every level of alcohol use
women are less likely than men to have escaped involvement in violence.
The question is: given level of alcohol use/abuse, do conduct disorder and the experience of
physical and sexual abuse before age 15 add to the ability to predict involvement in domestic
violence? The answer is to be found in Table 9, which displays the results of logistic regressions
of the same dependent variables that were used in Tables 6 and 7: whether informants struck and/or
were struck by their partners.
Within each sample, a history of physical abuse has a significant positive effect on both hitting
and being hit. The effect is particularly significant among NADC. Sexual abuse has an unclear
effect. It may have a small negative effect, given the level of physical abuse. It does not appear to
play a role by itself. Conduct disorder is only marginally significant in one analysis, among CAS
who struck their partners.

DISCUSSION

The major limitations of the data reported in this paper have to do with recall and reporting
biases of abuse in childhood. Reporting may be biased if people were unwilling to discuss or unable
to remember episodes of childhood abuse. Assuming under reporting occurred, it would result in
an underestimation of the prevalence of abuse, but it would not necessarily distort the analyses of
abuse as a risk factor unless underreporting was selective. That is, if NADC were more likely to
underreport abuse than DEP or CAS, a spurious association between alcohol dependence and abuse
would result. This could happen if people who had been in treatment had been encouraged or even
taught to recall episodes of childhood abuse.
To check this potential treatment bias, the histories of abuse and conduct disorder were compared
between DEPs who had been in treatment programs and those who had not. For neither men nor
women were there any significant differences. Thus we can find no effect of treatment bias,

Table 6. Regression of Has Struck Partner Onto Abuse, logASYES, and Square Root ALCUMSAB,
Given Stratification, Controls Only
Model 1 Model 2

Risk Factors Est. S.E. P Est. S.E. P

Physical Abuse ,15 0.441 0.140 0.0016 0.493 0.134 0.0002

Sexual Abuse ,15 20.0094 0.204 0.6463 - - -
logASYES 0.268 0.163 0.1082 - - -
Square Root ALCUMSAB 0.466 0.073 ,0.0001 0.485 0.071 ,0.0001
 Childhood abuse and subsequent alcohol dependence 1087

Table 7. Regression of Has Been Struck By Partner Onto Abuse, logASYES, and Square Root ALSUMSAB,
Given Stratification, Controls Only
Model 1 Model 2

Risk Factors Est. S.E. P Est. S.E. P

Physical Abuse ,15 0.457 0.134 0.0007 0.460 0.131 0.0004

Sexual Abuse ,15 20.323 0.200 0.1068 - - -
logASYES 0.195 0.162 0.2280 - - -
Square Root ALCUMSAB 0.295 0.070 ,0.0001 0.299 0.068 ,0.0001

although we think it likely that there has been underreporting across all samples, especially for men
who experience sexual abuse.
Questions of the transcultural validity of the DIS may be raised. We are unaware of any other
studies among Navajos that have used the conduct disorder or alcohol dependence scales. To avoid
assuming that these diagnostic categories are the same across all populations, we have done the
analyses using both dichotomous nominal variables (Conduct Disorder and Alcohol Dependence)
and continuous variables (logASYES and ALCSUMAB), the latter being counts of the number of
relevant behaviors in which people say they have engaged. The two different types of analyses give
identical results. Thus, the diagnoses of conduct disorder and alcohol dependence seem to be as
relevant to this population as they are to the population in which the DIS was originally used.
The results indicate the following. First, the prevalence of abuse before age 15 is within limits
described in other populations. With regard to sexual abuse, Finkelhor has written that, “Prevalence
estimates in community surveys range from 6 to 62% for females and 3 to 16% for males with the
better studies generally finding higher rates. . . . A rough expectation that at least 1 in 4 girls, and
1 in 10 boys will suffer victimization, gives the order of magnitude that professionals ought to
expect” (Finkelhor, 1993; see also Siegel, Sorenson, Golding, Burnam, & Stein, 1987). The
prevalence reported by the people in this study were 2.4% among men and 12.7% among women.
It has proven more difficult to find comparable figures for physical abuse, but the historical rates
reported in this study appear to be within the range from national studies. In the Second National
Family Violence Survey in 1985, Minor Violence against children was 619 per 1,000. Severe
Violence was 110 per 1,000 (Wolfner & Gelles, 1993). In other words, in the survey year 62% of
children experienced minor violence and 11% experienced major violence. This compares with a
lifetime prevalence of 12.7% reported by Navajo adults.

Table 8. Proportions of Men and Women Who Have Struck and/or Been Struck by Their Partners, By Sample (%)
CAS DEP NADC X2 p-value

A. Men
Both struck and was struck 39.9 28.0 14.4
Struck partner 11.3 18.4 6.4
Was struck 13.7 9.7 9.6
Neither 35.1 49.9 69.6
N 168 321 125 48.032 0.0000

B. Women
Both struck and was struck 57.6 44.8 21.5
Struck partner 6.5 12.1 9.2
Was struck 25.2 22.4 20.8
Neither 10.8 20.7 48.5
N 139 58 130 59.535 0.0000

Unknown and not applicables (never had a partner) are excluded.

1088 S. J. Kunitz, J. E. Levy, J. McCloskey, and K. R. Gabriel

Table 9. Regressions of Having Struck Partner and Having Been Struck By Partner Onto Abuse and logASYES,
Given Stratification, Within Each Sample
Has Struck Partner Has Been Struck by Partner

Sample Est. S.E. P Est. S.E. P

NADC
Physical abuse ,15 0.748 0.261 0.0041 0.859 0.268 0.0013
Sexual abuse ,15 20.098 0.340 0.7737 20.510 0.347 0.1419
logASYES 0.592 0.303 0.9504 0.320 0.288 0.2659
DEP
Physical abuse ,15 0.331 0.161 0.0395 0.381 0.163 0.0195
Sexual abuse ,15 20.205 0.258 0.4268 20.576 0.297 0.0520
logASYES 0.270 0.191 0.1565 0.178 0.199 0.3705
CAS
Physical abuse ,15 0.362 0.166 0.0287 0.380 0.182 0.0362
Sexual abuse ,15 20.476 0.219 0.0294 0.078 0.284 0.7842
logASYES 0.426 0.212 0.0443 0.131 0.230 0.5677

The analyses of abuse by sampling strata also indicate that there is no difference in prevalence
among those 50 and above and less than 50. Contrary to what has been assumed by other writers
(Hauswald, 1987), this suggests that there has not been an increase in the abuse of children over
the past several generations.
Second, physical and sexual abuse are risk factors for conduct disorder. This is not surprising and
is congruent with the results of studies cited in the Introduction.
Third, both physical abuse and conduct disorder are risk factors for alcohol dependence. The
analysis of the association between conduct disorder and alcohol dependence is described in detail
elsewhere (Kunitz et al., in press). In the present context what this finding suggests is that people
who were abused had a higher probability of becoming alcohol dependent than those people who
were not abused, whether or not they also manifested the behavior characteristic of conduct
disorder. Again this is congruent with much, but not all, that has been published previously
(Widom, Ireland, & Glynn, 1995). Nonetheless, conduct disorder, whatever its source, is a more
powerful predictor of alcohol dependence than is abuse.
Fourth, family violence appears to be more common in this population than in many others that
have been surveyed. The proportion of women who say they have been struck at least once is
28.6% among women 50 and above and 52.7% among women below age 50, compared to a range
of 9 –30% in other populations (Wilt & Olson, 1996). The fact that a substantially higher proportion
of younger than older women acknowledges being struck suggests that there may have been an
increase in recent decades, although differential survival and reporting biases cannot be ruled out.
The high rates are consistent with the high rates of homicides of Navajo and other Southwestern
Native American women resulting from domestic disputes (Levy, Kunitz, & Everett, 1969;
Arbuckle, Olson, Howard, Brillman, Anctil, & Sklar, 1996).
Fifth, alcohol dependence is a risk factor for family violence, and the more severe the
alcoholism, the more likely is violence to have occurred. This, too, is similar to observations made
elsewhere and reported in the Introduction, and it means that people who become severely alcohol
dependent with or without a history of childhood abuse are likely to participate in domestic
violence. It leaves unresolved, however, the vexed question of the means by which alcohol abuse
“causes” domestic violence: whether by lowering inhibitions, by stimulating aggression, as a
reflection of learned behavior and what is culturally acceptable, or some combination of all of them
(MacAndrew & Edgerton, 1969; Martin, 1993; Pernanen, 1991). Even in the presence of alcohol
dependence, however, a history of physical abuse is a significant predictor of being both a
perpetrator and a victim of family violence. This indicates that when both alcohol dependence and
 Childhood abuse and subsequent alcohol dependence 1089

a history of physical abuse in childhood are present, the likelihood of engaging in family violence
as an adult are greatly enhanced. And finally, these results demonstrate that in the absence of
alcohol dependence, a history of physical abuse is a particularly significant predictor of being both
a perpetrator and a victim of domestic violence.

Acknowledgement—The authors are grateful to Chena DuPuy, Eric Henderson, Wilson Howard, Kathleen Lampert, Gilbert
Quintero, Scott Russell, and Alan Vince; to the boards and CEOs of the Indian Health Service hospitals in Tuba City,
Arizona and Shiprock, New Mexico; and to the directors and staffs of the Four Winds Addiction Treatment Program in
Farmington, New Mexico and the Rehoboth-Christian Medical Services in Gallup, New Mexico.

REFERENCES

Arbuckle, J., Olson, L., Howard, M., Brillman, J., Anctil, C., & Sklar, D. (1996). Safe at home? Domestic violence and other
homicides among women in New Mexico. Annals of Emergency Medicine, 27, 210 –215.
American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed. rev.). Washington,
DC: American Psychiatric Association.
Bachman, R. (1992). Death and violence on the reservation: Homicide, family violence, and suicide in American Indian
populations. New York: Auburn House.
DeBruyn, L. M., Lujan, C. C., & May, P. A. (1992). A comparative study of abused and neglected American Indian children
in the Southwest. Social Science and Medicine, 35, 305–315.
Finkelhor, D. (1993). Epidemiological factors in the clinical identification of child sexual abuse. Child Abuse & Neglect,
17, 67–70.
Fischler, R. (1985). Child abuse and neglect in American Indian communities. Child Abuse & Neglect, 9, 95–106.
Hauswald, L. (1987). External pressure/internal change: Child neglect on the Navajo Reservation. In N. Scheper-Hughes
(Ed.), Child survival: Anthropological perspectives on the treatment and maltreatment of children (pp. 145–164).
Dordrecht, The Netherlands: D. Reidel Publishing Company.
Holmes, S. J., & Robins, L. N. (1988). The role of parental disciplinary practices in the development of depression and
alcoholism. Psychiatry, 51, 24 –36.
Jones, D. M. (1969). Child welfare problems in an Alaskan native village. Social Service Review, 43, 297–309.
Kantor, G. K., & Straus, M. A. (1995). The “drunken bum” theory of wife beating. In M. A. Straus, & R. J. Gelles, (Eds.),
Physical violence in American families (pp. 203–224). New Brunswick, NJ: Transaction Publishers.
Kunitz, S. J., & Levy, J. E. (1991). Navajo aging: The transition from family to institutional support. Tucson, AZ: The
University of Arizona Press.
Kunitz, S. J., & Levy, J. E. (1994). Drinking careers: A 25 year followup of three Navajo populations. New Haven, CT:
Yale University Press.
Kunitz, S. J., Levy, J. E., Gabriel, K. R., Henderson, E., Lampert, K., McCloskey, J., Quintero, G., Russell, S., & Vince,
A. (in press). Alcohol dependence and conduct disorder among Navajo Indians. Journal of Studies on Alcohol.
Levy, J. E., & Kunitz, S. J. (1974). Indian drinking: Navajo practices and Anglo-American theories. New York: John Wiley
and Sons.
Levy, J. E., Kunitz, S. J., & Everett, M. (1969). Navajo criminal homicide. Southwestern Journal of Anthropology, 25,
124 –152.
Lujan, C., DeBruyn, L. M., May, P. A., & Bird, M. E. (1989). Profile of abused and neglected American Indian children
in the Southwest. Child Abuse & Neglect, 13, 449 – 461.
MacAndrew, C., & Edgerton, R. (1969). Drunken comportment. Chicago, IL: Aldine.
Martin, S. (Ed.) (1993). Alcohol and interpersonal violence: Fostering multidisciplinary perspectives (NIH Publication No.
93-3496). Rockville, MD: US Department of Health and Human Services, Public Health Service, National Institutes of
Health, National Institute on Alcohol Abuse and Alcoholism.
McCauley, J., Kern, D. E., Kolodner, K., Dill, L., Schroeder, A. F., DeChant, H. K., Ryden, J., Derogatis, L. R., & Bass,
E. B. (1997). Clinical characteristics of women with a history of childhood abuse: Unhealed wounds. Journal of the
American Medical Association, 277, 1362–1368.
McGaha, J. E., & Leoni, E. L. (1995). Family violence, abuse, and related family issues of incarcerated delinquents with
alcoholic parents compared to those with nonalcoholic parents. Adolescence, 30, 473– 482.
Mullen, P. E., Martin, J. L., Anderson, J. C., Romans, S. E., & Herbison, G. P. (1996). The long-term impact of the physical,
emotional and sexual abuse of children: A community study. Child Abuse & Neglect, 20, 7–22.
Northern Navajo Medical Center. (1996). The Navajo health status report. Shiprock, NM: US Indian Health Service.
Oakland, L., & Kane, R. L. (1973). The working mother and child neglect on the Navajo reservation. Pediatrics, 51,
849 – 853.
Offord, D. R., Alder, R. J., & Boyle, M. H. (1986). Prevalence and sociodemographic correlates of conduct disorder. The
American Journal of Social Psychiatry, 6, 272–278.
Pernanen, K. (1991). Alcohol in human violence. New York: The Guilford Press.
Piasecki, J. M., Manson, S. M., Biernoff, M. P., Hiat, A., Taylor, S. S., & Bechtold, D. W. (1989). Abuse and neglect of
1090 S. J. Kunitz, J. E. Levy, J. McCloskey, and K. R. Gabriel

American Indian children: Findings from a survey of providers. American Indian and Alaska Native Mental Health
Research, 3, 43– 62.
Regier, D. A., & Robins, L. N. (1991). Introduction. In L. N. Robins, & D. A. Regier (Eds.), Psychiatric disorders in
America: The epidemiologic catchment area study (pp. 1–10). New York: The Free Press.
Salzinger, S., Feldman, R. S., Hammer, M., & Rosario, M. (1991). Risk for physical child abuse and the personal
consequences for its victim. Criminal Justice and Behavior, 18, 47– 63.
Siegel, J. M., Sorenson, S. B., Golding, J. M., Burnam, M. A., & Stein, J. A. (1987) The prevalence of childhood sexual
assault: the Los Angeles catchment area project. American Journal of Epidemiology, 126, 1141–1153.
Straus, M. A. (1995). Ordinary violence, child abuse, and wife beating: What do they have in common? In M. A. Straus,
& R. J. Gelles (Eds.), Physical violence in American families (pp. 403– 424). New Brunswick, NJ: Transaction Publishers.
US Congress Office of Technology Assessment. (1990). Indian adolescent mental health (OTA-H-446). Washington, DC:
US Government Printing Office.
Velleman, R. (1992a). Intergenerational effects—a review of environmentally oriented studies concerning the relationship
between parental alcohol problems and family disharmony in the genesis of alcohol and other problems. I: The
intergenerational effects of alcohol problems. The International Journal of the Addictions, 27, 253–280.
Velleman, R. (1992b). Intergenerational effects—a review of environmentally oriented studies concerning the relationship
between parental alcohol problems and family disharmony in the genesis of alcohol and other problems. II: The
intergenerational effects of family disharmony. The International Journal of the Addictions, 27, 367–389.
White, R. B., & Cornely, D. A. (1981). Navajo child abuse and neglect: A comparison group examination of abuse and
neglect of Navajo children. Child Abuse & Neglect, 5, 9 –17.
Wichlacz, C., Lane, J., & Kempe, C. (1978). Indian child welfare: A community team approach to protective services. Child
Abuse & Neglect, 2, 29 –35.
Widom, C. S., Ireland, T., & Glynn, P. J. (1995). Alcohol abuse in abused and neglected children followed-up: Are they
at increased risk? Journal of Studies on Alcohol, 56, 207–217.
Widom, C. S. (1989). The cycle of violence. Science, 244, 160 –166.
Wilt, S., & Olson, S. (1996). Prevalence of domestic violence in the United States. Journal of the American Medical
Womens Association, 51, 77– 82.
Wolfner, G. D., & Gelles, R. J. (1993). A profile of violence toward children: A national study. Child Abuse & Neglect, 17,
197–212.

RÉSUMÉ

Objectif: D’étudier chez les Navahos: (1) les mauvais traitements en tant que facteur important de risque menant aux
troubles de comportement; (2) l’importance des différents types de maltraitance en tant que facteurs de risque pour
développer des troubles d’alcool et de dépendance sur la drogue; et (3) l’importance relative de chaque type de mauvais
traitement, de désordre de comportement et de dependance sur l’alcool par rapport à la probabilité de devenir un agresseur
ou une victime de violence entre conjoints.
Méthode: Cette recherche s’est basée sur une étude de cas avec groupes contrǒles. On a interviewé 204 hommes et 148
femmes ǎgés de 21 à 65 ans, appareillés à deux groupes contrǒles: l’un composé de 304 hommes et 60 femmes souffrant
de dépendance sur l’alcool et un groupe contrǒle de 157 hommes et 143 femmes sans problèmes de dépendance. Prenant
en considération l’ǎge, le lieu de résidence et le sexe, les deux groupes contrǒles dans l’ensemble constituent un échantillon
représentatif d’hommes et de femmes navahos dans une population ǎgée de 21 à 65 ans.
Résultats: La prévalence des mauvais traitements physiques et sexuels avant l’ǎge de 15 ans ressemble à la population en
général. Chaque type de mauvais traitement représente un facteur de risque pour les désordres de comportement. Les
mauvais traitements physiques sont un facteur pour la dépendance sur l’alcool. Pour ce qui est de la violence entre conjoints,
les mauvais traitements physiques et la dépendance sur l’alcool sont des facteurs de risque qui sont indépendants, tant pour
l’agresseur que pour la victime. Il ne semble pas y avoir d’augmentation des cas de mauvais traitements d’une génération
à l’autre mais il semblerait que les cas de maltraitance entre conjoints soient plus fréquents.
Conclusions: Les mauvais traitements physiques sont un facteur important de risque menant à la dépendance sur l’alcool
et aussi à la violence entre conjoints indépendamment des effets de l’abus de l’alcool. Les conséquences des mauvais
traitements sexuels par rapport à la violence entre conjoints et la dépendance sur l’alcool ne semblent pas e˛ des facteurs
importants.

RESUMEN

Objetivo: Examinar en la población “Navajo” (1) la importancia del maltrato infantil como factor de riesgo para los
trastornos de conducta, (2) la importancia de cada forma de maltrato y de los trastornos de conducta como factores de riesgo
para la dependencia del alcohol, y (3) la importancia relativa de cada forma de maltrato, trastornos de conducta y
dependencia del alcohol como factores de riesgo para ser perpetrador y/o vı́ctima de violencia doméstica.
Método: El estudio se basa en un diseño caso-control. Los casos (204 varones y 148 mujeres) con edades comprendidas
entre 21 y 65 años fueron entrevistados en un programa de tratamiento del alcoholismo y fueron emparejados con el grupo
control formado por sujetos de la comunidad. Se formaron dos grupos de sujetos control: (1) sujetos dependientes del
 Childhood abuse and subsequent alcohol dependence 1091

alcohol (374 varones y 60 mujeres). Cuando todos los sujetos del grupo control quedaron ajustados de manera estratificada
por edad, barrie de recidonoia, y sexo se observó que constituı́an una muestra representativa de la población de mujeres y
hombres de etnia “Navajo” con edades comprendidas entre 21 y 65 años.
Resultados: La prevalencia de maltrato fı́sico y abuso sexual ocurrido antes de los 15 años se encuentra dentro de los lı́mites
observados en otras poblaciones. Cada forma de maltrato es un factor de riesgo para los trastornos de conducta. El maltrato
fı́sico, junto con los trastornos de conducta es un factor de riesgo para la dependencia del alcohol. El maltrato fı́sico y la
dependencia del alcohol son de riesgo independientes para la implicación en violencia doméstica tanto como vı́ctima como
perpetrador. Parece que no se observa una tendencia antigua a la incidencia del maltrato infantil a lo largo de las
generaciones pasadas, pero hay una cierta evidencia de que la violencia doméstica está haciéndose más frecuente.
Conclusión: El maltrato fı́sico es un factor de riesgo significativo para la dependencia del alcohol ası́ como para la violencia
doméstica independientemente de los efectos del abuso del alcohol. Los efectos del abuso sexual en relación tanto con la
violencia doméstica como con la dependencia del alcohol no parecen ser significativos.