Content uploaded by Ruth Ogden Halstead
Author content
All content in this area was uploaded by Ruth Ogden Halstead on Aug 24, 2018
Content may be subject to copyright.
FAMILY DYNAMICS AND SELF-INJURY BEHAVIORS:
A CORRELATION ANALYSIS
Ruth Ogden Halstead, Thomas W. Pavkov and Lorna L. Hecker
Purdue University Calumet
Michelle M. Seliner
S.A.F.E. ALTERNATIVES
®
This study tested the relationship between family dynamics and self-injury. A total of 189
participants responded to a web-based survey collecting information related to previous self-
injury behaviors and family dynamics. Participants were over 18 years old who had used
self-injury (intentionally harming themselves physically to relieve painful emotions without
suicidal intent), but who had not used self-injury for over a year. Results indicated that
healthy family dynamics were negatively correlated and associated with higher scores of self-
injury behaviors. This study offers some evidence that family dynamics influence self-injury
behaviors. The implications for family therapy are discussed.
Many factors influence the prevalence of self-injury, including societal, gender, familial, and
intrapersonal factors (Gratz, Conrad & Roemer, 2002; Selekman, 2002). Among the familial fac-
tors, clinicians report that self-injury is a symptom of family stress and lack of sound family coping
mechanisms (Conterio & Lader, 1998; Levenkron, 1998; Selekman, 2002; Walsh, 2006). However,
more research is needed to verify clinicians’ observations that families contribute to the mainte-
nance and cessation of self-injury (Eckenrode, 2006; Gratz et al., 2002; Jella, 2007; Selekman,
2002; Whitlock, 2006; Yip, Ngan & Lam, 2003). This study offers results that confirm that family
dynamics influence self-injury behaviors.
Self-injury occurs in all socioeconomic levels, in all cultures, and in all races (Favazza, 1996).
Briere and Gil (1998) estimated that 4% of the general adult population in the United States has
used self-injury to cope. Research indicates that about 20% of college students (Whitlock, Powers
& Eckenrode, 2006) and 17% to 20% of high school students have used self-injury (Lader, 2006).
Among the adolescent psychiatric inpatient samples, the percentage is as high as 40% to 61%
(Nock & Prinstein, 2004, 2005).
Self-injury is largely believed to be more prevalent among adolescent girls than in any other
population (Conterio & Lader, 1998; Favazza, 1998; Levenkron, 1998; Selekman, 2002). However,
Lader (2006) theorized that men probably self-injure in equal numbers, but they usually do not
seek treatment. Gratz et al. (2002) report no difference in the use of self-injury between male and
female college students. In addition, there is a high population among adult men in prison that
self-injure (Alderman, 1997; Favazza, 1996; Walsh, 2006).
Ruth Ogden Halstead, MS, LMFTA, gathered data for this study as a graduate student in the Marriage and
Family Therapy Program at Purdue University Calumet. Ruth Ogden Halstead has since graduated with her MS in
marriage and family therapy and is working as a family therapist in Lake County, Indiana.
Thomas W. Pavkov, PhD, Professor at Purdue University Calumet and Director of the Institute for Social and
Policy Research.
Lorna L. Hecker, PhD, Professor of the Marriage and Family Therapy Program at Purdue University Calumet.
Michelle M. Seliner, MSW, LCSW, is COO responsible for clinical and administrative operations of S.A.F.E.
ALTERNATIVES
®
(Self-Abuse Finally Ends) located in St. Louis, Missouri. Programs include S.A.F.E. Intensive
®
,
S.A.F.E. Expressions
®
and S.A.F.E. Choice
®
. www.selfinjury.com 800-DONTCUT
®
.
This study was presented at the INAMFT state conference, April 2009 and the AAMFT national conference,
October 2009.
Address correspondence to Ruth Ogden Halstead, the Institute for Social and Policy Research, 1247, 169th
Street, Hammond, Indiana 46322; E-mail: ruthogden@aol.com
246 JOURNAL OF MARITAL AND FAMILY THERAPY April 2014
Journal of Marital and Family Therapy
doi: 10.1111/j.1752-0606.2012.00336.x
April 2014, Vol. 40, No. 2, 246–259
Self-injury Behaviors
Definitions. “Self-injury,” “self-harm,” “deliberate self-harm,” “self-inflicted violence,” “self-
punishment,” “self-sabotage,” “self-injurious behavior,” and “self-mutilation” have been used
interchangeably in the literature. “Self-injury” is the most common term used in recent literature;
however, the term “non-suicidal self-injury” (NSSI) is also used to distinguish self-injury from sui-
cide attempts (Whitlock, 2006). In this study, the term “self-injury” is used. Favazza and
Rosenthal (1993) defined “self-mutilation” as the deliberate alteration or destruction of body tissue
without conscious suicidal intent. Selekman (2002) defined “self-harm” as hurting oneself to relieve
emotional pain or distress. The definition used in this study has been adapted from what
Levenkron (1998, p. 25), and Favazza and Rosenthal (1993, p. xvii–xix) proposed as the criteria
should self-injury be added to the DSM-V as a diagnosis category (It is currently under consider-
ation [see www.DSM5.org]). The definition specifically includes:
•Preoccupation with harming oneself physically.
•Recurrent failure to resist impulses to harm oneself physically, resulting in the destruction
or alteration of one’s body tissue.
•A sense of tension present immediately before the act is committed.
•Relaxation, gratification, pleasant feelings, and/or numbness experienced with the physical
pain.
•Self-injury that is not associated with conscious suicidal intent or serious mental retardation.
Types of self-injury. Favazza (1996) lists two types of self-injury. One type is the culturally
sanctioned (i.e., tattoos and body piercing). The second type is deviant (i.e., the product of a men-
tal disorder or anguish). This study is about the deviant type. Favazza categorizes deviant self-
injury into three levels of severity: mild, moderate, and severe. Mild forms of self-injury may
include typical fingernail biting, scratching, or picking at scabs. Moderate forms of self-injury also
include cutting the skin surface or burning the skin with a cigarette which may leave scars. Severe
forms of self-harm involve breaking bones, eye enucleation (removal), or amputation of limbs.
According to Holmes (2000), the most common forms of self-injury are the following: cutting,
72%; burning, 35%; hitting and head banging, 30% (this percentage may include those with seri-
ous mental retardation); hair pulling, 10%; scratching and interfering with wounds healing, 22%.
Conterio and Lader (1998) also list biting/chewing the inside of the mouth and ingestion of chemi-
cals. Conterio and Lader state that the methods of self-harm vary widely across individuals.
Frequency and periodicity. Favazza (1998) noted that some of his clients used self-injury repeat-
edly and others used self-injury episodically. Episodic self-injury use typically involves an isolated
number of occurrences throughout the individual’s life, while repetitive use is more habitual and is
commonly associated with addiction. Favazza mentioned that episodic users of self-injury often
become repetitive users. Gratz (2001) found that individuals averaged a frequency of 150 self-injuries.
Muehlenkamp (2005) reports individuals using self-injury an average of 50 times with some reporting
as many as 400 times. Ross and Heath (2002) found that among high school students, 13% of the stu-
dents who were using self-harm were doing this on a daily basis, 27% had self-injured on a weekly
basis, 19% had self-injured a couple of times a month, and 20% percent had self-injured only once.
Severity. Levenkron (1998) first noted a correlation between the severity of tissue damage
inflicted by self-injury and the amount of trauma individuals have endured. Juzwin’s (2004) devel-
oped the Self-Injury Self-Report Inventory (SISRI) which assesses severity of self-injury by
whether an individual received medical treatment or should have received medical treatment. Juz-
win also looked at the kind of scaring or other physical damage resulting from self-inflicted
wounds. Gratz (2001) suggested that frequency and duration of self-harm should be calculated into
severity measures. This study asked participants about the type, frequency, duration, and medical
history of their self-injury behaviors.
Family Dynamics
Family dynamics are the patterns of interactions among two or more family members. These
patterns of interaction often span generations. Olson and colleagues (Olson, Gorall & Tiesel, 2006;
Olson, Sprenkle & Russell, 1979) categorized family dynamics into two continuums of family
“cohesion” and family “flexibility.” Cohesion is defined as “the emotional bonding that family
April 2014 JOURNAL OF MARITAL AND FAMILY THERAPY 247
members have toward one another” (Kaslow, 1996, p. 60). Too little cohesion creates family inter-
action and relationships that are considered “disengaged,” while too much cohesion creates family
interactions and relationships that are considered “enmeshed.” When there is a healthy amount of
family emotional bonding among family members, cohesion is considered “balanced” on the cohe-
sion continuum (See Table 1). Family flexibility is defined as a family’s ability to adapt to change.
Too little flexibility among family rules and sharing of power is considered “rigid,” while too much
flexibility in family interactions and rules is considered “chaotic.” The right amount of flexibility in
family interactions is healthy or “balanced” on the family flexibility continuum. The extremes on
the cohesion and flexibility continuums represent family interactions that are dysfunctional and
lead to individual pathology (Fishman, 1993).
Enmeshed families tend to have extreme forms of proximity and intensity in family interac-
tions that result in poorly differentiated boundaries (Fishman, 1993). Disengaged families show
characteristics that include strict boundaries where affection and information do not flow easily
among family members (Olson, 1999). Chaotic families are characterized by a lack of structure,
rules, and leadership with little order (Olson, 1999). Rigid family characteristics include resistance
to change whereby individual family members who attempt change are often sabotaged to main-
tain the status quo (Levenkron, 1998). Clinicians today have noticed these dysfunctional family
characteristics among those they treat for self-injury (Conterio & Lader, 1998; Levenkron, 1998;
Levitt, Sansone & Cohn, 2004; Selekman, 2002; Walsh, 2006; Wedig & Nock, 2007; White, 1995).
Self-Injury and family functioning. Levenkron (1998) stated that “the fuel that drives self-
injury is the way family members relate to each other.” Selekman (2002) hypothesized that self-
injury is related to a person’s “disconnect” from significant others. He explains that adolescents
who use self-harm may feel they have no negotiation room with overbearing authoritarian parents,
or that they may have parents who are passive and provide little attention and emotional support.
Parenting styles that do not allow for the genuine expression of love or frustration also increase
the chances that their child will injure themselves or others (Levenkron, 1998; Levitt et al., 2004;
White, 1995). Conterio and Lader (1998) provided common familial scenarios among their self-
injuring clients, which included: instability (a lot of moves, illnesses, or death), abuse (physical, sex-
ual, or neglect), rigid values (applied hypocritically or inconsistently), and structure breakdown
(demands on a child to take on adult responsibility).
Three studies give some evidence of how family dynamics influence the development of self-
injury. First, Gratz et al. (2002) observed that family relationships may either exacerbate or offer
protection from the consequences of traumatic childhood experiences (such as self-injury) and that
care-giving relationships may influence the etiology of self-harm. Second, Wedig and Nock (2007)
found that adolescents’ self-criticism paired with low levels of parental criticism was not associated
with the adolescent using self-injury. But, adolescent self-criticism paired with high parental criti-
cism was associated with the daughters using self-injury to cope. Third, Jella (2007) found higher
levels of family conflict in the self-injury group as well as lower levels of family cohesion, indepen-
dence, organization, and control than in the non-self-injuring groups. These studies offer some evi-
dence that family dynamics influence the development, maintenance, and prevention of self-injury
behaviors. This study will attempt to offer more statistical support for these claims.
Table 1
Family Cohesion and Family Flexibility Continuums
Family cohesion
Very low cohesion Balanced cohesion Very high cohesion
Disengaged Balanced Enmeshed
Family flexibility
Very low flexibility Balanced flexibility Very high flexibility
Rigid Balanced Chaotic
248 JOURNAL OF MARITAL AND FAMILY THERAPY April 2014
Hypotheses
Given that clinicians and researchers have noticed a relationship between family dynamics
and self-injury behaviors, the question posed by this study is: What is the relationship between
self-injurious behaviors and family dynamics? From this question, four hypotheses were assessed.
(H1) –Duration: There is a negative correlation between how long someone uses self-
injury and healthy family dynamics.
(H2) –Frequency: There is a negative correlation between how many times someone uses
self-injury and healthy family dynamics.
(H3) –Periodicity: There is an association between how often someone uses self-injury
and family dynamics.
(H4) –Severity: There is a negative correlation between the severity of self-injuries and
healthy family dynamics.
METHODS
Subjects
This study included those who were 18 years or older, had at any time in the past used any
type of deviant self-injury (defined as intentionally harming one’s self physically to relieve painful
emotions without suicide intent) and who had not used self-injury in over a year. Participants who
did not fit these criteria were excluded from the study. A total of 189 participant surveys were
usable with 144 surveys completed.
Twenty-three states were represented in the sample with five countries including, Australia,
Canada, England/UK, and the United States. The age range of participants was 18–55 years
old. Fifty percent were between the ages of 18 and 23. There were 128 female participants and
18 male participants. One participant identified as transgendered. The majority of the sample at
66% identified as Caucasian-white, 8% as Latino, 3% as African American, 3% as Asian/Paci-
fic Islander, 1% identified as bi-racial, and the remaining 5% of participants identified as
“other.”
The average age participants began using self-injury was 13 years old with a range of onset
from 4 to 30 years old. The average age of cessation was 22 years old with a range from 13 to
51 years old.
Collecting Data
Participants were found through on-line links, college campuses, and acquaintances of the
investigators. The vast majority of participants responded on-line. Web-site managers and moder-
ators of on-line self-injury support groups were contacted and given the survey link and informa-
tion from: yahoo.com, facebook.com, myspace.com, twloha.com, recoveryourlife.com, and
S.A.F.E. Alternatives (selfinjury.com). S.A.F.E. Alternatives is an organization dedicated to the
prevention and treatment of self-injury. S.A.F.E. also disseminated the survey link and informa-
tion to their 2000 person email list that included clients, customers, educators, and clinicians on
the opportunities and products that S.A.F.E. provides.
The investigators also visited colleges and universities to disseminate information about
the study. Information was posted using flyers, distributed through multiple classrooms, and
university student email lists. The investigators also asked family and acquaintances to pass on
the survey link and information through e-mails, facebook.com, myspace.com, and business
cards.
On-line data were collected using computer-assisted web interviewing technology at the Insti-
tute for Social and Policy Research at Purdue University Calumet. All survey data were collected
anonymously from participants accessing the survey via the link provided on websites, social net-
work sites, campus flyers, email messages, and information cards disseminated to potential partici-
pants. All study protocols were reviewed and approved by the Institutional Review Board of
Purdue University.
April 2014 JOURNAL OF MARITAL AND FAMILY THERAPY 249
Measures
The measures used in this study include the Deliberate Self-Harm Inventory (Gratz, 2001), the
Family Adaptability and Cohesion Scale IV (Olson et al., 2006), and an array of demographic
variables described earlier.
Deliberate self-harm inventory. Gratz (2001) created the “Deliberate Self-Harm Inventory”
(DSHI) to measure the frequency, duration, and severity of those using self-harming behaviors
within the general population. The DSHI is a seventeen item, behaviorally based, self-report ques-
tionnaire. The DSHI includes the definition of deliberate self-harm as “intentionally harming one’s
self to relieve painful emotions without suicidal intent.” The measure assesses the various aspects
of deliberate self-harm, including frequency (how many times the participants used self-injury),
severity (how severe were the injuries), duration (how many years the participant used self-injury),
and the various types of self-harming behavior (such as cutting and burning). The DSHI demon-
strates high internal consistency (a=.82), as well as adequate test–retest reliability over an average
period of 3 weeks (u=.68, p<.001). The DSHI also demonstrated adequate construct, discrimi-
nate, and convergent validity (Gratz, 2001).
Family Adaptability and Cohesion Evaluation Scale. Olson and colleagues (Olson et al., 1979,
2006) created the “Family Adaptability and Cohesion Evaluation Scale” (FACES IV) to assess
family dynamics. This assessment measures the dimensions of family cohesion and family flexibil-
ity (see Table 1). There are two balanced scales called Balanced Cohesion and Balanced Flexibility
and there are four unbalanced scales designated as Disengaged, Enmeshed, Rigid, and Chaotic
that measure “unbalanced” family dynamics. For each of the two balanced scales, scores range
from 7 to 35 with higher scores reflecting healthier and more balanced family dynamics. For each
of the four unbalanced scales, scores range from 7 to 35 with higher scores indicating unhealthy
and more unbalanced patterns of family dynamics.
The six scales are used to calculate a Total Cohesion Ratio score and a Total Flexibility Ratio
score (See Table 2). The higher the ratio scale scores above 1.0, the healthier the family dynamics.
The lower the ratio scale scores, the more dysfunctional the family dynamics.
Recent additions to the FACES IV assessment include a Family Communication and a Family
Satisfaction scale. The Communication and Satisfaction scales have ten questions each. The scores
range from 10 to 50. The higher the scores on the communication scale, the less concern participants
have with their family communication. The higher the scores on the satisfaction scale, the more satis-
fied the participants are with their overall family functioning (which includes constructs such as fam-
ily coping, emotional reactivity, flexibility, cohesion, and communication). (Olson et al., 2006).
The FACES IV scales have adequate levels of reliability and validity. Reliabilities of the six
FACES IV scales are as follows: Disengaged =.87, Enmeshed =.77, Rigid =.83, Chaotic =.85,
Balanced Cohesion =.89, Balanced Flexibility =.80. Alpha reliability for the validation scales
range from .91 to .93. (Olson et al., 2006) For the purpose of this study, the questions were chan-
ged from present tense to past tense because the participants needed to recall what their family
dynamics were like when they first used self-injury.
Data Analysis
Various statistical tests were run to determine the relationships between family dynamics and
self-injury behaviors. Two-tailed Pearson correlation tests were used to calculate the strength and
Table 2
FACES IV Ratio Score Calculations
Total cohesion ratio =balanced cohesion/{(disengaged +enmeshed)/2}
Total flexibility ratio =balanced flexibility/{(rigid +chaotic)/2}
Total family ratio =cohesion ratio +flexibility ratio/2
(Olson et al., 2006, p. 15).
250 JOURNAL OF MARITAL AND FAMILY THERAPY April 2014
correlations between ratio scale variables. Independent sample t-tests were used to assess differ-
ences between groups. One-way ANOVA accompanied by post hoc tests were used to assess differ-
ences between multiple groups.
RESULTS
The participants were asked which type of self-injury they had used in the past. Nine types
found in the literature were listed. Cutting was the most common type of self-injury, with 89% of
the participants having cut themselves to relieve painful emotions. Some 70% had hit themselves,
64% had interfered with wounds healing, 59% had bit or chewed on themselves, 41% had burnt
themselves, 29% had pulled their hair out, 7% had ingested sharp or toxic substances, and 4% had
broken a bone. Ten percent of the participants had used only one type of self-injury, 15% had used
two types, 21% had used up to four different types, and one participant had used eight different
types of injury. No one in the study had amputated a body part. The average number of self-injury
types used was three.
(H1) –Duration and Family Dynamics
Duration of self-injury was calculated from the difference in the participants’ age at onset to
cessation. The minimum duration was less than a year and the maximum duration was 42 years.
The average duration of using self-injury was 8.5 years. A correlation analysis was used to test the
relationship between the duration of self-injury and the scores of the Total Cohesion Ratio, the
Total Flexibility Ratio, and the unbalanced scale scores as well as the Family Communication and
Family Satisfaction scores.
Duration and total flexibility were negatively correlated (r=.234, p=.005). Duration and
family satisfaction were negatively correlated (r=.236, p=.005). A statistically significant cor-
relation was not observed between Duration and the Total Cohesion Ratio. A negative correlation
approached statistical significance between Duration and Family Communication scores
(r=.159, p=.058). With regard to the unbalanced scales, a positive correlation was observed
between Duration and the scores on the Rigidity scale (r=.292, p=.000). This correlation indi-
cates that participants with longer duration of self-injury behaviors also reported higher levels of
rigidity in their family dynamics. No statistically significant correlations were observed between
Duration and the Enmeshed, Chaotic, or Disengaged scores.
(H2) –Frequency and Family Dynamics
The participants were asked how many times they had used self-injury. Categories of fre-
quency included: once, 2–5 times, 6–20 times, 21–50 times, 51–100 times, and 100 plus times. This
frequency question was asked for every type of self-injury listed. Of the participants who used cut-
ting (n=168), biting (n=105), or interfering with wounds healing (n=114), most indicated that
they used these types of self-injury over 100 times. Among those who had burned themselves
(n=75), hit themselves (n=130), or pulled their hair (n=52), most had used these types of self-
injury six to 20 times. Participants who had ingested sharp or toxic substances (n=13), most per-
formed these behaviors between two and five times. Those who had intentionally broken a bone
(n=7), half had done this only once, and the other half had done so two to five times.
Because many of the participants used more than one type of self-injury, the sum of the aver-
age frequency categories for each type of self-injury was totaled to create a Total Frequency Score
(See Table 3). The average minimum number of times that someone used self-injury was two and
the maximum average was 575 times. The total mean for how frequently self-injury was used was
176 times.
A correlation analysis was used to test the relationship between the frequency of self-injury
and the Total Cohesion Ratio, the Total Flexibility Ratio, and the unbalanced scale scores, as well
as the Family Communication and Family Satisfaction scores. The frequency of how many times
the participants used self-injury was statistically significant and negatively correlated with the
Total Cohesion Ratio (r=.203, p=.015) and Total Flexibility Ratio scores (r=.201,
p=.016), as well as, Family Communication (r=.189, p=.023) and Family Satisfaction scores
(r=.246, p=.003). All of these correlations suggest that increased use of self-injury is associated
April 2014 JOURNAL OF MARITAL AND FAMILY THERAPY 251
with less healthy family dynamics. With regard to the relationship between frequency of self-injury
and the unbalanced scale scores, both disengaged (r=.234, p=.003) and chaotic (r=.202,
p=.014) scale scores were positively associated with the frequency of self-injury. No statistically
significant relationships were observed between either the Enmeshed or Rigid unbalanced scale
scores and the frequency of self-injury. The correlation results indicated that how many times a
participant used self-injury was negatively correlated with better family communication (Family
Communication: r=.235, p=.005) and the participant’s overall satisfaction with family func-
tioning (Family Satisfaction: r=.246, p=.003). Results from the correlation tests support (H2)
that the frequency of times that someone uses self-injury is negatively correlated with healthy fam-
ily dynamics.
(H3) –Periodicity and Family Dynamics
Periodicity refers to how often self-injury was used. The categories for periodicity included
daily, weekly, monthly, yearly, and less than once a year. Participants were asked on average how
periodically they had used each type of self-injury. Among those who had used cutting (n=189),
hitting (n=130), or biting (n=105), the majority of participants used these types of injury on a
weekly basis. Participants who used burning (n=75), or pulling hair (n=52), the majority used
these types on a monthly basis. Those who ingested sharp or toxic substances did this less than once
a year. The type of self-injury that had the highest periodicity was interfering with wounds healing
(n=114); the majority of those who interfered with wounds healing did this on a daily basis.
To facilitate the analysis, the categories of periodicity were reduced from five categories (i.e.,
daily, weekly, monthly, yearly, and less than once a year) to a dichotomous variable. The dichoto-
mous variable distinguished between participants who had used any type of self-injury on a daily
basis (n=92) and those who had never used self-injury on a daily basis (n=97). An independent
samples t-test indicated that there was a significant difference in the family dynamic mean scores
between these two groups (See Table 4). Those who injured themselves daily had lower Total
Cohesion Ratio scores (Total Cohesion: t(140) =2.61, p=.010) and lower mean scores on the
Total Flexibility Ratio scale (Total Flexibility: t(142) =3.89, p=.000) than those who never
used self-injury daily. The t-test analysis indicated that daily users of self-injury perceived their
families as significantly more disengaged (t(145) =2.667, p=.009) and more chaotic (t
(138) =4.093, p=.000). No statistically significant differences emerged between the groups related
to levels of enmeshment and rigidity. Participants who used self-injury daily were also more con-
cerned with their family communication (Family Communication: t(143) =2.52, p=.031) and
more disappointed in their overall family functioning (Total Satisfaction: t(141) =3.02,
p=.003) than those who never self-injured daily.
Results of t-tests indicate statistical significant differences between how often someone uses
self-injury and how healthy their family dynamics were. Therefore, there is support for (H3) that
how often someone uses self-injury is associated with family dynamics.
Table 3
Total Frequency Calculation
Frequency
categories
(FC)
Averaged
frequency
(AF)
Total frequency score (TFS) =(AF) for one type +(AF) for
another type
Once 1
2–5 times 3 i.e., if a participant used cutting, and biting. They
indicated they cut 21–50 times and bit 51–100 times
(AF for cutting, 36) +(AF for biting, 75) =(TFS) of 111.
6–20 times 13
21–50 times 36
51–100 times 75
101 plus 100
252 JOURNAL OF MARITAL AND FAMILY THERAPY April 2014
(H4) –Severity and Family Dynamics
Participants indicated for each type of injury they used, if that type of injury ever resulted in
bleeding, scars, bruises, or medical attention. To test the correlation among every types of self-
injury used by one participant, a Total Severity Score was created (See Table 5). A value of 1.0
was assigned whenever bleeding, bruises, scars, or medical attention was a result of any type of
self-injury. The sum of all the 1.0s was tallied for a Total Severity Score. It was assumed that inten-
tionally breaking a bone in and of itself deserved a higher severity score. Therefore, the partici-
pants who had broken a bone received a value of 1.0 for bleeding, scaring, bruises, and medical
attention each, for a total of 4.0 just within the broken bone category. The minimum possible
severity score was 0; the maximum possible severity score was 32 (four severity questions multi-
plied by eight types of self-injury). The Total Severity Score from the participants in the present
study ranged from one to 21 with 6.5 being participants average Total Severity Score.
Table 4
Family Dynamic Mean Scores of Daily and NonDaily Self-Injury Use
Mean SD n
Cohesion ratio
Daily .90 .54 75
Nondaily 1.15 .58 67
Disengaged (unbalanced scale score)
Daily 24.63 6.76 82
Nondaily 21.76 5.77 75
Enmeshed (unbalanced scale score)
Daily 15.61 6.52 76
Nondaily 14.93 5.21 68
Flexibility ratio
Daily .72 .36 76
Nondaily .98 .44 68
Rigid (unbalanced scale score)
Daily 20.62 7.46 79
Nondaily 21.19 7.07 70
Chaotic (unbalanced scale score)
Daily 22.18 7.32 79
Nondaily 17.86 6.006 70
Communication
Daily 19.59 9.18 74
Nondaily 23.62 9.99 71
Satisfaction
Daily 16.64 7.32 73
Nondaily 20.60 8.29 70
Notes: Ratio Scores: The higher the ratio scores above 1.0, the more balanced (healthy) the
family system. The lower the ratio score below 1.0, the more unbalanced (problematic) the
family system (Olson et al., 2006, p. 15). Unbalanced Scale Scores: The higher the score, the
less balanced the family dynamics. Communication: 44–50 =felt very positive about quantity
and quality of family communication, 38–43 =felt good with a few concerns, 33–37 =feel
generally good but have concerns, 29–32 =have several concerns, 10–28 =have many con-
cerns about family communication. Satisfaction: 45–50 =very satisfied and really enjoy most
aspects of their family, 40–44 =satisfied with most aspects, 36–39 =somewhat satisfied and
enjoy some aspects, 30–35 =somewhat dissatisfied and have some concerns, 10–29 =very
dissatisfied and are concerned about their family (Olson et al., 2006, p. 16–17).
April 2014 JOURNAL OF MARITAL AND FAMILY THERAPY 253
Correlations were computed between the Total Severity Score and Total Cohesion Ratio,
Total Flexibility Ratio, Family Communication, and Family Satisfaction scores. The results indi-
cated a statistically significant negative correlation between severity of self-injuries and healthy
family dynamics. For example, the higher participants scored on the Total Severity Ratio score,
the lower they scored on the Total Cohesion Ratio (r=.215, p=.010). Also, the higher partici-
pants scored on the Total Severity score, the lower they scored on the Total Flexibility Ratio
Score (Total Flexibility: r=.207, p=.013). The correlation between severity and unbalanced
family dynamics indicated that participants who had higher severity scores also reported family
dynamicsthat were more disengaged (r=.234, p=.003). The analysis revealed no correlation
between severity and the other unbalanced scales (enmeshed, rigid, chaotic). Participant’s lower
severity scores were also negatively correlated with higher levels of concern for their family com-
munication (r=.229, p=.001) as well as high severity scores being negatively correlated with
how satisfied the participants were with their overall family functioning (r=.278, p=.001). In
other words, the more severely participant’s injured themselves, the more concern participants
had with family communication and the less satisfied they were with family functioning. These
results give support to (H4) –that the severity of injuries is negatively correlated with healthy
family dynamics.
DISCUSSION
Participants who had not used self-injury for over a year answered questions similar to the
Deliberate Self-Harm Inventory (Gratz, 2001) to measure how long (Duration), how frequently
(Frequency), how often (Periodicity), and how severely (Severity) they used self-injury to cope with
psychological stress. Self-injury was defined as intentionally harming one’s self to relieve painful
emotions without suicide intent. Using FACES IV, participants also answered questions about
their family dynamics during the time when they first used self-injury to cope. In this section, the
unbalanced scales of Disengaged and Enmeshed (on the Cohesion Continuum) and Rigid and
Chaotic (on the Flexibility Continuum) are discussed with the self-injury measures they were sig-
nificantly correlated or associated with. The Family Communication and Family Satisfaction
scales (also from FACES IV) are included in this discussion. This discussion is important to help
family therapists understand how family dynamics influence self-injury behaviors.
Disengaged and Enmeshed
Disengaged family characteristics include family members who do not have common goals,
recreation, or emotional connections. Participants who scored high in disengaged family character-
istics also scored high in self-injury Frequency, Periodicity, and Severity measures. The present
Table 5
Example Computation of Total Severity Score
Result of injury
Bleeding
Yes =1.0
No =.0
Scaring
Yes =1.0
No =.0
Bruising
Yes =1.0
No =.0
Med. attention
Yes =1.0
No =.0 Sum
Types of self-injury
Cutting 1.0 1.0 1.0 .0 3.0
Burning ———— —
Hitting 1.0 .0 1.0 .0 2.0
Biting or chewing ———— —
Interfering with wounds 1.0 1.0 .0 .0 2.0
Ingesting substances 1.0 .0 .0 1.0 2.0
Broken bones 1.0 1.0 1.0 1.0 4.0
Total severity score 13.0
254 JOURNAL OF MARITAL AND FAMILY THERAPY April 2014
studies’ participants indicated similar family dynamics to what Selekman (2002) described in his
self-injuring clients. He stated that his clients had a “disconnect” from family members. Other cli-
nicians that treat self-injury agree that their clients long for validation and connection from others
(Alderman, 1997; Conterio & Lader, 1998; Levenkron, 1998; Walsh, 2006). Researchers have
found higher measures of loneliness in those who self-injure (Briere & Gil, 1998; Guertin, Lloyd-
Richardson, Spirito, Donaldson & Boergers, 2001). If family members have low cohesion, then it
is not surprising that feelings of loneliness are a result. Jella (2007) found that the group of adoles-
cent girls who were self-injuring rated their families lower in cohesion than the control groups of
clinically distressed non-self-injuring girls. Results of the present study offer statistical support to
these clinicians’ and researchers’ observations that families with disengaged characteristics are cor-
related with how often and how severely a family member injures themselves to cope.
The measure of enmeshed family characteristics was not correlated with any of the self-injury
behaviors. These results were surprising because many clinicians in the family therapy literature
state that individual pathological coping takes place with enmeshed families. Enmeshed character-
istics are described as having few boundaries that protect and honor individuality as well as having
little boundaries to protect individuals’ privacy (Fishman, 1993; Olson et al., 2006; Minuchin,
Rosemon & Baker, 1978; Minuchin, Nichols & Lee, 2007). Conterio and Lader (1998) stated that
their clients who self-injure often struggle with too little separation and too few boundaries
between themselves and their family members. They stated that in these families, expression of
individuality is interpreted as a family betrayal. Still, in the present study, families with enmeshed
characteristics were not statistically correlated with self-injury behaviors. An explanation may be
that participants did not have the ability to self-injure as easily due to emotional and physical prox-
imity of family members due to enmeshment. Perhaps self-injury is more likely a symptom of dis-
engaged family dynamics than enmeshed family dynamics.
Rigid and Chaotic
Rigid family characteristics are described as families who have a strong resistance to change.
Participants who had higher self-injury duration scores also reported more rigid family characteris-
tics. Clinicians such as Conterio and Lader (1998) observed that families with rigid characteristics
are common in their clients who use self-injury. Selekman (2002) suggested that self-injuring ado-
lescents often feel they have no negotiation room with overbearing authoritarian parents. As such,
families with rigid characteristics may contribute to a family member using self-injury as a means
of coping.
Chaotic family characteristics include a lack of structure, rules, and leadership with little order
and safety measures. Participants who scored high on chaotic family characteristics also had ele-
vated scores of self-injury frequency and periodicity. Clinicians noticed that chaotic families, with-
out rules and sound family structures in place, add to their self-injuring clients’ sense of being out
of control. Selekman (2002) stated that parents can be passive to the point of neglect and therefore
offer little attention and support. Conterio and Lader (1998) noticed that common situations
among their clients involved instability, such as many moves or family rules applied inconsistently.
They also stated that overall structural breakdown is common in the family environments of their
clients. Jella (2007) found that self-injuring adolescents reported more daily family conflicts than
every day functioning adolescents. Results of the present study offer statistical support to clini-
cians’ observations that families with chaotic characteristics influence how many times and how
often a family member uses self-injury to cope.
Family Communication
Participants in the present study rated their family communication during the time when they
first started to use self-injury. The level of concern about their family communication was positively
correlated with how frequently and how severely the participants injured themselves. Clinicians
have noticed the relationship between family communication patterns and self-injury. Wedig and
Nock (2007) stated that working on family communication, specifically parental criticism, may help
prevent the use of self-injury. Selekman (2002) noticed that when some of his clients’ family commu-
nication improved, his clients stopped using self-injury. Alderman (1997) saw self-injury as an out-
ward symbol of internal emotional stress. This outward symbol most likely would not be necessary
April 2014 JOURNAL OF MARITAL AND FAMILY THERAPY 255
if family members were able to express their frustrations directly. This correlation between family
communication and frequency of self-injury seems to support clinicians’ observations that family
communication has an impact on the development, maintenance, and cessation of self-injury.
Family Satisfaction
Participants in the present study shared how satisfied they were with their families overall
functioning during the time when they first started to use self-injury. The results indicated par-
ticipant’s satisfaction with their families overall functioning was negatively correlated with
every self-injury measure. In other words, the participants who scored high on their satisfaction
with overall family functioning also scored low on self-injury duration, frequency, periodicity,
and severity measures. Wedig and Nock (2007) studied high emotional reactivity in parents of
adolescent girls and found that high parental emotional reactions were related to the develop-
ment of self-injury. Past users of self-injury have explained that the way their family members
reacted to their self-injury episodes helped them feel either validated and understood, or invali-
dated and judged. Those who stated they felt invalidated or judged also stated that they
increased the frequency of their self-harming behaviors (Redheffer & Brecht, 2005; Schneider,
2007; Yip et al., 2003). Levenkron (1998) stated that when parents respond to self-injury by
keeping their own emotions in check, they offer empathy and calm that reduces the stress that
often triggers self-injury use. Prest and Prest (2007) worked from Bowen’s transgenerational
model, which identifies how families transmit anxieties and coping strategies. The way that
families deal with stress seems to influence how individuals cope. The present study has offered
statistical support to clinician’s observations that family functioning has strong influences on
the use of self-injury.
Clinical Implications
The traditional treatments for self-injury include psychotherapy (Alderman, 1997; Connors,
2000), cognitive-behavioral therapy (Conterio & Lader, 1998), dialectic behavioral therapy (DBT)
(Linehan, 1993), and psychotropic medication management (Connors, 2000; Levitt et al., 2004).
Group therapy, play therapy, psychodrama, music therapy, and other forms of creative expression
are often added to traditional treatments (Bowen & Randall, 2005; Gratz & Gunderson, 2006). All
treatment programs address one or more of the many important aspects of self-injury, including:
identity development, self-expression, impulse control, invading thought patterns, emotional regu-
lation, self-care, stress management, and biological/chemical imbalances. These treatments are
geared toward the individuals because ultimately it is the individual who is responsible for the ces-
sation of their self-injury behaviors (Conterio & Lader, 1998). The purpose of this study is to give
the field of MFT an empirical basis to request that family therapy be integrated into the treatment
of this disorder. This study is a call to include family therapy as a viable component of a multi-
modal therapy for those who self-injure.
Family therapists have the theoretical and systemic training to offer a safe environment
where families can get the education, support, and training they need to improve their family
dynamics and communication which will in turn help their loved ones feel secure enough to
overcome self-injury. The present study found correlations and associations with family dynam-
ics and every self-injury measure (duration, frequency, periodicity, and severity). The implica-
tion is that family-based treatment is expected to be a viable treatment for self-injurious clients.
Family therapy has found to be a superior treatment for numerous psychosomatic disorders
(Fishman, 2006). It is expected that families where someone is self-injuring may benefit
similarly.
In the present study, family dynamics were measured using the FACES IV assessment (Olson
et al., 2006) which was based on Salvador Minuchin’s structural family therapy model (Minuchin
et al., 1978). Therapists can use the FACES IV to assess family dynamics where a family member
is using self-injury to cope. A family diagnosis of Disengaged, Enmeshed, Rigid, or Chaotic will
guide the therapist in the interventions necessary to balance the Cohesion and Flexibility of the
family. The most apparent treatment choice would be to use Structural Family Therapy (Minuchin
et al., 1978), because it closely corresponds to the assessment instrument. Hierarchy, rules, roles
and boundaries can be addressed in a traditional family therapy model, while integrating clinical
256 JOURNAL OF MARITAL AND FAMILY THERAPY April 2014
knowledge of behavioral and crisis management around self-injury behaviors. However, cohesion
and adaptability can be addressed in all major family therapy theories. Further research is needed
to understand how cohesion and adaptability can be therapeutically manipulated in an effective
way to decrease non-suicidal self-injurious behavior.
Limitations
The present study was conducted with self-report, retroactive survey measures. The main limi-
tations of the present study were that participants’ perceptions were the only measure of family
dynamics. Also, participants were asked to remember their family dynamics at the time when they
first used self-injury. For some participants this was decades ago. The participant sample may not
represent the general population as 66% of the sample was Caucasian-White and 84% were
female. The study did not distinguish between the clinical and nonclinical population of those who
had used self-injury in the past. Another limitation was in the severity measure. For example, the
severity score did not represent how many times bleeding, bruises, or scaring resulted—just that
they had occurred. Therefore, the present study used a very conservative severity measure. In addi-
tion, some participants indicated that staying overnight at a medical facility was more for suicide
prevention watch than an indication of the severity of their wounds. Therefore, some of the sever-
ity questions were not used in the Total Severity Score.
Future Research
Studies that examine family dynamics at the onset and cessation of self-injury use would offer
more insight into what changes in family dynamics influence overcoming self-injury. Such studies
should take into account the family life cycle as participants’ significant family members may
change from their family-of-origin at self-injury onset to their family-of-creation (romantic partner
and children) at self-injury cessation. Also, there is limited research on adult use of self-injury,
especially on how adult self-injury affects family dynamics, for example when a parent is cutting
(Conterio & Lader, 1998). Further research on the differences of family dynamics between the clin-
ical and non-clinical populations of those who self-injure and the implications for treatment
among the two populations.
CONCLUSION
Family dynamics were correlated with the duration, frequency, periodicity, and severity of
self-injury behaviors as reported by study participants. The present study offers statistical evidence
to support clinicians’ observations that family dynamics influence self-injury behaviors. Families
who work to balance their family cohesion and flexibility, along with improving communication
and overall family functioning, give their self-injuring family member a better chance to overcome
their self-injury tendencies sooner. Family therapists are specifically trained in working on the
complex family dynamics that lead to individual pathological behaviors. Family therapy is a viable
treatment option to be added to the traditional treatments for clients who strive to overcome these
life threatening behaviors.
REFERENCES
Alderman, T. (1997). The scarred soul: Understanding & ending self-inflicted violence. Oakland, CA: New Harbinger.
Bowen, S., & Randall, K. (2005). See my pain. Chapin, SC: Youth Light.
Campbell, T. L., & Patterson, J. M. (1995). The effectiveness of family interventions in the treatment of physical
illness. Journal of Marital and Family Therapy,21, 545–583.
Connors, R. E. (2000). Self-injury: Psychotherapy with people who engage in self-inflicted violence. North Bergen, NJ:
Book-mart Press.
Conterio, K., & Lader, W. (1998). Bodily harm. New York: Hyperion.
Eckenrode, J. (Ed.). (2006). Proceedings from: International Network of Self-Injury Prevention, Intervention,
Research, and Education. Conference at Cornell University, July 13–14, 2006. Retrieved October 2006, from
http://www.crpsib.com/people.as is active and usable
Favazza, A. R. (1996). Bodies under siege: Self-mutilation and body modification in culture and psychiatry (2nd ed.).
Baltimore and London: The John Hopkins University Press.
April 2014 JOURNAL OF MARITAL AND FAMILY THERAPY 257
Favazza, A. R. (1998). The coming of age of self-mutilation. The Journal of Nervous and Mental Disease,186, 259
–268.
Favazza, A. R., & Rosenthal, R. J. (1993). Diagnostic issues in self-mutilation. Hospital and Community Psychiatry,
44, 134–140.
Fishman, C. H. (1993). Intensive structural therapy: Treating families in their social context. New York: Basic Books.
Fishman, C. H. (2006). Juvenile anorexia nervosa: Family therapy’s natural niche. Journal of Marital and Family
Therapy,32(4), 505–514.
Briere, J., & Gil, E. (1998). Self-mutilation in clinical and general population samples: Prevalence, correlates, and
functions. American Journal of Orthopsychiatry,68, 609–620.
Gratz, K. L. (2001). Measurement of deliberate self-harm: Preliminary data on the deliberate self-harm inventory.
Journal of Psychopathology and Behavioral Assessment,239, 253–263.
Gratz, K. L., Conrad, S. D., & Roemer, L. (2002). Risk factors for deliberate self-harm among college students.
American Journal of Orthopsychiatry,72, 128–140.
Gratz, K., & Gunderson, J. (2006). Preliminary data on an acceptance-based emotion regulation group interven-
tion for deliberate self-harm among women with borderline personality disorder. Behavior Therapy,37,25
–35.
Guertin, T., Lloyd-Richardson, E., Spirito, A., Donaldson, D., & Boergers, J. (2001). Self-mutilative behavior in ado-
lescents who attempt suicide by overdose. Journal of the American Academy of Child and Adolescent Psychiatry,
40, 1062–1069.
Holmes, A. (2000). Cutting the pain away: Understanding self-mutilation. New York: Chelsea House.
Jella, S. H. (2007). The family environments of self-injuring female adolescents. (Dissertation, Alliant International
University). Received as an email attachment from S. Jella, February, 2009.
Juzwin, K. R. (2004). An assessment tool for self-injury: The Self-Injury Self-Report Inventory (SISRI). In J. L.
Levitt, R. A. Sansone & L. Cohn (Eds.), Self-harm behavior and eating disorders (pp. 105–118). New York: Brun-
ner-Rutledge.
Kaslow, F. W. (Ed.) (1996). Handbook of relational diagnosis and dysfunctional family patterns. New York: John
Wiley & Sons.
Lader, W. (Ed.) (2006). Proceedings from: Understanding and Treating the Self-Injurer. Aurora, IL: Provena Medical
Center.
Levenkron, S. (1998). Cutting: Understanding and overcoming self-mutilation. New York: W.W. Norton & Norton.
Levitt, J. L., Sansone, R. A., & Cohn, L. (Eds.) (2004). Self-harm behavior and eating disorders. New York: Bruner-
Rutledge.
Linehan, M. M. (1993). Cognitive-behavioral treatment of borderline personality disorder. New York: Guilford Press.
Minuchin, S., Nichols, M. P., & Lee, W. (2007). Assessing families and couples: From symptom to system. Boston,
MA: Pearson Education.
Minuchin, S., Rosemon, B. L., & Baker, L. (1978). Psychosomatic families: Anorexia nervosa in context. Cambridge,
MA: Harvard University.
Muehlenkamp, J. J. (2005). Self-injurious behavior as a separate clinical syndrome. American Journal of Orthopsychi-
atry,75, 324–333.
Nock, M. K., & Prinstein, M. J. (2004). A functional approach to the assessment of self-mutilative behavior. Journal
of Consulting and Clinical Psychology,72, 885–889.
Nock, M. K., & Prinstein, M. J. (2005). Contextual features and behavioral functions of self-mutilation among ado-
lescents. Journal of Abnormal Psychology,114, 140–146.
Olson, D. (1999). Circumplex model of marital and family systems. Journal of Family Therapy,22, 144–167.
Olson, D. H., Gorall, D. M., & Tiesel, J. W. (2006). FACES IV package: Administration manual. (March 2007
Edition). Life Innovations, Minneapolis. FACES IV Resource CD ordered from www.facesiv.com, December,
2007.
Olson, D. H., Sprenkle, D. H., & Russell, C. S. (1979). Circumplex model of marital and family systems: I. Cohesion
and adaptability dimensions, family types, and clinical applications. Family Process,18,3–28.
Prest, K. M. & Prest, L. A. (Eds). (2007). Relational & DBT Treatment of Personality. Proceedings from the annual
American Association for Marriage and Family Therapy Conference.
Redheffer, J., & Brecht, S. (2005). Beyond the razor’s edge. Lincoln, NE.: iUniverse.
Ross, S., & Heath, N. (2002). A study of the frequency of self-mutilation in a community sample of adolescents.
Journal of Youth and Adolescence,1,67–77.
Schneider,W.(2007). Cut: Teensand self-injury[Motion Picture]. (Available from S.A.F.E. Alternatives, selfinjury.com).
Selekman, M. D. (2002). Living on the Razor’s Edge. New York: W.W. Norton & Company.
Walsh, B. (2006). Treating self-injury: A practical guide. New York: Guildford Press.
Wedig, M. M., & Nock, M. K. (2007). Parental expressed emotion and adolescent self-injury. Journal of Academy of
Child and Adolescent Psychiatry,46, 1171–1178.
White, M. (1995). Re-authoring lives: Interviews and essays. Adelaide, SA: Dulwich Center Publications.
258 JOURNAL OF MARITAL AND FAMILY THERAPY April 2014
WhitlockJ. L. (Ed.). (2006). Proceedings from International Network of Self-Injury Prevention, Intervention,
Research, and Education. Conference at Cornell University, July 13–14, 2006. Retrieved October 2006, from
http://www.crpsib.com/people.asp
Whitlock, J. L., Powers, J. L., & Eckenrode, J. E. (2006). The virtual cutting edge: The internet and adolescent self-
injury. Developmental Psychology,42, 407–417.
Yip, K., Ngan, M., & Lam, I. (2003). A qualitative study of parental influence on and response to adolescent’s self-
cutting in Hong Kong. Families in Society,84, 405–416.
April 2014 JOURNAL OF MARITAL AND FAMILY THERAPY 259