Families, Systems, & Health © 2011 American Psychological Association

2011, Vol. 29, No. 3, 232–252 1091-7527/11/$12.00 DOI: 10.1037/a0025341

Autism Spectrum Disorder:

Parenting Stress, Family Functioning and Health-Related
Quality of Life
NORAH JOHNSON, PHD
MARILYN FRENN, PH D
SUZANNE FEETHAM, PHD
PIPPA SIMPSON, PHD

The prevalence of Autism Spectrum Disor- ing were significantly larger than male par-
der (ASD) is 1 in 110 persons in the U.S. ents, p ⫽ .002. Results of stepwise linear
Both parents of children with ASD are un- regression for the male-female partners
der stress that may impact their health- showed that (1) higher female caregiving
related quality of life (HRQL) (physical and stress was related to lower female physical
mental health). The purpose of the current health (p ⬍ .001), (2) a higher discrepancy
study was to explore the relationship of par- score in family functioning predicted lower
enting stress, support from family function- mental health (p ⬍ .001), accounting for
ing and the HRQL (physical and mental 31% of the variance for females and (3)
health) of both parents. Female (n ⫽ 64) male parent personal and family life stress
and male (n ⫽ 64) parents of children with (p ⬍ .001) and family functioning discrep-
ASD completed Web-based surveys examin- ant (D) score (p ⬍ .001) predicted poor men-
ing parenting stress, family functioning, tal health, with the discrepancy score ac-
and physical and mental health. Results of
counting for 35% of the variance. These
a Wilcoxon signed-ranks test showed that
findings suggest that there may be differ-
female parent discrepant (D) scores between
ences in mothers’ and fathers’ perceptions
“what is” and “should be” family function-
and expectations about family functioning
and this difference needs to be explored and
applied when working with families of chil-
Norah Johnson, PhD, and Marilyn Frenn, PhD, Col-
lege of Nursing, Marquette University; Suzanne dren with ASD.
Feetham, PhD, College of Nursing, University of Wis-
consin–Milwaukee and Children’s National Medical Keywords: Autism Spectrum Disorder,
Center, Washington, DC; Pippa Simpson, PhD, Chil- family functioning, parenting stress,
dren Research Institute, Milwaukee, Wisconsin. health-related quality of life
Funding from the Nurse Practitioner Health Care
Foundation/Pfizer and the Milwaukee District Nurs-
ing Association Lamplight Scholarship. Recruitment
he focus of this study is to understand
assistance from the Kennedy Krieger Institute’s In-
teractive Autism Network.
Correspondence concerning this article should be
T the relationship of parenting stress,
support from family functioning and
addressed to Norah Johnson, PhD, College of Nurs-
ing, Marquette University, Milwaukee, WI 53201. health-related quality of life: physical and
E-mail: Norah.johnson@marquette.edu mental health, for both parents of children
232
 AUTISM SPECTRUM DISORDER 233

with Autism Spectrum Disorder (ASD). In the present study, support is concep-
ASD is a collective term used to represent tualized as family functioning. The family
three Pervasive Developmental Disorders: science literature cites multiple functions
(1) Autistic Disorder, (2) Pervasive Devel- of the family (Friedman, Bowden, & Jones,
opmental Disorder, not otherwise specified, 2003). Family functioning includes a com-
and (3) Asperger’s Syndrome (Tanguay, mitment to support the functions of the
2006). The prevalence of ASD is reported family that include: economic, safety, child
as one out of every 110 Americans (Rice, rearing, caregiving, and communication
2009) resulting in approximately 2.8 mil- Feetham (2007). Support from family func-
lion people with ASD in the United States. tioning is defined as supportive relation-
Symptoms of ASD appear before 3 years of ships among parents, family members,
age (American Psychiatric Association friends and community members (Roberts
[APA], 1994). With the diagnostic criteria & Feetham, 1982). The social relationships
of the DSM–IV (APA, 1994), a person is in turn influence health. Parents are part
diagnosed with ASD if they exhibit quali- of a family system that functions by ap-
tative impairments in social interaction praising, operating, and behaving in cer-
and communication, and restricted, repet- tain ways (McCubbin & Thompson, 1991).
itive, and stereotyped patterns of behavior, Family functioning includes generational
activities, and interests (APA, 2000). The relationships, unique sets of rules, priori-
symptoms of the three disorders range in ties, and ethics. The family is also consid-
severity with Autistic Disorder being the ered to be a mediator between the individ-
most severe. The least severe, Asperger’s ual and the environment (Andrews,
syndrome, is also referred to as high func- Bubolz, & Paolucci, 1980).
tioning autism. Knafl and Deatrick (2003) found that
Family functioning is an important parents’ joint effort to manage their child’s
source of support for the family. The per- chronic illness resulted in different views
ceptions of family functioning of parents of on how to manage the situation. Although
children with ASD merits direct research. the parents work together, they may have
A study that gathers data from female and different future expectations or a plan on
male parents is most helpful for under- how to get there, which is a potential
standing differences in what parents find source of distress (Knafl & Deatrick, 2003).
stressful, as well as how support from fam- Within families, there may be differences
ily functioning dynamics are related to pa- in expectations about family functioning.
rental physical and mental health. This However, in most studies, the mother is the
study will help guide the development of only source of data (Cassano, Adrian,
appropriate interventions targeted to re- Veirs, & Zaman, 2006). With only one
duce parental stress, and improve parental source of data, the differences in expecta-
physical and mental health outcomes. tions, between mother and father, about
Parents of children with ASD are family functioning are missed.
stressed with caring for their child and the Improved quality of life is one recog-
decision making that accompanies the nized metric of a successful outcome (Lan-
care. The children with ASD are dependent tos, 2007), which is crucial in the case of
on their parents for diagnostic assessment, ASD, as it has no cure. Health-related
treatment selection and support (AHRQ, quality of life (HQRL) includes physical
2009). Parents who are mentally and phys- and mental health. HRQL refers to the ef-
ically healthy are best prepared to cope fects of health, illness and treatment on
with the ramifications of the diagnosis of quality of life (Ferrans, Zerwic, Wilbur, &
ASD, and be supportive advocates for their Larson, 2005). These include biological
child (Feetham, 2011). function, symptoms, functional status, and
234 JOHNSON, FRENN, FEETHAM, AND SIMPSON

general health perceptions (Ferrans et al., Our model was developed to test theoreti-
2005). Parenting stress impacts the mental cal relationships (see Figure 1) of the three
and physical health of parents of children basic concepts we believe to be fundamen-
with ASD in differing degrees; there can be tal: parenting stress, supportive family
a negative impact on mental health (Baker, functioning, and health-related quality of
Blacher, & Olsson, 2005; Phetrasuwan & life (physical and mental health).
Miles, 2009), or both physical and mental
health (Allik, Larsson, & Smedge, 2009; REVIEW OF LITERATURE
Lee et al., 2006). A systematic review of scientific litera-
ture included the following databases: CI-
THEORETICAL FRAMEWORKS NAHL, Health Sciences in Proquest, Social
This study examined the relationships Sciences in Proquest, PSYCH info, ERIC,
among the three variables of parenting MEDLINE, and Dissertations and theses
stress, supportive family functioning and abstracts. The search terms were: autism,
the HRQL (physical and mental health) for dyads, social support, stress, family func-
parents of children with ASD. The concept tioning, health-related quality of life, and
that a person’s appraisal of their own well-being.
stress will have an impact on their physical
and mental health (HRQL) derives from PARENTING STRESS
the Transactional Model of Stress and Cop- Prior research revealed three foci of
ing (Lazarus, 1999; Lazarus & Folkman, stress in parents of children with ASD (1)
1984). The concept of family functioning in parenting a child with ASD is more stress-
a supportive role to mediate the potential ful than parenting other children; (2) the
effect of stress on parental HRQL (physical behaviors of children with ASD are the
and mental health) comes from Social Sup- principal stressors; (3) mothers of children
port Theory (Lakey & Cohen, 2000). Sup- with ASD perceive more stress than fa-
portive actions of others are effective in thers. Parents of children with ASD are
reducing the effects of a stressor when the more stressed than parents of children
form of the assistance matches the de- without ASD, as reported in three studies
mands of the stressor (Cohen & McKay, with a range of participants (n ⫽ 18 to n ⫽
1984; Cutrona & Russell, 1990). The effects 151) and stress measures. Herring, Gray,
of social support include a person’s adher- Taffe, Sweeny, and Einfeld (2006) used a
ence to health behaviors and a person’s 0 – 4 Likert scale with the poles “no stress”
immune response (Lakey & Cohen, 2000). to “very stressed.” Konstantareas and Pa-

Demographics Parenting Stress Support HRQL (PCS, MCS)

Age of child Parent stress Male parent

Age of parent Male parent Family SF-36 :
Gender of child (PSS:A) functioning PCS, MCS
Diagnosis of child M l parentt
Male
Number of children (FFFS)
Marital status
Length of parent Family
relationship functioning Female
Parent Female pparent
Age parent suspected
stress parent
ASD (FFFS) SF-36 :
Age Diagnosed ASD Female
parent PCS, MCS
(PSS:A)
Figure 1. Conceptual Model: Johnson Parenting Health-Related Quality of Life Model.
 AUTISM SPECTRUM DISORDER 235

pageorgiou (2006) used a Clarke modifica- amount of time a parent spent with the
tion of Holroyd’s Questionnaire on Re- child was not reported in either study.
sources and Stress (QRS). Brobst, Clopton,
and Hendrick (2009) measured parenting
SUPPORT FROM FAMILY FUNCTIONING
stress with the Parenting Stress Index-
Family functioning for parents of chil-
Short Form (PSI-SF) (Abidin, 1995). Her-
dren with ASD was examined in two stud-
ring and Brobst included mothers and
ies. Parents reported struggles with their
fathers, while the other two studies only
extended family members (n ⫽ 26 mother-
included mothers.
father dyads) (Altiere & Von Kluge, 2009)
Severity of the symptoms of ASD (func-
and more mothers than fathers perceived
tional impairments and child behavior
that the family accessed support during a
problems) was associated with parental
crisis. Family functioning was measured
stress in four studies, using different mea-
with the Family Adaptability and Cohesion
sures. Tobing and Glenwick (2006) (n ⫽ 97
Evaluation Scales-III (FACES-III) (Olson
mothers of children with ASD) measured
et al., 1985). Bromley, Hare, Davison, and
stress with the PSI-SF (Abidin, 1995).
Emerson (2004) reported that single moth-
Hastings and Johnson (2001) (n ⫽ 141)
ers of male children with ASD and those
used the Questionnaire on Resources and
living in poor housing (n ⫽ 68), perceived
Stress (QRS)-Friedrich Short Form lower levels of social support than parents
(Friedrich, Greenberg, & Crnic, 1983). Her- of female children or those living in better
ring et al. (2006) (n ⫽ 151) reported both housing conditions. The Family Support
behavior and emotional problems contrib- Scale, developed by the authors for the
uted to parenting stress, as measured with study, measured social support. Therefore,
a 0 – 4 Likert scale. Konstantareas and Pa- a parent’s access to support may be related
pageorgiou (2006) used the Clarke Modi- to socioeconomic or relationship variables.
fication of Holroyd’s QRS: 43 mothers Parents of children with ASD may lack
reported more stress with parenting non- the support they expect from their ex-
verbal, low-functioning, in-flexible, and ac- tended family. For example, Easter Seals
tive children with ASD, than higher func- (2009) surveyed 1652 (8% male, 92% fe-
tioning children. A Parenting Stress Scale: male) parents of children age 30 or younger
Autism was developed to assess the partic- with ASD. The majority (80%, n ⫽ 3745) of
ular stressors of parents of children with the parents reported that they received lit-
ASD (Phetrasuwan, 2003). In the study of tle support from their extended families,
103 mothers of children with ASD (Phetra- but 40% (n ⫽ 1872) did receive some sup-
suwan, 2003), the unusual fears of children port from the government such as special
and trouble adapting to change were re- education and Medicaid. For the children
ported as stressing mothers the most. in the Study 18 years of age or younger
Mothers of children with ASD perceived (n ⫽ 1496), 78% (n ⫽ 1167) of their parents
more stress than fathers in two studies. were extremely or very worried about their
Herring et al. (2006) (n ⫽ 72 fathers and 34 financial future. Furthermore, in a Web-
mothers) accessed fathers of children with based study, 4,682 parents (88% mothers)
ASD early in the diagnostic evaluation pro- of children with ASD, parents rated the
cess. Little (2002) reported an Internet- impact of having a child with ASD on the
recruited sample of mothers and fathers couple’s relationship as somewhat or very
(n ⫽ 103 couples) of children with Asperg- negative for 60% (n ⫽ 2472) of mothers and
er’s Syndrome and nonverbal learning dis- 54% (n ⫽ 303) of fathers, with the division
orders. Mothers reported more stress, as of labor being the most common contribu-
measured by the QRS, than fathers. The tor to marital conflict (IAN, 2009).
236 JOHNSON, FRENN, FEETHAM, AND SIMPSON

HRQL: PHYSICAL AND MENTAL teristics, and social support. Mental health
HEALTH was measured by the Center for Epidemi-
Parents of children with ASD may be ologic Studies Depression Scale (CES-D)
less mentally and physically healthy than (Randolph, 1977). Interestingly, the child’s
parents of other children. Physical and overall level of functioning did not predict
mental health was measured in two studies any measure of parental well-being (Kersh
for parents of children with ASD. Lee et al. et al., 2006) for mothers or fathers, sug-
(2009) (n ⫽ 65 mothers, n ⫽ 24 fathers) gesting that marital quality may be an im-
reported significantly lower physical portant source of support for parents car-
health summary component scores and ing for the most challenging children with
mental health component scores than par- ASD.
ents of typically developing children (n ⫽ Differences in expectations about sup-
46), as measured by the MOS 36-item port from family functioning may mediate
short-form health survey (Ware & Sher- parenting stress and physical and/or men-
tal health. No studies looked at differences
bourne, 1992). In a study of 31 mothers and
in family functioning for both parents of
30 fathers of children with Asperger’s syn-
children with ASD. In the present study,
drome, Allik, Larsson, and Smedge (2006)
data on the differences in supportive family
found that mothers reported poorer physi-
functioning will be reported for both par-
cal health, as measured by the 12-item
ents. The support from the parental rela-
Short Form Health Survey (SF-12) (Ware,
tionship is of particular interest based on
Kosinski, & Keller, 1996), than the control
the review of the literature that demon-
group mothers of typically developing chil-
strates the greater perceived parenting
dren or the fathers of the children with
stress for mothers over fathers.
Aspergers (Allik, Larsson, & Smedje,
2006). Therefore, there may be gender dif-
CURRENT RESEARCH
ferences related to physical and mental
health outcomes in parents of children with This study was designed to address the
ASD. gaps in the science concerning the relation-
Research has focused on the mental ships of support from family functioning,
health of mothers of children with ASD and parenting stress and parental HRQL
on the mediators of parental mental (physical and mental health) in families
health. Mental health for mothers of chil- with children with ASD. Specifically the
dren with ASD was reported in three stud- aims of the study were to investigate the
following research questions for both male
ies using different measures. Baker et al.
and female parents of children with ASD:
(2005) (n ⫽ 214) used the Family Impact
Questionnaire (Donenberg & Baker, 1993). 1. Is parenting stress associated with
Phetrasuwan and Miles (n ⫽ 103) used the mental or physical health?
Psychological Well Being Scale (Ruff & 2. Is a wide discrepancy in expectations
Keyes, 1995), and Bromley et al. (2004) about family functioning associated
(n ⫽ 68) used the 1999 ONS Survey (Melt- with lower mental or physical
zer, Gatward, Goodman, & Ford, 2000). health?
Kersh, Hedvat, Hauser, and Warfield 3. Does discrepancy in expectations of
(2006) reported that for mothers and fa- family functioning mediate the rela-
thers of children with developmental dis- tionship of stress on mental or phys-
abilities (n ⫽ 67 dyads), greater marital ical health?
quality predicted lower parenting stress
and fewer depressive symptoms above and Two specific hypotheses for this study
beyond socioeconomic status, child charac- were: (1) Highly discrepant parents’ scores
 AUTISM SPECTRUM DISORDER 237

within spouse perceptions of “what is” and working sites, autism blogs, hospital clin-
“what should be” in family functioning will ics, and autism conferences. Subjects were
be associated with lower mental and phys- also recruited with the assistance of the
ical health; (2) Female parents will have Interactive Autism Network (IAN) Re-
significantly greater discrepancy scores in search Database at the Kennedy Krieger
family functioning than fathers. Institute and Johns Hopkins Medicine Bal-
timore, sponsored by the Autism Speaks
METHOD Foundation. A recruitment flyer directed
participants to the Internet Web site for
Procedure
access to the study. Inclusion criteria were
A nonexperimental, cross sectional, cor- being the biological parent of one or more
relational design was used in the study. children, 2–18 years old, with a diagnosis
Both parents were asked to fill out ques- of ASD (Autism, PDD-NOS, or Asperger’s
tionnaires. The objective of the study was syndrome), Internet access, residence in
to gain a better understanding of the rela- the U.S., and the ability to identify a
tionship of parenting stress, support from spouse or significant other. There were no
family functioning and HRQL (physical exclusion criteria. Demographics were col-
and mental health) for both parents of chil- lected to describe the sample of partici-
dren with ASD. Approval for the study was pants who reported being a parent.
obtained from the human research review
boards at both the University and a free- Measures
standing children’s hospital in the Mid- In addition to the demographic ques-
west. Data were collected over the Internet tions, participants were also asked to
via a Web-based site (www.surveymonkey- complete three surveys to measure the
.com) that housed the electronic version of key constructs of the study: Parenting
the questionnaires and a description of the stress, supportive family functioning,
study. The researcher’s e-mail address was and HRQL (physical health and mental
provided in the call for participants as well health). The data collected from both par-
as log in information to go to the study’s ents were used for this report.
online Web site. The participant read the
cover letter describing the research study, Parenting Stress
accepted the terms of a consent form and Parenting stress was measured with the
then completed the demographic form and Parenting Stress Scale: Autism (Phetrasu-
each questionnaire. Confidentiality was wan, 2003). This 28-item questionnaire
maintained by not using personal identifi- captures stressful parental experiences in
ers on the questionnaire. Participants cre- four subscales (Phetrasuwan, 2003). The
ated their own numeric identifier. The Behaviors and Communication subscale
consent form wording included text that includes questions related to the stress
instructed the participant to ask their dealing with behaviors problems in pub-
spouse to fill out a set of questionnaires, lic, helping the child learn how to be with
however, verification that questionnaires other children, learning how to best com-
were completed by both parents separately municate with the child, helping the child
was not possible. Spouses were matched communicate with others and trying to
based on the numeric identifiers. figure out what their child needs or wants
during a tantrum. The parental caregiv-
Sample ing (caring, protecting) subscale ques-
The 64 female and male partner parents tions relate to feelings about the child not
were recruited from across the United wanting to be touched, problems related to
States via autism list-serves, social net- eating, managing sleep problems, bathing
238 JOHNSON, FRENN, FEETHAM, AND SIMPSON

and dressing difficulties, how to discipline lationships between the family and its sub-
the child, overcoming feelings of protective- systems, such as the division of labor; (3)
ness, keeping life as normal as possible, relationships between the family and the
and keeping the child on a regular routine. individual, focusing on the reciprocal rela-
The advocating for the child’s needs sub- tionships between husband and wife and
scale questions relate to communicating to between parents and children (Roberts &
the day care, school, or health care profes- Feetham, 1982).
sionals about the child’s special needs. The Each item includes a stem item of a
personal and family life subscale questions family function, followed by three ques-
include items such as financial problems, tions: (1) how much there is (2) how much
and finding time for their own activities. there should be, and (3) how important is
For each question, parents rated their this to you? An example of an item is:
stress from 1 ⫽ not stressful, to 5 ⫽ ex- “The amount of time you spend with your
tremely stressful. Responses were summed spouse/partner.” Each item has seven re-
to form a total score, and reported as a sponse options; where 1 represents “lit-
mean total score, ranging from 0 to 140. tle” and 7 represent “much.” There are
Higher scores indicated higher stress. also two open ended questions “What is
Phetrasuwan (2003) examined the content most difficult for you now?” and “What is
and face validity of the PSS:A by asking most helpful to you now?” Scoring for the
parents and professionals working with FFFS is as follows: First the discrepancy
children to review the tool and comment on score is determined for each item. The
its comprehensiveness, clarity, appropri- difference between the (1) and (2) score
ateness, and level of understandability. for each item is converted to an absolute
Cronbach’s alpha for the entire scale in the score. Next, the discrepant scores from
one study (n ⫽ 108) was .94 (Phetrasuwan each item are summed for an instrument
& Miles, 2009). In the present study Cron- score. The total family functioning survey
bach’s alpha was 0.95 for males and 0.94 discrepancy score can range from a pos-
for females. sible score of 0 to 126. A high score D
score indicates a higher discrepancy be-
Support From Family Functioning tween the respondents’ perception of
The Feetham Family Functioning Sur- what is, from their perception of what
vey (FFFS) was used to measure support should be. The importance score is not
from family functioning (Feetham & Hu- included in the instrument score (Roberts
menick, 1982). The FFFS includes dimen- & Feetham, 1982). The scoring allows the
sions of family relationships and outside assessment of individual dissonance or
supports. Since 1984, The FFFS has been dissatisfaction among or within the three
used in 70 published reports in Japan, major areas of family relationships (Rob-
U.S., U.K., South Africa and Australia. The erts & Feetham, 1982). A higher score
FFFS measures the family members’ per- indicates higher discrepancy between de-
ception of their response to any child, re- sired and present functioning. Cron-
gardless of their condition. The survey con- bach’s alpha reliability coefficients for
sists of 25 items, and is designed to be the present study were .91 for males and
self-administered in approximately 10 min .90 for females.
(Roberts & Feetham, 1982; Sawin & Har-
rigan, 1995). Three areas of family func- HRQL: Physical and Mental Health
tions are assessed and supported through The Rand SF 36-Item Health Survey
factor analysis: (1) relationship between (Version 1.0) was used to measure physical
the family and broader social units, such as and mental health as two distinct compo-
schools and work outside the home; (2) re- nents. The tool has 36 items that cover
 AUTISM SPECTRUM DISORDER 239

eight health concepts. These concepts are: month of respondent’s birthday, month of
physical functioning (10 items), bodily pain spouse/partner’s birthday, initial of respon-
(2 items), role limitations because of health dent’s first given name, initial of spouse/
problems (4 items), role limitations be- partner’s first given name and the current
cause of personal or emotional problems (3 state of residence. Male-female parent cou-
items), emotional well being (5 items), so- ples were identified visually by the statis-
cial functioning (2 items), energy/fatigue (4 tician by looking at the data set for the
items), general health perceptions (5 seven matches. Participants (n ⫽ 128) who
items), and a single item that provides an completed all four surveys were used for
indication of perceived health. Cronbach’s the data analysis.
alpha reliability coefficients for subscales Checks for agreement between male
in the present study ranged from .81 to .93. and female parents were performed on gen-
First, eight separate scores were calcu- der of their child, age of diagnosis, and the
lated. Twenty-nine questions were rated on diagnosis. Gender of the child did not
a Likert scale ranging from 3 to 6 points, match in one case. Diagnosis of child dif-
and seven questions are yes/no. All ques- fered in one case. Age of diagnosis differed
tions were recoded to 0 –100% representa- in five cases. For discrepant information
tions according to a scoring guide (Ware, the females’ viewpoint was used.
Kosinski, & Keller, 1994). Next, an addi- Descriptive statistics were used in ana-
tional scoring technique for the HRQL in- lyzing demographic data, the severity of
strument yielded two scores: the physical ASD, family functioning, and for each scale
component score (PCS) and the mental or survey. Discrepancies scores (D scores)
component score (MCS). The eight subscale were calculated for the items on the FFFS.
scores were standardized according to the The D score is equal to the absolute score of
formulas in the user’s manual. Psychomet- the (a) “what is score” minus the (2) ‘how
rically based summary measures aggre- much there should be’ score. The D score
gate the eight concepts with different was computed for each item and summed
weights to two components, the PCS (Cron- for each respondent.
bach’s ␣ ⫽ .91) and the MCS (Cronbach’s Pearson r correlations were used to ex-
␣ ⫽ .92), without substantial loss of infor- amine zero order correlations for all of the
mation (Ware, Kosinski, & Keller, 1994). study variables: demographics, stress, fam-
The mean scores for both the PCS and the ily functioning D score, and parental
MCS of the general U.S. population are 50 HRQL (PCS and MCS separate). Signifi-
(SD ⫽ 10) (Ware, Kosinski, & Keller, 1994). cant correlations were used in further anal-
yses. Multiple regressions for male and
Statistical Analysis female SF 36 subscales will all variables
Participants completed the surveys at included with and without D score were
the www.surveymonkey.com Web site. The run first. Then, stepwise linear regression
data were entered by the participant and was used to develop models to predict men-
automatically entered into an Excel spread tal and physical component scores of the
sheet. The Excel file was imported to SPSS HRQL measure. Backwards linear regres-
16.0 (SPSS, 2008). Each participant had sion was done as a check.
the opportunity to fill out demographics A Wilcoxon signed-ranks test was per-
and the three surveys. Each survey began formed to detect differences between the
with the same seven questions for match- two related groups (males’ and females’
ing purposes, for identification of the male- scores for the discrepancy between what is
female parent couples, while allowing for and should be FFFS). Baron and Kenney’s
anonymity. The questions included: gen- (1986) mediation steps were followed in es-
der, birth month of oldest child with ASD, tablishing mediation models (X ⫽ parent-
240 JOHNSON, FRENN, FEETHAM, AND SIMPSON

ing stress; M ⫽ D score of the FFFS survey; Descriptive Statistics

and Y ⫽ PCS, MCS). If the effect of the Results of parenting stress, family func-
stress parameters decreased, the D scores tioning discrepancy (D) score, and MCS
would be considered a mediator for the and PCS are reported in Tables 1 and 2.
stress effect on HRQL (PCS and MCS). Overall, both male and female parents re-
ported somewhat to moderate amounts of
RESULTS parenting stress. The highest mean stress
Study Participants subscale scores were on the personal and
The number of respondents who com- family life subscale for both males and fe-
pleted all aspects of the study was 64 male males. Both male and female parents had a
and 64 female parents. The majority of the high discrepancy (D) score for ‘how much
respondents were white (n ⫽ 60 males, there is now’ and ‘how much there should
95%, n ⫽ 60 females, 95%), married (n ⫽ be’ family functioning. Both males and fe-
60, 94%), college graduates (n ⫽ 40 males, males had low levels on the mental health
63%, n ⫽ 44 females 69%). The length of component and normative scores of 50
the relationship for the parents ranged (Ware et al., 1994) on the physical health
from 2–31 years (M ⫽ 14.3, SD ⫽ 5.8). The component of the Rand 36 item Health In-
number of the children in the home ranged ventory. Females (MCS ⫽ 34.21 ⫾ 13.11)
from 1–5 (M ⫽ 2.2, SD ⫽ 0.8). Fathers in had lower scores than males (MCS ⫽
this study were between 25 and 64 years of 41.60 ⫾ 12.93).
age (M ⫽ 42, SD ⫽ 7.8). Mothers were
between 29 and 69 years of age (M ⫽ 39.2, Correlations
SD ⫽ 7.2). The children in the study Correlations of the key study variables
ranged in age from 2 to 24 years of age are found in Table 3. For parenting stress,
(M ⫽ 7.8, SD ⫽ 3.9). Most were male (81%), all four subscales of the PSS:A correlated
and had a diagnosis of Autism (62%), or for the male-female partners: advocating
PDD-NOS (21%), or Aspergers (17%). for child’s needs (r ⫽ .32, p ⫽ .05), behavior

Table 1
Means and SDs for the Female Parent Stressor Scale: Autism Scale, Feetham Family
Functioning Survey (FFFS), and Physical Health Component (PCS) and Mental Health
Component (MCS) of the Rand 36-Item Heath Survey
Mean (SD) Range
ⴱ
Parental Stressor Scale: Autism (N)
Total scale (59) 83.27 (24.97) 20–137
Subscales
Behavior and communication (62) 19.95 (5.34) 6–30
Advocating for needs (64) 11.88 (4.41) 4–20
Parental care giving (63) 20.19 (7.85) 5–40
Personal and family life (62) 31.71 (11.12) 5–49
Feetham Family Functioning Survey (N ⫽ 64)ⴱ
Total scale C score 131.58 (22.33) 71–174
Total D score (A-B) 41.97 (21.14) 2–95
PCS and MCS of Rand 36-item heath (N ⫽ 59)ⴱ
PCS 49.10 (11.22) 23.30–69.80
MCS 34.21 (13.11) 6.00–58.90
Note. A ⫽ How much there is family functioning, B ⫽ How much there should be family
functioning, C ⫽ How important it is, D ⫽ A-B.
ⴱ
Complete data on all measures was not available.
 AUTISM SPECTRUM DISORDER 241

Table 2
Means and SDs for the Male Parent Stressor Scale: Autism Scale, Feetham Family Functioning
Survey (FFFS), and Physical Health Component (PCS) and Mental Health Component (MCS) of
the Rand 36-Item Heath Survey
Mean (SD) Range
Parental Stressor Scale: Autism (N)ⴱ
Total scale (61) 71.18 (25.70) 25–130
Subscales
Behavior and communication (64) 17.78 (5.54) 6–29
Advocating for needs (63) 9.73 (4.09) 0–19
Parental care giving (63) 18.16 (7.67) 5–37
Personal and family life (62) 25.74 (11.68) 0–49
Feetham Family Functioning Survey (N ⫽ 64)ⴱ
Total scale C score 117.84 (23.57) 50–168
Total D score (A-B) 33.06 (20.11) 0–94
PCS and MCS of Rand 36-item (N ⫽ 61)ⴱ
PCS 51.16 (9.56) 20.50–65.10
MCS 41.60 (12.93) 8.20–62.30
Note. A ⫽ How much there is family functioning, B ⫽ How much there should be family
functioning, C ⫽ How important it is, D ⫽ A-B.
ⴱ
Complete data on all measures was not available.

and communication (r ⫽ .43, p ⫽ .01), pa- family functioning (D score), was associ-
rental caregiving (r ⫽ .39, p ⫽ .01), and ated with lower mental and physical health
personal and family life (r ⫽ .45, p ⫽ .01). component scores. Males and females were
considered separately. A higher discrep-
Question 1 ancy score in family functioning (D score)
To identify if parenting stress was asso- was associated with both male and female
ciated with mental and physical health, the mental health (p ⬍ .001), accounting for
most statistically significant stress sub- 31% of the variance for females and 35% of
scales for female mental health (stress of the variance for males (see Table 4).
personal and family life) and the most sta-
tistically significant stress subscale for fe- Question 3
male physical health (stress of caregiving), Question 3 asked if discrepancy in ex-
from the correlation table, were entered in pectations of family functioning mediates
a stepwise regression with the outcome the relationship of stress (subscales) on
variables female physical and mental com- mental (MCS) or physical health (PCS).
ponent scores. The stress of caregiving was Males and females were considered sepa-
associated with female PCS (R2 ⫽ .20, p ⬍ rately.
.001, B ⫽ ⫺18) but not the MCS. For males, only the stress of personal
For males, only the stress of personal and family life subscale was associated
and family life subscale was associated with the discrepancy (D) score (R2 ⫽ .33,
with the MCS (R2 ⫽ .29, p ⬍ .001, B ⫽ p ⬍ .001, B ⫽ ⫺.38) and the MCS (R2 ⫽ .29,
⫺.63). Only the stress of personal life sub- p ⬍ .001, B ⫽ ⫺.63). Only the stress of
scale was associated with male PCS (R2 ⫽ personal life subscale was associated with
.07, p ⫽ .00). male PCS (R2 ⫽ .07, p ⫽ .00), but there was
no decrease in the effect on the PCS with
Question 2 the addition of the discrepancy (D) score.
Next, we tested whether or not a wide Therefore, the discrepancy (D) score medi-
discrepancy (D score) in expectations about ated the effect of personal and family life
 242

Table 3
Correlational Analysis of All Key Variables for Males and Females in Hypothesized Model
1 2 3 4 5 6 7 8 9 10 11 12 13 14
1. Stress advocate male 1
2. Stress advocate female .32ⴱ 1
3. Stress B/C male .60ⴱⴱ .23 1
4. Stress B/C female .34ⴱⴱ .68ⴱⴱ .43ⴱⴱ 1
5. Stress care give male .52ⴱⴱ .27ⴱ .74ⴱⴱ .39ⴱⴱ 1
6. Stress care give female .21 .70ⴱⴱ .27ⴱ .66ⴱⴱ .45ⴱⴱ 1
7. Stress PL male .60ⴱⴱ .20 .74ⴱⴱ .32ⴱ .76ⴱⴱ .34ⴱⴱ 1
8. Stress PL female .39ⴱⴱ .69ⴱⴱ .32ⴱ .65ⴱⴱ .41ⴱⴱ .76ⴱⴱ .45ⴱⴱ 1
9. Total male D score .29ⴱ .10 .49ⴱⴱ .25 .29ⴱ .14 .49ⴱⴱ .21 1
10. Total female D score .34ⴱⴱ .58ⴱⴱ .50ⴱⴱ .57ⴱⴱ .49ⴱⴱ .57ⴱⴱ .50ⴱⴱ .70ⴱⴱ .44ⴱⴱ 1
11. MCS male ⫺.23 ⫺.20 ⫺.25 ⫺.07 ⫺.32ⴱ ⫺.03 ⫺.38ⴱⴱ ⫺.21 ⫺.55ⴱⴱ ⫺.27ⴱ 1
12. MCS female ⫺.17 ⫺.42ⴱⴱ ⫺.25 ⫺.34ⴱ ⫺.21 ⫺.48ⴱⴱ ⫺.28ⴱ ⫺.52ⴱⴱ ⫺.52ⴱⴱ ⫺.58ⴱⴱ .31ⴱ 1
13. PCS male .12 .06 .00 .22 .18 .24 .15 .27ⴱ ⫺.19 .08 ⫺.13 ⫺.14 1
14. PCS female ⫺.11 ⫺.41ⴱⴱ ⫺.01 ⫺.36ⴱⴱ ⫺.03 ⫺.41ⴱⴱ ⫺.00 ⫺.30ⴱ ⫺.05 ⫺.37ⴱⴱ ⫺.26 .12 ⫺.00 1
Note. B/C ⫽ Behavior and communication; PL ⫽ Personal Life; MCS ⫽ Mental health component of the Rand 36 item Health Inventory
JOHNSON, FRENN, FEETHAM, AND SIMPSON

Version 1.0; PCS ⫽ Physical health component of the Rand 36 item Health Inventory Version 1.0.
ⴱ
Correlation is significant at the 0.05 level two-tailed. ⴱⴱ Correlation is significant at the 0.01 level two-tailed.
 AUTISM SPECTRUM DISORDER 243

stress on male mental health (MCS) but

.002

⬍.001

⬍.001

⬍.001

⬍.001
p
not physical health (PCS).
For females only the stress of caregiving
predicted the female physical health (PCS).
For female mental health, only the stress of
Standardized B

personal and family life was associated

with the female MCS. For females, the D
Variables statistics

⫺.43

⫺.60

⫺.61

⫺.45

⫺.57
score mediated the effect of the stress of
personal and family life (p ⬍ .001) on MCS.
While the stress of caregiving was a signif-
icant variable impacting physical health
(p ⬍ .001), the D score did not mediate this
effect (see Table 4).
We also entered demographics in the
SE B

.15

.08

.08
.05

.20

.07
regression. For male parents, in Step 1 of
the model (see Table 4) only the older age of
the parent was associated with male phys-
ical health (PCS), accounting for 17% of the
⫺.49

⫺.43

⫺.43
.23

⫺.68

⫺.34

variance. In the second model, the number

B

of children living at home and discrepancy

score for “how much there is” - “how much
there should be” family functioning ac-
counted for 39% of the variance in male
R2 ⫽ .19, adjusted R2 ⫽ .17

R2 ⫽ .36, adjusted R2 ⫽ .35

R2 ⫽ .42, adjusted R2 ⫽ .39

R2 ⫽ .20, adjusted R2 ⫽ .18

R2 ⫽ .32, adjusted R2 ⫽ .31

F(1, 48) ⫽ 11.05, p ⫽ .002

mental health (MCS). For female parents,

F(1, 48) ⫽ 27.34, p ⬍.001

F(2, 47) ⫽ 16.69, p ⬍.001

in the first model, parental caregiving

stress accounted for 18% of the variance for
female physical health (PCS). In the second
model, the discrepancy score for “how much
there is” - “how much there should be” fam-
3.48 1.70

ily functioning accounted for 31% of the

variance in female mental health (MCS). In
Model 2, Step 2 the discrepancy score along
with more children accounted for 39% of
the variance in MCS.
Model of Male and Female PCS and MCS

Number of children live with fulltime

Hypothesis Testing
Hypothesis A stated that parent percep-
tions between “what is” and “what should
be” family functioning will be associated
with lower HRQL (physical and mental
Stress parent care giving
Models for female parents
Variables

health). The relationship was partially

Models for male parents

Total female D score

Model 2: Female MCS

supported. Using Pearson’s r, males’ D

Model 1: Female PCS
Model 1: Male PCS

Total male D score

Total male D score

Model 2: Male MCS

scores correlated ⫺.55 (p ⬍ .01) with their

MCS, but did not correlate significantly
with PCS. Females’ D scores correlated
⫺.58 (p ⬍ .01) with their MCS and ⫺.37
Step 1

Step 2

(p ⬍ .01) with PCS.

Table 4

Age

Hypothesis B stated that mothers will

have significantly greater discrepancy
244 JOHNSON, FRENN, FEETHAM, AND SIMPSON

scores in family functioning than fathers.

Upper bound
The results supported hypothesis B. For

95.0% Confidence interval

81.613
13.486
.626
3.163
.594
⫺.204
1.761
1.576
the FFFS there was a significant difference
(z ⫽ ⫺3.12, p ⫽ 002) for the discrepancy
scores on the FFFS comparing males and

for B
females.

Lower bound
Additional Female Models

42.040
⫺4.041
⫺2.552
⫺4.294
⫺.752
⫺1.014
⫺.908
⫺11.069
Additional females models of PCS (see
Table 5) and MCS (see Tables 6 and 7),
were developed with variables that were
significantly correlated with the outcome
variable (PCS or MCS). We used regression

.000
.283
.228
.761
.814
.004
.523
.137
to see which variables were significantly

Sig.
related once the other variables were con-
sidered. Because of the high intercorrela-
tion of variables we noted that we had ho-

6.297
1.086
⫺1.222
⫺.306
⫺.237
⫺3.028
.644
⫺1.513
moscedascity with for example coefficients

Standardized

t
coefficients
of the parenting stress subscales alternat-
ing in sign when we simply entered all
variables. Therefore, we used backward

.150
⫺.212
⫺.043
⫺.042
⫺.428
.129
⫺.217
Beta
and stepwise regression with an entry cri-
teria for a variable of p ⬍ .05 and a removal
criterion of 0.1 to find what was associated
with the outcomes. Table 5 and 6 displays
9.818
4.348
.788
1.850
.334
.201
.662
3.137
Unstandardized

the results of the amount of variance ex-

SE
coefficients

plained by each variable.

DISCUSSION
61.827
4.722
⫺.963
⫺.566
⫺.079
⫺.609
.426
⫺4.746
Findings for parenting stress totals on
B

the PSS:A included somewhat to moderate

amounts of parenting stress, with a total
scale mean of 71.18 (SD ⫽ 25.7) for the
Model of Female PCS of the SF-36 (Version 1.0)

Child gender R2 ⫽ .223, Adjusted R2 ⫽ .10

male and female parents. These means are

similar to the total scale mean for a female
sample reported by Phetrasuwan (2003) of
Number of children full or part time
(Constant) F(7, 51) ⫽ 1.808, p ⫽ .11

75.01 (SD ⫽ 22.65). The large standard

Length in years of relationship

deviations suggest that the data may be

bimodal or multimodal. Future analysis
could be done to assess for this phenome-
Stress parent caregiving

non. The Easter Seals (2009) study re-

Child age at diagnosis

ported that a primarily female sample was

stressed by financial concerns for their
Child diagnosis

ASD-affected child. Financial concerns are

captured in the personal/family life sub-
Child age

scale of the PSS:A. That subscale was the

only stress subscale that was associated
Table 5

with poor male and female mental health

in the present study. Both male and female
1
 AUTISM SPECTRUM DISORDER 245

parents perceived that dealing with the

Upper bound
95.0% Confidence interval
personal and family life was the most

59.695
10.626

1.044
4.943
.674
3.029
9.611
.294
⫺.006
.239
stressful, and the personal and life issues
stress was associated with poorer mental
health outcomes. Past researchers reported
for B

parents perceived they were most stressed

by their child’s behaviors (Clark, 2007;
Lower bound

15.832
⫺10.287

⫺2.648
⫺3.253
⫺.797
⫺.045
⫺4.520
⫺.579
⫺.517
⫺.203
Estes et al., 2009; Hastings, 2003; Herring
et al., 2006; Hoffman et al., 2008).
In the present study, measuring stress
with the PSS:A, which has a personal and
family life stress subscale advances the
state of the state of the science on under-
sign.

.001
.974

.386
.679
.867
.057
.471
.512
.045
.871
standing stressors associated with mental
health in this population. Parents would
benefit from a discussion focused on plan-
3.475
.033

⫺.877
.416
⫺.168
1.960
.727
⫺.661
⫺2.065
.163

ning how to deal the personal and family

t

life stressors captured in the subscale of

the PSS:A such as differences in opinions,
how to balance the needs of other family
Standardized
Model of Female Mental Health Component Score (MCS) of the SF-36 (Version 1.0)

coefficients

members and work responsibilities, and

.004

⫺.144
.051
⫺.026
.367
.094
⫺.119
⫺.392
.024
Beta

managing finances.
For family functioning, females in the
present study had a wider range for dis-
crepancy scores than males. In addition,
male D score and female D scores were
10.868
5.181

.915
2.031
.364
.761
3.501
.216
.127
.110
Unstandardized

positively related. When males had a wide

SE
coefficients

discrepancy between expectations and per-

ceived functioning, the females also had a
wide discrepancy in expectations. A differ-
37.763
.170

⫺.802
.845
⫺.061
1.492
2.545
⫺.143
⫺.262
.018

ence in what was valued could account for

B

the difference. Past research notes that

males tend to value financial contributions
and females value emotional support
Child age diagnosed with ASD R2 ⫽ .436,

Number of children full-time or part-time

Child gender F(9, 51) ⫽ 3.611, p ⫽ .002

(Deris, 2005).
Males and females differed in their ex-
pectations for supportive family function-
Variables model statistics

Length in years of relationship

ing. The total mean discrepancy score for

Stress: personal life subscale

the FFFS (see Table 1) was 33.06 for men

and 41.97 for women. The discrepancy
scores for men match findings of past stud-
Total female D score

ies. Feetham et al. (2007) reported that

Total male D score
adjusted R2 ⫽ .315

discrepancy scores tend to range from 17–

35. The females’ scores were higher than
the norm indicating that women caring for
(Constant)

Diagnosis
Child age

a child with ASD are not getting what they

expect from supportive family functioning.
Table 6

In the present study HRQL was mea-

sured by the Rand 36-Item Health Survey
1
 246

Table 7
Model of Female Mental Health Component Score (MCS) of SF-36 (Versions 1.0)
Unstandardized Standardized 95.0% Confidence interval
coefficients coefficients for B
Variables model statistics B SE Beta t Sig. Lower bound Upper bound
1
(Constant) F(12, 48) ⫽ 2.45, p ⫽ .019 30.071 12.994 2.314 .026 3.718 56.425
Child gender R2 ⫽ .45, Adjusted R2 ⫽ .266 2.472 5.771 .062 .428 .671 ⫺9.233 14.177
Child age diagnosis ⫺.841 1.040 ⫺.155 ⫺.809 .424 ⫺2.950 1.268
Number of children full/part-time ⫺.229 2.314 ⫺.014 ⫺.099 .922 ⫺4.923 4.464
Length of relationship .044 .409 .019 .107 .915 ⫺.785 .872
Stress care giving ⫺.430 .371 ⫺.241 ⫺1.160 .254 ⫺1.182 .322
Stress personal life ⫺.032 .269 ⫺.027 ⫺.119 .906 ⫺.577 .514
Stress advocating needs .289 .677 .090 .427 .672 ⫺1.084 1.663
Stress behavior and communication .427 .529 .166 .806 .425 ⫺.646 1.499
Child age 1.372 .818 .351 1.677 .102 ⫺.287 3.032
Total male D score .015 .121 .021 .125 .901 ⫺.230 .260
Total female D score ⫺.313 .146 ⫺.484 ⫺2.147 .039 ⫺.608 ⫺.017
JOHNSON, FRENN, FEETHAM, AND SIMPSON

Diagnosis 1.350 3.803 .050 .355 .725 ⫺6.363 9.064

 AUTISM SPECTRUM DISORDER 247

1.0, which allowed for two components manage care of the child at home. Male
(MCS and PCS) (Table 1 and 2). The phys- parents may not be able to find time to
ical health scores for both men and women exercise and that may impact their physi-
were close to the reported average of 50 cal health.
(Ware et al., 1994). Conversely, the MCS
scores were much lower than the expected Question 2
norm of 50 (Ware et al., 1994). Females Higher discrepancy scores for family
(MCS ⫽ 34.21 ⫾ 13.11) had lower scores functioning were related to lower mental
than males (MCS ⫽ 41.60 ⫾ 12.93). Past health for both males and females. This
studies with mostly female participants finding extends to male parents the results
have also found psychological distress in of previous studies (Dunn et al., 2001;
parents of children with ASD (Allik, 2006; Hastings et al., 2005) that found a discrep-
Phetrasuwan, 2003). Score comparisons for ancy in expectations negatively affects
parents of children with ASD to other stud- both mental and physical health for fe-
ies were not possible because of the use of males. As the discrepancy in how much
different measures; Lee et al. (2009) used there is and how much there should be for
the MOS 36-Item Short Form Survey the FFFS rose, so too did parenting stress.
(Ware & Sherbourne, 1992); Allik et al. Knafl and Deatrick (2003) found joint effort
(2006) used the SF-12 (Ware, Kosinski, & to manage illness with different views on
Keller, 1996). how to manage the situation can affect
family functioning. Although the parents
Question 1 work together they may have different fu-
The stress of caregiving was associated ture expectations or a plan on how to get
with female physical health but not the there, which is a potential source of dis-
mental health. Previous studies found par- tress.
ents of children with ASD were more
stressed than other parents (Brobst et al., Question 3
2009; Herring et al., 2006; Konstantareas For males, only the stress of personal
& Papageorgiou, 2006) with mothers per- and family life subscale was associated
ceiving more stress than fathers (Herring with the discrepancy (D) score and mental
et al., 2006; Little, 2002). However, the health. Only the stress of personal life sub-
stress of caregiving was associated with scale was associated with male PCS, but
lower female physical health, accounting there was no decrease in the effect on the
for 18% of the variance. The stress of care- PCS with the addition of the discrepancy
giving subscale of the PSS:A includes man- (D) score. Therefore the discrepancy (D)
aging sleep problems, bathing, and dress- score mediated the effect of personal and
ing difficulties, which are physically drain- family life stress on male mental health
ing for female parents. The physical work (MCS) but not physical health (PCS). Ad-
of caregiving impacts women most likely ditional analysis of demographic variables
because they are often the main caregiver found that the older age of the father was
(Little, 2002). associated with male PCS.
For males, only the stress of personal For females the only the stress of care-
and family life subscale was associated giving predicted the female physical health
with the mental and physical health MCS. (PCS). For female mental health, only the
Items on this scale include financial prob- stress of personal and family life was asso-
lems, and finding time for their own activ- ciated with the female MCS. For females,
ities. Male parents are the traditional the D score mediated the effect of the stress
wage-earner in the family (Easter Seals, of personal and family life on mental
2009). Often mothers of children with ASD health. While the stress of caregiving was a
248 JOHNSON, FRENN, FEETHAM, AND SIMPSON

significant variable impacting physical the family, division of labor was reported
health the D score did not mediate this as being the most common contributor to
effect. The findings support the two theo- marital conflict (IAN, 2009).
ries that framed the study. In the present
study, the social network is conceptualized Limitations
as the male–female partners’ expectations There are admitted limitations in the
about supportive family functioning (that study that include a one-time data collec-
may include members from a network out- tion and not confirming the diagnosis of
side the immediate family) that are ASD. Lack of verification that both parents
thought to mediate stress. Networks also completed questionnaires separately is
have a positive impact on parental health also a potential weakness. ASD tends to be
according to this model. Without the sup- diagnosed in children of well-educated par-
portive family functioning, both males and ents (Meter et al., 2010). Mulvihill et al.
females are at risk for poorer mental (2009) report that the race and ethnicity
health. distribution in Wisconsin is 65.7% White,
The findings in the present study are nonhispanic; 18.1% Black, non-Hispanic;
also consistent with past research where 12.2% Hispanic, and 4% Other. Therefore,
support is conceptualized as a result of a the recruited sample of parents in the pres-
social process. Mothers with psychological ent study is congruent with a population
distress perceived low social relationship with high diagnostic incidence. The re-
support in past studies with parents of chil- cruitment strategy was a strength in ac-
dren with ASD (Bromley et al., 2004; Tob- cessing those with a diagnosis, although
ing & Glenwick, 2006) and had children there are thought to be many children of
with problem behaviors (Allik et al., 2006). other races and ethnicities with character-
istics of ASD. In the present study, fathers
Hypotheses Testing and mothers of children with ASD partici-
Hypothesis A was partially supported. pated. The literature review highlighted
Parent perceptions between “what is” and the historic challenge of recruiting fathers
“what should be” family functioning were of children with ASD and other health con-
correlated mental health and female ditions in research studies. Placing flyers
health for females and mental health only on the Internet at sites and in clinics that
for males. Hypothesis B, was supported as parents of children with ASD were likely to
mothers had a significantly greater dis- access, worked well in the present study.
crepancy scores in family functioning than As the SF 36 proved cumbersome to use as
fathers. Because mothers had the wider an outcome measure, the SF-12 is recom-
discrepancy in expectations, and the dis- mended for future research. Dyadic analy-
crepancy was correlated to both mental sis is recommended as a basis for interven-
and physical health in mothers, additional tions with dyads (Knafl et al., 2009; Knafl,
female models were developed. Knafl, & McCorkle, 2005). A manuscript on
this is in progress (Johnson, Feetham,
Additional Female Models Knafl, & Simpson, 2011).
Additional females models of PCS and
MCS showed that the stress of caregiving Clinical Applications
was associated with female physical health The main finding of the study was that
(see Table 5) and the female D score was there were significant differences in expec-
associated with mental health (see Tables tations between parents related to support-
6 and 7). In past studies parents reported ive family functioning. The discrepancy
receiving little support from extended fam- between expectations about supportive
ily members (Easter Seals, 2009). Within family functioning was associated with
 AUTISM SPECTRUM DISORDER 249

poor mental health for parents. Wider dis- child with ASD as stressful, the stress of
crepancy in expectations about supportive personal and family life (finances, work re-
family functioning was also associated with sponsibilities, balancing the needs of sib-
parenting stress. lings) and the discrepancy in expectations
The significant difference in expecta- related to family functioning were nega-
tions suggests needs that must be ad- tively associated with parents’ mental
dressed by clinicians. It may be useful for health. Furthermore, the discrepancy in
clinicians to facilitate a conversation be- expectations mediated the mental health
tween parents to allow parents time to dis- in both parents. We know that family func-
cuss and understand their expectations. tioning is a process amenable to change.
Once the expectations are understood it Health care professionals should encour-
may be possible for parents to reach con- age family members to reflect on their dif-
sensus on alternate ways to make decisions ferent expectations. Central to a family as-
and provide care for their child. Possible sessment for quality health care is knowing
approaches include: (1) assess mothers’ the structure of a family and knowing what
and fathers’ expectations, then foster dis- family functions are being met and who is
cussion about areas of similarity and assisting the meeting of the functions
difference; (2) affirm the normality of dif- (Feetham, 2005). The reflection affords the
ferent expectations; (3) encourage consid- family an opportunity to negotiate the un-
eration of alternate ways expectations met expectations.
could be met and discussion of different Females’ physical health was impacted
expectations; and (4) foster problem- by the physical tasks related to caring for
solving and resource identification. Fami- children, which becomes more physically
lies may need more external support to demanding as a child ages. Future re-
help (particularly) mothers have better search should look at the amount of time
mental and physical functioning in addi- spent with the child and the age of the
tion to the focus on internal family dynam- child: both may increase stress as the child
ics and relationships. and family age.
More children living full with the par- New research should relate to the cur-
ents, indicating possible extra sibling rent state of the science. Gender differ-
stress and financial burdens, were also as- ences were reported, but the amount of
sociated with poor male mental health. Fe- time spent with the child was not reported.
males had lower PCS and MCS scores than Future research could assess if differences
males, and higher stress scores on all the were a result of mothers spending more
stress subscales than the males. The time with and having more responsibility
amount of variance that the discrepancy for the child. Effective strategies to recruit
score accounts for is large. Interventions and retain fathers or male partners are
addressing the discrepancy in expectations also needed. The expectation that even par-
are likely to contribute to improved mental ents of the same child may well have dif-
health for parents. ferent experiences and views could also be
opened up for discussion and thereby nor-
CONCLUSIONS malized with parents of children with ASD.
In conclusion, the findings of the study
guide health care professionals in their as-
REFERENCES
sessment and intervention planning re-
lated to parenting stress, family function- Abidin, R. (1995). Parenting Stress Index: Pro-
fessional manual. Odessa, FL: Psychological
ing, and physical and mental health for Assessment Resources.
parents of children with ASD. While past AHRQ. (2009). Research Reviews: Comparative
studies highlight behavior problems of the effectiveness of therapies for children with Au-
250 JOHNSON, FRENN, FEETHAM, AND SIMPSON

tism Spectrum Disorders. U.S. Department of Dissertation, University of Kentucky, Lex-

Health and Human Services Agency for ington, KY.
Healthcare Research and Quality, Retrieved Cohen, S., & McKay, G. (1984). Social support,
from http://effectivehealthcare.ahrq.gov/ stress, and the buffering hypothesis: A The-
healthInfo.cfm?infotype⫽rr&processID⫽106& oretical analysis. In A. Baum, S. E. Taylor, &
status⫽d J. E. Singer (Eds.), Handbook of psychology
Allik, H., Larsson, J.-O., & Smedje, H. (2006). and health (Vol. 4, pp. 253–267). Hillsdale,
Health-related quality of life in parents of NJ: Erlbaum.
school-age children with Asperger Syndrome Cutrona, C. E., & Russell, D. W. (1990). Type of
or High-Functioning Autism. Health & Qual- social support and specific stress: Toward a
ity of Life Outcomes, 4, 1. doi:10.1186/1477- theory of optimal matching. In B. R. Sarason,
7525-4-1 I. G. Sarason, & G. Pierce (Eds.), Social sup-
Altiere, M. J., & Von Kluge, S. (2009). Family port: An interactional view (pp. 319 –336).
functioning and coping behaviors in parents New York, NY: Wiley.
of children with autism. Journal of Child Deris, A. R. (2005). Social supports among par-
and Family Studies, 18, 83–92. doi:10.1007/ ents of children recently diagnosed with au-
s10826-008-9209-y tism: Comparisons between mothers and fa-
American Psychiatric Association. (2000). Diag- thers. Unpublished Dissertation, University
nostic and statistical manual of mental dis- of New Orleans, New Orleans, LA.
orders (4th edition ed. TR). Washington, DC: Donenberg, G., & Baker, B. L. (1993). The im-
American Psychiatric Association. pact of young children with externalizing be-
Andrews, M. P., Bubolz, M., & Paolucci, B. haviors on their families. Journal of Abnor-
(1980). An ecological approach to study of the mal Child Psychology, 21, 179 –198.
family. Marriage and Family Review, 31, 29 – Dunn, M. E., Burbine, T., Bowers, C. A., &
49. Tantleff-Dunn, S. (2001). Moderators of
Baker, B. L., Blacher, J., & Olsson, M. (2005). stress in parents of children with autism.
Preschool children with and without develop- Community Mental Health Journal [NLM-
mental delay: Behaviour problems, parents’ MEDLINE], 37, 39.
optimism and well-being. Journal of Intellec- Easter Seals. (2009). Easter Seals’ living with au-
tual Disability Research, 49, 575–590. doi: tism study. Retrieved from http://www
10.1111/j.1365-2788.2005.00691.x .easterseals.com/site/PageServer?pagename⫽
Baron, R. M., & Kenny, D. A. (1986). The mod- ntlc8_living_with_autism_study_home
erator-mediator variable distinction in social Estes, A., Munson, J., Dawson, G., Koehler, E.,
psychological research: Conceptual, strategic Zhou, X., & Abbott, R. (2009). Parenting
and statistical considerations. Journal of stress and psychological functioning among
Personality and Social Psychology, 51, 1173– mothers of preschool children with autism
1182. doi:10.1037/0022-3514.51.6.1173 and developmental delay. Autism, 13, 375–
Brobst, J., Clopton, J., & Hendrick, S. (2009). Par- 387. doi:10.1177/1362361309105658
enting children with Autism Spectrum Disor- Feetham, S. L. (2005). Family nursing: Chal-
ders: The couple’s relationship. Focus on Autism lenges and opportunities: Providing leader-
and Other Developmental Disabilities, 24, 38– ship in family nursing from local to global
49. doi:10.1177/1088357608323699 health. Journal of Family Nursing, 11, 327–
Bromley, J., Hare, D. J., Davison, K., & Emer- 331. doi:10.1177/1074840705280820
son, E. (2004). Mothers supporting children Feetham, S. (2011). The relationship of family
with autistic spectrum disorders: Social sup- to health: Historical overview. In M. C.
port, mental health status and satisfaction Rosenberg (Ed.), Encyclopedia of family
with services. Autism, 8, 409 – 423. doi: health (pp. xxxi–xxxvi). London: Sage Publi-
10.1177/1362361304047224 cations.
Cassano, M., Adrian, M., Veirs, G., & Zaman, J. Feetham, S., & Humenick, S. (1982). The
(2006). The Inclusion of fathers in empirical Feetham Family Functioning Survey. In S.
investigation of child psychopathology: An Humenick (Ed.), Analysis of current assess-
update. Journal of Clinical Child and Ado- ment strategies in the health care of young
lescent Psychology, 35, 583–589. doi:10.1207/ children and childbearing families (pp. 259 –
s15374424jccp3504_10 268). New York, NY: Appleton-Century
Clark, T. A. (2007). Mediation of parents stress Crofts.
through parent attribution and self-efficacy: Ferrans, C. E., Zerwic, J., Wilbur, J., & Larson,
Implications for the parents of children with J. (2005). Conceptual model of health-related
Autism Spectrum Disorders. Unpublished quality of life. Journal of Nursing Scholar-
 AUTISM SPECTRUM DISORDER 251

ship, 37, 336 –342. doi:10.1111/j.1547- Knafl, G. J., Dixon, J., O’Malley, J., Grey, M.,
5069.2005.00058.x Deatrick, J., Gallo, A., & Knafl, K. A. (2009).
Friedman, M., Bowden, V. R., & Jones, E. G. Analysis of cross-sectional univariate mea-
(2003). Family nursing: Research, theory, surements for family dyads using linear
and practice (5th ed.). Upper Saddle River, mixed modeling. Journal of Family Nursing,
NJ: Prentice Hall. 15, 130 –151. doi:10.1177/1074840709331641
Friedrich, W. N., Greenberg, M. T., & Crnic, K. Knafl, G., Knafl, K., & McCorkle, R. (2005).
(1983). A Short Form of the Questionnaire on Mixed models incorporating intra-familial
Resources and Stress. American Journal of correlation trough spatial autoregression.
Mental Deficiency, 88, 41– 48. Research in Nursing & Health, 28, 348 –356.
Hastings, R. P. (2003). Child behaviour prob- doi:10.1002/nur.20082
lems and partner mental health as correlates Knafl, K. A., & Deatrick, J. A. (2003). Further
of stress in mothers and fathers of children refinement of the family management style
with autism. Journal of Intellectual Disabil- framework. Journal of Family Nursing, 9,
ity Research, 47, 231–237. doi:10.1046/ 232–256. doi:10.1177/1074840703255435
j.1365-2788.2003.00485.x Konstantareas, M. M., & Papageorgiou, V.
Hastings, R., Kovshoff, H., Brown, T., Ward, N., (2006). Effects of temperament, symptom se-
Espinosa, F., & Remington, B. (2005). Coping verity and level of functioning on maternal
strategies in mothers and fathers of preschool stress in Greek children and youth with
and school-age children with autism. Autism, 9, ASD. Autism, 10, 593– 607. doi:10.1177/
377–391. doi:10.1177/1362361305056078 1362361306068511
Hastings, R. P., & Johnson, E. (2001). Stress in Lakey, B., & Cohen, S. (2000). Social support
UK families conducting intensive home- theory and measurement. In S. Cohen, L.
based behavioral intervention for their Underwood, & B. Gottlieb (Eds.), Social sup-
young child with autism. Journal of Autism port measurement and intervention (pp. 29 –
and Developmental Disorders, 31, 327–336. 52). Oxford: Oxford University Press.
doi:10.1023/A:1010799320795 Lantos, J. D. (2007). The edge of the known
Herring, S., Gray, J., Taffe, B., Sweeny, D., & world. Health Affairs, 26, 510 –514.
Einfeld, S. (2006). Behaviour and emotional Lazarus, R. S. (1999). Stress and emotion: A new
problems in toddlers with pervasive develop- synthesis. New York, NY: Springer Publish-
mental disorders and developmental delay: ing Company.
Associations with parental mental health Lazarus, R. S., & Folkman, S. (1984). Stress,
and family functioning. Journal of Intellec- appraisal, and coping. New York, NY:
tual Disability Research, 50, 874 – 882. doi: Springer Publishing Company Inc.
10.1111/j.1365-2788.2006.00904.x Lee, G. K., Lopata, C., Volker, M., Thomeer, M. L.,
Hoffman, C., Sweeney, D., Lopez-Wagner, M., Nida, R. E., Toomey, J. A., . . . Smerbeck, A. M.
Hodge, D., Nam, C., & Botts, B. (2008). Chil- (2009). Health related quality of life of parents of
dren with Autism: Sleep problems and moth- children with high functioning autism spectrum
er’s stress. Focus on Autism and Other De- disorder. Focus on Autism and Other Disablities,
velopmental Disabilities, 23, 155–165. doi: 24, 227–239. doi:10.1177/1088357609347371
10.1177/1088357608316271 Little, L. (2002). Differences in stress and cop-
IAN. (2009). IAN Research report #11. Family ing for mothers and fathers of children with
Stress-Part 3: Relationships and resilience. Aspergers syndrome and mom-verbal learn-
Retrieved from http://iancommunity.org/cs/ ing disorders. Pediatric Nursing, 28, 565–
ian_research_reports/ian_research_report_ 570.
sept_2009. McCubbin, H., & Thompson, A. (Eds.). (1991).
Johnson, N., Feetham, S., Knafl, G., & Simpson, Family assessment inventories for research
P. (2011). Health related quality of life for and practice (2nd ed.). Madison, WI: Univer-
parents of children with Autism Spectrum sity of Wisconsin.
Disorder: A dyadic analysis. Manuscript in Meltzer, H., Gatward, R., Goodman, R., & Ford,
process. T. (2000). Mental health of children and ad-
Kersh, J., Hedvat, T. T., Hauser-Cram, P., & olescents in Great Britain. London: Statio-
Warfield, M. E. (2006). The contribution of nery Office.
marital quality to the well-being of parents of Mulvihill, B., Wingate, M., Kirby, R. S., Petty-
children with developmental disabilities. grove, S., Cunniff, C., Meaney, F. J., . . .
Journal of Intellectual Disability Research, Ratchford, A. (2009). Prevalence of autism
50, 883– 893. doi:10.1111/j.1365-2788.2006 spectrum disorders: Autism and Develop-
.00906.x mental Disabilities Monitoring Network Sur-
252 JOHNSON, FRENN, FEETHAM, AND SIMPSON

veillance Year 2006, Centers for Disease established self-report instruments:

Control and Prevention (CDC). MMWR Sur- Feetham Family Functioning Survey
veillance Summary, 58, 1–20. (FFFS). In K. J. Sawin & M. P. Harrigan
Olson, D., McCubbin, H., Barnes, H., Lansen, (Eds.), Measuring family functioning for re-
A., Muxen, H., & Wilson, M. (1985). Family search and practice (pp. 42– 49). New York,
inventories. St. Paul, MN: University of Min- NY: Springer Publishing Company.
nesota. SPSS. (2008). Statistical software package for
Phetrasuwan, S. (2003). Psychological adjust- the social sciences 16.0. Chicago, IL: SPSS.
ment in mothers of children with Autism Tanguay, P. (2006). Autism and pervasive de-
Spectrum Disorder. Chapel Hill, NC: Univer- velopment disorder. In P. Jensen, P. Knapp,
sity of North Carolina, Chapel Hill. & D. Mrazek (Eds.), Toward a new diagnostic
Phetrasuwan, S., & Miles, M. S. (2009). Parent- system for child psychopathology (pp. 150 –
ing stress in mothers of children with Autism 161). New York, NY: The Guilford Press.
Spectrum Disorder. Journal for Specialists Tobing, L., & Glenwick, D. (2006). Predictors
in Pediatric Nursing, 14, 157–165. doi: and moderators of psychological distress in
10.1111/j.1744-6155.2009.00188.x mothers of children with pervasive develop-
Randolph, l. (1977). The CES-D scale: A self mental disorders. Journal of Family Social
report depression scale for research in the Work, 10, 1–22. doi:10.1300/J039v10n04_01
general population. Applied Psychological Van Meter, K. C., Christiansen, L. E., Delwiche,
Measurement, 1, 385– 401. L. D., Azari, R., Carpenter, T. E., & Hertz-
Rice, C. (2009). Prevalence of Autism Spectrum Picciotto, I. (2010). Geographic distribution
Disorders: Autism and Developmental Dis- of autism in California: A retrospective birth
abilities Monitoring Network, United States, cohort analysis. Autism Research, 3, 19 –29.
2006. CDC MMWR Surveillance Summaries, Ware, J., Kosinski, M., & Keller, S. (1994).
58. Retrieved from http://www.cdc.gov/ SF-36 physical and mental heath summary
mmwr/preview/mmwrhtml/ss5810a1.htm scales: A users manual. Boston, MA: New
Roberts, C. S., & Feetham, S. (1982). Assessing England Medical Center.
family functioning across three areas of rela- Ware, J., Kosinski, M., & Keller, S. (1996). A
tionships. Nursing Research, 31, 231–235. 12-item short form health survey: Construc-
doi:10.1097/00006199-198207000-00011 tion and scales and preliminary tests of reli-
Ryff, C. D., & Keyes, A. (1995). The Structure of ability and validity. Medical Care, 34, 220 –
psychological well being revisited. Journal of 233. doi:10.1097/00005650-199603000-00003
Personality and Social Psychology, 69, 719 – Ware, J., & Sherbourne, C. D. (1992). The MOS
727. doi:10.1037/0022-3514.69.4.719 36-Item Short-Form Health Survey (SF-36).
Sawin, K. J., & Harrigan, M. P. (1995). Well- Medical Care, 30, 473– 483.