J Autism Dev Disord (2010) 40:1274–1284

DOI 10.1007/s10803-010-0986-y

ORIGINAL PAPER

Optimism, Social Support, and Well-Being in Mothers of Children

with Autism Spectrum Disorder
Naomi V. Ekas • Diane M. Lickenbrock •

Thomas L. Whitman

Published online: 2 March 2010

Ó Springer Science+Business Media, LLC 2010

Abstract This study used structural equation modeling to services (Whitman 2004). As a result of coping with these
examine the relationship between multiple sources of and many other challenges associated with raising a child
social support (e.g., partner, family, and friends), optimism, with ASD, parents report greater levels of depression
and well-being among mothers of children with ASD. (Olsson and Hwang 2001), higher levels of stress (Benson
Social support was examined as a mediator and moderator 2006), and generally lower overall well-being (Ekas et al.
of the optimism-maternal well-being relationship. More- 2009).
over, the role of optimism as a mediator of the social Social support has been identified as a critical factor that
support-maternal well-being relationship was also evalu- reduces the negative psychological effects of raising a child
ated. Results revealed that family support was associated with ASD as well as other disabilities (Bishop et al. 2007;
with increased optimism that, in turn, predicted higher Bromley et al. 2004; Hassall et al. 2005). In particular,
levels of positive maternal outcomes and lower levels of informal support, such as that provided by friends and
negative maternal outcomes. In addition, partner and friend family, has been shown to be effective in reducing stress
support were directly associated with maternal outcomes. among mothers of children with ASD. For example,
Implications for the development of interventions directed mothers of children with autism who perceive receiving
at increasing the quality of social support networks are higher levels of support, especially from spouses and rel-
discussed. atives, report lower levels of depression-related somatic
symptoms and fewer marital problems (Dunn et al. 2001).
Keywords Autism
Social support
Optimism
Several studies have also shown that mothers first turn to
Mediation
Psychological well-being their spouse for support, then to their immediate family,
and finally to other parents of children with disabilities (see
Boyd 2002 for a review).
Introduction Although different types of informal support have been
associated with increased well-being, research has not
Parents of children with autism spectrum disorder (ASD) systematically examined whether one source of informal
face numerous challenges, including obtaining a diagnosis, support is more effective than another in helping mothers
finding appropriate treatment and educational programs, of children with ASD cope with stress. For example,
and struggling with the financial burden of paying for although mothers of children with ASD may turn to their
spouse first (e.g. Boyd 2002), their spouse may not be the
most effective support. It is possible that their spouse may
N. V. Ekas
D. M. Lickenbrock
T. L. Whitman be equally distressed and unable to provide effective sup-
University of Notre Dame, Notre Dame, IN, USA port (Coyne et al. 1990) and that extended family (e.g.,
parents, brothers, sisters, etc.) or close friends may be more
N. V. Ekas (&)
beneficial in promoting mothers’ well-being. Indeed,
University of Miami, PO Box 248185, Coral Gables,
FL 33124-0751, USA research suggests that women turn to individuals outside
e-mail: [email protected] their marriage (e.g., friends and family) when they do not

123
J Autism Dev Disord (2010) 40:1274–1284 1275

receive adequate support from their spouse (Julien and supported, an individual’s cognitive representations about
Markman 1991). Moreover, among mothers of critically ill the future may be changed. Indeed, Karademas (2006)
children, social support from friends and family has been found support for a model in which optimism mediated the
shown to buffer the adverse effects of having low spousal social support-psychological well-being relationship in a
support (Rini et al. 2008). The current study investigated community adult sample, but not for a model in which
the relative effects of three types of informal social support social support mediated the optimism-psychological well-
(support from partners, other family members, and friends) being relationship.
on the overall well-being of mothers of children with ASD. Although there are limited studies examining the sepa-
In addition to the impact of extrinsic influences, such as rate effects of social support and optimism among mothers
social support, characteristics of the individual may also of children with ASD (e.g., Bristol 1984; Greenberg et al.
impact well-being. Previous research has found that 2004), no studies to date have specifically examined the
personality characteristics, such as locus of control and dynamic relationship between these two constructs and
hardiness, contribute to the well-being of mothers of chil- how the combination of the two impact well-being. In the
dren with ASD (Dunn et al. 2001; Gill and Harris 1991; present study two competing models were examined: (a)
Weiss 2002). One personality characteristic that may be social support as a mediator of the optimism-maternal well-
especially important in promoting better psychological being relationship, and (b) optimism as a mediator of the
functioning in mothers of children with ASD is optimism. social support-maternal well-being relationship. In addition
Optimism refers to the tendency of an individual to expect to the direct and indirect benefits of social support, it is also
positive outcomes in life (Scheier and Carver 1985). There possible that social support serves a moderating function
is growing evidence that optimism is associated with better (Cohen and Wills 1985), with social support buffering the
physical and mental health outcomes in a variety of pop- negative effects of low optimism. Shelby et al. (2008)
ulations (Achat et al. 2000; Conway et al. 2008; Scheier found that, among women with breast cancer, high social
and Carver 1992; Segerstrom et al. 1998). Optimistic support moderated the adverse influences of low optimism
patients exhibit faster recovery from cardiac bypass sur- on psychological distress, well-being, and psychosocial
gery (Scheier et al. 1989), and are more likely to employ functioning. For this reason, the present study also exam-
positive health practices (Scheier and Carver 1992). ined the role of social support as a potential moderator of
Among mothers of adults with autism, optimism has been the optimism-maternal well-being relationship.
associated with greater psychological well-being (Green-
berg et al. 2004). To date, there have been no studies Present Study and Hypotheses
investigating the effects of optimism on the well-being of
mothers raising children with ASD. For that reason, the The present study examined the relationships between
present study also examined the effects of optimism on optimism, informal social support, and well-being in a
maternal well-being. sample of mothers of children with ASD. More specifi-
Increasing research indicates that social support and cally, the role of three types of informal social supports,
optimism are positively related (e.g. Dougall et al. 2001; spouse, friends, and family, were examined as mediators of
Park and Folkman 1997). Moreover, social support has the optimism-maternal well-being relationship. In mea-
been proposed as an important mediator of the relationship suring maternal well-being, both negative maternal out-
between optimism and psychological adjustment (Dougall comes, such as depression, negative affect, and parenting
et al. 2001); specifically because optimistic individuals stress, and positive maternal outcomes, such as positive
appear to be able to establish greater social support affect, life satisfaction, and overall psychological well-
resources and build more effective social networks in times being were included. Past research has suggested that
of stress. Several studies have found evidence for this positive and negative aspects of psychological functioning,
mediational relationship among breast cancer survivors under typical conditions, may be separate dimensions with
(Shelby et al. 2008), survivors of traumatic events (Dougall different influences, whereas in times of elevated stress
et al. 2001), and college students (Brissette et al. 2002). these two dimensions become highly related (Ekas et al.
These studies, however, have not examined the differential 2009; Zautra et al. 2001). For this reason, it seems espe-
impact of multiple types of social support. cially important for studies to include both negative and
Although social support has typically been considered as positive dimensions in order to provide a more complete
a mediating variable, it is also possible that optimism may picture of well-being in mothers of children with ASD,
mediate the social support-psychological well-being rela- who are susceptible to experiencing high levels of stress.
tionship. That is, a strong social support network may Consistent with previous research (e.g., Dougall et al.
promote a more optimistic outlook on life that, in turn, 2001), we hypothesized that optimism would be positively
fosters well-being. As a result of feeling loved and associated with social support which, in turn, would predict

123
1276 J Autism Dev Disord (2010) 40:1274–1284

increased maternal well-being. Furthermore, consistent Measures

with previous research (e.g., Dunn et al. 2001) we expected
that partner support would be more effective than friend or Optimism
family support in this process. We also explored the role of
optimism as a possible mediator of the social support- The Life Orientation Test (LOT; Scheier and Carver 1985)
maternal well-being relationship. Finally, we examined was designed to measure optimism as it relates to outcome
whether a higher level of social support buffered the neg- expectancies. The LOT contains 8 items asking participants
ative impact of having low optimism. Consistent with to indicate their agreement using a 4-point Likert-type
previous research (e.g., Shelby et al. 2008), we predicted scale (1 = strongly agree to 4 = strongly disagree). Sam-
that social support, particularly spousal support, would ple items included: ‘‘In uncertain times, I usually expect
indeed serve as a buffer against the adverse effects of low the best,’’ and ‘‘I hardly ever expect things to go my way.’’
optimism. Items were reverse coded and a total score was created with
a high score indicating greater optimism. Adequate internal
consistency and test–retest reliability has been previously
Method established (Scheier and Carver 1985). Cronbach’s alpha
for this measure in the current study was .86, indicating
Participants high internal consistency.

Participants consisted of 119 mothers who were part of a Informal Social Support
larger study examining stress and well-being in families of
children with ASD. Mothers included in the present study Informal social support was examined for the following
had at least one child younger than 18 years of age who relationships: family (not including partner), friends, and
had been diagnosed with ASD. The majority of mothers partner. Whalen and Lachman (2000) created the scales
were married (82.9%); the remaining mothers were sepa- using items that were adapted from a previous study
rated, divorced, or widowed (11.9%) or single (5.1%). (Schuster et al. 1990). Mothers answered each question
Mothers were predominantly Caucasian (95%) and middle using a 4-point Likert-type scale, (1 = a lot to 4 = not at
class; specifically, 8.7% had annual household incomes all). Social support was assessed through 4 items that were
below $24,999, 55.7% earned $25,000–$74,999, and similar for family, friends, and partner. The four items
35.7% made $75,000 or more annually. The majority of were: (a) How much do they (family, friends, partner)
mothers had some college or completed college (75.4%) or really care about you?; (b) How much do they understand
completed postgraduate training (14.4%), while a smaller the way you feel about things?; (c) How much can you rely
percentage of mothers had a high school degree or less on them for help if you have a serious problem?; and (d)
(10.1%). Mothers ranged in age 23–61 (M = 40.13, How much can you open up to them if you need to talk
SD = 7.38). Children were between the ages of 2 and 18 about your worries? The partner support scale included two
(M = 9.45, SD = 4.08), and were predominantly male additional items that asked: (a) How much does he
(82.9%). Ten families had at least one other child who was appreciate you?; and (b) How much can you relax and be
also diagnosed with ASD. yourself around him? Items were reverse coded so that a
higher score indicates high support, and a separate score
Procedure was computed for each type of social support. Good
internal consistency has been found in previous studies
Participants were recruited with the assistance of local (Whalen and Lachman 2000). Cronbach’s alphas in the
autism support groups and a regional autism service center. current study were .87, .92, and .89 for friend, partner, and
A total of 311 families, who indicated preliminary interest family, respectively.
in being part of a research project, were contacted and sent
packets with a cover letter explaining the study, consent Parenting Stress
forms, and the survey material. Families who did not wish
to participate in the study returned a prepaid postcard This measure of parenting stress was adapted from the
indicating their decision. A total of 123 families completed Parental Stress Items scale (Pearlin and Schooler 1978).
and returned the questionnaire in a prepaid envelope, The original 7-item scale instructed mothers to indicate the
resulting in a 39.5% response rate. Four families were not extent to which they experienced distressed feelings about
included in the final data analysis because they did not parenting. Sample items included: ‘‘How worried do you
meet the study requirements of having a child under the age feel?’’ and ‘‘How unhappy do you feel?’’ In an adapted
of 18 who was diagnosed with ASD. version used in this study, four positively worded items

123
J Autism Dev Disord (2010) 40:1274–1284 1277

were added to reduce response bias and increase the reli- most ways my life is close to my ideal,’’ and ‘‘If I could
ability of the scale (Bonds et al. 2002). A sample item live my life over, I would change almost nothing.’’ A high
asked ‘‘How satisfied do you feel?’’ Mothers responded to score indicated higher satisfaction. High internal consis-
all items on a 4-point Likert-type scale (1 = not at all to tency and test–retest reliability have been previously
4 = very much so). Positively worded items were reverse reported (Pavot and Diener 1993). Cronbach’s alpha in the
coded so that a higher score indicated higher levels of current study was .85, indicating high internal consistency.
parenting stress. An internal consistency of .87 for the
adapted version of the scale has been reported (Bonds et al. Psychological Well-Being
2002). Cronbach’s alpha in the current study for this scale
was .89. The Psychological Well-Being Scale (PWB; Ryff and
Keyes 1995) is a 42-item self-report measure that assesses
Positive and Negative Affect well-being in multiple domains. A Personal Growth scale
assesses feelings of continued development, openness to
The Positive and Negative Affect Schedule (PANAS; new experiences, and improvement in self and behavior
Watson et al. 1988) is a 20-item questionnaire designed to over time. An Environmental Mastery scale includes items
measure two dimensions of mood, 10 items reflecting that evaluate the individual’s sense of mastery and com-
positive mood and 10 items reflecting negative mood. petence in managing the environment and controlling
Participants were instructed to indicate the extent to which external activities. A Purpose in Life scale assesses the
they experienced each mood state, using a 5-point Likert- extent to which the individual has goals in life and a sense
type scale (1 = not at all to 5 = extremely). A total score of directedness, feels there is meaning to present and past
was created for the two mood dimensions with a high score life, and has aims and objectives for living. Participants
indicating high levels of that emotion. High internal con- indicated their agreement with each statement using a
sistency, adequate test–retest reliability, and external 5-point Likert-type scale (1 = strongly agree to 5 =
validity with measures of distress and psychopathology strongly disagree). Selected items were reverse coded so
have been reported (Watson et al. 1988). In the present that a high score reflected a higher degree of overall well-
study, Cronbach’s alphas were .86 for negative mood and being. An overall score reflecting psychological well-being
.89 for positive mood. was employed in this study. Previous studies have found
moderate to strong associations between this measure and
Depression measures of positive affect, negative affect, life satisfac-
tion, and depression (Ryff and Keyes 1995). Cronbach’s
The Center for Epidemiologic Studies Depression inven- alpha for this measure in the current study was .92 for
tory (CES-D; Devins et al. 1988; Radloff 1977) is a overall psychological well-being.
20-item self-report questionnaire designed to assess depres-
sive symptoms in adults. Mothers were asked to indicate Analytic Strategy
how frequently they experienced various symptoms during
the last week, using a 4-point Likert-type scale (1 = rarely Structural equation modeling (SEM) was used to test our
or none of the time to 4 = most or all of the time). Posi- mediational hypotheses regarding pathways between social
tively worded items were reverse coded so that a high score support, optimism, and maternal well-being. All models
indicated greater overall depression. Previous studies have were estimated using the M-Plus 5.2 statistical package,
reported high internal consistency, adequate test–retest using the maximum likelihood estimation method. Multiple
reliability, and good criterion and discriminant validity fit indices will be reported for each model in order to
coefficients (Devins et al. 1988; Radloff 1977). Cronbach’s evaluate the degree to which our models are an appropriate
alpha for the current study was .92, indicating high internal fit for the data. In addition to the traditional chi-square
consistency. statistic, the v2/df ratio indicates good fit if values are close
to 1, and an acceptable fit if values are less than 5 (Hu and
Life Satisfaction Bentler 1999). Additionally, a comparative fit index (CFI)
greater than .95 indicates a very good fit, although a CFI
The Satisfaction with Life Scale (SWLS; Pavot and Diener value greater than .90 also indicates the model is an ade-
1993) is a 5-item questionnaire developed to mea- quate fit to the data (Hu and Bentler 1999). Finally, when
sure global cognitive judgments about one’s life. Using a the root mean square error of approximation (RMSEA) is
7-point Likert-type scale, mothers indicated the extent to less than .06 then the hypothesized model is a good fit for
which they agreed with each statement (1 = strongly dis- the observed data; however, if the RMSEA is less than .08
agree to 7 = strongly agree). Sample items included: ‘‘In then the hypothesized model is an adequate fit for the data

123
1278 J Autism Dev Disord (2010) 40:1274–1284

(Hu and Bentler 1999). These alternative fit indices control ASD reported less friend support (F(1, 117) = 5.80,
for sample size (Hu and Bentler 1999). When deciding p \ .05), partner support (F(1, 108) = 6.67, p \ .05), and
whether a model is a good fit, each of these fit indices are family support (F(1, 116) = 9.57, p \ .01). In addition,
considered. mothers of more than one child with ASD also reported
The Baron and Kenny (1986) causal steps approach to higher levels of depression (F(1, 111) = 7.67, p \ .01),
mediation has been one of the most popular and frequently higher levels of negative affect (F(1, 114) = 9.59,
cited mediation analyses. Recently, however, studies have p \ .01), and lower levels of life satisfaction (F(1, 116)
shown that this approach results in an increase of the Type = 4.28, p \ .05). For these reasons, the number of addi-
II error rate. In addition, this approach does not provide a tional children with ASD was included as a covariate in all
simultaneous test of all paths being estimated, a direct mediation and moderation models.
estimate of the total indirect effect of the independent
variable on the dependent variable, or provide standard Descriptive Statistics and Correlations
errors to construct confidence limits (MacKinnon et al.
2002; Shrout and Bolger 2002). In addition, when models All variables were examined for skewness and kurtosis,
incorporate multiple mediators and multiple outcomes the and were found to be normally distributed. Therefore, the
traditional approach to mediation has difficulty evaluating data in the present study does not violate the normality
each of the individual effects (MacKinnon et al. 2002). In assumption of maximum likelihood estimation. For the
order to determine the significance of mediation, bias- variable of partner support, however, the kurtosis value was
corrected confidence intervals for the indirect effects were slightly higher than accepted values. Multicollinearity tests
constructed using a bootstrap approach with 500 resamples (e.g. variance inflation factor) between independent vari-
(MacKinnon et al. 2004). Resampling methods may pro- ables were conducted. No evidence for multicollinearity
vide more accurate results when sample sizes are low was found. Table 1 presents descriptive statistics for each
(Mackinnon 2008 p. 342). of the measures. In general, in comparison to low risk
Multiple linear regression analyses were conducted to samples, mothers reported higher degrees of stress (Bonds
test our moderational hypotheses regarding the effects of et al. 2002), higher negative affect (Watson et al. 1988),
the interaction between informal social support and opti- and levels of depression that were severe enough to warrant
mism on maternal well-being. Separate regression analyses a clinical diagnosis of depression (Radloff 1977), lower
were conducted for each form of social support (friend, positive affect (Watson et al. 1988), lower mean life sat-
spouse, and family), as well as for each outcome. Each isfaction scores (Pavot and Diener 1993), and lower levels
multiple regression model included control variables, of optimism (Scheier and Carver 1985). For the other
optimism, social support, and the optimism x social support measures, listed in Table 1, normative comparison data
interaction term. All variables were mean-centered (Aiken was not available.
and West 1991). Problems of missing data were minimal, As shown in Table 2, correlations support our hypoth-
as follow-up correspondence was sent to participants to eses concerning the relation between optimism, informal
obtain missing information. For individual items missing social support, and maternal outcomes. Specifically, a
on a scale score, a mean score was imputed if 90% of the
items on the scale had been completed. Table 1 Descriptive statistics
Mean SD Range Possible range
Results Predictors/mediators
Optimism 22.43 3.62 14–32 8–32
Demographic Relationships Friend support 13.25 2.74 4.5–16 4–16
Partner support 20.88 4.06 6–24 6–24
Relationships between demographics and all variables of Family support 12.70 2.99 5–16 4–16
interest were examined in order to determine whether any Negative outcomes
demographic variables needed to be included as covariates. Depression 33.60 9.90 20–67 20–80
Correlational analyses revealed that maternal age was not Negative affect 20.67 6.58 10–38 10–50
significantly related to any of the variables. In addition, Parenting stress 24.80 6.03 12–39 11–44
there was no significant pattern of associations between Positive outcomes
child age and any of the variables in the current study.
Positive affect 33.95 6.96 18–49 10–50
There were, however, significant differences between
Life satisfaction 21.99 6.63 5–35 5–35
mothers of one child with ASD and mother of more than
Psychological well-being 157.36 19.33 111–210 42–210
one child with ASD. Mothers of more than one child with

123
J Autism Dev Disord (2010) 40:1274–1284 1279

Table 2 Correlations between optimism, social support, and maternal outcomes

1 2 3 4 5 6 7 8 9 10

1. Friend support –
2. Partner support .29** –
3. Family support .38*** .08 –
4. optimism .25** .22* .29** –
5. Depression -.33*** -.40*** -.28** -.54*** –
6. Negative Affect -.26** -.31** -.21* -.39*** .59*** –
7. Parenting stress -.29** -.23* -.21* -.52*** .69*** .69*** –
8. Positive affect .27** .19 .07 .49*** -.48*** -.25** -.40*** –
9. Life satisfaction .32*** .50*** .18 .53*** -.63*** -.53*** -.56*** .36*** –
10. Psychological well-being .32*** .37*** .21* .59*** -.62*** -.50*** -.62*** .64*** .58*** –
* p \ .05; ** p \ .01; *** p \ .001

significant negative association was found between the informal social support and maternal outcomes. As shown
informal social support variables and maternal negative in Fig. 2, we tested two separate hypothesized models, one
outcomes. Mothers who reported higher levels of informal containing negative maternal outcomes and a second con-
social support also reported lower levels of parenting sisting of positive maternal outcomes. Figure 3 presents the
stress, negative affect, and depression. Similar results were results of the hypothesized model test. For the sake of
found for the relation between optimism and maternal clarity, only significant paths are shown. The results of the
negative outcomes. Further, consistent with our hypothe- SEM containing negative maternal outcomes indicated a
ses, a significant positive association was found between good fit with the data, v2 (2, n = 119) = .46, p = .79,
both optimism and informal social support variables and v2/df ratio = .23, CFI = 1.0, and RMSEA = 0 (90% CI:
maternal positive outcomes. Mothers who reported higher 0–.12). The results of the SEM containing positive mater-
levels of optimism and informal social support also nal outcomes also indicated a good fit with the data, v2
reported higher levels of positive affect, life satisfaction, (3, n = 119) = .58, p = .90, v2/df ratio = .19, CFI = 1.0,
and psychological well-being. In addition, we also found a and RMSEA = 0 (90% CI: 0–.07).
significant positive association between optimism and each As hypothesized, several links between informal social
of the informal social support variables. support, optimism, and maternal negative outcomes were
found (see Fig. 3a). Partner support was associated with
Mediation Analyses lower levels of depression. Optimism was significantly
associated with decreased depression, negative affect, and
In the first set of analyses, we used SEM to test whether parenting stress. In addition, an indirect relationship
informal social support mediated the relationship between between family support and negative maternal outcomes
optimism and maternal outcomes. As shown in Fig. 1, we was found. Family support was associated with increased
tested two separate hypothesized models, one containing optimism; this increased optimism, in turn, predicted lower
negative maternal outcomes and a second consisting of levels of depression, negative affect, and parenting stress.
positive maternal outcomes. Based upon the consideration No direct effect between family support and any of the
of multiple fit indices, the results of the SEM containing negative outcomes was found, suggesting that optimism
negative maternal outcomes indicated poor model fit, v2 completely mediated the relationship between family sup-
(4, n = 119) = 12.73, p = .01, v2/df ratio = 3.18, CFI = port and maternal negative outcomes. Confidence intervals
.96, and RMSEA = .14 (90% CI: .06–.22). Similarly, the of the total indirect effects of family support on maternal
results of the SEM containing positive maternal outcomes depression (95% CI: -.18, -.02) and parenting stress
also indicated a poor model fit, v2 (5, n = 119) = 12.73, (95% CI: -.20, -.01) based on 500 bootstrap sample did
p = .02, v2/df ratio = 2.55, CFI = .96, and RMSEA = .11 not include zero, further indicating significant complete
(90% CI: .04–.19). Due to the poor overall model fit the mediation. The confidence interval of the total indirect
individual paths will not be discussed further. Thus, the effects of family support on negative affect (95% CI: -.14,
results suggest that social support did not mediate the 0) terminated at zero, indicating that this indirect effect
effects of optimism on any of the outcomes. may not be as strong.
In the next set of mediation analyses, we used SEM to Several significant associations between informal social
test whether optimism mediated the relationship between support, optimism, and maternal positive outcomes were

123
1280 J Autism Dev Disord (2010) 40:1274–1284

Partner
Support
A Friend
Support
Depression

Optimism Negative Other Child

Affect

Parenting
Family Stress
Support

Partner
Support
B Friend
Support Positive
Affect

Optimism Life Other Child

Satisfaction

Psychological
Family Well-being
Support

Fig. 1 Hypothesized models depicting social support as a mediator of the relationship between optimism and maternal well-being. Models
predicted negative maternal outcomes [a] and positive maternal outcomes [b]

A Optimism
Friend Depression
Support

Partner Negative Other Child

Support Affect

Family Parenting
Support Stress

B Optimism
Friend Positive
Support Affect

Partner Life Other Child

Support Satisfaction

Family Psychological
Support Well-being

Fig. 2 Hypothesized models depicting optimism as a mediator of the relationship between informal social support and maternal well-being.
Models predicted negative maternal outcomes [a] and positive maternal outcomes [b]

also found (see Fig. 3b). Friend support was associated positive maternal outcomes were also found.. Specifically,
with increased positive affect. Partner support was associ- family support was associated with increased optimism;
ated with higher levels of life satisfaction and psycholog- this increased optimism was, in turn, associated with higher
ical well-being. Optimism was significantly associated with levels of positive affect, life satisfaction, and psychological
greater positive affect, life satisfaction, and psychological well-being. There was no direct effect found between
well-being. Indirect effects between family support and family support and any of the positive outcomes,

123
J Autism Dev Disord (2010) 40:1274–1284 1281

Fig. 3 Model results depicting

optimism as a mediator of the A Optimism -.42
relationship between informal
Friend Depression
social support and maternal
Support
well-being. Models predicted
negative maternal outcomes [a] .28 -.29
and positive maternal outcomes -.27
.38 Partner Negative
[b]. Standardized path Support Affect
coefficients are presented. Only
significant paths are shown
.24 -.45
Family Parenting
Support Stress

B Optimism .47
Friend .20 Positive
Support Affect
.26 .43
.39 Partner .38 Life
Support Satisfaction
.23 .50
.24
Family Psychological
Support Well-being

suggesting that optimism completely mediated the rela- provided by partners, friends, and family may be particu-
tionship between family support and maternal positive larly beneficial in promoting maternal well-being in this
outcomes. Confidence intervals of the total standardized population (Boyd 2002). Moreover, research with other
indirect effects of family support on maternal positive populations suggests that optimism and social support are
affect (95% CI: .02, .21), life satisfaction (95% CI: .02, positively related to each other and that social support
.19), and psychological well-being (95% CI: .02, .22) based mediates the optimism-psychological well-being relation-
on 500 bootstrap sample did not include zero, further ship (Brissette et al. 2002; Dougall et al. 2001, Shelby et al.
indicating significant complete mediation. 2008). An emerging literature, however, suggests that
optimism may mediate the social support-psychological
Moderator Analyses well-being relationship (Karademas 2006). Little is known,
however, about how these processes unfold in mothers of
These analyses tested whether there was a significant children with ASD. The present study was directed toward
interaction between the various forms of informal social understanding how these intrinsic and extrinsic factors
support and optimism. Based on the results of the demo- promote well-being among this population which is espe-
graphic analyses previously discussed, we controlled for cially susceptible to experiencing elevated levels of stress
the presence of another child with ASD in the models (Ekas et al. 2009). Three different sources of informal
containing depression, negative affect, and life satisfaction. social support, including partner, other family members,
Results from these analyses found no significant interac- and friends were examined as mediators as well as mod-
tions between the informal social support variables and erators of the optimism-maternal well-being relationship.
optimism when predicting maternal negative and positive One of the initial goals of this study was to examine the
outcomes. simple relationships between informal social supports and
maternal well-being as well as that between optimism and
maternal well-being. We were especially interested in
Discussion examining whether three different sources of social support
(partner, other family members, and friends) were differ-
Past research suggests that optimism and social support are entially related to maternal outcomes. As seen in Table 2,
associated with increased well-being among mothers of correlational data indicated that each source of social
children with ASD (Bishop et al. 2007; Greenberg et al. support was associated with lower levels of depression,
2004). Although not systematically investigated, research negative affect, and parenting stress. With regard to posi-
has also suggested that informal supports, such as those tive maternal outcomes, social support received from

123
1282 J Autism Dev Disord (2010) 40:1274–1284

friends was associated with increased life satisfaction, personality, Cervone (2004) argued that behavior is based
positive affect, and psychological well-being, whereas upon knowledge, an enduring structural feature of per-
partner support was associated with increased life satis- sonality, and appraisals. Cervone (2004) further suggests
faction and psychological well-being. Finally, social sup- that the knowledge structures contribute to the appraisal
port received from other family members was associated processes through which individuals assign meaning to
only with increased psychological well-being. The corre- particular situations. We similarly argue that social sup-
lational data also indicated that optimism was negatively ports could be conceptualized as providers of knowledge as
associated with each of the maternal negative outcomes, well as providers of instrumental and emotional support.
and positively associated with each positive maternal out- Social supports empower the mothers by giving them hope
come. In other words, higher levels of optimism were and leading them to positively appraise the future, in turn
associated with increased positive outcomes and decreased reducing negative affect and engendering positive affect.
negative outcomes. These preliminary results are consistent These findings also have possible implications for the
with previous research highlighting the benefits of having a development of interventions directed at increasing well-
good social support network (e.g. Dunn et al. 2001) and being in mothers raising children with ASD. One possible
having an optimistic outlook (e.g. Achat et al. 2000). implication is that such an intervention should focus on
In the present study, another goal was to examine the increasing optimism. For example, cognitive-behavior
relationship between social support, optimism, and mater- therapy (CBT) techniques, such as self-monitoring and
nal well-being. While the correlation results indicated that cognitive restructuring, have been shown to increase opti-
optimism was positively related to each form of social mism (Pretzer and Walsh 2001) through getting individuals
support, social support was not found to either mediate or to monitor the positive aspects, changes and possibilities in
moderate the relationship between optimism and well- their life situation. In one study, Matthews and Cook
being. This finding contrasted with those of previous (2009) reported that self-transcendence, defined as main-
research (e.g. Dougall et al. 2001; Trunzo and Pinto 2003). taining an interest in the future and planning for the years
Optimism, however, was found to mediate the relationship ahead, was positively associated with optimism. Interven-
between social support received from family members and tions utilizing CBT techniques have been consistently
both positive and negative maternal outcomes (see Fig. 3). effective among families of children with developmental
More specifically, greater family support was associated disabilities (Singer et al. 2007). It would be beneficial for
with increased optimism. In turn, optimism was associated researchers to adapt, and implement, these established
with decreased depression, parenting stress, and negative intervention programs with families of children with ASD.
affect. Moreover, optimism was also directly associated Although focusing on increasing optimism may be
with increased positive affect, life satisfaction, and psy- effective in reducing distress and promoting positive out-
chological well-being. Taken together, the mediation comes, the results of the present study suggest that opti-
models indicate that each form of social support (partner, mism alone is not the only factor to consider in this
other family members, and friends) is important for moth- equation. Given that increased social support is associated
ers’ well-being, but that they operate in different ways. with increased optimism, it seems important to increase
Partners and friends appear to directly impact some aspects both the amount and quality of various informal, as well as
of maternal well-being, whereas family support operates in formal, social supports available to mothers. More specif-
an indirect fashion by helping mothers become or remain ically, it seems likely that increases in social support that
optimistic as they confront the challenges associated with provides helpful information along with other types of
raising a child with ASD. Based on these results, it would instrumental and emotional assistance will increase opti-
seem advisable for interventions to focus on strengthening mism and enhance maternal well-being. One possible
each of the informal social supports specifically because model that could be utilized in this population is the Par-
each of the sources are associated, either directly or indi- ent-to-Parent model (e.g. Singer et al. 1999). This is a
rectly, with unique sources of benefit for the mother. program designed to match parents of children with dis-
The role of optimism as a mediator has not been abilities with parent supporters (i.e. individuals who have
extensively studied in past research and, in fact, only one experience caring for children with disabilities). The sup-
previous study involving a community sample of adults has porter provides emotional support and shares general
reported similar results to that obtained in the present study information with the parent. This model could be easily
(Karademas 2006). The current study extended this pre- implemented with parents of children with ASD. Indeed,
liminary research by examining the relationship of social the regional autism support centers would be a logical
support, optimism, and well-being in a population experi- choice to provide parent supporters. Overall, there are
encing high levels of distress, mothers of children with several intervention programs that could be implemented
ASD. In discussing his cognitive architecture of with families of children with ASD that would focus on

123
J Autism Dev Disord (2010) 40:1274–1284 1283

imparting the skills necessary to increase optimism as well Acknowledgments This research was supported in part by an
as increasing parent’s social support networks. NIMH training grant (2 T32 HD007184-28) and by the Institute for
Scholarship in the Liberal Arts at the University of Notre Dame. We
Although the current study provides important new thank the various parent support groups for their help and support in
information about social support, optimism, and well-being participant recruitment. We are also indebted to the families who gave
in mothers of children with ASD, there are several limi- their time to participate in this research. We are indebted to Cindy
tations to the current study. First, because the sample Bergeman and Anthony Ong for their help in the development of this
project.
consisted of mainly Caucasian, upper-middle class fami-
lies, the generalizability of the results is limited. Second,
this study provides a static picture of the mothers at only a
References
single time point. Without longitudinal data, it is not pos-
sible to make stronger inferences about the relationships Achat, H., Kawachi, I., Spiro, A., DeMolles, D. A., & Sparrow, D.
between social supports, optimism, and maternal well- (2000). Optimism and depression as predictors of physical and
being. For example, we are unable to determine whether mental health functioning: The normative aging study. Annals of
optimism influenced maternal outcomes or vice versa. A Behavioral Medicine, 22, 127–130.
Aiken, L., & West, S. (1991). Multiple regression: Testing and
longitudinal design would help in understanding how these interpreting interactions. Newbury Park, CA: Sage.
relationships change across time and which social supports Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator
provide short-term or long-term benefits. Finally, the variable distinction in social psychological research: Conceptual,
present study also relied on maternal self-report for all strategic, and statistical considerations. Journal of Personality
and Social Psychology, 51, 1173–1182.
measures employed. Therefore, the issue of shared source Benson, P. R. (2006). The impact of child symptom severity on
variance is a valid concern. Future research should include depressed mood among parents of children with ASD: The
both questionnaire and observational components. mediating role of stress proliferation. Journal of Autism and
In general, the present study expands upon previous Developmental Disorders, 36, 685–695.
Bishop, S. L., Richler, J., Cain, A. C., & Lord, C. (2007). Predictors
research by delineating the various mechanisms by which of perceived negative impact in mothers of children with autism
social support influences maternal well-being and the role spectrum disorder. American Journal on Mental Retardation,
maternal optimism plays in this process. Although the 112, 450–461.
current study is an important first step to untangling the Bonds, D. D., Gondoli, D. M., Sturge-Apple, M. L., & Salem, L. N.
(2002). Parenting stress as a mediator of the relation between
complex nature of the social support-psychological well- parenting support and optimal parenting. Parenting Science and
being relationship among mothers of children with ASD, Practice, 2, 409–435.
there is still much to be studied in this area. For example, Boyd, B. A. (2002). Examining the relationship between stress and
future research might follow a family systems perspective lack of social support in mothers of children with autism. Focus
on Autism and Other Developmental Disabilities, 17, 208–215.
(Cox and Paley 1997), including both fathers and mothers Brissette, I., Scheier, M. F., & Carver, C. S. (2002). The role of
of children with ASD as well as other members of the optimism in social network development, coping, and psycho-
family, assessing their perceptions of support given and logical adjustment during a life transition. Journal of Personality
received, and how congruence between family members or and Social Psychology, 82, 102–111.
Bristol, M. M. (1984). Family resources and successful adaptation to
lack thereof contributes to the well-being of each member. autistic children. In E. Schopler & G. B. Mesibov (Eds.), The
Moreover, the relative and interactive contributions of effects of autism on the family (pp. 290–308). New York: Plenum
formal and informal supports need to be evaluated. For Press.
example, it may be that formal supports help structure the Bromley, J., Hare, D. J., Davison, K., & Emerson, E. (2004). Mothers
supporting children with autism spectrum disorders—Social
informal support system which promotes greater optimism support, mental health status and satisfaction with services.
in the family and, in turn, greater well-being. Future Autism, 8, 409–423.
research might also examine more systematically how Cervone, D. (2004). The architecture of personality. Psychological
other personality characteristics, such as locus of control, Review, 111, 183–204.
Cohen, S., & Wills, T. A. (1985). Stress, social support, and the
engender or are engendered by various sources of social buffering hypothesis. Psychological Bulletin, 98, 310–357.
support. Future studies also need to examine the conditions Conway, F., Magai, C., Springer, C., & Jones, S. C. (2008). Optimism
under which social support mediates the optimism-psy- and pessimism as predictors of physical and psychological health
chological well-being relationship versus those conditions among grandmothers raising their grandchildren. Journal of
Research in Personality, 42, 1352–1357.
in which optimism mediates the social support-psycho- Cox, M. J., & Paley, B. (1997). Families as systems. Annual Review
logical well-being relationship. Longitudinal research of Psychology, 48, 243–267.
might also investigate whether the mediational relation- Coyne, J. C., Ellard, J. H., & Smith, D. A. F. (1990). Social support,
ships change over time. Together, the results of the present interdependence, and the dilemmas of helping. In B. R. Sarason,
I. G. Sarason, & G. R. Pierce (Eds.), Social support: An
study, and the future research suggested, may provide a interactional view (pp. 129–149). New York: Wiley.
useful framework for developing interventions tailored to Devins, G. M., Orme, C. M., Costello, C. G., Binik, Y. M., Frizzell,
meet the needs of families of children with ASD. B., Stam, H. J., et al. (1988). Measuring depressive symptoms in

123
1284 J Autism Dev Disord (2010) 40:1274–1284

illness populations: Psychometric properties of the center for Radloff, L. S. (1977). The CES-D scale: A self report depression scale
epidemiologic studies depression (CES-D) scale. Psychology for research in the general population. Applied Psychological
and Health, 2, 139–156. Measurement, 1, 385–401.
Dougall, A. L., Hyman, K. B., Hayward, M. C., McFeeley, S., & Rini, C., DuHamel, K., Ostroff, J., Boulad, F., Martini, R., Mee, L.,
Baum, A. (2001). Optimism and traumatic stress: The impor- et al. (2008). Social support from family and friends as a buffer
tance of social support and coping. Journal of Applied Social of low spousal support among mothers of critically ill children:
Psychology, 31, 223–245. A multilevel modeling approach. Health Psychology, 27,
Dunn, M. E., Burbine, T., Bowers, C. A., & Tantleff-Dunn, S. (2001). 593–603.
Moderators of stress in parents of children with autism. Ryff, C. D., & Keyes, C. L. M. (1995). The structure of psychological
Community Mental Health Journal, 37, 39–52. well-being revisited. Journal of Personality and Social Psychol-
Ekas, N. V., Whitman, T. L., & Shivers, C. (2009). Religiosity, ogy, 69, 719–727.
spirituality, and socioemotional functioning in mothers of Scheier, M. F., & Carver, C. S. (1985). Optimism, coping, and health:
children with autism spectrum disorder. Journal of Autism and Assessment and implications of generalized outcome expectan-
Developmental Disorders, 39, 706–719. cies. Health Psychology, 4, 219–247.
Gill, M. J., & Harris, S. L. (1991). Hardiness and social support as Scheier, M. F., & Carver, C. S. (1992). Effects of optimism on
predictors of psychological discomfort in mothers of children psychological and physical well-being: Theoretical overview and
with autism. Journal of Autism and Developmental Disorders, empirical update. Cognitive Therapy and Research, 16, 201–228.
21, 407–416. Scheier, M. F., Matthews, K. A., Owens, J., Magovern, G. J.,
Greenberg, J. S., Seltzer, M. M., Krauss, M. W., Chou, R. J., & Hong, Lefebvre, R. C., Abbott, R. A., et al. (1989). Optimism and
J. (2004). The effect of quality of the relationship between recovery from coronary artery bypass surgery: The beneficial
mothers and adult children with schizophrenia, autism, or down effects on physical and psychological well-being. Journal of
syndrome on maternal well-being: The mediating role of Personality and Social Psychology, 57, 1024–1040.
optimism. American Journal of Orthopsychiatry, 74, 14–25. Schuster, T. L., Kessler, R. C., & Aseltine, R. H. (1990). Supportive
Hassall, R., Rose, J., & McDonald, J. (2005). Parenting stress in interactions, negative interactions, and depressive mood. Amer-
mothers of children with an intellectual disability: The effects of ican Journal of Community Psychology, 18, 423–438.
parental cognitions in relation to child characteristics and family Segerstrom, S. C., Taylor, S. E., Kemeny, M. E., & Fahey, J. L.
support. Journal of Intellectual Disability Research, 49, (1998). Optimism is associated with mood, coping, and immune
405–418. change in response to stress. Journal of Personality and Social
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in Psychology, 74, 1646–1655.
covariance structure analysis: Conventional criteria versus new Shelby, R. A., Crespin, T. R., Wells-Di Gregorio, S. M., Lamdan, R.
alternatives. Structural Equation Modeling, 6, 1–55. M., Siegel, J. E., & Taylor, K. L. (2008). Optimism, social
Julien, D., & Markman, H. J. (1991). Social support and social support, and adjustment in African American women with breast
networks as determinants of individual and marital outcomes. cancer. Journal of Behavioral Medicine, 31, 433–444.
Journal of Social and Personal Relationships, 8, 549–568. Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and
Karademas, E. C. (2006). Self-efficacy, social support and well-being: nonexperimental studies: New procedures and recommendations.
The mediating role of optimism. Personality and Individual Psychological Methods, 7, 422–445.
Differences, 40, 1281–1290. Singer, G. H. S., Ethridge, B. L., & Aldana, S. I. (2007). Primary and
MacKinnon, D. P. (2008). Introduction to statistical mediation secondary effects of parenting and stress management interven-
analysis. New York, NY: Lawrence Erlbaum Associates. tions for parents of children with developmental disabilities: A
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & meta-analysis. Mental Retardation and Developmental Disabil-
Sheets, V. (2002). A comparison of methods to test mediation ities Research Reviews, 13, 357–369.
and other intervening variable effects. Psychological Methods, 7, Singer, G. H. S., Marquis, J., Powers, L. K., Blanchard, L., Divenere,
83–104. N., Santelli, B., et al. (1999). A multi-site evaluation of parent to
MacKinnon, D. P., Lockwood, C. M., & Williams, J. (2004). parent programs for parents of children with disabilities. Journal
Confidence limits for the indirect effect: Distribution of the of Early Intervention, 22, 217–229.
product and resampling methods. Multivariate Behavioral Trunzo, J. J., & Pinto, B. M. (2003). Social support as a mediator of
Research, 39, 99–128. optimism and distress in breast cancer survivors. Journal of
Matthews, E. E., & Cook, P. F. (2009). Relationships among Consulting and Clinical Psychology, 71, 805–811.
optimism, well-being, self-transcendence, coping, and social Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and
support in women during treatment for breast cancer. Psycho- validation of brief measures of positive and negative affect: The
Oncology, 18, 716–726. PANAS scales. Journal of Personality and Social Psychology,
Olsson, M. B., & Hwang, C. P. (2001). Depression in mothers and 54, 1063–1070.
fathers of children with intellectual disability. Journal of Weiss, S. (2002). Hardiness and social support as predictors of stress
Intellectual Disability Research, 45, 535–545. in mothers of typical children, children with autism, and children
Park, C. L., & Folkman, S. (1997). Meaning in the context of stress with mental retardation. Autism, 6, 115–130.
and coping. Review of General Psychology, 1, 115–144. Whalen, H. R., & Lachman, M. E. (2000). Social support and strain
Pavot, W., & Diener, E. (1993). Review of the satisfaction with life from partner, family and friends: Costs and benefits for men and
scale. Psychological Assessment, 5, 164–172. women in adulthood. Journal of Social and Personal Relation-
Pearlin, L. I., & Schooler, C. (1978). The structure of coping. Journal ships, 17, 5–30.
of Health and Social Behavior, 19, 2–21. Whitman, T. L. (2004). The development of autism: A self-regulatory
Pretzer, J. L., & Walsh, C. A. (2001). Optimism, pessimism, and perspective. London: Jessica Kingsley Publishers.
psychotherapy: Implications for clinical practice. In E. C. Chang Zautra, A. J., Smith, B., Affleck, G., & Tennen, H. (2001).
(Ed.), Optimism and pessimism, implications for theory, Examination of chronic pain and affect relationships: Applica-
research, and practice (pp. 321–346). Washington, DC: Amer- tions of a dynamic model of affect. Journal of Counseling and
ican Psychological Association. Clinical Psychology, 69, 786–795.

123