Content uploaded by Stephen J. Lepore
Author content
All content in this area was uploaded by Stephen J. Lepore on Dec 05, 2020
Content may be subject to copyright.
Social Constraints are Associated with Negative
Psychological and Physical Adjustment
in Bereavement
Vanessa Juth and Joshua M. Smyth*
The Pennsylvania State University, USA
Michael P. Carey
The Miriam Hospital and Brown University, USA
Stephen J. Lepore
Temple University, USA
Losing a loved one is a normative life event, yet there is great variability in
subsequent interpersonal experiences and adjustment. The Social-Cognitive
Processing (SCP) model suggests that social constraints (i.e. limited opportu-
nities to disclose thoughts and feelings in a supportive context) impede emo-
tional and cognitive processing of stressful life events, which may lead to
maladjustment. This study investigates personal and loss-related correlates of
social constraints during bereavement, the links between social constraints and
post-loss adjustment, and whether social constraints moderate the relations
between loss-related intrusive thoughts and adjustment. A community sample
of bereaved individuals (n=238) provided demographic and loss-related infor-
mation and reported on their social constraints, loss-related intrusions, and
psychological and physical adjustment. Women, younger people, and those
with greater financial concerns reported more social constraints. Social con-
straints were significantly associated with more depressive symptoms, perceived
stress, somatic symptoms, and worse global health. Individuals with high social
constraints and high loss-related intrusions had the highest depressive symp-
toms and perceived life stress. Consistent with the SCP model, loss-related
social constraints are associated with poorer adjustment, especially psychologi-
cal adjustment. In particular, experiencing social constraints in conjunction
with loss-related intrusions may heighten the risk for poor psychological health.
Keywords: adjustment, bereavement, death, intrusions, social-cognitive pro-
cessing, social constraints
* Address for correspondence: Joshua M. Smyth, Department of Biobehavioral Health, The
Pennsylvania State University, University Park, PA 16802-6507, USA. Email: jms1187@psu.edu
bs_bs_banner
APPLIED PSYCHOLOGY: HEALTH AND WELL-BEING, 2015
doi:10.1111/aphw.12041
© 2015 The International Association of Applied Psychology
INTRODUCTION
The loss of a loved one is ubiquitous; nonetheless, this experience is typically
challenging, with individuals highly variable in their adjustment (e.g.
Wortman & Silver, 1989). For some, their entire worldview is shattered (e.g.
Marris, 1982), leading to a difficult bereavement process or period following
a loss. Others successfully integrate or re-frame the loss (e.g. Stroebe & Schut,
2010), facilitating adaptation to their new circumstances. This heterogeneity
in bereavement experiences is not fully understood, but the Social-Cognitive
Processing (SCP) model (Lepore, 2001; Lepore, Silver, Wortman, &
Waymant, 1996; Lepore & Revenson, 2007) suggests that people’s interper-
sonal experiences may be an important factor in post-loss adjustment. The
present study uses the SCP model as a framework for analyzing cross-
sectional data to investigate the role of social constraints (i.e. actual or
perceived interpersonal hindrances of emotional disclosure) in self-reported
(mal)adjustment to bereavement. It focuses on three aims: (1) to identify the
personal and contextual conditions under which social constraints on disclo-
sure are likely to occur among the bereaved; (2) to examine how social
constraints relate to adjustment (psychological and physical health) during
bereavement (independent of perceived social support); and (3) to test
whether social constraints moderate the relation between loss-related intru-
sive thoughts and adjustment in a community sample of bereaved persons as
has been previously observed (in bereaved mothers; Lepore et al., 1996).
Social Constraints and (Mal)adjustment
Disclosing and discussing thoughts and feelings about a significant event can
be helpful, and may even be integral, for adjustment (Lindstrom, 2002; Smyth
& Greenberg, 2000). The benefits of loss-related disclosure have been docu-
mented in close personal relationships (e.g. family, friends), professional
settings (e.g. clinical therapy), and therapeutic interventions (e.g. Segal,
Chatman, Bogaards, & Becker, 2001). Unfortunately, bereaved persons’
social networks may provide limited access to, or altogether lack, sympa-
thetic listeners (Lepore et al., 1996). For example, some people may react
negatively to repeated discussions about the loss, or inadvertently provide
unhelpful advice that trivialises it (Wortman & Silver, 1989). Others may
avoid talking about the loss due to benevolent intentions, such as not wanting
to encourage rumination or to avoid “making it worse” (Zakowski et al.,
2003). These unsatisfying interpersonal experiences can leave bereaved indi-
viduals feeling unsupported, misunderstood, and alienated (e.g. Lepore &
Ituarte, 1999), in turn inhibiting their willingness or ability to freely disclose
their emotions and thoughts. Such actual or perceived social contexts that
hinder individuals’ expression of their thoughts, feelings, or concerns are
known as social constraints (Lepore & Revenson, 2007; Lepore et al., 1996).
2JUTH ET AL.
© 2015 The International Association of Applied Psychology
Some evidence suggests that social constraints can lead to worse post-loss
adjustment (Wortman & Boerner, 2007). The SCP model proposes three
pathways through which disclosure may facilitate, and social constraints may
impede, adjustment (Lepore, 2001). First, disclosure may provide comforting
social validation and affirmation of trauma-related thoughts and feelings.
Second, others’ responses may generate new perspectives, facilitate meaning
making, and provide helpful coping advice. Further, the process of attempt-
ing to articulate one’s thoughts and feelings to others may bring greater
personal insight into these thoughts and feelings, in turn facilitating psycho-
logical assimilation and accommodation processes. Third, disclosure can
reduce arousal that often accompanies trauma-related thoughts and feelings;
this can reduce stress. When disclosure attempts are met with social con-
straints, however, these adaptive processes can be disrupted, slowing or
obstructing emotional recovery, leading to maladjustment (Lepore &
Helgeson, 1998). We use the SCP model to guide our investigation of
responses to bereavement; if the mechanistic pathways proposed in the model
are plausible, we should detect significant cross-sectional associations
between loss-related social constraints and self-reported adjustment consist-
ent with the processes outlined by the SCP framework.
A growing literature supports the SCP framework for understanding the
interplay of social, cognitive, and emotional processes that facilitate adjust-
ment. In coping with chronic illness, social constraints are associated with
greater depressive symptoms, negative mood states and behaviors, worse
general well-being, and non-adherence to health regimens among patients
(Adams, Winger, & Mosher, 2014; Badr, Pasipanodya, & Laurenceau, 2013;
Braitman et al., 2008; Mosher et al., 2012), as well as worse negative affect
among their spouses (Pasipanodya et al., 2012; Sheridan, Sherman, Pierce, &
Compas, 2010). Long-term implications have also been observed: mothers
who lost a child to sudden infant death syndrome reported more depressive
symptoms over time if they experienced greater social constraints (Lepore
et al., 1996). Left unaddressed, interpersonal contexts that hinder social-
cognitive processing may give way to severe adjustment problems that are
difficult to treat (e.g. depression, anxiety, posttraumatic stress disorder;
Belsher, Ruzek, Bongar, & Cordova, 2012; Green, Ferguson, Shum, &
Chambers, 2013).
Most prior studies on bereavement have either not examined interpersonal
factors (see Stroebe & Schut, 2010, for discussion) or have limited their
investigations to examining the role of social support (i.e. actual or perceived
positive interpersonal interactions that facilitate coping and adjustment; e.g.
Dyregrov & Dyregrov, 2008). The distinction between perceived constraints
and perceived support from one’s social environment is an important one
because they reflect two unique aspects of one’s interpersonal experiences.
Individuals who have had cancer often report that while their family and
SOCIAL CONSTRAINTS AND ADJUSTMENT IN BEREAVEMENT 3
© 2015 The International Association of Applied Psychology
friends are highly supportive—typically in ways that are practical or help
them to get back into “normal” life routines—they can also constrain disclo-
sure such as conveying discomfort or changing the subject when the topic of
cancer is raised (see Lepore & Revenson, 2007). These two perceptions of
one’s social environment may be related (i.e. perceptions of fewer constraints
likely coincide with perceptions of greater support), but contexts or interven-
tions that increase social support do not necessarily reduce social constraints.
Recent studies have demonstrated effects of social constraints beyond those
attributable to social support (Kaynak, Lepore, & Kliewer, 2011; Nenova,
DuHamel, Zemon, Rini, & Redd, 2013), but we are not aware of such
evidence in the context of bereavement. The present study will thus examine
how social constraints relate to self-reported adjustment during bereavement
independent of social support, and may present evidence to set social con-
straints apart from support as a unique aspect of one’s post-loss social
context.
Personal Factors, Loss Characteristics, and
Intrusive Thoughts
Knowing whether social constraints are more prevalent among some
bereaved persons has important implications for identifying individuals who
may be at risk for maladjustment. It may also present insight on differences
in interpersonal interactions among different groups of people or different
types of loss. To date, factors that may predict social constraints during
bereavement, such as personal characteristics (e.g. sex, age, ethnicity, educa-
tion) and/or loss characteristics (e.g. relationship to the decedent, time since
loss), are not well understood. Therefore, this study aims to extend the
understanding of personal factors (demographics) and contextual factors
(loss characteristics) that might give rise to social constraints.
Some bereaved persons experience loss-related intrusions (thoughts, emo-
tions, mental images, and physical effects provoked by their loss) that are
more severe than those typically experienced during bereavement (Lepore
et al., 1996). Social constraints may be particularly damaging for these indi-
viduals as constraints may interfere with their capacity to engage in social-
cognitive processing of these intrusions. Thus, individuals who experience
both loss-related intrusions and social constraints may be at heightened risk
for maladjustment. Prior research has linked intrusive thoughts to worse
psychological health, with the strongest associations among people who have
a relatively high level of social constraints (Belsher et al., 2012; Kliewer,
Lepore, Oskin, & Johnson, 1998). For example, high social constraints exac-
erbated the positive relation between intrusive thoughts and depressive symp-
toms among bereaved mothers (Lepore et al., 1996). Yet, we are not aware of
any research examining social constraints’ influence on the relation between
4JUTH ET AL.
© 2015 The International Association of Applied Psychology
loss-related intrusions and perceived stress or somatic symptoms. This study
tests whether these factors have a synergistic effect on a broad range of
self-reported adjustment outcomes, and examines these associations in a
mixed gender sample.
Current Study
The complexity of, and variability in, the bereavement process highlights the
need for a better understanding of how different people cope with loss, the
social responses they encounter, and how this might influence their adjust-
ment (e.g. Stroebe, Schut, & Stroebe, 2007). The current study was designed
to address these aims. Aim 1 presents a descriptive account of personal (sex,
age, ethnicity, education, income, financial concerns since loss, and living
alone or with others) and loss characteristics (time since loss, relationship to
deceased, cause of death, and length of illness prior to death) associated with
level of social constraints in a community sample of bereaved persons. As
there is not a strong empirical base that identifies which of these character-
istics give rise to social constraints during bereavement, we approached this
aim in an exploratory (rather than hypothesis-driven) fashion. Aim 2 inves-
tigates the potential associations between loss-related social constraints and
adjustment, independent of perceived social support. We hypothesised that
highly socially constrained bereaved persons would report worse psychologi-
cal health (greater depressive symptoms and perceived life stress) and worse
self-reported physical health (more somatic symptoms and poorer global
health), even when controlling for perceived social support. Aim 3 examines
whether social constraints moderate the relation between loss-related intru-
sions and adjustment. We hypothesised that maladjustment would be great-
est among bereaved persons with greater intrusions and relatively high versus
low levels of social constraints.
METHOD
Procedures and Participants
Survey data were collected from a sample of community-dwelling adults
living in a medium-sized urban city. Public service and paid announcements
were placed in local newspapers, bereavement newsletters, radio stations,
online, and cable television stations. Announcements invited persons who
had lost a loved one to complete a confidential survey to help researchers
better understand the experience of loss and were asked to call a toll-free
number for information about the “Loss and Health” project. Persons who
called were given a description of the study by a trained research assistant and
screened for the following eligibility criteria: adults over 18 years of age, and
SOCIAL CONSTRAINTS AND ADJUSTMENT IN BEREAVEMENT 5
© 2015 The International Association of Applied Psychology
reporting a human loss. Eligible persons interested in participating were sent
a study packet including a letter that further explained the study, consent
forms, instructions, the survey, and a pre-addressed stamped envelope to
return the completed consents and surveys. All procedures and materials
were approved by Institutional Review Boards.
Measures
Demographic and Loss Characteristics. Participants reported their sex,
age at time of loss, ethnicity, years of education, annual household income,
major financial concerns since the loss, and whom they live with (if anyone).
Six items gathered information on the loss: length of time since loss, relation-
ship of the decedent, cause of death, and length of sickness before death.
Social Constraints. The Social Constraints Scale (SCS; Lepore &
Ituarte, 1999) consists of 12 items that assess the degree to which discussing
trauma-related thoughts and feelings with others was perceived to be con-
strained over the past month. Items were rated on a 5-point scale (1 =never
to 5 =very often). A sample item is: “How often did your friends and/or
family act uncomfortable when you talked about your loss?” All items were
summed for a total score ranging from 12 to 60; higher scores indicate greater
social constraints. Cronbach’s alpha for this sample was 0.95.
Perceived Social Support. The Perceived Social Support Scale
(Procidano & Heller, 1983) consists of two 20-item subscales that assess
perceived social support from family members and from friends. Participants
indicated whether they encountered specific thoughts and experiences with
family members and friends with “Yes”, “No”, or “I don’t know”. A sample
item is: “My family is sensitive to my personal needs.” Responses that rep-
resent positive social support are coded 1 (others coded 0) and summed for a
total score. As no research suggests that variable sources of support would be
differentially related to social constraints, the subscales (r=.32, p<.001)
were combined to form a single total score ranging from 0 to 40; higher scores
indicate greater perceived social support from family members/friends.
Cronbach’s alphas for this sample was 0.93 (friends subscale =0.93; family
subscale =0.90).
Loss-Related Intrusions. The intrusive symptoms subscale of the Impact
of Event Scale (IES; Horowitz, Wilner, & Alvarez, 1979) assessed intrusions
related to the loss within the past week. Seven items ask about thoughts,
emotions, mental images, and physical effects provoked by the loss. Partici-
pants rated the extent to which they experienced each symptom on a 4-point
6JUTH ET AL.
© 2015 The International Association of Applied Psychology
scale (0 =not at all to 3 =often). All items were summed for a total score
ranging from 0 to 21; higher scores indicate greater intrusions. Cronbach’s
alpha for this sample was 0.85.
Psychological Health. Depressive symptoms: The Center for Epidemio-
logical Studies-Depressed Mood Scale (CES-D; Radloff, 1977) is a 20-item
scale that measures depressive symptoms. Participants rated how often they
felt or behaved depressed during the past week. Items are rated on a 4-point
scale from 0 =rarely or none of the time (less than 1 day) to 3 =most or all
of the time (5–7 days). All items were summed for a total score ranging from
0 to 60; higher scores indicate greater depressive symptoms. Scores ≥16 have
been identified as the clinical depression cut-off point (Radloff, 1977).
Cronbach’s alpha for this sample was 0.92.
Perceived life stress: The 10-item version of the Perceived Stress Scale (PSS;
Cohen & Williamson, 1988) assessed the degree to which participants per-
ceived events or circumstances in their lives as stressful in the last month.
Items are rated on a 5-point scale (0 =never to 4 =very often). All items were
summed for a total score ranging from 0 to 40; higher scores indicate greater
levels of perceived life stress. Cronbach’s alpha for this sample was 0.91.
Physical Health. Somatic symptoms: A revised version of the
Pennebaker Inventory of Limbic Languidness (PILL; Pennebaker, 1982)
asked participants to rate 18 items assessing common somatic symptoms or
bodily sensations, including fever, cough, stuffed up nose, sinus pain, sneez-
ing, hoarseness, muscle aches, aching joints or bones, cramps, fainting or
dizziness, diarrhea, rash, constipation, ear ache, vomiting, indigestion,
abdominal or stomach pain, and racing heart. Participants rated the fre-
quency with which they experience these symptoms on a 5-point scale
(0 =never or almost never, 1 =less than 3 or 4 times per year, 2 =every
month or so, 3 =every week or so, 4 =more than once per week). All items
were summed for a total score ranging from 0 to 72; higher scores indicate
greater somatic symptoms. Cronbach’s alpha for this sample was 0.83.
Global health (self-reported): A single item measure, shown to be a reliable
indicator of health (e.g. Cunny & Perri, 1991), assessed participants’ overall
physical health during the past month. Participants were asked: “In general,
would you say your health is: poor(1), fair(2), good(3), very good(4), excel-
lent(5).” This item was reverse coded to match the direction of the other
measures; higher scores indicate worse health.
Analytic Plan
We obtained descriptive statistics on demographic and loss characteristics,
which were then tested as predictors of social constraints (Aim 1). The
SOCIAL CONSTRAINTS AND ADJUSTMENT IN BEREAVEMENT 7
© 2015 The International Association of Applied Psychology
residuals of the four adjustment outcomes were correlated and the Breusch-
Pagan test of independence indicated that they were not independent
(χ2(6) =205.786, p<.001); therefore, we conducted multivariate regression
analyses. Social constraints, perceived social support, age, financial concerns
since loss, time since loss, and each adjustment outcome were examined as
continuous variables. To examine whether social constraints was related to
adjustment independent of perceived social support in Aim 2, social con-
straints, the significant factors from Aim 1 (sex, age, financial concerns since
loss), and relevant theoretical factors (time since loss) were included in the
model; this model was then re-run controlling for perceived social support. A
separate model including an interaction term for social constraints and loss-
related intrusions was used to test for moderation in Aim 3 (using centered
variables to account for potential multicollinearity).
RESULTS
Recruitment and Descriptive Information
Inquiry calls were received from 339 persons, of whom 323 initially appeared
to meet eligibility criteria and requested mailed study materials. A total of 244
individuals (75% of eligible participants) completed and returned all
materials. Six cases were excluded from analysis due to: missing >20 per cent
of social constraints measure (n=4), non-human bereavement reported (loss
of a pet; n=1), missing substantial data on other measures (n=1). For 13
cases (∼5%) that were missing ≤20 per cent of items within a measure,
measure-level means were used to impute values for missing items (Little &
Rubin, 2002). Multivariate regression results showed the same pattern with
and without the 13 cases; thus, they were included for increased analytic
power. The final analyses included 238 participants (97% of enrolled partici-
pants). As shown in Table 1, participants consisted primarily of Caucasian
women ranging in age from 18 to 84 with relatively low annual household
income and educational attainment. The majority of losses were of an
immediate family member and due to a severe illness, with the average length
of time since the loss occurred being approximately 4 years (SD =∼
1
⁄
2
year).
Aim 1. Correlates of Social Constraints
The mean level of social constraints in this sample was 25.23 (SD =10.85,
range =12–58), similar to values reported among others experiencing stress-
ful life circumstance (e.g. cancer patients, M=25.0, SD =10.4) (Green et al.,
2013). Table 1 presents correlates of social constraints. Social constraints
tended to be higher, on average, among women, younger individuals, and
8JUTH ET AL.
© 2015 The International Association of Applied Psychology
TABLE 1
Demographic and Loss Characteristic Descriptive Statistics and Correlates of
Social Constraints (n= 238)
Sample descriptives Social constraints
Mor n(SD)or% Mor β(SD or SE)p
Demographics
Sex
Men 29 12% 21.23 (6.91) <.01
Women 209 88% 25.79 (11.19)
Age at time of loss 52.25 (14.99) −0.13 (0.05) <.01
Ethnicity
Caucasian 221 93% 25.58 (10.91) ns
Othersa17 7% 20.77 (9.15)
Education
High school or less 75 32% 26.02 (11.58) ns
Some college 72 30% 24.89 (10.11)
Bachelor’s degree 41 17% 26.12 (10.57)
Greater than bachelor’s degree 50 21% 23.81 (11.11)
Annual household income
<$10,000 25 11% 24.60 (10.67) ns
$10,000–$19,999 43 18% 27.01 (13.17)
$20,000–$40,000 84 36% 26.72 (11.18)
>$40,000 83 35% 23.33 (8.91)
Financial concerns since loss
Rarely 77 32% 23.30 (10.61)b<.05
Sometimes 109 46% 24.90 (9.75)
Often 52 22% 28.79 (12.62)b
Lives alone
Yes 102 57% 24.41 (10.66) ns
Noc135 43% 25.89 (11.02)
Loss characteristics
Time since loss (months) 41 (71) −0.01 (0.01) ns
Length of sickness (months) 32 (77) −0.01 (0.01) ns
Relationship to deceasedd
Immediate family 214 90% 25.13 (10.65) ns
Others 24 10% 26.10 (12.69)
Cause of deathe
Natural 146 82% 25.54 (10.47) ns
Unnatural 31 18% 25.26 (9.71) ns
Notes: Statistical tests for p-values varied based on predictor. Time since loss: range =0–600 months. Length
of sickness: range =0–720 months. Different ns due to missing data.
aOthers =African-American, Hispanic, Native American, Other.
bPost-hoc Scheffé: p=.02.
cNo =lives with partner/spouse and/or children and/or other.
dImmediate family =spouse/partner, child, parent, sibling; Others =grandchild, grandparent, aunt/uncle,
friend, other.
eNatural =illness-related; Unnatural =accident, homicide, suicide, other.
SOCIAL CONSTRAINTS AND ADJUSTMENT IN BEREAVEMENT 9
© 2015 The International Association of Applied Psychology
individuals with heightened financial concerns since the loss. None of the
other demographic factors or loss characteristics were significantly associated
with social constraints.
Aim 2. Social Constraints and Maladjustment
Depressive symptoms (M=20.11, SD =12.14, range =0–56), perceived life
stress (M=18.60, SD =7.50, range =2–38), somatic symptoms (M=20.26,
SD =10.46, range =2–59), and global health (M=2.88, SD =1.01,
range =1–5) were moderately correlated with each other (rs ranged from .32
to .70, ps<.001), with depressive symptoms and perceived life stress showing
the highest correlation. Given that overall adjustment tends to increase
(improve) over time, we tested time since loss as a covariate in the model,
along with the significant correlates from Aim 1 (sex, age, financial concerns).
Table 2 presents the multivariate regression results.
Psychological Health. Bereaved persons who reported more social con-
straints also reported worse psychological health. Social constraints were
significantly associated with higher levels of depressive symptoms and higher
perceived life stress, accounting for 26.5 per cent and 15.9 per cent of the
variance, respectively. Over half of the participants (59%) endorsed symp-
toms suggestive of clinical depression (scored ≥16 on the CESD; Radloff,
1977); these individuals reported significantly higher social constraints
(M=28.71, SD =11.13) than those who did not meet criteria (M=20.26,
SD =8.25; t(233.91) =−6.712, p<.001).
Physical Health. Bereaved persons who reported more social constraints
reported more somatic symptoms during the past month; social constraints
accounted for 9.1 per cent of the variance in somatic symptoms. Also,
bereaved persons with more social constraints reported having worse self-
reported global health, of which social constraints accounted for 5.0 per cent
of the variance.
Controlling for Perceived Social Support. Perceived social support
(n=238; M=27.52, SD =9.39, range =3–40) was moderately and inversely
correlated with social constraints (r=−.354, p<.001). Perceived social
support significantly predicted perceived life stress (b(SE)=−0.160 (0.050),
p=.002), but none of the other health outcomes (ps>.10). To test the inde-
pendence of the observed associations between social constraints and psycho-
logical and physical adjustment, perceived support was included in the
multivariate model (Wilks’ lambda =0.503; F(6, 224) =7.00, p<.001), which
did not change the significance and pattern of results.
10 JUTH ET AL.
© 2015 The International Association of Applied Psychology
Aim 3. Loss-Related Intrusions
As perceived social support did not alter the results in Aim 2, it was excluded
from the model testing interaction in order to use a more parsimonious model.
Loss-related intrusions (n=237; M=13.61, SD =4.84, range =0–21) were
moderately and positively correlated with social constraints (r=.326,
p<.001). More intrusions were related to more depressive symptoms
(b(SE)=0.972(0.137), p<.001), perceived life stress (b(SE)=0.318(0.095),
p=.001), and somatic symptoms (b(SE)=0.384(0.143), p=.008), but were
TABLE 2
Multivariate Regressions of Each Adjustment Outcome on Social Constraints
Controlling for Sex, Age, Financial Concerns Since Loss, and Time Since Loss
(n= 231)
Adjustment outcome Predictors Beta SE t-test 95% CI
Depressive symptomsaConstant 6.73 3.41 1.97 0.01, 13.46
Sex 0.86 2.05 0.42 −3.18, 4.90
Age −0.04 0.04 −0.99 −0.12, 0.04
Financial concerns 3.57 0.95 3.74*** 1.69, 5.45
Time since loss −0.03 0.01 −2.74** −0.05, −0.01
Social constraints 0.51 0.06 8.09*** 0.38, 0.63
Perceived life stressbConstant 16.07 2.21 7.26*** 11.71, 20.43
Sex 1.82 1.33 1.37 −0.80, 4.44
Age −0.11 0.03 −4.10*** −0.17, −0.06
Financial concerns 1.39 0.62 2.25* 0.18, 2.61
Time since loss −0.01 0.01 −1.43 −0.02, <0.01
Social constraints 0.23 0.04 5.65*** 0.15, 0.31
Somatic symptomscConstant 11.46 3.29 3.48** 4.97, 17.94
Sex −0.61 1.98 −0.31 −4.51, 3.28
Age −0.01 0.04 −0.24 −0.09, 0.07
Financial concerns 4.17 0.92 4.54*** 2.36, 5.98
Time since loss <−0.01 0.01 −0.07 −0.02, 0.02
Social constraints 0.24 0.06 4.00*** 0.12, 0.36
Physical healthdConstant 1.69 0.33 5.08*** 1.03, 2.34
Sex 0.02 0.20 1.97 −.37, 0.41
Age 0.01 <0.01 3.27 <0.01, 0.02
Financial concerns 0.30 0.09 0.98** 0.12, 0.49
Time since loss <0.01 <0.01 3.06 <−0.01, <0.01
Social constraints 0.02 0.01 5.08** 0.01, 0.03
Notes: Wilks’ lambda =0.53; F(5, 225) =7.81, p<.001. Beta =standardised beta. *p<.05; **p<.01;
***p<.001.
aRoot Mean Square Error (RMSE) =10.05, R2=0.34, Model F=22.89***.
bRMSE =6.52, R2=0.26, Model F=15.58***.
cRMSE =9.69, R2=0.17, Model F=9.39***.
dRMSE =0.98, R2=0.10, Model F=5.15***.
SOCIAL CONSTRAINTS AND ADJUSTMENT IN BEREAVEMENT 11
© 2015 The International Association of Applied Psychology
unrelated to self-reported health (p=.107). The main effects for social
constraints remained significant across all outcomes. Social constraints
moderated the relation between loss-related intrusions and depressive
symptoms (b(SE)=0.026(0.011), p=.024) and perceived life stress
(b(SE)=0.025(0.008), p=.002). As shown in Figures 1a and 1b, the positive
relations between intrusions and psychological maladjustment were stronger
among bereaved individuals with high rather than low levels of social con-
straints. No moderation effects were found for physical adjustment (ps>.080).
There was substantial variability in participants’ time since loss (see
Table 1). To investigate whether observed results were influenced by outliers,
we re-ran the multivariate regression analyses for Aims 2 and 3 replacing the
continuous time since loss variable with several more conservative time since
loss estimates (e.g. excluding four outliers >2SDs above the mean, excluding
24 outliers above the 3rd quartile); these more stringent tests did not change
the significance or pattern of results, indicating that subsets of extreme scores
on time since loss were not unduly influencing the results.
DISCUSSION
Although the loss of a loved one is a universal stressor, the impact of such
losses can be influenced by characteristics of the loss, characteristics of the
individual, and characteristics of the individual’s social environment. We
FIGURE 1A. Interactive effects of social constraints and loss-related intrusions
predict more depressive symptoms. Simple effects test of slopes significant at:
***p<.001.
12 JUTH ET AL.
© 2015 The International Association of Applied Psychology
extend the literature on disclosure and social support in bereavement by
examining how social constraints on disclosure of loss-related thoughts and
emotions may relate to bereaved persons’ adjustment. The sample appeared
broadly typical, in that participants reported moderate levels of social con-
straints, and generally normative physical and mental health; thus, the
sample did not report extreme symptoms related to bereavement and may be
broadly representative of the bereavement process (Bonanno, 2004).
Demographic and Loss-Related Correlates of
Social Constraints
Little is known about whether different types of individuals and/or loss
contexts may result in differences in social constraints during bereavement.
Of the demographic predictors we tested, sex of respondent (being female)
predicted (higher) social constraints. Bereaved women may have more poten-
tial for encountering unreceptive responses and experiencing social con-
straints (Turner, 1994) because they are more likely to express their emotions
(e.g. Gross & John, 2003) and share their feelings with more people
(Harrison, Maguire, & Pitceathly, 1995). Nonetheless, men report greater
distress due to social constraints than women in certain social contexts
(e.g. spousal relationships; Zakowski et al., 2003), and decades of research
FIGURE 1B. Interactive effects of social constraints and loss-related intrusions
predict more perceived life stress. Simple effects test of slopes significant at:
**p<.01; ***p<.001.
SOCIAL CONSTRAINTS AND ADJUSTMENT IN BEREAVEMENT 13
© 2015 The International Association of Applied Psychology
demonstrate the importance of social contexts for coping with life events for
both sexes (see Charuvastra & Cloitre, 2008).
Younger age was related to the report of more social constraints. Bereave-
ment at a later stage of life may seem more normative, and thus be accom-
panied by few social constraints. In contrast, as loss early on in one’s life may
be a more unexpected event (cf. Lazarus & Folkman, 1984), it may represent
a more difficult topic for others to discuss, yielding greater constraints.
Younger people may also have fewer opportunities to deal with loss (i.e. they
know fewer people who have died), and so have greater needs that are harder
for others to fulfill/manage.
Individuals with greater financial concerns reported more social con-
straints, perhaps because they are preoccupied with managing loss-related
expenses or changes in their income. It may also be uncomfortable to talk
with family and friends about financial worries, particularly for individuals
who have been financially independent and/or do not want to be a burden to
others. Further enquiry into the factors contributing to financial concerns in
relation to social constraints (during bereavement and other stressful events)
appears warranted.
Social Constraints and Maladjustment
Psychological Health. The SCP model suggests that social constraints
may hinder interpersonal validation and cognitive integration of a loss event,
impeding bereaved persons’ coping efforts. This may, in turn, lead them to
internalise their concerns and experience greater distress and dysphoria. In
support of this view, our results show a strong positive association between
social constraints and depressive symptoms. Experiencing these symptoms in
tandem with the dysphoric affect common in bereaved persons (e.g. despair,
dejection, loneliness; Stroebe et al., 2007) may make individuals susceptible
to developing clinically complicated grief (Burke, Neimeyer, &
McDevitt-Murphy, 2010). Individuals reporting more social constraints also
report higher perceived stress, perhaps resultant from increased vulnerability
to daily hassles/stressors or because actual social constraints experienced
during interpersonal interactions with friends or family members are causing
additional stress.
Physical Health. The SCP model predicts that social constraints disrupt
physiological regulatory processes and lead to physical health problems.
Indeed, persons with higher social constraints reported more somatic com-
plaints and worse global health (a similar pattern as was observed with
psychological health, albeit smaller in magnitude). This finding may reflect
socially constrained individuals’ withdrawal from potentially rewarding
social interactions (i.e. self-imposed isolation), which may sensitise them to
14 JUTH ET AL.
© 2015 The International Association of Applied Psychology
physical symptoms (e.g. headaches, dizziness, chest pain). It may also be
related to loneliness, which is higher among individuals with high social
constraints (Mosher et al., 2012), and has been linked to a wide range of
health problems (e.g. Cacioppo et al., 2002). Importantly, social constraints
may yield feelings of rejection or shame that prevent individuals from seeking
out required medical/health care, with neglected ailments potentially devel-
oping into serious illness or chronic conditions (Stroebe et al., 2007).
Perceived Social Constraints and Social Support
Perceived constraints are not merely a lack of support; in fact, the two are
only moderately (negatively) correlated. Yet, importantly, the links between
social constraints and all health outcomes remain when controlling for per-
ceived social support (even for perceived stress, which was directly predicted
by social support). Extending prior evidence (e.g. Kaynak et al., 2011), this
suggests that social constraints test something conceptually and practically
distinct from perceived support. The implications of this are twofold. For
research, it suggests that we need to assess and examine social constraints and
social support as distinct constructs and that they appear to play unique roles
in adjustment. With respect to therapeutic interventions, it suggests that
providing social support may not be sufficient for promoting recovery if
social constraints are persistent. Educational components that teach
bereaved persons how best to elicit positive and/or receptive responses from
those they rely on may be useful. Moreover, social network members should
be informed on how to both offer support and minimise negative (socially
constraining) responses. In addition, alternative avenues for “safe” emotional
expression, such as online support groups or expressive writing, may be
beneficial for individuals experiencing a high level of social constraints from
family and friends (although evidence in support of the latter during bereave-
ment is mixed; see Lichtenthal & Creuss, 2010; Range, Kovac, & Marion,
2000; Stroebe, Stroebe, Schut, Zech, & van den Bout, 2002).
Moderation Effects
As expected, social constraints moderated the relation between loss-related
intrusions and depressive symptoms as well as perceived life stress. These
findings are consistent with the SCP model (Lepore, 2001), which maintains
that intrusive thoughts are most distressing when people are unable to safely
discuss the stressful experiences fueling those thoughts. Notably, the interac-
tion patterns for depressive symptoms and perceived life stress are slightly
different. Figure 1a shows that among people with very low social con-
straints, intrusive thoughts still has a significant positive association with
depressive symptoms. In contrast, Figure 1b shows that among people with
SOCIAL CONSTRAINTS AND ADJUSTMENT IN BEREAVEMENT 15
© 2015 The International Association of Applied Psychology
very low social constraints, there is no significant relation between loss-
related intrusive thoughts and perceived life stress. The implication is that a
receptive, non-constraining social environment may buffer the adverse effects
of loss-related intrusions on perceived life stress, but not provide sufficient
protection from the adverse effects of loss-related intrusions on depressive
symptoms. In the event that bereaved persons experience both distressing
intrusions and other types of psychological maladjustment (e.g. depressive
symptoms), interventions may need to address sources of social constraints as
well as strategies for coping with intrusive thoughts.
Overall, these findings suggest that social constraints and loss-related
intrusions have a synergistic negative effect on psychological adjustment.
Inhibited loss-related disclosure may hinder individuals from productive cog-
nitive processing on their own. This may lead to repeated activation of
stress-related systems and physiology via perseverative feedback loops (e.g.
an intrusion occurs and upsets the individual, they do not feel able to share it
with others, they attempt to suppress it, and this cycle repeats; Smyth,
Zawadzki, & Gerin, 2013), resulting in chronic stress that places the individ-
ual at risk for a wide range of maladaptive outcomes. Inconsistent with this
view, however, we did not find evidence of synergistic effects of constraints
and intrusions on self-reported physical symptoms. Future research should
continue to explore the interactive relationship between social constraints
and loss-related intrusions as a potential risk factor (e.g. for severe grief or
major depressive disorder).
Limitations
There are several notable limitations to our study. First, the cross-sectional
design does not enable us to make causal inferences about the link between
social constraints and adjustment, leaving open alternative explanations for
the findings on psychological and physical adjustment and social constraints.
Bereaved persons with greater depressive symptoms or stress may encounter
more social constraints because others want them to avoid negative affect. It
may also be the case that individuals with more stressful bereavement experi-
ences desire to disclose more, but encounter greater social constraints (e.g. if
the cause of death is due to a stigmatised illness) or because the death also
affected others in their social network (making it difficult to get support from
usual sources). Understanding the dynamic processes that relate social con-
straints to specific physical and mental health outcomes requires further
research. Second, we relied on self-reported data, which may conflate report-
ing biases with observed results. Third, the study used a self-selected sample
that included primarily Caucasian women and individuals with relatively
high variability in their loss characteristics (e.g. time since loss); as such, care
must be taken in generalising these results, particularly to bereaved men and
16 JUTH ET AL.
© 2015 The International Association of Applied Psychology
non-Caucasian ethnicities, and to specific loss contexts (e.g. the immediate
post-loss period). The failure to observe reliable relationships between loss
characteristics and health may be due to the heterogeneity in loss character-
istics in this sample (e.g. cause of and time passed since the death).
Clinical Implications
Our findings have implications for education and practice, particularly for
those who work with the bereaved (e.g. clinical and pastoral care settings,
nursing homes and long-term care facilities, and hospices). Given that we
observed relatively high social constraints regardless of demographic and loss
characteristics, inhibited loss-related disclosure may be a common risk factor
for maladjustment during bereavement. Moreover, results from moderation
analyses suggest that post-bereavement risk needs to be evaluated in the
context of both social constraints and loss-related intrusions. In contexts
where loss-related intrusions are low, there may be little risk of—and, thus,
little need to address—high levels of social constraints. Conversely, when
loss-related intrusions and social constraints are both high, risk (of depressive
symptoms, perceived life stress) is magnified; as such, provision of bereave-
ment or other counseling services to individuals in such circumstances may be
warranted. These data suggest that efforts to reduce social constraints may be
useful (in conjunction with, or independent of, specific opportunities for
disclosure; Smyth, Nazarian, & Arigo, 2008), although we note that disclo-
sure does not always result in positive outcomes (e.g. Bonanno, 2004; Stroebe
et al., 2002). Finally, outreach services may consider making extended care
available for individuals who have experienced a loss, particularly for those
who have a limited or small social network, verbalise the need to talk about
their loss, and/or demonstrate persistent loss-related intrusions and psycho-
logical or physical dysfunction.
ACKNOWLEDGEMENT
This project was conducted at Syracuse University with support from the
Center for Health and Behavior.
REFERENCES
Adams, R.N., Winger, J.G., & Mosher, C.E. (2014). A meta-analysis of the relation-
ship between social constraints and distress in cancer patients. Journal of
Behavioral Medicine, published online ahead of print. doi: 10.1007/s10865-014-
9601-6
Badr, H., Pasipanodya, E.C., & Laurenceau, J.-P. (2013). An electronic diary study of
the effects of patient avoidance and partner social constraints on patient momen-
tary affect in metastatic breast cancer. Annals of Behavioral Medicine,45, 192–202.
SOCIAL CONSTRAINTS AND ADJUSTMENT IN BEREAVEMENT 17
© 2015 The International Association of Applied Psychology
Belsher, B.E., Ruzek, J.I., Bongar, B., & Cordova, M.J. (2012). Social
constraints, posttraumatic cognitions, and posttraumatic stress disorder in
treatment-seeking trauma survivors: Evidence for a social-cognitive pro-
cessing model. Psychological Trauma: Theory, Research, Practice, and Policy,4,
386–391.
Bonanno, G.A. (2004). Loss, trauma, and human resilience: Have we underestimated
the human capacity to thrive after extremely aversive events? American Psycholo-
gist,59, 20–28.
Braitman, A.L., Derlega, V., Henson, J., Robinett, I., Saadeh, G., Janda, L. et al.
(2008). Social constraints in talking about diabetes to significant others and dia-
betes self-care: A social-cognitive processing perspective. Journal of Social and
Clinical Psychology,27, 949–969.
Burke, L.A., Neimeyer, R.A., & McDevitt-Murphy, M.E. (2010). African American
homicide bereavement: Aspects of social support that predict complicated grief,
PTSD, and depression. Omega,61, 1–24.
Cacioppo, J.T., Hawkley, L.C., Crawford, E., Ernst, J.M., Burleson, M.H.,
Kowalewski, R.B. et al. (2002). Loneliness and health: Potential mechanisms.
Journal of Psychosomatic Medicine,64, 407–417.
Charuvastra, A., & Cloitre, M. (2008). Social bonds and posttraumatic stress disor-
der. Annual Review of Psychology,59, 301–328.
Cohen, S., & Williamson, G. (1988). Perceived stress in a probability sample of the
United States. In S. Spacapan & S. Oskamp (Eds.), The social psychology of health
(pp. 31–67). Newbury Park, CA: Sage.
Cunny, K.A., & Perri, M. (1991). Single-item vs. multiple-item measures of health-
related quality of life. Psychological Reports,69, 127–130.
Dyregrov, K., & Dyregrov, A. (2008). Effective grief and bereavement support: The
role of family, friends, colleagues, schools and support professionals. London: Jessica
Kingsley Publishers.
Green, H.J., Ferguson, M., Shum, D.H.K., & Chambers, S.K. (2013). Prospective
individual and social predictors of changes in adjustment for patients attending a
regional cancer service. Quality of Life Research,22, 1–12.
Gross, J.J., & John, O.P. (2003). Individual differences in two emotion regulation
processes: Implications for affect, relationships, and well-being. Journal of Person-
ality and Social Psychology,85, 348–362.
Harrison, J., Maguire, P., & Pitceathly, C. (1995). Confiding in crisis: Gender differ-
ences in pattern of confiding among cancer patients. Social Science & Medicine,9,
1255–1260.
Horowitz, M.J., Wilner, N.R., & Alvarez, W. (1979). Impact of Event Scale: A
measure of subjective stress. Psychosomatic Medicine,41, 209–218.
Kaynak, Ö., Lepore, S.J., & Kliewer, W.L. (2011). Social support and social con-
straints moderate the relation between community violence exposure and depres-
sive symptoms in an urban adolescent sample. Journal of Social and Clinical
Psychology,30, 250–269.
Kliewer, W., Lepore, S.J., Oskin, D., & Johnson, P.D. (1998). The role of social and
cognitive processes in children’s adjustment to community violence. Journal of
Consulting and Clinical Psychology,66, 199–209.
18 JUTH ET AL.
© 2015 The International Association of Applied Psychology
Lazarus, R.S., & Folkman, S. (1984). Stress, appraisal, and coping. New York:
Springer Publishing.
Lepore, S.J. (2001). A social-cognitive processing model of emotional adjustment to
cancer. In A. Baum & B. Anderson (Eds.), Psychosocial interventions for cancer
(pp. 99–118). Washington, DC: American Psychological Association.
Lepore, S.J., & Helgeson, V.S. (1998). Social constraints, intrusive thoughts and
mental health after prostate cancer. Journal of Social and Clinical Psychology,17,
89–106.
Lepore, S.J., & Ituarte, P.H.G. (1999). Optimism about cancer enhances mood by
reducing negative social interactions. Cancer Research, Therapy and Control,8,
165–174.
Lepore, S.J., & Revenson, T.A. (2007). Social constraints on disclosure and adjust-
ment to cancer. Social and Personality Psychology Compass: Health,1, 313–333.
Lepore, S.J., Silver, R.C., Wortman, C.B., & Waymant, H.A. (1996). Social con-
straints, intrusive thoughts, and depressive symptoms among bereaved mothers.
Journal of Personality and Social Psychology,70, 271–282.
Lichtenthal, W., & Creuss, D. (2010). Effects of directed written disclosure on grief
and distress symptoms among bereaved individuals. Death Studies,34, 475–499.
Lindstrom, T.C. (2002). “It ain’t necessarily so. . .”: Challenging mainstream thinking
about bereavement. Family & Community Health. Special Issue: Bereavement in
families and the community,25, 11–21.
Little, R.J.A., & Rubin, D.B. (2002). Statistical analysis with missing data (2nd edn.).
New York: John Wiley.
Marris, P. (1982). Attachment and society. In C.M. Parkes, & J. Stevenson-Hinde
(Eds.), The place of attachment in human behavior (pp. 185–201). New York: Basic
Books.
Mosher, C.E., Lepore, S.J., Wu, L., Austin, J., Valdimarsdottir, H., Rowley, S. et al.
(2012). Social correlates of distress following hematopoietic stem cell transplanta-
tion: Exploring the role of loneliness and cognitive processing. Journal of Health
Psychology,17, 1022–1023.
Nenova, M., DuHamel, K., Zemon, V., Rini, C., & Redd, W.H. (2013). Posttraumatic
growth, social support and social constraint in hematopoietic stem cell transplant
survivors. Psycho-Oncology,22, 195–202.
Pasipanodya, E.C., Parrish, B.P., Laurenceau, J.P., Cohen, L.H., Siegel, S.D.,
Graber, E.C. et al. (2012). Social constraints on disclosure predict daily well-being
in couples coping with early-stage breast cancer. Journal of Family Psychology,26,
661–667.
Pennebaker, J.W. (1982). The psychology of physical symptoms. New York: Springer-
Verlag.
Procidano, M., & Heller, K. (1983). Measures of perceived social support from friends
and from family: Three validation studies. American Journal of Community Psy-
chology,11, 1–24.
Radloff, L.S. (1977). The CES-D Scale: A self-report depression scale for research in
the general population. Applied Psychological Measurement,1, 385–401.
Range, L., Kovac, S., & Marion, M. (2000). Does writing about the bereavement
lessen grief following sudden, unintentional death? Death Studies,24, 115–134.
SOCIAL CONSTRAINTS AND ADJUSTMENT IN BEREAVEMENT 19
© 2015 The International Association of Applied Psychology
Segal, D.L., Chatman, C., Bogaards, J.A., & Becker, L.A. (2001). One-year follow-up
of emotional expression intervention for bereaved older adults. Journal of Mental
Health and Aging,7, 465–472.
Sheridan, M.A., Sherman, M.L., Pierce, T., & Compas, B.E. (2010). Social support,
social constraint, and affect in spouses of women with breast cancer: The role of
cognitive processing. Journal of Social and Personal Relationships,27, 5–22.
Smyth, J., & Greenberg, M. (2000). Scriptotherapy: The effects of writing about
traumatic events. In J. Masling & P. Duberstein (Eds.), Psychodynamic perspec-
tives on sickness and health (pp. 121–154). Washington, DC: American Psychologi-
cal Association Press.
Smyth, J., Nazarian, D., & Arigo, D. (2008). Expressive writing in the clinical context.
In J. Denollet, I. Nyklicek, & A. Vingerhoets (Eds.), Emotion regulation: Concep-
tual and clinical issues (pp. 215–233). New York: Springer.
Smyth, J., Zawadzki, M., & Gerin, W. (2013). Stress and disease: A structural and
functional analysis. Social and Personality Psychology Compass,7, 217–227.
Stroebe, M., & Schut, H. (2010). The dual process model of coping with bereavement:
A decade on. Death Studies,61, 273–289.
Stroebe, M., Schut, H., & Stroebe, W. (2007). Health outcomes of bereavement.
Lancet,370, 1960–1973.
Stroebe, M., Stroebe, W., Schut, H., Zech, E., & van den Bout, J. (2002). Does
disclosure of emotions facilitate recovery from bereavement? Evidence from two
prospective studies. Journal of Consulting and Clinical Psychology,70, 169–178.
Turner, H.A. (1994). Gender and social support: Taking the bad with the good. Sex
Roles,30, 521–541.
Wortman, C.B., & Boerner, K. (2007). Beyond the myths of coping with loss: Pre-
vailing assumptions versus scientific evidence. In H.S. Friedman & R.C. Silver
(Eds.), Foundations of health psychology (pp. 285–324). New York: Oxford Uni-
versity Press.
Wortman, C., & Silver, R.C. (1989). The myths of coping with loss. Journal of
Consulting and Clinical Psychology,57, 349–357.
Zakowski, S.G., Harris, C., Krueger, N., Laubmeier, K., Garrett, S., Flanigan, R.
et al. (2003). Social barriers to emotional expression and their relations to distress
in male and female cancer patients. British Journal of Health Psychology,8, 271–
286.
20 JUTH ET AL.
© 2015 The International Association of Applied Psychology
