Social Network Structure and Social Support in
HIV-Positive Inner City Mothers
Edythe S. Hough, EdD, RN
Morris A. Magnan, PhD, RN
Thomas Templin, PhD
Hesham F. Gadelrab, PhD
It has been documented that social support influ-
ences health outcomes of persons with chronic ill-
nesses. The incidence of HIV and AIDS among mi-
nority women is growing at an alarming rate, but
little is known about social support in this vulnerable
population, and even less is known about the social
network conveying that support. Guided by the con-
voy of social networks model, this study describes the
social networks in a sample of HIV-positive, urban-
dwelling mothers (N ⫽ 147) by stage of disease (i.e.,
asymptomatic, symptomatic, AIDS) and examines re-
lationships between social network structure and so-
cial support. Hierarchical linear modeling showed
that women’s social networks were disproportion-
ately populated by children, and network members of
women with AIDS were significantly older than net-
work members of HIV-positive women with or with-
out symptoms. Profile analyses showed that women’s
perceptions of the quality of social support differed
according to the proportion of family members pop-
ulating different segments of the social network.
Key words: social support, social network, HIV-
positive mothers
An increasing proportion of newly diagnosed HIV
and AIDS cases are being reported among minority
urban women. The majority of women in the United
States with HIV are women of color. African Amer-
ican and Hispanic women account for 80% of the
women with HIV, and, in 2000, over 63% of all
women with AIDS were African American (Centers
JOURNAL OF THE ASSOCIATION OF NURSES IN AIDS CARE, Vol. 16, No. 4, July/August 2005, 14-24
doi:10.1016/j.jana.2005.05.002
Copyright © 2005 Association of Nurses in AIDS Care
for Disease Control and Prevention [CDC], 2003).
AIDS was the third leading cause of death among
African American women ages 25 to 44 (CDC,
2003), many of whom are mothers. Consequently, in
addition to suffering from a life-threatening condi-
tion, many HIV-infected women are faced with ad-
ditional challenges such as coping with single par-
enthood, poverty, and racism.
Social support has been found to be a critical
factor in coping with HIV (Hays, Chauncey, & To-
bey, 1990; Kyle & Sachs, 1994; Rodgers, 1995).
However, research examining the importance of so-
cial support has focused primarily on White gay/
bisexual men, and results may have limited general-
izability to poor urban African American women.
Hough, Brumitt, Templin, Salzt, and Mood (2003),
in a study of HIV-infected African American moth-
ers, found that effective coping (i.e., characterized by
a woman’s ability to make meaning of her experi-
ences) was positively related to her social support.
Mothers with adequate social support were more
likely to use effective coping strategies and to exhibit
Edythe S. Hough, EdD, RN, is a professor in the College
of Nursing at Wayne State University, Detroit. Morris A.
Magnan, PhD, RN, is an associate professor in research in
the College of Nursing at Wayne State University, Detroit.
Thomas Templin, PhD, is an associate professor in re-
search in the College of Nursing at Wayne State Univer-
sity, Detroit. Hesham F. Gadelrab, PhD, at the time of the
study, was a doctoral candidate and research assistant in
the College of Nursing at Wayne State University.
 Hough et al. / Social Network and Support in HIV⫹ Mothers
decreased mood disturbance. These results point to
the importance of social support as a buffer against
the deleterious effects of HIV-related stress on a
mother’s ability to cope with her diagnosis. How-
ever, to obtain a more complete picture of the social
support system and an understanding of the sources
of support, information is needed regarding the char-
acteristics of the social networks of these women.
Social Networks
The term social networks refers to the structural
characteristics of an individual’s informal support
network as opposed to a formal network of paid
support providers. A model that provides a frame-
work for the study of social networks across the life
span is the Social Convoy Model (Kahn & Anto-
nucci, 1980). The social convoy is represented as a
series of three concentric circles surrounding an in-
dividual. Inclusion of persons in a convoy circle is
determined by the individual’s emotional attachment
to the person as well as by the role of the person in
relation to the focal individual. Thus, persons occu-
pying the innermost circle of an individual’s social
convoy are likely to be important support providers
with whom the focal individual feels very close.
Membership in the second circle suggests a degree of
closeness to the focal individual as well as relation-
ships that are based on more than just fulfillment of
a role. Finally, members of the third circle are
thought to be close to the focal individual but usually
in a very role-prescribed manner. In its entirety, the
convoy model evokes an image of a protective layer
of family, friends, and others who “surround the
individual and help in the successful negotiation of
life’s challenges” (Antonucci & Akiyama, 1987, p.
516) as the individual moves through life. Changes in
the convoy are hypothesized to occur across norma-
tive life transitions, in response to nonnormative
events, or as a result of individual maturation.
The relationship of social network characteristics
to perceptions of social support, psychological well-
being, and health-related outcomes has not been
studied extensively in HIV-positive persons. How-
ever, the literature suggests that the size of one’s
social network seems to be less important to health
outcomes than its composition. For example, re-
15
searchers incorporating specific providers of social
support (e.g., family, friends, partner) have reported
that family members are not viewed as particularly
helpful to HIV-positive persons. In fact, the majority
of studies suggest that friends provide more support
than family to HIV-infected individuals (Burgoyne &
Saunders, 2000; Friedland, Renwick, & McColl,
1996; Hays, Catania, McKusick, & Coates, 1990;
Hays et al., 1990; Hays, Magee, & Chauncey, 1994;
Johnston, Stall, & Smith, 1995; Namir, Woolcott,
Fawzy, & Alambaugh, 1987; Schwarzer, Dunkel-
Schetter, & Kemeny, 1994; Stowe, Ross, Wodak,
Thomas, & Larson, 1993). Explanations offered for
friends being more supportive than family have in-
cluded both the social stigma associated with HIV
(Barroso, 1997) and issues stemming from problem-
atic life styles before contracting HIV (Owens,
2003). Epidemiological studies (Cohen, Doyle,
Skoner, Rabin, & Gwaltney, 1997; Flaskerud & Win-
slow, 1998) suggest that individuals who engage in
diverse social networks live longer than those with
fewer types of social relationships, and Cohen (1988)
proposed that the more isolated HIV-positive women
are, the more susceptible they might be to delayed
healthcare and disease progression.
To date, no published work details the character-
istics of the social networks of poor urban minority
HIV-positive mothers. There is a great need for in-
formation about the social networks of HIV-infected
mothers because of the potential for the illness to
exert extreme burdens on a woman’s social support
system. First, there is the issue of stigma which,
despite over two decades of media coverage of the
successful treatment of the disease, is still an issue,
especially among African Americans. Stigma leads
to secrecy and its main consequence: the presence of
few knowledgeable people that the person can rely
on in times of needed support (Armistead & Fore-
hand, 1995). Additionally, HIV-infected mothers,
like HIV-infected gay men, may enter this phase of
their life with weakened social networks because of
their own past behavior and transgressions (Owens,
2003). Finally, as a mother gets increasingly more ill,
the social network may simply become more de-
pleted because of her decreased ability to maintain
reciprocal relationships (Hudson, Lee, Miramontes,
& Portillo, 2001) leaving the woman with limited
and perhaps inadequate sources of social support at a
16 JANAC Vol. 16, No. 4, July/August 2005
time when it is needed most for her and her children
(Dorsey, Chance, Forehand, Morse, & Morse, 1999;
Dorsey, Klein, & Forehand, 1999).
Social Support
Social support is the functional imperative of the
social network. Social support, an omnibus concept,
is broadly defined as the subjective feeling of belong-
ing; of being accepted, loved, esteemed, valued, and
needed for oneself, not for what one can do for others
(Moss, 1973). Several dimensions of social support,
namely emotional, informational, and instrumental
support, are seen recurrently in the social support
literature. Emotional support has been defined as
providing esteem and affiliation to others (Norbeck,
Lindsey, & Carrieri, 1981, 1983). Instrumental sup-
port includes financial or housing assistance (Gielen,
McDonnel, Wu, O’Campo, & Faden, 2001), and
informational support has been reported as giving
advice and information on social, health, or employ-
ment matters (Turner, Hays, & Coates, 1993). As-
sessments of social support often include the function
it serves (i.e., emotional, instrumental, informational)
as well its availability, desirability, and frequency of
occurrence. The data clearly indicate that quality of
social support influences coping, psychological well-
being, and health-related outcomes in persons with a
chronic illness (DiMatteo & Hays, 1981; Schwarzer
& Leppin, 1989; Taylor & Aspinwall, 1990; Wort-
man & Conway, 1985), including in persons with
AIDS (Britton, Zarski, & Hobfall-Li, 1993; Grum-
mon, Rigby, Orr, Procidan, & Reznikoff, 1994;
Hays, Turner, & Coates, 1992; Linn, Lewis, Cam-
eron, & Kimbrough, 1993; Nott, Vedhara, & Power,
1995; Pakenham, Dadds, & Terry, 1994).
Despite a strong tendency toward secrecy among
HIV-infected individuals, the beneficial effect of so-
cial support on psychological distress has been dem-
onstrated consistently among samples of gay men
with AIDS (Leserman, Perkins, & Evans, 1992;
Namir et al., 1987; Namir, Alumbaugh, Fawzy, &
Wolcott, 1990; Ostrow et al., 1989; Zich & Temo-
shok, 1990) and HIV-infected women (Clark et al.,
1995; Demi et al., 1995; Nannis, Semple, & Patter-
son, 1995). Clark et al. reported that HIV-infected
women who were less depressed were more likely to
have a confidante. Also, greater availability of social
support has been reported to be associated with less
depressive symptomatology in women (Bluestone,
1995; Demi et al., 1995; Nannis et al., 1995;
O’Leary, Wadhwani, Gebelt, & Frenkel, 1995).
However, Bluestone cautioned that, although social
support may buffer depression when there are few
infection-related symptoms, social support may be a
less effective buffer as the illness progresses.
One study of social support in poor minority HIV-
positive women (Klein et al., 2000) reported that
HIV-infected women received less social support
than a sociodemographically similar group of unin-
fected women. Research examining the specific pro-
viders of social support has suggested that for many
with HIV, family members are not viewed as being
as helpful or supportive as friends. This result has
been particularly salient among those who contracted
the virus through homosexual sex or drug use
(Johnston et al., 1995; Stowe et al., 1993). In other
words, relationships with family members may have
been strained before the diagnosis. It is not clear how
this process may apply to heterosexual women. Hud-
son et al. (2001), from a study of a group of ethni-
cally diverse women in San Francisco, reported lim-
ited social interactions with family and friends and a
low level of perceived social support. Limited per-
ceived social support was a significant predictor of
distress in this sample of women.
The use of active behavioral coping strategies has
been positively associated with both the size and
satisfaction with one’s social support network
(Leserman et al., 1992; Namir et al., 1987, 1990;
Nannis et al., 1995). Hough et al. (2003) tested a
model of mother and child coping and adjustment in
relation to maternal HIV in a sample of inner-city
African American women. Accounting for 36% of
the variance in child adjustment, the maternal vari-
ables reported to be significant predictors in the
model were maternal HIV-associated stressors, ma-
ternal emotional distress, child social support, child
coping, and quality of the parent-child relationship.
In particular, mothers with adequate social support
were more likely to use effective coping strategies
such as seeking information and using positive reap-
praisal; they also exhibited decreased emotional dis-
tress. Hough et al. (2003) also reported that the
 Hough et al. / Social Network and Support in HIV⫹ Mothers
amount of social support experienced decreased as
HIV symptoms increased.
Although the literature clearly demonstrates the
importance of social support in the lives of HIV-
positive persons including HIV-infected women, we
have little knowledge regarding the characteristics of
the social network conveying the needed support.
Moreover, theory and research suggest that the diag-
nosis of HIV as well as disease progression—from
asymptomatic to symptomatic to AIDS—precipitate
a reconfiguration of the social network, which may
adversely affect the quality of social support pro-
vided. To the best of the intestigators’ knowledge,
the social networks of HIV-infected mothers have
not been studied and the relationship between social
network structure and social support in this vulnera-
ble population has yet to be determined. This report
will (a) describe the structural characteristics of the
social networks of HIV-positive mothers at different
stages of the disease (i.e., asymptomatic, symptom-
atic, and AIDS), and (b) examine the relationship
between social network structure and social support
characteristics (i.e., desirability, availability, fre-
quency, helpfulness) reported by women in this par-
ticularly vulnerable population. This information will
provide caregivers with important information nec-
essary for the design and implementation of nursing
interventions to strengthen and improve the social
support systems of HIV-positive women.
Method
Sample
The sample for this study consisted of 147 HIV-
positive women living in the metropolitan Detroit
area. These women were young (M ⫽ 36.1 years; SD
⫽ 7.1), predominantly African American (86%), and
single (78%). Although most of the women were
receiving public assistance (97%), they were very
poor with an average monthly household income of
only $712. The average time since diagnosis was a
little more than 5 years (M ⫽ 62.8 months; SD ⫽ 39).
Using the CDC’s classification system for HIV status
(Osmond, 1998), 57% of the women were asymp-
tomatic, 18% were symptomatic, and 25% had
AIDS.
17
Procedure and Measures
Eligible women were clients of a participating
clinic/agency who identified themselves as being
HIV-infected with confirmation by agency records.
The women were recruited from community-based
AIDS service organizations, case management agen-
cies, and medical clinics in the metropolitan Detroit
area to participate in a larger study investigating the
effects of mother-child coping and adjustment to
HIV (Hough et al., 2003). Contact between the re-
cruitment coordinator and the subjects occurred ei-
ther face to face during a regularly scheduled agency
visit or by a subject-initiated telephone call after
referral by a caseworker. During the initial contact,
subjects were asked a number of screening questions
to determine their eligibility and willingness to par-
ticipate in the study. The recruitment coordinator
then scheduled an appointment for subjects to come
to Wayne State University to be interviewed. Trans-
portation and childcare were provided to facilitate
subject participation in the study. Subjects were paid
$50 on completion of the interview.
Social network was assessed using the convoy
model of social networks described by Antonucci
(1986) and Antonucci and Akiyama (1987). Network
membership was established by giving subjects a
target diagram consisting of three concentric circles
with the word you written in the innermost circle.
Subjects were told that each circle should be thought
of as including “people who are important in your
life right now” but who are not equally close to you.
To identify memberships in the innermost circle,
subjects were told to think about “people to whom
you feel so close that it is hard to imagine life without
them.” Such persons were listed in the innermost
circle of the network diagram. The same procedure
was followed to determine memberships for the mid-
dle circle, described as including “people you do not
feel quite that close to but who are still very impor-
tant to you,” and for the outer circle, described as
including “people you feel less close to but who are
still important to you.”
Structural characteristics of the social networks
were determined by asking subjects to provide addi-
tional information about the first 10 people listed in
their social network diagram. Thus, structural char-
acteristics of networks included not only circle place-
18 JANAC Vol. 16, No. 4, July/August 2005
ment (i.e., inner, middle, outer) but also relationship
to the focal person (e.g., child, family, friend), age,
distance between the residence of the focal person
and the network member, frequency of telephone
calls, and frequency of visits. Scores for these addi-
tional network variables were computed by averag-
ing values across memberships within circles. For
example, for respondent 1, an inner-circle age score
was obtained by determining the average age of
network members listed in the innermost circle. Mid-
dle-circle and outer-circle age scores were computed
in the same way. This yielded from 1 to 3 average
age scores per subject depending upon how the first
10 network members were distributed across circles.
This procedure was repeated for other structural
characteristics.
Social support, a component of network function,
was assessed using a questionnaire developed for a
study of social support and physical and psycholog-
ical distress among gay men diagnosed with AIDS or
AIDS-related complex (Zich & Temoshok, 1990).
Using this eight-item instrument, subjects provided
self-ratings on four components of social support
(i.e., desirability, availability, frequency, and help-
fulness) on a 5-point Likert-type scale. A ninth item
was added to assess satisfaction with support for
child caretaking responsibilities (“he/she helped me
by taking care of my children when I couldn’t”). In
this sample, alpha coefficients for desirability, avail-
ability, frequency, and helpfulness were .84, .89, .84,
and .83 respectively.
Data Analysis
All data were analyzed using Statistical Package
for the Social Sciences 11.5 (SPSS, Inc., Chicago,
IL). Our data were inherently unbalanced because of
the fact that network memberships across circles
differed from subject to subject. Therefore, hierar-
chical linear modeling was used to analyze social
network characteristics by HIV status and circle
placement because it offers the distinct advantage of
making it possible to analyze unbalanced data at any
level. Profile analysis was used to determine whether
the pattern of social support characteristics (i.e., de-
sirability, availability, frequency, and helpfulness)
differed by circle composition (e.g., high or low
percentage of children in the circle).
Analyses of the size of the social network convoy
was based on the total number of network members
(N ⫽ 1,979) identified by the respondents. However,
detailed information about network members was
obtained only on the first 10 members identified by
each respondent. Consequently, analyses involving
network characteristics other than size were based on
a network membership of N ⫽ 1,329.
Results
Social Network Structure
Network size. The 147 mothers nominated a total
of 1,979 network members, creating an average net-
work of 13.46 members (Table 1). There was a
significant main effect [F (2, 288.5) ⫽ 69.86, p ⬍
.01] in the number of network members across cir-
cles. As shown in Table 1, the innermost circle was
more densely populated than the middle circle or the
outer circle. Network size did not differ significantly
by HIV status (i.e., asymptomatic, symptomatic,
AIDS), and there was no significant interaction effect
of HIV status by circle placement
Age. The age of network members ranged from
less than 1 year to 98 (M ⫽ 31.1; SD ⫽ 20.8). There
were significant main effects of circle placement [F
(2, 239.04) ⫽ 35.73, p ⬍ .01] and HIV status [F (2,
131.41) ⫽ 3.50, p ⬍ .05]. As shown in Table 1,
members of the inner circle were younger than those
of the middle or outer circles. Network members of
subjects with AIDS were older than the network
members of subjects who were asymptomatic or
symptomatic. No significant HIV status by circle
placement interaction was observed.
Distance. Geographic distance from the focal
person ranged from “live with” to about 200 miles
away from the respondent, with an average of 25.5
miles in distance. A significant main effect for circle
placement [F (2, 193.58) ⫽ 11.03, p ⬍ .01] was
reported, indicating that network members in the
inner circle lived closer than network members in the
middle and outer circles. Distance did not differ
significantly by HIV status and there was no signif-
icant HIV status by circle interaction.
 Hough et al. / Social Network and Support in HIV⫹ Mothers 19

Table 1. Mean Values of Network Characteristics by Circle Placement and HIV Classification

Circle Placement
Network Characteristic HIV Class Inner Middle Outer Marginal Mean
Size of network Asymp. 7.15 4.20 2.23 13.59
Symp. 8.07 3.70 2.16 13.81
AIDS 7.00 4.22 1.67 12.89
All 7.41 4.04 2.02 13.46
Age of network members Asymp. 25.18 40.52 40.59 35.43
Symp. 27.66 38.28 36.72 34.22
AIDS 28.79 41.36 47.46 39.20
All 27.21 40.06 41.59 36.29
Distance in miles Asymp. 22.65 39.80 23.31 28.59
Symp. 9.50 49.15 51.08 36.60
AIDS 11.30 24.39 12.97 16.22
All 14.50 38.91 29.12 27.13
Telehone calls per month Asymp. 16.34 11.33 11.93 13.20
Symp. 19.31 7.61 5.78 10.90
AIDS 17.62 10.74 10.11 12.83
All 17.76 9.90 9.27 12.31
Visits per month Asymp. 20.50 8.51 7.45 12.15
Symp. 20.57 6.01 7.50 11.36
AIDS 20.71 9.72 11.49 13.97
All 20.59 8.08 8.82 12.50

NOTE: Asymp. ⫽ asymptomatic, symp. ⫽ symptomatic.

Frequency of telephone calls. Frequency of con-
tact between mothers and their network members
ranged from 0 to 30, with a mean of 13.7 calls per
month. Frequency of telephone calls differed signif-
icantly by circle placement [F (2, 177.53) ⫽ 19.71, p
⬍ .01] but not by HIV status. Members of the inner
circle made telephone contact with focal subjects
more than members in the other two circles. HIV
status and circle placement did not interact signifi-
cantly to influence the frequency of telephone calls.
Frequency of visits. The average number of visits
per month between mothers and their network mem-
bers was 12.5. A significant effect of circle place-
ment was reported [F (2, 210.31) ⫽ 67.95, p ⬍ .01].
Totals for each circle (see Table 1) showed that
members of the inner circle made substantially more
visits than members of the other two circles. There
was no significant difference in the number of visits
reported by persons in the asymptomatic, symptom-
atic, and AIDS groups. The interaction of HIV status
and circle placement had no significant effect on the
number of monthly visits.
Relationship to mothers. Table 2 presents the
relationship of network members to the mothers. In
the entire network, nearly 84% (83.9%) of the net-
work members were family; 4.7% were husbands,
33.7% were children, and 45.5% were family other
than children. The remaining 16.1% were friends
(13.2%), health care professionals (2.1%), and higher
beings (i.e., God, spiritual beings; .8%). A majority
(69.9%) of the network members were nominated to
the inner circle. Eighty-six percent of the inner circle
members were children (47.1%) and family other
than children (38.8%). Children were the most salient
support members in the inner circle. In the middle
and outer circles, family other than children were the
most prominent members; 61.4% and 60.9%, respec-
tively. The proportion of friends in the inner circle
was only 6%, whereas the proportion of friends in the
middle and outer circles was 28.6% and 34.8%,
respectively.
The investigators explored the role of four types of
relationships (i.e., husband, children, family other
than children, and friends) by HIV status and circle
placement. There were no significant main effects for
20 JANAC Vol. 16, No. 4, July/August 2005

Table 2. Percentages for Social Network Relationships by Circle Placement and HIV Classification

Number Relationship to Focal Person

of Family
Circle HIV Network Other Than Health Care Higher
Placement Class Members Husband Children Children Friends Professionals Being
Inner circle Asymp. 515 4.7 51.5 36.9 4.9 1.2 1.0
Symp. 175 6.3 41.7 39.4 7.4 4.0 1.1
AIDS 239 6.3 41.8 42.3 7.5 0.8 1.3
ALL 929 5.4 47.1 38.8 6.0 1.6 1.1
Middle circle Asymp. 184 2.7 2.7 60.3 29.9 4.3 0.0
Symp. 57 7.0 1.8 64.9 26.3 0.0 0.0
AIDS 67 3.0 4.5 61.2 26.9 4.5 0.0
ALL 308 3.6 2.9 61.4 28.6 3.6 0.0
Outer circle Asymp. 62 1.6 0.0 59.7 35.5 3.2 0.0
Symp 10 0.0 10.0 60.0 30.0 0.0 0.0
AIDS 20 0.0 0.0 65.0 35.0 0.0 0.0
ALL 92 1.1 1.1 60.9 34.8 2.2 0.0
Entire network Asymp 761 3.9 35.5 44.4 13.4 2.1 0.7
Symp 242 6.2 31.0 46.3 12.8 2.9 0.8
AIDS 326 5.2 31.6 47.5 13.2 1.5 0.9
ALL 1,329 4.7 33.7 45.5 13.2 2.1 0.8

NOTE: Asymp. ⫽ asymptomatic, symp. ⫽ symptomatic.

circle placement or HIV status for husbands, and
there were no significant interaction effects from
HIV status by circle placement. However, significant
main effects of circle placement were reported for
children [F (2,221) ⫽ 128.3, p ⬍ .01], for family
other than children [F (2,237) ⫽ 14.9, p ⬍ .01], and
for friends [F (2,223) ⫽ 17.3, p ⬍ .01]. Thus, the
inner circles of the women were more likely to be
occupied by children, whereas their outer circles
were populated by friends and family other than
children.
Social Network Structure and Social Support
To investigate the relationship between social net-
work characteristics and social support, three net-
work variables were created to represent what ap-
peared to be the most prominent features of the
networks of these women, specifically: (a) the pro-
portion of children in the inner circle, (b) the propor-
tion of family other than children in the inner circle,
and (c) the proportion of family other than children in
the middle circle. Each of these variables was then
dichotomized and coded as either high proportion or
low proportion using the median as the cut point.
Profile analyses were performed using each of these
dichotomous variables as the independent variable
and the functional characteristics of social support
(i.e., desirability, availability, frequency, and help-
fulness) as the dependent variable.
Results of profile analysis for the proportion of
children in the inner circle revealed a significant
difference in profiles [F (3, 435) ⫽ 2.96, p ⬍ .01].
Mothers falling above the median in the proportion
of children in their inner circle rated their social
support as less available, less frequent, and less help-
ful than those mothers who fell below the median.
However, the two groups of mothers did not differ in
their ratings of the desirability of social support.
In terms of family other than children in the inner
circle, a significant difference in profiles was re-
ported [F (3, 435) ⫽ 2.7, p ⬍ .05]. Mothers above the
median who reported family members other than
children as their primary source of support in the
inner circle rated their support as more available but
less desirable. Social support frequency and helpful-
ness were rated almost the same by both groups of
mothers.
Mothers who fell above the median in reporting
predominantly family other than children in the mid-
 Hough et al. / Social Network and Support in HIV⫹ Mothers
dle circle reported experiencing significantly less so-
cial support [F (3,324) ⫽ 81.86, p ⬍ .01] on all
dimensions (i.e., desirability, availability, frequency,
and helpfulness).
Discussion
This report provides a detailed description of the
social networks of a group of poor HIV-positive,
largely African American, inner-city mothers. It ex-
amines the social network characteristics of these
women according to the stage of illness (i.e., asymp-
tomatic, symptomatic, and AIDS) and relates social
network structure to mother’s report of her extant
social support. This study is a preliminary step in
increasing our understanding of the interpersonal
context (i.e., both social network structure and social
support) of this highly vulnerable population of
women.
The women had no difficulty responding to the
instructions on how to complete the convoy diagram
and did so readily. This sample of 147 women named
a total of 1,979 network participants, or an average of
13.46 per person. Although we could find no other
report on the network size of HIV-positive persons,
this is somewhat larger than those reported by
women who abuse drugs, as was the case with the
majority of the women in our sample. For example,
Falkin and Strauss (2003) reported an average of 9
persons in their sample of drug-using women. Also,
by comparison, Antonucci and Akiyama (1987) re-
ported an average of 9 network members in a large
sample of elderly persons. Consequently, on the basis
of size it seems that this particular sample of HIV-
positive women had fairly robust social networks.
When we examined various network characteris-
tics such as age, distance, frequency of telephone
calls, and frequency of visits, we noted that the inner
circles of these women were populated with signifi-
cantly more younger people than were the middle
and outer circles and that women with AIDS reported
significantly more older people in their entire net-
work than women who were either asymptomatic or
symptomatic. Our result that network members of
women with AIDS were older than network members
of women who were asymptomatic or symptomatic is
consistent with the theoretical claim that changing
21
life events precipitate a reconfiguration of one’s so-
cial network. One explanation for this latter result
might be that as a woman progresses to full-blown
AIDS, her social network becomes more populated
with older people (i.e., parents, aunts, grandparents)
taking on caregiving roles for either the mother or her
children (Dorsey et al., 1999). As one might expect,
women reported that inner-circle members lived
closer and called and visited more than members of
either the middle or outer circles. The fact that other
characteristics of network structure (i.e., size, dis-
tance, and frequency of telephone calls) did not differ
across different stages of the disease suggests that
women were able to maintain these aspects of net-
work structure despite changes in their disease status.
Consistent with the result showing a greater num-
ber of younger people in the mother’s inner circle,
we noted a preponderance of children (47.1%) in this
circle. Taken together with the fact that very few
friends were nominated for the inner circles and
made up only about one-third of both the middle and
outer circles, it appears that the social networks of
these mothers contained relatively few friends and
were somewhat top-heavy with children. This result
is consistent with reports from qualitative studies in
which HIV-positive women reported relying heavily
on their children for a variety of supportive functions
(Andrews, Williams, & Neil, 1993; Barosso, 1997).
The preponderance of children populating the inner
circles is worrisome and raises questions about how
and to what extent children are able to meet the social
support needs of their HIV-positive mothers. The
relative paucity of friends and the preponderance of
family in the social networks of the women also raise
concerns, because of research suggesting that family
members are not viewed as particularly helpful to
HIV-positive persons (Johnston et al., 1995; Stowe et
al., 1993).
Further data analyses were conducted to examine
the relations between social network structure and
the mother’s overall satisfaction with the level of
social support she was receiving. We found that
mothers falling above the median in the proportion of
their inner circle made up of children rated their
social support as significantly less available, fre-
quent, and helpful. This result has a number of im-
portant implications. Although children may provide
mothers with HIV a reason to live, help them main-
22 JANAC Vol. 16, No. 4, July/August 2005
tain a healthier life style, and provide love and af-
fection (Andrews et al., 1993), they also present a
complex set of developmental requirements for such
things as nurturance, material goods, and safety, just
to mention a few. Perhaps these developmental re-
quirements of children outweigh their ability to pro-
vide social support to their mothers. Contrary to
Barosso’s (1997) suggestion that HIV-positive moth-
ers who rely on their family do not need to reassem-
ble their social networks, our results suggest that a
reassembly is very much needed. Mothers may re-
ceive some degree of support from their children, but
it may not provide them with either the type or
amount of support needed to deal with the multiple
stressors associated with being poor, having HIV,
and trying to raise a family in an impoverished social
environment. These mothers are in great need of
assistance, not only to maintain their current network
that places their children at the forefront but also to
enhance their networks with adult friends and family
members who can relieve both the mother and her
children of some of their burdens.
The result showing an almost complete lack of
professional caregivers in the lives of these women is
disappointing and suggests that professional nurses,
among others, have not established themselves as
potential sources of support for this group of women.
A great deal of attention needs to be paid by nurses
and other health professionals to learning how to
communicate and intervene in the social support
systems of HIV-positive women and their children.
One starting point for nurses is to advocate for the
assessment of the social network and social support
available to HIV-positive women as a part of their
ongoing care. Mothers who are relying mainly on
their children for social and emotional support need
to be assisted in maintaining a delicate balance be-
tween their role as mother and their role as an indi-
vidual with a chronic illness. One area of assistance
relates to the process of disclosure of her HIV status
to her child, because this has been identified as an
important threat to the stability of their relationship
(Vallerand, Hough, Brumitt, Pittiglio, & Marviscin,
2004). HIV-positive mothers are also in need of
respite care for their children as well as parenting
advice and counseling. The availability of respite
care is particularly important because of evidence
suggesting that, in general, women exhibit more so-
matic complaints as a result of being worn down by
nurturing role demands.
These results point to a number of directions for
future research. First, since little is known about the
social network structures of impoverished inner-city
African American mothers, the results of this study
need to be validated with women of a similar socio-
economic background who are HIV negative. In ad-
dition, further work is needed to identify the deter-
minants of the mother’s social network composition,
whether these factors change over the course of her
illness, and to what extent the children have an im-
pact on the mother’s ability to form and maintain
supportive relationships with adults.
The children in our study were key providers of
social support for their HIV-positive mothers, but
little is known about the effect this has on the indi-
vidual child. How do these children manage this
burden? Are there gender differences or differences
related to the child’s knowledge and understanding
of HIV? And finally, what kinds of support services
do mothers and children need and find acceptable?
Because professionals played such a minute role in
the interpersonal lives of this sample of women,
future research needs to explore the effectiveness of
intervening at the extended family and community
levels as well as with the individual.
In summary, we described the social network
structure of HIV-positive mothers across a number of
dimensions and at various stages of their disease. In
addition, we examined the mothers’ level of satisfac-
tion with the social support conveyed by their social
network. Our results suggest that there is much to be
learned from studying social networks and social
support concurrently rather than in isolation. An im-
portant implication of our study has to do with the
level measurement of constructs. In this sample of
women, the relationship between network structure
and social support was observed only when a distinc-
tion was made between family members who are
adults and family members who are children. Using
this approach, it became evident that social support
varied according to the proportion of children occu-
pying specific segments of social networks of HIV-
positive mothers.
 Hough et al. / Social Network and Support in HIV⫹ Mothers 23

Acknowledgement Program and abstracts of the HIV infection in women

conference: setting a new agenda: February 22-24, 1995,
Sheraton Washington Hotel, Washington, D.C. (p. 110).
This research was supported by National Institute Richmond, VA: Philadelphia Sciences Group.
of Nursing Research Grant No. NR04349. DiMatteo, M. R., & Hays, R. (1981). Social support and serious
illness. In Gotlieb (Ed.), B. H. Social networks and social
support: Vol. 4. (pp. 117-148). Beverly Hills CA: Sage.
References Dorsey, S., Chance, M. W., Forehand, R., Morse, E., & Morse,
P. (1999). Children whose mothers are HIV infected: Who
resides in the home and is there relationship to child psycho-
Andrews, S., Williams, A. B., & Neil, K. (1993). The mother-
social adjustment? Journal of Family Psychology, 13, 103-
child relationship in the HIV-positive family. IMAGE: Jour-
117.
nal of Nursing Scholarship, 25, 193-198.
Dorsey, S., Klein, K., & Forehand, R. (1999). Parenting self-
Antonucci, T. C. (1986). Measuring social support networks:
efficacy of HIV-infected mothers: The role of social support.
Hierarchical mapping technique. Generations, X(4), 10-12.
Journal of Marriage and the Family, 61, 295-305.
Antonucci, T. C., & Akiyama, H. (1987). Social networks in
Falkin, G. P., & Strauss, S. M. (2003). Social supporters and drug
adult life and a preliminary examination of the convoy model.
use enablers: A dilemma for women in recovery. Addictive
Journal of Gerontology, 42, 519-527.
Behavior, 28, 141-155.
Armistead, L., & Forehand, R. (1995). For whom the bell tolls:
Flaskerud, J. H., & Winslow, B. J. (1998). Conceptualizing
Parenting decisions and challenges faced by mothers who are
vulnerable populations health-related research. Nursing Re-
HIV seropositive. Clinical Psychology: Science and Practice,
search, 47, 69-78.
2, 239-250.
Friedland, J., Renwick, R., & McColl, M. (1996). Coping and
Barroso, J. (1997). Social support and long-term survivors of
social support as determinants of quality of life in HIV/AIDS.
AIDS. Western Journal of Nursing Research, 19, 554-573.
AIDS Care, 8, 15-31.
Bluestone, H. H. (1995). Factors affecting distress in European
Gielen, A. C., McDonnel, K. A., Wu, A. W., O’Campo, P., &
American women with HIV disease [Abstract]. Program and
Faden, R. (2001). Quality of life among women living with
abstracts of the HIV infection in women conference: setting a
HIV: The importance of violence, social support, and self-
new agenda: February 22-24, 1995, Sheraton Washington
care behaviors. Social Science and Medicine, 52, 315-322.
Hotel, Washington, D.C. (p. 124). Richmond, VA: Philadel-
Grummon, K., Rigby, E. D., Orr, D., Procidano, M., & Rezni-
phia Sciences Group.
koff, M. (1994). Psychosocial variables that affect the psy-
Britton, P. J., Zarski, J. J., & Hobfall-Li, S. E. (1993). Psycho-
chological adjustment of IVDU patients with AIDS. Journal
logical distress and the role of significant others in a popula-
of Clinical Psychology, 50, 488-502.
tion of gay/bisexual men in the era of HIVAIDS Care, 5(1),
Hays, R. B., Catania, J. A., McKusick, L., & Coates, T. J. (1990).
43-54.
Help-seeking for AIDS-related concerns: a comparison of gay
Burgoyne, R. W., & Saunders, D. S. (2000). Perceived support in
men with various HIV diagnoses. American Journal of Com-
newly registered HIV/AIDS clinic outpatients. AIDS Care,
munity Psychology, 18, 743-755.
12, 643-650.
Hays, R. B., Chauncey, S., & Tobey, L. A. (1990). The social
Centers for Disease Control and Prevention (2003). HIV/AIDS
support networks of gay men with AIDS. Journal of Com-
among U.S. women: Minority and young women at continu-
munity Psychology, 18, 374-385.
ing risk. Retrieved October 21, 2004, from http://www.ced.
Hays, R. B., Magee, R. H., & Chauncey, S. (1994). Identifying
gov/hiv/pubs/facts/women.htm
helpful and unhelpful behaviors of loved ones: The PWA’S
Clark L Schoenbaum, E., Solomon, L., Schuman, P., Fiore, T., &
perspective. AIDS Care 6, 379-392.
Moore, J. (1995). Concealment of HIV infection and psycho-
Hays, R. B., Turner, H., & Coates, T. J. (1992). Social support,
logical distress [Abstract]. Program and abstracts of the HIV
AIDS-related symptoms, and depression among gay men.
infection in women conference: setting a new agenda: Feb-
Journal of Consulting and Clinical Psychology, 60, 463-469.
ruary 22-24, 1995, Sheraton Washington Hotel, Washington,
Hough, E. S., Brumitt, G., Templin, T., Salzt, E. & Mood, D.
D.C. (S26). Richmond, VA: Philadelphia Sciences Group.
(2003). A model of mother-child coping and adjustment to
Cohen, S. (1988). Psychosocial models of the role of social
HIV. Social Science & Medicine, 56, 643-655.
support in the etiology of physical disease. Health Psychol-
Hudson, A. L., Lee, K. A., Miramontes, H., & Portillo, C. J.
ogy, 7, 269-297.
(2001). Social interactions, perceived support, and level of
Cohen, S., Doyle, W. J., Skoner, D. P., Rabin, B., & Gwaltney,
distress in HIV-positive women. The Journal of the Associ-
J. M. (1997). Social ties and susceptibility to the common
ation of Nurses in AIDS Care, 12(4), 68-76.
cold. Journal of the American Medical Association, 277,
Johnston, D., Stall, R., & Smith, K. (1995). Reliance by gay men
1940-1944.
and intravenous drug users on friends and family for AIDS-
Demi, A., Bakeman, R., Moneyham, L., Sowell, R., Cohen, L.,
related care. AIDS Care, 7(3), 307-319.
Seals, B., et al. (1995). The relationship of family structure to
psychological distress of HIV infected women [Abstract].
24 JANAC Vol. 16, No. 4, July/August 2005
Kahn, R. L., & Antonucci, T. C. (1980). Convoys over the life
course: Attachment, roles and social support. In P. B. Baltes
& O. G. Brim (Eds.), Life-span development and behavior
(pp. 253-286). New York: Academic Press.
Klein, K., Armistead, L., Devine, D., Kotchick, B., Forehand, R.,
Morse, E., et al., (2000). Socioemotional support in African
American families coping with maternal HIV: An examina-
tion of mothers’ and children’s psychosocial adjustment. Be-
havior Therapy, 31, 1-26.
Kyle, R. D., & Sachs, L. G. (1994). Perceptions of control and
social support in relation to psychosocial adjustment to HIV/
AIDS. AIDS Patient Care, 8, 322-327.
Leserman, J., Perkins, D. O., & Evans, D. L. (1992). Coping with
the threat of AIDS: The role of social support. American
Journal of Psychiatry, 149, 1514-1520.
Linn, J. G., Lewis, F. M., Cameron, V. A., & Kimbrough, G. A.
(1993). HIV-illness, social support, sense of coherence, and
psychological well-being in a sample of help-seeking adults.
AIDS Education and Prevention, 5, 254-262.
Moss G. E. (1973). Immunity and social interactions. New York,
NY: Wiley.
Namir, S., Alumbaugh, M. J., Fawzy, F. I., & Wolcott, D. L
(1990). Implications of different strategies for coping with
AIDS. In L. Temoshok & A. Baum (Eds.), Psychosocial
perspectives on AIDS: Etiology, prevention, and treatment
(pp. 173-190). Hillsdale, NJ: Lawrence Erlbaum Associates.
Namir, S. Woolcott, D. L, Fawzy, F. I., & Alambaugh, M. J.
(1987). Coping with AIDS: Psychological and health impli-
cations. Journal of Applied Social Psychology, 17, 309-328.
Nannis, E., Semple, S., & Patterson, T. (1995). HIV⫹ women
and coping: Psychosocial predictors of “fighting spirit” [Ab-
stract]. Program and abstracts of the HIV infection in women
conference: Setting a new agenda: February 22-24, 1995
Sheraton Washington Hotel, Washington, D.C. (S32). Rich-
mond, VA: Philadelphia Sciences Group.
Norbeck, J. S., Lindsey, A. M., & Carrieri, V. L. (1981). The
development of an instrument to measure social support.
Nursing Research, 30, 264-269.
Norbeck, J. S., Lindsey, A. M., & Carrieri, V. L. (1983). Further
development of the Norbeck Social Support Questionnaire:
Normative data and validity testing. Nursing Research, 32,
4-9.
Nott, K. H., Vedhara, K., & Power, M. J. (1995). The role of
social support in HIV infection. Psychological Medicine, 25,
971-983.
O’Leary, A., Wadhwani, S., Gebelt, J., & Frenkel, L. (1995).
Predictors of depression in the primary caregivers of HIV-
infected children [Abstract]. Program and abstracts
of the HIV infection in women conference: setting a new
agenda: February 22-24, 1995, Sheraton Washington Hotel,
Washington, D.C. (S102). Richmond, VA: Philadelphia Sci-
ences Group.
Osmond D. H. (1998). Classification and staging of HIV infec-
tion. HIV InSite Knowledge Base. Retrieved August 15, 1998,
from www.hivinsite.ucsf.edu
Ostrow, D. G., Monjan, A., Joseph, J., VanRaden, M., Fox, R.,
Kingsley, L., et al. (1989). HIV-related symptoms and psy-
chological functioning in a cohort of homosexual men. Amer-
ican Journal of Psychiatry, 146, 737-742.
Owens, S. (2003). African American women living with HIV/
AIDS: Families as sources of support and of stress. Social
Work, 48, 163-173.
Pakenham, K. I., Dadds, M. R., & Terry, D. J. (1994). Relation-
ships between adjustment to HIV and both social support and
coping. Journal of Consulting and Clinical Psychology, 62,
1194-1203.
Rodgers, A. Y. (1995). The relationship between changes in
sexual support and adjustment to AIDS in gay males. Social
Work in Health Care, 20, 37-49.
Schwarzer, R., Dunkel-Schetter, C., & Kemeny, M. (1994). The
multidimensional nature of received social support in gay
men at risk of HIV infection and AIDS. American Journal of
Community Psychology, 22, 319-339.
Schwarzer, R., & Leppin, A. (1989). Social support and health:
A meta-analysis. Psychology and Health: An International
Journal, 3, 1-15.
Stowe, A., Ross, M. W., Wodak, A., Thomas, G. V., & Larson,
S. A. (1993). Significant relationships and social supports of
injecting drug users and their implications for HIV/AIDS
services. AIDS Care, 5, 23-33.
Taylor, S. E., & Aspinwall, L. (1990). Psychosocial aspects of
chronic illness. In Costa, Jr., & P. T. Vandenbos (Eds.), G. R.
Psychosocial aspects of serious illness: Chronic conditions,
fatal diseases and clinical care (pp. 3-60). Washington, DC:
American Psychological Association.
Turner, H. A., Hays, R. B., & Coates, T. J. (1993). Determinants
of social support among gay men: The context of AIDS.
Journal of Health and Social Behavior, 34, 37-53.
Vallerand A. H., Hough E., Brumitt G., Pittiglio L., & Marviscin
D. (2004, March). The process of disclosure between HIV
positive mothers and their children. Competitive symposium.
Symposium conducted at MNRS 2004 Annual Research Con-
ference, March 28, St. Louis, MO.
Wortman, C. B., & Conway, T. L. (1985). The role of social
support in adaptation and recovery from serious illness. In
Cohen & S. Syme (Eds.), S. L. Social support and health (pp.
281-302). New York: Academic Press.
Zich, J., & Temoshok, L. (1990). Perceptions of social support,
distress, and hopelessness in men with AIDS and ARC:
Clinical implications. In Temoshok, & L. Baum (Eds.), A.
Psychosocial perspectives on AIDS: Etiology, prevention,
and treatment (pp. 173-190). Hillsdale, NJ: Lawrence Erl-
baum Associates.