Sex Roles, Vol. 37, Nos.

3/4, 1997

Men's Adjustment to Prostate Cancer: The

Role of Agency and Unmitigated Agency1
Vicki S. Helgeson2 and Stephen J. Lepore
Carnegie Mellon University

We examined how two male gender-related traits, agency (focus on self) and
unmitigated agency (focus on self to the exclusion of others), were related to
physical and emotional functioning in 162 men (92% Caucasian) treated for
prostate cancer. As predicted, unmitigated agency was associated with worse
functioning and more cancer-related difficulties. By contrast, agency was
associated with better functioning and fewer cancer-related difficulties. We
tested whether difficulties expressing emotions explained these relations.
Unmitigated agency was associated with difficulty expressing emotions, and
agency was associated with the ability to express emotions. Structural equation
modeling was used to show that emotional expressiveness mediated the
relations of unmitigated agency and agency to adjustment to prostate cancer.
The implications of these results for interventions to enhance men's adjustment
to prostate cancer are discussed.

Prostate cancer is the most prevalent cancer among men and the second
leading cause of cancer death among men in our country (American Cancer
Society, 1996). It is one of the slower progressing cancers, so many men
live with prostate cancer for some time. Because it is a disease that afflicts
elderly men (80% of prostate cancer occurs in men over 65), many men
have other illnesses that pose a greater life threat and eventually die from
something else besides prostate cancer. The fact that men live with, rather
1We are grateful to the following physicians and their staff who facilitated this research:
Thomas Hakala, Robert Bahnson, Ronald Hrebinko, Hugh Flood, Carlos Vivas, Mark
Jordan, Melvin Deutsch, Richard Finegold, Stephen Campanella, Leonard Stept, Daniel Gup,
John Franz, Arnold Sholder, and Walter O'Donnell. We also thank Pamela Snyder, Justin
Ktemmer, Rocco Mercurio, and Pamela Blair for their competent research assistance. This
research was partly supported by NIMH grants 1-54217,1-49432, and CA-61303.
2To whom correspondence should be addressed at Department of Psychology, Carnegie
Mellon University, Pittsburgh, PA 15213-3890.

251
0360-0025/97/0800-0251$12.50/0 c 1997 Plenum Publishing Corporation
252 Helgeson and Lepore

than die from, prostate cancer does not alleviate the emotional, social, sex-
ual, and physical impairments associated with prostate cancer or obviate
the need to study these outcomes. Yet, psychosocial issues in prostate can-
cer have been virtually ignored (Green, 1987; Sharp, Blum, & Aviv, 1993).
Prostate cancer can be construed as a victimization experience that
threatens one's view of the self (Janoff-Bulman, 1992). There is one par-
ticular self that may be especially threatened by prostate cancer—the mas-
culine self. Becoming ill, in and of itself, is a threat to the traditional male
role as illness implies weakness and a lack of control over one's body.
Strength and control are central features of traditional masculinity (Bran-
non & Juni, 1984; Spence, Helmreich, & Stapp, 1974) and physical inade-
quacy is a situation that men appraise as stressful (Eisler & Skidmore,
1987). Chronic illness, in particular, is threatening to the male gender role
because the physical limitations are continuous and often progressive. The
future of a chronic illness is often uncertain. Adherence to the male gender
role may especially impede adjustment to an illness such as prostate cancer,
because the treatment for prostate cancer affects sexuality and control over
bodily functions. In our culture, sexual functioning is viewed as one way
to validate masculinity. The two most common side effects of treatment
for prostate cancer are impotence and incontinence. The majority of pa-
tients report at least some problems with urine leakage and/or erection
ability years after surgery and radiation (Bagshaw, Cox, & Ray, 1988;
Fowler, Barry, Lu-Yoa, Wasson, Roman, & Wennberg, 1995; Kornblith,
Herr, Ofman, Scher, & Holland, 1994).
Researchers have taken a number of approaches to studying the tra-
ditional male gender role. In the 1970s, both clinical and research psycholo-
gists began to suggest that characteristics of the male gender role could
have hazardous health consequences (e.g., Goldberg, 1976; Nichols, 1975).
At the time, the primary approach to studying the male gender role was
the trait approach. Researchers identified personality traits that were more
common among men than women and constructed trait masculinity scales,
the most common of which are the Bern Sex Role Inventory (BSRI; Bern,
1974) and the Personal Attributes Questionnaire (PAQ; Spence et al, 1974).
Research was not successful, however, in showing harmful consequences of
possessing these traits. In fact, higher scores on the masculinity scales of
the PAQ and BSRI have been consistently related to reduced psychological
distress (Bassoff & Glass, 1982; Whitley, 1984) and higher self-esteem
(Whitley, 1983) in healthy populations.
More recently, researchers recognized that the traits on these mascu-
linity scales reflect only one set of male gender-related traits, specifically
an instrumental or agentic orientation (Spence, 1984). The agency scale of
the PAQ is said to reflect a focus on the self and includes items such as
Men's Adjustment to Prostate Cancer 253

independent, active, competitive, and self-confident. Researchers recog-

nized that agency is a positive male gender-related trait that needs to be
distinguished from its negative counterpart, unmitigated agency (Helgeson,
1994; Spence, Helmreich, & Holahan, 1979). Unmitigated agency reflects
a focus on the self to the exclusion of others. An unmitigated agency scale
was developed and added to the PAQ (Spence et al., 1979); scale items
include arrogant, dictatorial, greedy, cynical, and hostile.
Research that has distinguished unmitigated agency from agency has
shown that unmitigated agency is associated with adverse health outcomes,
whereas agency is not. For example, in a cross-sectional study of men and
women with heart disease, higher scores on unmitigated agency were re-
lated to more severe heart attacks, whereas higher scores on agency were
related to less severe heart attacks, when traditional coronary risk factors
(including sex) were statistically controlled (Helgeson, 1990). In that same
study, unmitigated agency was related to increased Type A behavior and
longer delays before seeking help for symptoms, whereas agency was not.
Instead, agency was associated with increased social support. In a longitu-
dinal study of adjustment to heart disease, agency predicted better adjust-
ment among patients, whereas unmitigated agency predicted worse
adjustment among spouses over a one-year period (Helgeson, 1993). Fi-
nally, in a study of male psychiatric inpatients, agency was associated with
less pathology and unmitigated agency was related to more pathology
(Evans & Dinning, 1982). Thus, unmitigated agency and agency appear to
have divergent relations to well-being.
The two constructs typically reveal small positive correlations or are
not correlated (Helgeson, 1993; Spence et al., 1979). It is important to note
that in the studies described above, the two constructs were never inversely
related. A negative correlation might suggest a social desirability response
bias with respondents endorsing positive traits and rejecting negative traits.
Thus, we argue that agency reflects a positive focus on the self. It is
not inconsistent with good relationships and should not prevent men from
turning to others in times of stress. Unmitigated agency, by contrast, reflects
a focus on the self to the exclusion of others, which may preclude men
from turning to others for support. Thus, unmitigated agency may be det-
rimental to men's adjustment to prostate cancer for two reasons. First, phy-
sicians and family can be sources of informational and emotional support,
which may not be available to the unmitigated agency individual. Second,
the treatment for prostate cancer is associated with side-effects that will
have an impact on social relationships (Lepore & Helgeson, in press). To
the extent that the unmitigated agency individual ignores this impact, dif-
ficulties in emotional and social functioning may persist. Thus, one pathway
254 Helgeson and Lepore

by which unmitigated agency may be associated with poor health outcomes

is the failure to utilize social support or the inability to access it.
A second and more recently adopted approach to the study of the tra-
ditional male gender role is to examine normative beliefs about or attitudes
towards men (i.e., prescriptions for men, such as "the husband should make
more money than his wife" or "men should not cry") or domains of conflict
for men (e.g., being subordinate to women, expressing emotions). This is a
fundamentally different approach to the study of the traditional male gender
role than the trait approach. The normative approach or masculine ideology
perspective (Fleck, Sonestein, & Ku, 1993; Thompson & Pleck, 1995) assesses
beliefs that men should have certain characteristics rather than an assessment
of whether men do have those characteristics. As such, a masculine ideology
approach will be bound by the culture's conception of masculinity. As evi-
dence of the distinction between the two approaches, Thompson, Pleck and
Ferrerra (1992) reviewed the literature on male gender-related constructs and
found that the male gender-related trait, agency, is not related to traditional
attitudes towards men (Downs & Engleson, 1982) or areas of male gender-
role conflict (Eisler & Skidmore, 1987).
However, it is possible that men who possess the gender-related trait,
unmitigated agency, may hold prescriptive beliefs about how men should
be and be more likely to suffer from gender-role conflict. We hypothesize
that men who score high on unmitigated agency may be more likely to
suffer from a domain of gender-role conflict that could affect their adjust-
ment to prostate cancer: difficulties expressing emotions. The self-absorp-
tion of the unmitigated agency individual may prevent that person from
expressing himself to others, and the ability to express emotions may be
necessary for positive adjustment to prostate cancer.
Restricted emotionality is a norm of the traditional male gender role
and difficulties with emotional expressiveness is a domain of gender-role
conflict that men face. Although researchers do not entirely agree on all
of the prescriptions for the traditional male gender role, inventories that
assess norms to which men should adhere universally contain a scale that
reflects concealing or restricting one's emotions (Brannon & Juni, 1984;
Downs & Engleson, 1982; Doyle & Moore, 1978; Levant etal.,1992; Snell,
1986). The masculine ideology scale also contains an item that men should
not disclose their problems to others (Pleck et al, 1993). Similarly, male
gender-role conflict scales recognize expressing emotions is a stressful situ-
ation men face (Eisler & Skidmore, 1987; O'Neil, Helms, Gable, David, &
Wrightsman, 1986). The ability to express emotions is particularly important
when faced with a traumatic event, such as prostate cancer. Inhibiting one's
emotions has been associated with poor mental health in a wide range of
trauma survivors (Harber & Pennebaker, 1992; Pennebaker, 1989). Emo-
Men's Adjustment to Prostate Cancer 255

tional inhibition has been associated with a reduced willingness to self-dis-

close among men (Snell, 1986). Others can not provide help to men who
are not able to articulate their needs. In a study of testicular cancer sur-
vivors, the concealing emotions subscale from the Brannon Masculinity
Scale (Brannon & Juni, 1984) was related to greater sexual impairments
(Rieker, Edbril, & Garnick, 1985).
The primary purpose of the present study is to determine whether men
who score high on unmitigated agency have greater difficulty adjusting to
prostate cancer. The secondary purpose of the study is to examine a mecha-
nism for this association that involves a domain of gender-role conflict for
men. We expect that one reason unmitigated agency men will poorly adjust
to prostate cancer is an inability to express emotion, especially to other
men. This domain of conflict may be particularly problematic for men with
prostate cancer because there are not many open forums for men to discuss
prostate cancer. Prostate cancer is an illness that affects parts of the body
that people do not publicly discuss. The illness is only recently beginning
to receive attention from the media. Thus, we expect that difficulty ex-
pressing emotions will mediate the association of unmitigated agency to
poor adjustment. We expect that the association of unmitigated agency to
adjustment will be strongest for domains that tap emotional and social func-
tioning because these are the areas that will suffer most from the inability
to express emotions. By contrast, we predict that agency will be positively
or not at all related to adjustment.

Hypotheses

1. Unmitigated agency will be associated with poor adjustment,

whereas agency will not.
2. Unmitigated agency will be associated with difficulties expressing
emotions, whereas agency will not.
3. Difficulties expressing emotions will mediate the association of un-
mitigated agency to poor adjustment, particularly adjustment out-
comes that involve relationships.

METHOD

Participants

We mailed questionnaires to 258 men who had been diagnosed with

prostate cancer. Letters and questionnaires were sent to patients from the
256 Helgeson and Lepore

offices of four groups of urologists. Physicians asked patients to complete

the questionnaire and return it to us in the enclosed self-addressed,
stamped envelope. All of the information patients provided was completely
confidential. There were no names attached to the questionnaires. After
two weeks, patients were sent a reminder notice. Neither the physician nor
the present investigators recruited patients by phone or in person.
Seventy percent (n = 181) of the patients returned the questionnaires.
This response rate is excellent, given that no personal contact was made
with participants by either the physician or the investigators, the question-
naire was lengthy, and the incentive was small (we donated $1 to the
American Cancer Society for each returned survey). Three questionnaires
with extensive missing data had to be discarded, thereby reducing the sam-
ple to 178.
Eight men who had been diagnosed with prostate cancer more than
2.5 years ago, and eight men who did not specify their year of diagnosis
were removed from the analyses to create a more homogenous sample of
patients with respect to recency of diagnosis. The final sample of men was
162, ranging from less than 1 month to 30 months since diagnosis, with a
mean and median of 13 months. The data provided in the rest of this paper
are based on these 162 men.
The majority of patients were treated with surgery, radical prostatec-
tomy (83%). A smaller portion of the patients were treated with radiation
(29%), and few were treated with hormonal therapy (7%) or chemotherapy
(4%). Twelve percent of the patients reported having had another kind of
cancer, but nearly half of those were skin cancer (8 of the 19 cases). We
did not have access to patients' medical records to obtain information about
the stage of disease. The high rate of surgery, however, suggests that the
majority had localized disease.
Ages ranged from 48 to 84, with a mean and median of 66. The ma-
jority of patients were married (85%) and Caucasian (92%). About half of
the sample (54%) was employed prior to the illness and one-third were
currently working (36%). Highest education level included less than high
school (10%), graduated high school (29%), some college (19%), graduated
college (17%), and post-graduate training (26%). Patients classified them-
selves as Catholic (46%), Protestant (39%), Jewish (8%), atheist (2%), and
other (6%).

Questionnaire

Agency and Unmitigated Agency. We used the agency and unmitigated

agency scales from the Extended Version of the Personal Attributes Ques-
Men's Adjustment to Prostate Cancer 257

tionnaire (Spence et al, 1979). These scales were originally referred to as

positive masculinity and negative masculinity. The two scales are modestly
positively correlated but associated with different outcomes (Spence et al.,
1979). Respondents rated themselves on 5-point, bipolar adjective scales.
For example, an agency item ranges from 1 "not at all self-confident" to
5 "very self-confident"; an unmitigated agency item ranges from 1 'not at
all cynical" to 5 "very cynical." Each scale consists of eight items. In this
study, the internal consistency for agency was .69 and for unmitigated
agency was .75. The two scales were uncorrelated.
Emotional Expressiveness. We developed a scale to measure difficulties
expressing emotions, especially to other men, using existing instruments on
male gender role conflict. The scale was composed of seven items (e.g.,
"It is difficult for me to comfort a male friend," "It is difficult to admit
that I am afraid"). Two items were taken from Eisler and Skidmore's (1987)
emotional inexpressiveness subscale of the Male Gender Role Stress scale.
The other five items were selected from O'Neil et al.'s (1986) restricted
emotionality and restricted affectionate behavior subscales of the Gender
Role Conflict Scale for men. All items tapped difficulty with expressing
emotions and were rated on 5-point scales, ranging from 1 'definitely true'
to 5 'definitely false.' Higher numbers indicate more comfort expressing
emotions. The internal consistency of the scale was .84.
Cancer-Related Problems. We used abbreviated subscales from the Can-
cer Rehabilitation Evaluation System (Schag & Heinrich, 1989) to measure
cancer-specific difficulties. This instrument measures problems stemming
from cancer in several different domains: physical, psychosocial, marital,
sexual, and interaction with medical staff. Time constraints did not permit
us to include the full scales. We reduced the size of the instrument in order
to optimize the chance of men completing the questionnaire. Thus, we se-
lected the items that we thought would be most relevant to men with pros-
tate cancer. The number of items used from the original scale are shown
in parentheses: communication with doctor (alpha = .87, 3 of 6 items),
communication with friends (alpha = .85, 6 of 7 items), sexual interest
(alpha = .63, 2 of 4 items), sexual dysfunction (alpha = .57, 3 of 4 items),
cancer-related worries (alpha = .84, all 4 items), spouse intimacy (alpha
= .82, 4 items), and spouse communication difficulties (alpha = .78, 3
items). We added two items to tap financial difficulties (alpha = .91) and
four items to tap work difficulties (alpha = .82). Higher numbers on all
scales indicate greater problems. Because we did not include the full ver-
sion of these scales, the reader should interpret findings for individual
scales with caution. Results that are consistent across the scales might be
more informative.
258 Helgeson and Lepore

Global Functioning. We used a global quality of life instrument, the

Health Status Questionnaire (HSQ; Stewart, Hays, & Ware, 1988), from
the Medical Outcomes Study, to assess multiple domains of adjustment.
The short form of the HSQ is a 36-item instrument that contains eight
multi-item scales: (1) general health perceptions (alpha = .62), (2) physical
functioning (alpha = .91), (3) role limitations due to physical health prob-
lems (alpha = .90), (4) bodily pain (alpha = .81), (5) general mental health
(alpha = .83), (6) vitality (alpha = .84), (7) role limitations due to personal
or emotional problems (alpha = .84), (8) social functioning (alpha = .71).
The time frame for the majority of the items is the past four weeks. An
additional item assesses perceived change in health status but was not in-
cluded in the present study. Scale reliabilities ranged from .81 to .88 in a
sample of 9385 community-based adults (Stewart et al, 1989; Stewart et
al., 1988); our reliabilities are in parentheses above. The HSQ has been
used successfully to evaluate functional status in over 20,000 depressed,
chronically ill, and healthy patients (Wells et al, 1989). The current version
has been tested in over 60 populations and provides a brief but compre-
hensive measure of functional status. Scale scores are converted to f-scores,
ranging from 0 to 100, with higher numbers indicating better functioning.
Factor analytic studies suggested that the eight subscales define two
distinct components of health: mental health and physical health (Ware,
Kosinski, & Keller, 1994). The first four scales outlined above reflect physi-
cal health and the last four scales reflect mental health. Thus, in addition
to the specific subscales, we examine the mental health composite score
(MCS) and the physical health composite score (PCS). We calculated these
indices according to the authors' procedures (Ware et al, 1994), which uses
factor coefficients derived from a variety of populations, including people
with cancer.

RESULTS

Statistical Controls
Prior to testing our specific hypotheses, we needed to determine if
there were any demographic or health-related background variables that
might account for relations between independent and dependent variables.
The two primary treatments, surgery and radiation, were related to the ma-
jority of the outcome variables (i.e., cancer-related problems and global
functioning). In each case, patients who had surgery were having fewer dif-
ficulties and patients who had radiation were having more difficulties. We
decided to control for both treatments in all our analyses. Among the
Men's Adjustment to Prostate Cancer 259

demographic variables, race was associated with outcome variables, such

that Caucasian respondents fared better than minority respondents. Edu-
cation was related to better adjustment on some of the outcome variables
and also was positively associated with agency. Thus, race and education
also were statistically controlled in all analyses. None of the other back-
ground variables (age, religion, time since diagnosis, marital status, chemo-
therapy, hormonal therapy) were related to independent or dependent
variables.

Relations to Adjustment

The partial correlations between unmitigated agency and global func-

tioning outcomes are shown in Table I and the partial correlations between
unmitigated agency and cancer-related problems are shown in Table II. Sur-
gery, radiation, race, and education were statistically controlled in all of
these analyses. We note, however, that the significant relations shown in
Tables I and II remained significant without the use of statistical controls.
Unmitigated agency was related to worse functioning for six of the eight
HSQ domains. The relations of unmitigated agency to functioning seemed
to be stronger for the emotional and social domains than the physical do-
mains. Thus, it was not surprising that unmitigated agency was associated
with the MCS (mental component) but not the PCS (physical component).
Unmitigated agency also was positively associated with five of the nine do-
mains of cancer-related difficulties (see Table II). The strongest associations
involved communication and relationship domains.
By contrast, agency was related to better functioning for four of the
eight HSQ domains, as shown in Table I. Again, the associations were more
consistent across the emotional than the physical domains of functioning.
And, agency was significantly associated with the MCS but only marginally
associated with the PCS. Agency also was inversely related to three of the
ten domains of cancer-related difficulties, specifically communication with
doctor, spouse intimacy, and interest in sex (see Table II).

Mediation

We hypothesized that difficulties expressing emotions would mediate

the association of unmitigated agency to poor adjustment, especially when
the adjustment domains involved relationships. To evaluate mediation, we
first had to demonstrate that unmitigated agency was associated with emo-
tional inexpressiveness. Unmitigated agency was associated with greater dif-
ficulties expressing emotions, r = -.31, p < .001. Interestingly, agency was
260 Helgeson and Lepore
Men's Adjustment to Prostate Cancer 261
262 Helgeson and Lepore

associated with fewer difficulties with emotional expressiveness, r = .27, p

< .001. Thus, comfort in expressing emotions could be construed as a me-
diator of the relations of both unmitigated agency and agency to adjust-
ment.
We used EQS Structural Equation Modeling (Bentler, 1995) to test a
model that shows the relations of unmitigated agency and agency to both
global functioning and cancer-related problems are mediated through emo-
tional expressiveness. The model is shown in Fig. 1. Because unmitigated
agency and agency were only related to the MCS and not the PCS, we
used the four MCS subscales to construct a latent general mental health
variable. Unmitigated agency and/or agency were related to five of the nine
cancer difficulties; thus, only these difficulties were included in the model
and were used to create a second latent variable. We did not include the
four control variables in the model as their inclusion would have made the
model unnecessarily complicated, and they did not alter the relations of
unmitigated agency and agency to cancer-related problems or mental
health.
Model fit is typically evaluated by a non-significant chi-square and fit
indices that exceed .90 (Bentler, 1990). We evaluate the non-normed fit

Fig. 1. Model of the relations of unmitigated agency and agency to adjustment to prostate
cancer. Starred numbers are standardized parameter coefficients, such that "p < .01 and
***p < .001. Unstarred numbers are standardized factor loadings.
Men's Adjustment to Prostate Cancer 263

index (NNF1) and the comparative fit index (CFI). The chi-square statistic
is limited in its informational value because it is sensitive to sample size;
in small samples it may not be accurate and in lagre samples minor de-
viations of the data from the model will lead to rejection of the model
(Hu & Bentler, 1995). Instead, it has been suggested that one evaluate the
chi-square/degrees of freedom ratio, such that a 2:1 ratio or less indicates
good fit.
The test of whether the model provided an adequate fit to the data
revealed a significant chi-square (x2 = 105.80, df = 53, p < .001), a ratio
that equaled the 2:1 ideal (2.00), but fit indices below .90 (NNFI = .88;
CFI = .89). The Lagrange Multiplier Test suggested that correlating the
two factors would improve the model's fit. Because the factors are depend-
ent variables in this model, only their error terms (indicated by D's) could
be correlated. The inclusion of this additional path provided an adequate
fit for the data. Although the chi-square was still significant (X2 = 92.51,
df = 52, p < .001), the X2/df ratio was less than 2 (ratio = 1.78) and the
fit indices exceeded .90 (NNFI = .91; CFI = .93). Because the data were
not normally distributed as indicated by a standardized Mardia's kurtosis
coefficient of 14.14, we also applied the Satorra-Bentler scaling correction
for chi-square (S-BX2; Hu, Bentler, & Kano, 1992). This correction revealed
a lower chi-square (S-BX2 = 74.91, df = 52, p = .02) a better ratio (1.44),
and a higher fit index (ROBUST CFI = .94). All of the parameters in the
model shown in Fig. 1 were significant. Unmitigated agency was associated
with difficulties expressing emotions and agency was associated with com-
fort expressing emotions. The ability to express emotions was positively re-
lated to general mental health and negatively associated with cancer-related
problems. In addition, mental health was inversely related to cancer-related
problems. Note that direct paths of agency and unmitigated agency to the
latent adjustment variables were not included in the model, suggesting that
the mediational model provided an adequate fit for the data.

DISCUSSION

The purpose of this study was to determine whether men who possess
personality traits that our culture associates with a masculine or agentic
orientation have more difficulties adjusting to an illness that poses a threat
to their masculinity. It is not the case that men who possess a positive agen-
tic orientation have greater difficulties adjusting to prostate cancer. In fact,
the agency scale (which has been referred to in the literature as masculin-
ity) was associated with better adjustment to prostate cancer. It is the case,
264 Helgeson and Lepore

however, that men who possess features of unmitigated agency have greater
difficulties adjusting to prostate cancer.
Men who score high on agency may be said to have a healthy sense of
self (recall that agency is positively correlated with self-esteem), and may
be more likely to successfully adapt to an illness that threatens their view
of themselves. Agentic men do not feel uncomfortable revealing emotions
to others and their ability to express emotions has positive health implica-
tions. Men who score high on unmitigated agency, however, may be insecure
in their sense of masculinity and threatened by disclosing weaknesses and
vulnerabilities to others. The unmitigated agency man's complete self-reli-
ance may be maladaptive when faced with a situation that he cannot solve
by himself. These results underscore the importance of distinguishing a posi-
tive agentic orientation from a negative one (i.e., unmitigated agency).
We also examined a primary mechanism that might explain why un-
mitigated agency would be associated with more difficulties adjusting to an
illness such as prostate cancer: inability or discomfort expressing emotions.
Men with a high level of unmitigated agency had greater difficulties ex-
pressing emotions and these emotional constraints accounted for the rela-
tion of unmitigated agency to poor adjustment. Note that the emotional
expressiveness measure did not tap specific difficulties with discussing can-
cer or cancer-related problems; it was a generic measure of feeling uncom-
fortable expressing feelings and emotions, especially to other men. The
adjustment difficulties most strongly associated with unmitigated agency
and expressiveness seemed to involve emotional functioning and social
functioning. This is not surprising and, in fact, was predicted. Taken col-
lectively, the man at risk for adverse psychological health outcomes follow-
ing prostate cancer is one who scores high on unmitigated agency and has
difficulty expressing feelings, both of which lead to problems talking about
the illness with network members (physicians, family, friends), problems
with intimacy in the marital relationship, and poor overall mental health.
Future researchers should aim to help men overcome their difficulties
expressing emotions. Men who have difficulties expressing emotions may
not reveal that they are bothered by physical problems to physicians, es-
pecially those problems related to sexual dysfunction and incontinence.
These men's problems are then more likely to remain untreated. Men who
have difficulties expressing emotions also may fail to disclose problems to
friends or family members. This wall of silence may then adversely affect
those relationships, as well as preclude network members from being
sources of support
Interventions could be designed to equip men with skills in commu-
nication with physicians, partners, and friends about their illness and its
effects on their lives. Interventions also could provide a warm and accepting
Men's Adjustment to Prostate Cancer 265

atmosphere for men to practice expressing emotions and cancer-related

concerns, perhaps to other men. The majority of past intervention research
has focused on women (see Helgeson & Cohen, 19%, for a review) and
support groups in the community are more often frequented by women
than men (Taylor, Falke, Mazal, & Hillsberg, 1988). Researchers should
consider testing the effectiveness of supportive services that take the form
of education (i.e., informational support) and group discussions (i.e., emo-
tional support) to men. Given that men have more difficulties expressing
feelings and emotions than women and are more likely than women to
suffer from a lack of social support (Barbee et al, 1993), the effects of
such services on men's psychological and physical functioning could be
large.

REFERENCES

American Cancer Society. (1996). Cancer facts and figures. Atlanta, GA: American Cancer
Society.
Barbee, A. P., Cunningham, M. R, Winstead, B. A., Derlega, V J., Gulley, M. R., Yankeelov,
P. A., & Druen, P. B. (1993). Effects of gender role expectations on the social support
process. Journal of Social Issues, 49, 175-190.
Bagshaw, M. A, Cox, R. S., & Ray, G. R. (1988). Status of radiation treatment of prostate
cancer at Stanford University. In: NCI Monographs, No. 7. Washington, DC: Government
Printing Office (NIH Pub. No. 88 3005): 127-131.
Bassoff, E., & Glass, G. (1982). The relationship between sex roles and mental health: A
meta-analysis of 26 studies. The Counseling Psychologist, 10, 105-112.
Bem, S. (1974). The measurement of psychological androgyny. Journal of Consulting and Clini-
cal Psychology, 42, 165-174.
Bentler, P. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107,
238-246.
Bentler, P. (1995). EQS Structural equations program manual Encino, CA: Multivariate Soft-
ware, Inc.
Brannon, R., & Juni, S. (1984). A scale for measuring attitudes toward masculinity. Psycho-
logical Documents, 14, 6 (Ms. 2012).
Downs, A. C, & Engleson, S. A. (1982). The attitudes toward men scale (AMS): An analysis
of the role and status of men and masculinity. JSAS Catalog of Selected Documents in
Psychology, 12, 45 (Ms. 2505).
Doyle, J. A., & Moore, R. J. (1978). Attitudes toward the male role scale: An objective in-
strument to measure attitudes towards the male's sex role in contemporary society. JSAS
Catalog of Selected Documents in Psychology, 8, 35-36 (Ms. 1678).
Eisler, R. M., & Skidmore, J. R. (1987). Masculine gender role stress: Scale development and
component factors in the appraisal of stressful situations. Behavior Modification, 11, 123-
136.
Evans, R. G., & Dinning, W. D. (1982). MMPI correlates of the Bern Sex Role Inventory
and Extended Personal Attributes Questionnaire in a male psychiatric sample. Journal of
Clinical Psychology, 38, 811-815.
Fowler, F. J., Barry, M. J., Lu-Yoa, G.,Wasson,J., Roman, A., & Wennberg, J. (1995). The
effect of radical prostatectomy for prostate cancer on patient quality of life: Results from
a medicare survey. Urology, 44, 600-607.
266 Helgeson and Lepore

Goldberg, H. (1976). The hazards of being male. New York: Sanford J. Greenburger Associates,
Inc.
Green, R. L. (1987). Psychosocial consequences of prostate cancer—my father's illness and
review of the literature. Psychiatric Medicine, 5, 315-327.
Harber, K. D., & Pennebaker, J. (1992). Overcoming traumatic memories. In S. A. Christian-
son (Ed.), The handbook of emotion and memory. Hillsdale, NJ: Erlbaum.
Helgeson, V S. (1990). The role of masculinity in a prognostic predictor of heart attack se-
verity. Sex Roles, 22, 755-774.
Helgeson, V S. (1993). Implications of agency and communion for patient and spouse adjust-
ment to a first coronary event. Journal of Personality and Social Psychology, 64, 807-816.
Helgeson, V S. (1994). Relation of agency and communion to well-being: Evidence and po-
tential explanations. Psychological Bulletin, 116, 412-428.
Helgeson, V S., & Cohen, S. (1996). Social support and adjustment to cancer: Reconciling
descriptive, correlational, and intervention research. Health Psychology, 15, 135-148.
Hu, L., & Bentler, P. M. (1995). Evaluating model fit. In R. H. Hoyle (Ed.), Structural equation
modeling: Concepts, issues, and applications. Thousand Oaks, CA: Sage Publications.
Hu, L., Bentler, P. M., & Kano, Y. (1992). Can test statistics in covariance structure analysis
be trusted? Psychological Bulletin, 112, 351-362.
Janoff-Bulman, R. (1992). Shattered assumptions: Towards a new psychology of trauma. The
Free Press: New York.
Kornblith, A. B., Herr, H. W, Ofman, U. S., Scher, H. L, & Holland, J. C. (1994). Quality
of life of patients with prostate cancer and their spouses. Cancer, 73, 2791-2802.
Lepore, S. J., & Helgeson, V S. (in press). Social constraints, intrusive thoughts, and mental
health after prostate cancer. Journal of Social and Clinical Psychology.
Levant, R. F., Hirsch, L., Celentano, E., Cozza, T, Hill, S., MacEachern, M., Marty, N., &
Schnedeker, J. (1992). The male role: An investigation of contemporary norms. Journal
of Mental Health Counseling, 14, 325-337.
Nichols, J. (1975). Men's liberation: A new definition of masculinity. Middlesex, England: Pen-
guin Books.
O'Neil, J. M., Helms, B. J., Gable, R. K., David, L., & Wrightsman, L. S. (1986). Gender-role
conflict scale: College men's fear of femininity. Sex Roles, 14, 335-350.
Pennebaker, J. W. (1989). Confession, inhibition, and disease. In L. Berkowitz (Ed.), Advances
in experimental social psychology (Vol. 22). Orlando, FL: Academic Press.
Pleck, J. H., Sonenstein, F. L., & Ku, L. C. (1993). Masculinity ideology and its correlates.
In S. Oskamp & M. Costanzo (Eds.), Gender issues in contemporary society. Newbury
Park, CA: Sage Publications.
Rieker, P. P., Edbril, S. D., & Garnick, M. B. (1985). Curative testis cancer therapy: Psychoso-
cial sequelae. Journal of Clinical Oncology, 3, 1117-1126.
Schag, C. C, & Heinrich, R. L. (1989). Cancer Rehabilitation Evaluations System (CARES)
Manual (Ed. 1). Los Angeles: CARES Consultants.
Sharp, J. W., Blum, D., & Aviv, L. (1993). Elderly men with cancer: Social work interventions
in prostate cancer. Social Work in Health Care, 19, 91-107.
Snell, W. E. (1986). The masculine role inventory: Components and correlates. Sex Roles, 15,
443-455.
Spence, J. T. (1984). Masculinity, femininity, and gender-related traits: A conceptual analysis
and critique of current research. In B. A. Maher & W. B. Maher (Eds.), Progress in
experimental personality research (Vol. 13). San Diego, CA: Academic Press.
Spence, J. T., Helmreich, R., & Holahan, C. K. (1979). Negative and positive components of
psychological masculinity and femininity and their relationship to self-reports of neurotic
and acting out behaviors. Journal of Personality and Social Psychology, 37, 1673-1682.
Spence, J. T., Helmreich, R., & Stapp, J. (1974). The Personal Attributes Questionnaire: A
measure of sex-role stereotypes and masculinity-femininity. JSAS Catalog of Selected
Documents in Psychology, 4, 43.
Stewart, A. L., Greenfield S., Hays, R. D., Wells, K., Rogers, W. H., Berry, S. D., McGlynn,
E. A., & Ware, J. E. (1989). Functional status and well-being of patients with chronic
Men's Adjustment to Prostate Cancer 267

conditions. The MOS Short-Form General Health Survey. Journal of the American Medical
Association, 262, 907-913.
Stewart, A. L., Hays, R. D., & Ware, J. E. (1988). The MOS Short-Form General Health
Survey. Medical Care, 26, 724-735.
Taylor, S. E., Falke, R. L., Mazal, R. M, & Hilsberg, B. L. (1988). Sources of satisfaction
and dissatisfaction among members of cancer support groups. In B. H. Gottlieb (Ed.),
Marshaling social support: Formats, processes and effects. Newbury Park, CA: Sage.
Thompson, E. H., & Pleck, J. H. (1995). Masculine ideology: A review of research instru-
mentation on men and masculinities. In R. F. Levant & W. S. Pollack (Eds.), A new
psychology of men. New York: Basic Books.
Thompson, E. H., Jr., Pleck, J. H., & Ferrera, D. L. (1992). Men and masculinities: Scales
for masculinity ideology and masculinity-related constructs. Sex Roles, 27, 573-607.
Ware, J. E., Kosinski, M., & Keller, S. D. (1994). SF-36 Physical and Mental Health Summary
Scales: A user's manual. Boston, MA: The Health Institute.
Wells, K. B., Stewart, A., Hays, R. D., Burnam, M. A., Rogers, W., Daniels, M., Berry, S.,
Greenfield, S., & Ware, J. (1989). Detection of depressive disorder for patients receiving
prepaid or fee-for-service care: Results from the Medical Outcomes Study. Journal of the
American Medical Association, 262, 914-919.
Whitley, B. E. (1983). Sex role orientation and self-esteem: A critical meta-analytic review.
Journal of Personality and Social Psychology, 44, 765-778.
Whitley, B. E. (1984). Sex role orientation and psychological well-being: Two meta-analyses.
Sex Roles, 12, 207-225.