Ij

This material is protected by U.S. copyright law. Unauthorized reproduction is prohibited. To purchase quantity reprints,
please e-mail reprints@ons.org or to request permission to reproduce multiple copies, please e-mail pubpermissions@ons.org.
ONLINE EXCLUSIVE
Risk Factors for Ovarian Cancer:
Lesbian and Heterosexual Women
Suzanne L. Dibble, RN, DNSc, Stephanie A. Roberts, MD,
Patricia A. Robertson, MD, and Steven M. Paul, PhD
Key Points . . .
Purpose/Objectives: To compare the distribution of risk
factors for developing ovarian cancer in lesbian and het-
erosexual women.
Approximately 23,400 women were diagnosed with ovarian
Design: Secondary analysis of a retrospective medical
cancer in 2001, and an estimated 1,451 of them were lesbian.
record review.
Setting: Urban health clinic with special outreach to les-
Lesbians are a diverse group of women from all ethnic, reli-
bians.
gious, cultural, economic, and age groups.
Sample: Typical participant (N = 1,019) was 42.9 years
Women of all sexual orientations undoubtedly are being
old and white (70%). Most were without health insurance,
treated in oncology practices.
and 99% were poor (< $15,780 annual income). The major-
ity (58%, n = 586) described themselves as heterosexual;
Research suggests that lesbians may have a risk profile that
42% (n = 433) said they were lesbian.
would indicate a higher rate of ovarian cancer than hetero-
Methods: Data were collected from medical records
sexual women.
and analyzed using analysis of covariance and logistic re-
gression techniques.
Main Research Variables: Ovarian cancer risk factors
(parity, exogenous hormone use, smoking, body mass in-
If this is true, one can estimate that approximately 1,451 les-
dex [BMI], and tubal ligation/hysterectomy).
bians may be diagnosed with ovarian cancer in the United
Findings: Lesbians had a higher BMI; heterosexual
States in 2001. These figures may be conservative if the rate
women had higher rates of current smoking and a higher
of ovarian cancer is eventually found to be higher among les-
incidence of the protective factors of pregnancy, children,
bians.
miscarriages, abortions, and use of birth control pills.
The term lesbian describes "not only sexual orientation, but
Conclusions: The results of this study indicate that lesbi-
also an identity based on psychological responses, cultural
ans may have an increased risk for developing ovarian
cancer. A study designed specifically to explore the risk
values, societal expectations, and a woman's own choices in
factors of lesbian and heterosexual women for developing
identity formation" (White & Levinson, 1995, p. 463). Lesbi-
ovarian cancer must be undertaken to confirm these find-
ans are a diverse group of women from every ethnic, religious,
ings.
economic, cultural, and age group. Because homosexuality is
Implications for Nursing Practice: Differences in risk lev-
stigmatized and because lesbians often defy stereotypes, les-
els may exist for lesbians; therefore, healthcare providers
bians may remain a hidden population in their interactions
must become comfortable asking questions about sexual
orientation and behavior.
Suzanne L. Dibble, RN, DNSc, is a professor at the Institute for Health
and Aging in the School of Nursing at the University of California, San
n the year 2001, an estimated 23,400 women in the United
I
Francisco (UCSF); Stephanie A. Roberts, MD, is the medical direc-
States were diagnosed with ovarian cancer and 13,900
tor of Medical Weight Management in Walnut Creek, CA; Patricia A.
women died from the disease (Greenlee, Hill-Harmon,
Robertson, MD, is a professor of clinical obstetrics and gynecology
Murray, & Thun, 2001). The five-year survival for women
and the director of medical student education in the School of Medi-
diagnosed with ovarian cancer is only 50%. Some of the
cine at USCF; and Steven M. Paul, PhD, is a senior statistician in the
women included in these statistics are lesbians. The actual
School of Nursing at UCSF. (Submitted April 2001. Accepted for
publication August 10, 2001.) The authors gratefully acknowledge
number of lesbians is unknown; thus, any attempt to report the
primary financial support from the Breast Cancer Research Program,
distribution of sexual orientation in women is subject to some
University of California, Office of the President and additional sup-
bias and distortion (Solarz, 1999). In the National Health and
port from the Center for Lesbian Health Research at UCSF.
Social Life Survey, 6.2% of women reported same-sex behav-
ior or desire (Laumann, Gagnon, Michael, & Michaels, 1994).
Digital Object Identifier: 10.1188/02.ONF.E1E7
DIBBLE VOL 29, NO 1, 2002
E1
30.3 and another at 8084 years of 41.6; however, increasing
with healthcare providers and researchers (Eliason, Donelan,
& Randal, 1992). The assumption of heterosexuality is so
age resulted in more ovarian cancer for these women. Ameri-
can Indian/Alaskan native (AI/AN) women also had two peaks
prevalent (Denenberg, 1995; Rankow, 1995) that healthcare
in the incidence of ovarian cancer, one at 5054 years of age at
providers and researchers may perpetuate the invisibility of
a rate of 30.4 and another at 8084 years of 52.6 (NCI, 2001).
the lesbian experience. In 1999, the Institute of Medicine re-
Why the AI/AN group had a younger average diagnosis age is
ported on its work examining lesbian health issues and con-
unknown. The distribution of lesbians among the various age
cluded that the first priority for research was ". . . to better
groups is also unknown.
understand the physical and mental health status of lesbians
and to determine whether there are health problems for which
Ethnicity/Heritage Group
lesbians are at higher risk as well as conditions for which pro-
The risk of developing ovarian cancer varies by ethnic
tective factors operate to reduce their health risk" (Solarz,
group (ACS, 1999). For instance, using the SEER Incidence
1999, p. 156.)
Age-Adjusted Rates per 100,000 women for the 11 registries
Some believe that the risk factors for developing ovarian
from 19941998, white women had the highest rate of ovarian
cancer may be different between lesbian and heterosexual
cancer diagnoses at 15.2 (confidence interval [CI] = 14.9
women, resulting in higher rates of ovarian cancer among les-
15.5), Latinas had the next highest rate at 11.1 (CI = 10.5
bians. Even if the actual incidence of ovarian cancer is similar
11.7), Asian/Pacific Islanders had a rate of 10.7 (CI = 10.0
between the two groups, whether a difference exists in the pro-
11.3), black women had a rate of 10.3 (CI = 9.610.9), and
file of risk factors is not known. Studies of ovarian cancer in-
AI/AN women had a rate of 7.2 (CI = 5.49.5) (NCI, 2001).
cidence or the risk factors associated with developing ovarian
The percentage of self-identified lesbians in various ethnic
cancer among lesbians are noticeably absent in the literature.
groups is unknown.
Therefore, the purpose of this pilot study was to determine the
differences in the distribution of ovarian cancer risk factors in
Family History
lesbian and heterosexual women. Some of the modifiable ova-
Women with a family history of ovarian cancer in a first-
rian cancer risk factors that were compared were (a) parity, (b)
exogenous hormone usage, (c) smoking, (d) body mass index
degree relative (e.g., mother, sister) have a higher risk for de-
(BMI), and (e) tubal ligation/hysterectomy.
veloping ovarian cancer (Whittemore, 1994). Genetically de-
termined ovarian cancers probably comprise only 10% of the
Background and Significance
total number of ovarian cancers (Berchuck, Carney, & Fut-
real, 1999). Although some evidence suggests that sexual ori-
Data are needed to answer questions about risk status and
entation may be genetically determined (Bailey, Pillard,
various characteristics, including sexual orientation. How-
Neale, & Agyci, 1993), any potential or actual genetic link-
ever, in the cancer statistics of the National Cancer Institute's
ages between ovarian cancer and sexual orientation have not
Surveillance, Epidemiology, and End Results (SEER) pro-
been reported.
gram, no data are collected about sexual orientation. As a
Parity
result, the number of lesbians actually diagnosed with ovarian
cancer is unknown. Until researchers complete and publish
The risk of ovarian cancer is significantly higher among
studies comparing differences in the distribution of risk fac-
women who have not been pregnant and decreases with in-
tors by sexual orientation, the prevalence of established risk
creasing numbers of pregnancies (Hankinson et al., 1995). A
factors for the development of ovarian cancer among lesbians
common assumption is that lesbians do not have children;
can only be estimated by what is known about the differences
however, surveys indicate that 6%46% of lesbians do have
between lesbian and heterosexual women.
children (O'Hanlan, 1995). This is in contrast to parity esti-
The following is a description of the current state of knowl-
mates for women in general of 85% (Bachu, 1995).
edge about the differences between lesbian and heterosexual
Breastfeeding
women as they relate to established risk factors for ovarian
cancer. Although only some risk factors are modifiable,
The longer total time that women breastfeed, the more pro-
knowledge of every factor is necessary to develop appropri-
tected they are from developing ovarian cancer (Riman,
ate screening and interventions.
Persson, & Nilsson, 1998). Because lesbians theoretically
have decreased opportunities to breastfeed as a result of de-
Age
creased parity, an assumption is made that lesbians are at
The most important risk factor for the development of ova-
higher risk for the development of ovarian cancer. Again, no
rian cancer is age, because the rate of ovarian cancer increases
studies are available to support these inferences about differ-
with age (American Cancer Society [ACS], 1999). For in-
ential risk by sexual orientation relative to breastfeeding.
stance, using the National Cancer Institute's (NCI's) SEER
Exogenous Hormones
incidence rates per 100,000 women for the 11 registries from
19941998, the rate for white women increased with age to a
Oral contraceptives have been associated with a decreased
high of 60.7 for women between 7579 years. Black women
risk for developing ovarian cancer, and protection seems to
also followed the same pattern, with ovarian cancer rates in-
increase with the duration of use (Riman et al., 1998). Presum-
creasing with age to a high of 50.9 for women between 7579
ably, lesbians use oral contraceptives infrequently (Harrison
years. For Latina women, the rate of ovarian cancer also in-
& Silenzio, 1996); however, Johnson, Smith, and Guenther
creased with age to a high of 46.3 for women over 85 years of
(1987) reported that in a sample of 1,500 lesbians, 61% indi-
age. Asian/Pacific Islander women had two peaks in the inci-
cated past use of oral contraceptives. Another issue within the
dence of ovarian cancer, one at 6064 years of age at a rate of
lesbian community is the administration of fertility medica-
ONF VOL 29, NO 1, 2002
E2
tions to enhance the chances of pregnancy when donor sperm
Methods
are used. The use of fertility drugs and their association with
Design
ovarian cancer is yet to be resolved (Riman et al.).
Whether women are at increased risk for ovarian cancer
This descriptive study is a secondary analysis of a retro-
secondary to exposure to hormone replacement therapy
spective medical record review conducted to explore the dif-
(HRT) is not clear. The prevalence of HRT usage among les-
ferences in risk factors for developing breast cancer between
bians is unknown.
lesbian and heterosexual women (Roberts et al., 1998).
Tubal Ligation and Hysterectomy
Setting
Undergoing surgical tubal occlusion or hysterectomy may
The setting for this study was the Lyon-Martin Women's
reduce one's ovarian cancer risk (Riman et al., 1998). Presum-
Health Services (LMWHS) in San Francisco, CA. This clinic
ably lesbians do not undergo tubal ligation as often as hetero-
was established in 1979 to ensure that lesbians had access to
sexual women, but no studies have been conducted to confirm
health care. Currently, LMWHS provides health care to
this. Some preliminary evidence indicates that lesbians may
underserved women of all sexual orientations, but serves pri-
have a high rate of hysterectomies (Harrison & Silenzio,
marily young, low-income women residing in urban San
1996), but this study has not been replicated or expanded to
Francisco. It remains the only health clinic in San Francisco
include a comparison group of heterosexual women. No direct
with significant outreach to the lesbian community.
comparison studies have been reported to suggest clinically
Sample
significant differences by sexual orientation in the rate of tu-
bal ligation or hysterectomies.
To be included in the medical record review, women had to
be 35 years or older and have been seen at LMWHS in 1995,
Other Risk Factors
1996, or 1997. Charts that did not contain an intake form and
provider notes (e.g., patients seen only for tuberculosis test-
Other factors with conflicting evidence for increased risk of
ing) were excluded. Also excluded were the charts of trans-
ovarian cancer include smoking (Engeland, Andersen,
gender patients, both male-to-female and female-to-male. The
Haldorsen, & Tretli, 1996; Purdie et al., 1995), high BMI
intake form asked women to identify themselves as lesbian,
(Purdie et al.), use of talc in the perineal region (Purdie et al.;
heterosexual, bisexual, or celibate (usually regarded as sexual
Wong, Hempling, Piver, Natarajan, & Mettlin, 1996), high di-
behavior, but included in the LMWHS intake form under
etary galactose intake (Westhoff, 1996), and antidepressant use
identity, because many women had written it in on past intake
(Harlow, Cramer, Baron, Titus-Ernstoff, & Greenberg, 1998).
forms). Only women who checked lesbian or heterosexual
If cigarette smoking is proven to be a risk factor for some
women, then older lesbians should be at a higher risk because
were included in these analyses, resulting in a sample size of
1,019.
they have been found to be three times more likely to smoke
than heterosexual women (Bradford, Ryan, & Rothblum,
Instruments
1994). Studies have reported conflicting evidence between les-
bian and heterosexual women in body size. In Maine, research-
The chart audit form was created, pilot-tested, and revised
ers compared BMI (a relationship between height and weight)
prior to being used in this study. A multidisciplinary panel of
between 71 lesbian and 77 heterosexual women and were not
experts established content validity. A copy of the chart audit
able to demonstrate a significant difference (Patton et al., 1998);
form can be found in Roberts et al. (1998). Data were ab-
whereas in another study, lesbians did have a significantly
stracted from each chart using the audit form. Nearly all
higher BMI than heterosexual women (Roberts, Dibble,
LMWHS charts contained an intake form for self-report data.
Scanlon, Paul, & Davids, 1998). If higher BMI is found to be
Data were collected from both the intake form and the provid-
associated with higher rates of ovarian cancer, lesbians may be
ers' notes. For example, to determine "yes" or "no" on history
at greater risk. Differences in talc use and dietary galactose in-
of smoking, one of the two graduate student research assis-
take are not clear. Antidepressant use also is not clear, although
tants reviewed charts, considered a "yes" answer to any of the
Cochran and Mays (2000) found no significant differences in
four intake form questions regarding cigarette use as "yes,"
depression between women with same-gender sex partner(s)
and also reviewed the provider's notes looking for references
and heterosexual women during the previous year.
to smoking cessation or evidence that the patient was seeking
help or had been referred to a smoking cessation program, all
Summary
of which would be considered as "yes." An interrater reliabil-
An analysis of the limited data available suggests that lesbians
ity of greater than 95% was achieved between each research
may have different risk factor distributions for developing ova-
assistant and one of the authors. One of the authors performed
rian cancer than do heterosexual women. The question remains
additional monthly quality assurance checks.
whether differences exist between lesbian and heterosexual
Procedure
women that would result in a higher risk for either lesbians or
heterosexual women in developing ovarian cancer. Research is
After approval by the Human Subjects Committee, research
needed to explore the differences in risk factor distributions as-
assistants reviewed the LMWHS charts. One of the authors
sociated with the development of ovarian cancer between les-
trained them in data collection procedures. An alphabetical list
bian and heterosexual women. If lesbian or heterosexual
of patients (35 years or older seen at the clinic from 19951997)
women are at higher risk or if they have a different risk profile,
was generated using the LMWHS computer system. The re-
then targeted intervention programs to alert healthcare provid-
search assistants then selected patient charts from the storage
ers and the various communities about their differential risk
shelves starting at the beginning of the alphabetical list. If a
status need to be designed, implemented, and evaluated.
chart was not on the shelf, the name was skipped, and the re-
DIBBLE VOL 29, NO 1, 2002
E3
search assistants proceeded to the next name on the list. After
Most of the women were without health insurance, and 99%
the research assistants completed the entire list, the process was
had incomes at less than 200% of Federal Poverty Guidelines
(< $15,780 per year) (U.S. Federal Register, 1997). Of this
repeated with the same list to retrieve charts that were missing
sample, 58% (n = 586) identified themselves as heterosexual
the first time through. To protect patient confidentiality, each
audit form was given a study number, not the medical record
and 42% (n = 433) described themselves as lesbian. Table 1
contains a comparison of demographic information by sexual
number. Patient lists were locked in a file cabinet when not in
orientation. The sample sizes varied because of missing data.
use and were only accessible to project staff.
Significant differences were found in age, with the lesbian
Data Analysis
group approximately two years younger than the heterosexual
Data were analyzed using the CRUNCHTM Program Ver-
group. Significant differences were found in ethnicity, with
sion 4 (CRUNCH Software Corporation, Oakland, CA). De-
fewer black women and more white women in the lesbian
group. Significantly more lesbians were employed, and more
scriptive statistics were calculated to describe the demo-
heterosexuals were disabled. Most of the heterosexual women
graphic characteristics of the sample. Comparisons in the
(98%) had sex only with men, whereas 88% of the lesbians
demographics by sexual orientation were generated using t-
reported having sex with both men and women during their
tests or chi-square analyses as appropriate for the level of data.
lifetimes. The extent of missing data on lifetime sexual behav-
Because the age, ethnicity, employment status, and disability
ior for lesbians was extensive--32% for the lesbians com-
status of the sample were different by sexual orientation, the
pared with 3% for the heterosexuals.
c o m p a r i s o n s of risk factors were completed with age,
ethnicity, employment status, and disability status as covar-
Risk Factors
iates. Both analysis of covariance and logistic regression were
used to compare the risk factors; the choice depended on
After controlling for age, ethnicity, employment, and dis-
whether the outcome was continuous or dichotomous data.
ability status, some significant differences were found in risk
Significance was preset at p < 0.05.
factors for ovarian cancer between the two groups. The het-
erosexual women reported significantly more pregnancies
Results
(83% versus 37%), children (1.48 versus 0.87), miscarriages
(0.44 versus 0.25), abortions (1.42 versus 0.95), and birth
Sample
control pill use (60% versus 39%) than did the lesbians. The
The typical participant (N = 1,019) was 42.9 years old (SD
lesbians had higher BMIs (26.66 kg/m2 versus 25.52 kg/m2
+ 6.85, range 3575), white (70%), and employed (50%).
for heterosexual women). Heterosexual women were signifi-
Table 1. Comparison of Sample Demographics by Sexual Orientation
Lesbians
Heterosexuals
Variable
(N = 433)
Statistic
p
(N = 586)
Age (years)
t = 4.74
0.001
--
41.8
43.8
X
SD
6.3
7.1
Variable
n
%
n
%
χ  2 = 26.64
Ethnicity
0.001
Asian American
5
1
18
3
Black
36
8
105
18
Latina
32
7
51
9
White
323
75
366
63
Other
20
5
26
4
Missing data
17
4
20
3
χ  2 = 33.65
Employment status
0.001
Employed
230
53
220
38
Other
144
33
308
53
Missing data
59
14
58
10
χ2 = 8.53
Disability status
0.004
Disabled
58
13
125
21
Other
316
73
403
69
Missing data
59
14
58
10
χ2 = 825.32
Sexual behavior
0.001
Only women
35
8
0
0
Only men
0
0
560
96
Both
259
60
9
2
Celibate
1
<1
0
0
138
32
Missing data
17
3
Note. Percents do not equal 100 because of rounding.
ONF VOL 29, NO 1, 2002
E4
Table 2. Comparison of Risk Factors by Sexual Orientation After Controlling for Age, Ethnicity, Employment, and
Disability Status
Heterosexuals
Lesbians
95%
N = 586
N = 433
Confidence
n
%
Variable
Odds Ratio
Interval
n
%
Exogenous hormones
Ever used birth control pills*
Yes
354
60
3.03
2.274.17
168
39
224
38
No
255
59
8
1
Missing
10
2
Ever used hormone replace-
ment therapy
71
12
Yes
1.26
0.582.73
31
7
58
10
No
32
7
457
78
Missing
370
86
Smoking
Current smoking status**
224
38
Yes
1.45
1.032.04
124
29
238
41
No
213
49
124
21
Missing
96
22
Ever smoked
324
55
Yes
1.07
0.76 1.54
222
51
138
24
No
115
27
124
21
Missing
96
22
Family history
Family history of breast cancer
96
16
Yes
1.00
0.69 1.46
80
19
483
82
No
348
80
7
1
Missing
5
1
Hysterectomy
Undergone menopause earlier
than 45 years old
26
4
Yes
1.60
0.426.16
12
3
71
12
No
43
10
489
83
Missing
378
87
Endogenous hormones
Ever pregnant*
488
83
Yes
7.14
5.2610.00
161
37
98
17
No
271
63
0
0
Missing
1
<1
Heterosexuals
Lesbians
n = 586
n = 433
Adjusted
Adjusted
--
--
SD
n
SD
n
F
p
X
Variable
X
Endogenous hormones
(continued)
12.81
1.65
452
12.71
1.53
327
0.64
0.422
Age at menarche
1.48
1.39
392
0.87
1.08
122
18.88
0.0001
Number of children
0.44
0.92
381
0.25
0.68
119
4.18
0.042
Number of miscarriages
1.42
1.60
381
0.95
1.02
120
8.84
0.003
Number of abortions
47.2
5.61
70
47.0
4.93
48
0.03
0.872
Age at menopause
Body mass
64.99
2.58
416
65.03
2.65
304
0.04
0.844
Height (inches)
152.97
36.91
422
159.96
41.08
302
5.52
0.019
Weight (pounds)
25.52
5.97
409
26.66
6.15
327
5.66
0.018
Body mass index
* p < 0.05; ** p < 0.01
Note. Percents do not equal 100 because of rounding.
DIBBLE VOL 29, NO 1, 2002
E5
were found in HRT use between the two groups either, although
cantly more likely to currently smoke (38% versus 29%), al-
data about HRT use in both groups was incomplete. This was
though past smoking did not differ between the two groups.
probably, in part, because direct questions about the use of HRT
No significant differences were found in family history of
are not included in the LMWHS intake form. Also, given the
breast cancer (see Table 2).
relatively young age of the sample, the providers may not have
No statistically significant differences were found in age at
asked about these issues. Missing from the analysis is data
menarche (lesbian = 12.71; heterosexual = 12.81), age at
about the use of fertility drugs because this information was not
menopause (lesbian = 47.0; heterosexual = 47.2), or HRT use
collected in the original breast cancer risk study.
between the two groups (lesbian = 7%; heterosexual = 12%);
however, the extent of missing data for both menopause and
Limitations
HRT was extensive. Information about women who had gone
This study has significant limitations, primarily because of
through early menopause as a result of hysterectomy was not
the nature of medical chart review and because the charts were
readily available in the medical records. As a rough proxy
originally reviewed for breast cancer risk factors. Medical
estimate of potential hysterectomies, the authors explored the
record information may be limited by what the healthcare
differences in the number of lesbian and heterosexual women
providers believed should be recorded for future reference.
who had gone through early menopause (< 45 years of age).
Therefore, medical record information may be sparse for the
No significant differences were found.
categories of interest. For example, in this study, ascertaining
whether someone was a current smoker was possible, but the
Discussion
duration and amount of cigarettes smoked was not consis-
tently available in the record.
This is the first reported study comparing ovarian cancer
In addition, this sample consisted of relatively young women
risk factors between lesbian and heterosexual women. The
from one economic group residing in a single urban area. Al-
finding that more of the heterosexuals smoked than did the
most all of the women in this study lived at less than 200% of
lesbians was not expected. Data from the Women's Health
the poverty level (< $15,780 annual income, U.S. Federal Reg-
Initiative indicated significant differences in cigarette smok-
ister, 1997) in the San Francisco Bay area. Also, the average age
ing status by sexual orientation, with lesbians smoking more
of the sample was only 42 years and the sample was primarily
(Solarz, 1999). Bradford et al. (1994) previously reported
white. More women of color were represented in this sample
daily smoking rates for lesbians over age 35 to range from
(30%) than in other studies (Becker & Robison, 1996; Bradford
30%38%, with the highest percentage in the over 55 age
et al., 1994; Skinner & Otis, 1996). Nevertheless, these sample
group. Skinner and Otis (1996) reported smoking rates during
characteristics limit the generalizability of our findings to other
the prior month for lesbians over age 35 to be 38%, as com-
lesbian and heterosexual women.
pared with the national average of 22% in 1997 of women
over age 35 who smoked (American Lung Association, 1999).
Implications
This study's findings may reflect geographic differences in
for Research and Practice
smoking rates or reflect a true change within the lesbian com-
munities from the earlier studies.
The results of this study suggest that a basis exists for fu-
Previous contributions to the lesbian health literature have
ture research about the differences in the risk factor profiles
measured self-report of weight problems (Bradford et al., 1994)
between lesbian and heterosexual women. These studies
or weight (Herzog, Newman, Yeh, & Warshaw, 1992), but
should include a sample of lesbian and heterosexual women
none except the Houston Health Care Needs Assessment
from various economic groups, geographic regions, and ages.
(Becker & Robison, 1996) have measured both height and
Surveying older lesbians particularly is important because of
weight. Whether the finding of a significantly higher BMI in
the increased incidence of ovarian cancer with age. Including
lesbians is in the range to have adverse health consequences is
questions about sexual orientation in tumor registry data also
questionable. Neither the average BMI for lesbians or hetero-
would aid researchers in determining the numbers of lesbians
sexuals in this study was more than 27.3 kg/m2, a figure used
affected by ovarian cancer each year.
as a cut-point for overweight in the Second National Health and
Healthcare providers must understand that lesbians are part
Nutrition Examination Survey study (Kuczmarski, 1992). How-
of every racial, economic, religious, cultural, and age group.
ever, newer guidelines for a healthy BMI set the cut-off for
They are a part of every practice and have increased or, at
overweight at 25 kg/m2, which would mean that, on average,
least, different risk factors for developing ovarian cancer.
both groups were overweight (National Heart, Blood, and Lung
Therefore, sexual orientation and behavior is an essential part
Institute, 2000). Another factor to consider is that differences in
of the screening process.
age exist in the correlation of BMI with body fat, with BMI
Many healthcare professionals likely use assumptive lan-
more highly correlated with estimates of body fat in younger
guage that prevents the exchange of open and trusting com-
women and muscle mass in older adults (Micozzi & Harris,
munication. An example of this includes asking a woman if
1990). Because lesbians are reported to exercise more than their
she has a husband. Another way that this question could be
heterosexual counterparts (Becker & Robison), the differences
asked in the context of care is "Is your partner picking you
in BMI may reflect increased muscle mass in lesbians.
up?" or "Should we include your partner or friend in this dis-
As expected from previous reports, the lesbians had signifi-
cussion?". Questions such as these will signal to lesbian cli-
cantly fewer pregnancies, miscarriages, and abortions, and
ents that sharing information about their lives is safe, enabling
lower use of birth control pills. These variables place lesbians
nurses to provide better care.
at higher risk for developing ovarian cancer. However, no sig-
nificant differences were found in age at menarche or age at
Author Contact: Suzanne L. Dibble, RN, DNSc, can be reached at
menopause between the two groups. No significant differences
sdibble@itsa.ucsf.edu, with copy to editor at rose_mary@earthlink.net
ONF VOL 29, NO 1, 2002
E6
References
National Heart, Blood, and Lung Institute. (2000). Clinical guidelines on the
American Cancer Society, California Division, & Public Health Institute,
identification, evaluation, and treatment of overweight and obesity in
California Cancer Registry. (1999). California: Cancer facts and figures,
adults (NIH Publication #98-4083). Washington, DC: Author.
1999. Oakland, CA: American Cancer Society.
O'Hanlan, K.A. (1995). Lesbian health and homophobia: Perspectives for
American Lung Association, Epidemiology and Statistics Unit. (1999).
the treating obstetrician/gynecologist. Current Problems in Obstetrics,
Trends in cigarette smoking. Atlanta: Author.
Gynecology, and Fertility, 18(4), 96133.
Bachu, A. (1995). Fertility of American women. U.S. Bureau of Census,
Patton, C.L., Millard, P.S., Kessenich, C.R., Storn, D., Kinnicutt, E., &
current population reports (Report No. 20-482). Washington DC: U.S.
Rosen, C.J. (1998). Screening calcaneal ultrasound and risk factors for os-
Government Printing Office.
teoporosis among lesbians and heterosexual women. Journal of Women's
Bailey, J.M., Pillard, R.C., Neale, M.C., & Agyci, Y. (1993). Heritable fac-
Health, 7, 909915.
tors influence sexual orientation in women. Archives of General Psychia-
Purdie, D., Green, A., Bain, C., Siskind, V., Ward, B., Hacker, N., Quinn,
try, 50, 217233.
M., Wright, G., Russell, P., & Susil, B. (1995). Reproductive and other
Becker, B.L., & Robison, J. (1996). Synopsis: Houston area health care
factors and risk of epithelial ovarian cancer: An Australian case-control
needs assessment for lesbian, gay, bisexual, and transgender women.
study. International Journal of Cancer, 62, 678684.
Presented at the National Lesbian/Gay Health Association Conference,
Rankow, E.J. (1995). Breast and cervical cancer among lesbians. Women's
July 1996, Seattle, WA.
Health International, 5, 123129.
Berchuck, A., Carney, M.E., & Futreal, P.A. (1999). Genetic susceptibility
Riman, T., Persson, I., & Nilsson, S. (1998). Hormonal aspects of epithelial
testing and prophylactic oophorectomy. European Journal of Obstetrics,
ovarian cancer: Review of epidemiological evidence. Clinical Endocri-
Gynecology, and Reproductive Biology, 82, 159164.
nology, 49, 695707.
Bradford, J., Ryan, C., & Rothblum, E.D. (1994). National lesbian health
Roberts, S.A., Dibble, S.L., Scanlon, J., Paul, S., & Davids, H.R. (1998).
care survey: Implication for mental health care. Journal of Consulting
Differences in risk factors for breast cancer: Lesbians and heterosexual
Clinical Psychology, 62, 228242.
women. Journal of the Gay and Lesbian Medical Association, 2, 93101.
Cochran, S.D., & Mays, V.M. (2000). Relation between psychiatric syn-
Skinner, W.F., & Otis, M.D. (1996). Drug and alcohol use among lesbian
dromes and behaviorally defined sexual orientation in a sample of the
and gay people in a southern U.S. sample: Epidemiological, comparative,
U.S. population. American Journal of Epidemiology, 151, 516523.
and methodological findings from the Trilogy Project. Journal of Homo-
Denenberg, R. (1995). Report on lesbian health. Women's Health Interna-
sexuality, 30(3), 5991.
tional, 5(2), 8191.
Solarz, A.L. (Ed.). (1999). Lesbian health: Current assessment and direc-
Eliason, M., Donelan, C., & Randal, C. (1992). Lesbian stereotypes. Health
tions for the future. Washington DC: National Academy Press.
Care for Women International, 13, 131144.
U.S. Federal Register. (1997). Annual update of the Department of Health
Engeland, A., Andersen, A., Haldorsen, T., & Tretli, S. (1996). Smoking
and Human Services poverty guidelines, 62, 1085610859.
habits and risk of cancers other than lung cancer; 28 years' follow-up of
Westhoff, C. (1996). Ovarian cancer. Annual Review of Public Health, 17,
26,000 Norwegian men and women. Cancer Causes and Control, 7, 497
8596.
506.
White, J., & Levinson, W. (1995). Lesbian health care: What a primary care
Greenlee, R.T., Hill-Harmon, M.B., Murray T., & Thun, M. (2001). Cancer
physician needs to know. Western Journal of Medicine, 162, 463466.
statistics, 2001. CA: A Cancer Journal for Clinicians, 51, 1536.
Whittemore, A.S. (1994). Characteristics relating to ovarian cancer risk:
Hankinson, S.E, Colditz, G.A, Hunter, D.J., Willett, W.C., Stampfer, M.J.,
Implications for prevention and detection. Gynecologic Oncology,
Rosner, B., Hannekens, C.H., & Speizer, F.E. (1995). A prospective study
55(Suppl. 3, Pt 2), S15S19.
of reproductive factors and risk of epithelial ovarian cancer. Cancer, 76,
Wong, C., Hempling, R.E., Piver, M.S., Natarajan, N., & Mettlin, C.J.
284290.
(1996). Perineal talc exposure and subsequent epithelial ovarian cancer:
Harlow, B.L., Cramer, D.W., Baron, J.A., Titus-Ernstoff, L., & Greenberg,
A case-control study. Obstetrics and Gynecology, 93, 372376.
E.R. (1998). Psychotropic medication use and risk of epithelial ovarian
cancer. Cancer Epidemiology, Biomarkers, and Prevention, 7, 697702.
Harrison, A.E., & Silenzio, V.M. (1996). Comprehensive care of lesbian and
gay patients and families. Primary Care, 23, 3146.
Herzog, D.B., Newman, K.L., Yeh, C.J., & Warshaw, M. (1992). Body im-
age satisfaction in homosexual and heterosexual women. International
For more information . . .
Journal of Eating Disorders, 11, 391396.
Johnson, S.R., Smith, E.M., & Guenther, S.M. (1987). Comparison of gy-
Lesbians and Breast and Ovarian Cancer Issues
necologic healthcare problems between lesbians and bisexual women: A
www.annieappleseedproject.org/lesandbreasa.html
survey of 2,345 women. Journal of Reproductive Medicine, 32, 805811.
Kuczmarski, R.J. (1992). Prevalence of overweight and weight gain in the
National Ovarian Cancer Coalition
United States. American Journal of Clinical Nutrition, 55(Suppl. 2),
www.ovarian.org
495S502S.
Laumann, E.O., Gagnon, J.H., Michael, R.T., & Michaels, S. (1994). The
The Lesbian Cancer List
social organization of sexuality: Sexual practices in the United States.
How to subscribe: e-mail majordomo@Queernet.org with
Chicago: University of Chicago Press.
"subscribe lcl" as the body of the message.
Micozzi, M.S., & Harris, T.M. (1990). Age variations in the relation of body
mass indices to estimates of body fat and muscle mass. American Jour-
These Web sites are provided for information only. The hosts are re-
nal of Physical Anthropology, 81, 375379.
sponsible for their own content and availability. Links can be found
National Cancer Institute. (2001). Surveillance, epidemiology, and end re-
using ONS Online at www.ons.org.
sults: CanQues. Bethesda, MD: Author. Retrieved November 30, 2001
from the World Wide Web: http:\\canques.seer.cancer.gov
DIBBLE VOL 29, NO 1, 2002
E7