Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Cognitive functioning following ovarian removal before or after natural menopause
(USC Thesis Other)
Cognitive functioning following ovarian removal before or after natural menopause
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Running head: COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL
Cognitive Functioning Following Ovarian Removal
Before or After Natural Menopause
By
Keiko Kurita
University of Southern California
Degree Conferral Date: August 2014
This research was supported by the National Institutes of Health Grant Numbers R01 AG08724,
R01AG-17160, and F31AG040937 from the National Institute on Aging, U01-AT001653 from the
National Center for Complementary and Alternative Medicine, the Office of Dietary Supplements,
and the Office of Research on Women’s Health, and R01AG024154, P01 AG026572, P30 CA-
71789, and 5-T32-AG00037. The content is solely the responsibility of the authors and does not
necessarily represent the official views of the National Institute on Aging or the National Institutes
of Health.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 2
Table of Contents
Abstract 4
Introduction 6
Estrogen and Cognitive Function 9
Oophorectomy and Cognitive Decline 10
Limitations of Existing Studies 11
The current study 13
Possible Mediating, Moderating, or Confounding Factors 14
Affective symptoms 15
Health conditions including malignancy, cardiovascular history, and adiposity 16
Rationale and Hypotheses 18
Study I
Method 22
Participants and Procedures 22
Measures 23
Statistical Analysis 30
Results 34
Bilateral oophorectomy and cognitive impairment 34
Bilateral oophorectomy and cognitive screening score 36
Depressive symptoms as a mediator 37
Age of last menstrual period 38
Hormone therapy as a protective factor 38
Bilateral oophorectomy using co-twin control analyses 39
Differential survival 40
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 3
Study II
Method 42
Participants 42
Procedures 43
Measures 44
Statistical Analysis 48
Results 51
Post-hoc power analysis 51
Bilateral oophorectomy and cognitive performance 52
Depressive symptoms 52
Cognitive performance over time 53
Hormone therapy 54
Discussion 55
Limitations and strengths 61
Clinical implications 62
Future research 63
References 65
Study I Results Tables 80
Study II Results Tables 95
Study I and II Figures 122
Appendix 127
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 4
Abstract
Estrogen has been associated with cognitive function, and cognitive difficulties become
more prevalent in older age when estrogen is reduced. However, the long-term cognitive function of
older women who had reductions in estrogen through having their ovaries removed is not well
understood. This dissertation examined the extent to which bilateral oophorectomy and its timing
relative to natural menopause are associated with cognitive functioning through two complementary
studies. Depressive symptoms as a mediator and hormone therapy as a protective factor were also
examined. In Study I, twins over 65-years of age in the Swedish Twin Registry with linkages to the
national Inpatient Discharge Registry were used in classic, case-control (154 women with bilateral
oophorectomy and 7,600 women with ovaries intact) and discordant co-twin (n = 63 pairs) analyses
to examine associations between oophorectomy and both cognitive impairment and scores on a
cognitive screening test. In Study II, oophorectomy status and performance on global cognitive
function and five cognitive domains among healthy postmenopausal women in 3 clinical trials
conducted in Los Angeles County were analyzed in linear regression models both cross-sectionally (n
= 1,149) and on repeat cognitive assessment 2.5 years after baseline (n = 463). In both studies,
covariates included age, education, and health factors. In Study I, neither oophorectomy nor its
timing was associated with cognitive impairment. In Study II, oophorectomy after age 45 compared
to no history of oophorectomy was a risk factor for poorer performance in verbal learning (p = .03)
in the cross-sectional design and with a decrease in visual memory (p < .01) performance over time.
Oophorectomy status was associated with a decrease in visual memory (p < .05) over time. Neither
study showed a significant interaction of oophorectomy and hormone treatment. However, in Study
I, hormone treatment was protective against cognitive impairment (p < .02 for length of time on
hormone treatment, p < .01 for history of any hormone treatment). In both studies, depressive
symptoms were not found to be a mediator but were found to be associated with poorer cognitive
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 5
function. The dissertation findings suggest that ovarian removal may be associated with poorer
performance in specific cognitive domains but not with longer-term cognitive impairment.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 6
Cognitive Functioning Following Ovarian Removal Before or After Natural Menopause
Most women who are in their reproductive stage of life experience menses. The menstrual
cycle provides the body with important chemicals such as hormones to keep the body healthy.
Having regular menstrual cycles may be viewed as a sign that women’s bodies are working normally.
For the majority of women, the cessation of menses occurs naturally when they enter menopause,
signaling a stage in life when women can no longer become pregnant. Natural menopause typically
happens at approximately 51 years of age in the US and 50 in Sweden (Margolis, et al., 2005;
Weiderpass, et al., 1999). Women can also experience menopause if they undergo certain surgical
procedures before natural cessation of their menstrual cycles. If both ovaries are removed before
natural menopause, either through bilateral oophorectomy or salpingo-oophorectomy, which
includes removal of the fallopian tube, it leads to surgical menopause. These surgeries are sometimes
performed simultaneously with hysterectomy, a procedure that is the second-most common major
surgery for women of reproductive age in the US (Keshavarz, Hillis, Kieke, & Marchbanks, 2002).
Removal of ovaries may be significant because ovaries are typically the primary source of estrogen in
women of reproductive age.
One indication for oophorectomy is ovarian cancer treatment. It is also performed for risk-
reducing objectives such as the prevention of ovarian cancer and protection against breast cancer
despite the fact that it does not eliminate all of the risk for these cancers (Shoupe, et al., 2007;
American College of Obstetricians and Gynecologists, 2008). Women may also opt for
oophorectomy to minimize the effects of noncancerous tumors and cysts, endometriosis, and pelvic
pain. Approximately one half of women who have hysterectomy also have bilateral oophorectomy
(Keshavarz, et al., 2002). Of the bilateral oophorectomies performed in the U.S. annually, 71% of
them, or 300,000 procedures, were among women with benign primary diagnoses (Keshavarz et al.,
2002; Parker, Jacoby, Shoupe, & Rocca, 2009). Of the women with no known increased risk of
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 7
ovarian cancer who still had bilateral oophorectomy with hysterectomy, more than half were 49
years of age or younger, according to one nationally representative inpatient sample in the U.S.
(Asante, et al., 2010). More than 123,000 women between 15 and 44 years old undergo hysterectomy
annually with concomitant bilateral oophorectomy in the U.S. (Keshavarz, et al., 2002; see Appendix
Table A1). Because women who are under 51 years of age likely were not menopausal, the last
number is an estimate of women who may have experienced surgical menopause annually.
During menopause, ovaries generally reduce their production of estrogen. Some women are
prescribed estrogen or hormone therapy to “replace” the lower levels of hormones after surgical or
natural menopause. Ovarian removal also lowers levels of endogenous sex hormones. Levels of sex
hormones have been found to decrease immediately following bilateral oophorectomy and remain
lower long-term, even in older women who both used and were naïve to hormone therapy
compared to those who had at least one ovary (McTiernan, et al.; 2008). Also, among adult women
who were not on hormone therapy, those who had surgical menopause appear to have lower levels
of sex hormones compared to those who had natural menopause, even after adjusting for age and
time since menopause (Korse , Bonfrer, van Beurden, Verheijen, & Rookus, 2009). This decrease in
estrogen level may be the intended consequence that women seek when considering bilateral
oophorectomy, as lower levels may reduce their risks of certain cancers (Rebbeck, Levin, Eisen,
Snyder, Watson, Cannon-Albright, et al., 1999). However, lower levels of estrogen through bilateral
oophorectomy, especially before natural menopause, may not be without adverse effects, as
hormones have been correlated with better health outcomes such as cognitive functioning.
A few studies have demonstrated declines in cognitive functioning among premenopausal
women after bilateral oophorectomy (Phillips & Sherwin, 1992; Sherwin, 1988; Sherwin & Phillips,
1990). There is also some evidence of greater risk of cognitive impairment among women who had
oophorectomy at a young age (who likely experienced surgical menopause) compared to women of a
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 8
similar age who did not have oophorectomy. However, one qualitative review identified
observational and randomized controlled trials and reported both the specific type of reproductive
surgery (hysterectomy with or without bilateral salpingo-oophorectomy, surgical menopause) and
cognitive results (Vearncombe & Pachana, 2009). The authors concluded that although some smaller
trials (10 out of the 17 studies described had fewer than 100 participants) show detrimental cognitive
effects of surgical menopause, studies conducted among larger populations did not find significant
effects and definite conclusions were difficult to draw.
Studies examining the cognitive effects of surgical menopause have appropriately included
hormone therapy as a possible protective factor. Henderson and Sherwin (2007), who examined
cognitive domains and other cognitive outcomes in randomized clinical trials of hormone therapy,
stated that cognitive consequences of surgical menopause differ from those of natural menopause
and found that the effect of hormone therapy on cognitive functioning is undetermined, as some
studies suggest a beneficial effect towards reducing cognitive decline and others show no beneficial
effect. The authors recommended that longer-term cognitive consequences of hormone therapy use
near the time of menopause was needed as for most studies, the duration of hormone therapy was
less than 1 year. Other reviews (Hogervorst, Williams, Budge, Riedel, & Jolles, 2000; Hogervorst &
Bandelow, 2010) examined studies of hormone therapy effects on various cognitive domains and
concluded that women with surgical menopause appear to benefit from hormone therapy based on a
small number of studies.
The purpose of this dissertation is to build on the current research by further examining the
relations of ovarian removal, especially before natural menopause, use of hormone therapy, and
cognitive functioning in older women when the risk of cognitive impairment or decrease in function
are greater. Understanding these relations may be of critical importance for women considering
elective bilateral oophorectomy and may influence decisions regarding the use of hormone therapy.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 9
The findings may be relevant to all women entering menopause as surgical menopause serves as a
unique model for studying the effects of a relatively sudden decrease in hormone levels (Farrag,
Khedr, Abdel-Aleem, & Rageh, 2002).
Estrogen and Cognitive Function
In animal models of Alzheimer’s disease, the administration of hormones that mimic the
natural menstrual cycle in the early stages of the disease slows disease progression (Carroll, et al,
2007). A number of biological mechanisms seem to support this relation, such as the presence of
estrogen receptors in certain brain regions and estrogen’s ability to affect neural changes (McEwen,
2002; Sherwin, 2005; Sohrabji, 2005, Sohrabji, 2006).
In humans, the association is less clear. Although some studies have found that estrogen
therapy is protective against the risk of Alzheimer’s disease (Paganini-Hill & Henderson, 1994),
other studies have found that estrogen therapy either has no detectable effect or is a risk factor for
cognitive impairment such as Alzheimer’s disease (Amaducci, et al., 1986; Brenner, et al., 1994;
Maki, 2005; Shumaker, et al., 2003; Shumaker, et al., 2004).
Associations between estrogen levels and cognitive function have been found to be similarly
inconsistent. Although higher estrogen levels predicted no difference or poorer cognitive
performance in attention, executive function, and other domains among postmenopausal women
with mean ages over 70 who were not on estrogen therapy (Barrett-Connor & Goodman-Gruen,
1999; Yaffe, Grady, Pressman, & Cummings, 1998), higher estrogen levels in blood or serum
samples in slightly younger women have been found to be associated with better verbal memory
(Wolf & Kirschbaum, 2002).
There have been inconsistent findings regarding the beneficial, neutral, or harmful effects of
hormone therapy on cognitive performance from large-scale clinical trials and reviews. Hormone
therapy was generally considered to be neutral or beneficial for cognitive function in women,
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 10
regardless of type of menopause (LeBlanc, Janowsky, Chan, Nelson, 2001; Hogervorst, Williams,
Budge, Riedel, & Jolles, 2000). Some population-based studies found that estrogen use was
associated with improvements in cognitive domain performance, such as global cognitive function,
verbal memory, category fluency, and abstract reasoning (Rice, et al., 2000; Hogervorst, Williams,
Budge, Riedel, & Jolles, 2000). These findings are in contrast to the results of examining dementia.
The large-scale Women’s Health Initiative trial (Shumaker, et al., 2003) found that hormone therapy
increased the risk for probable dementia in older adult women. Partially to explain these conflicting
findings, researchers have hypothesized a “critical window” in which estrogen therapy’s
neuroprotection depends on age at the time of administration, type of menopause, and stage of
menopause. The theory proposes that women who had surgical menopause (or experienced early
natural menopause) may benefit from hormone treatment, but that perhaps after midlife, it may no
longer offer neuroprotection and that starting estrogen therapy many years after natural menopause
may be detrimental (Whitmer, Quesenberry, Zhou, & Yaffe, 2011; Rocca, Grossardt, & Shuster,
2011).
Oophorectomy and Cognitive Decline
Studies have examined associations between oophorectomy and cognitive decline, including
clinical outcomes, general cognitive function, and specific cognitive domains. Consistent
associations with oophorectomy and cognitive outcomes have not been found. Although one study
reported not finding an association with Alzheimer’s disease (Amaducci, et al., 1986), other studies
have found an association with dementia (Rocca, et al., 2007; Phung, et al., 2010). With respect to
measures of general cognitive function, 12-year declines on Mini-Mental State Examination scores
were largest among women who had menopause before 40 years old, most of whom who had
undergone surgical menopause (McLay, Maki, & Lyketsos, 2003).
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 11
Studies have also found associations between oophorectomy and a few cognitive domains,
including verbal and logical memory. There is evidence that surgical menopause is associated with
cognitive declines affecting verbal episodic memory (Henderson & Sherwin, 2007). Compared to
women who had natural menopause, women who had surgical menopause performed poorly on
verbal memory tests among women naïve to hormone therapy (Nappi, et al., 2000; Costa, et al.,
1997) and on mental flexibility and episodic memory among women on hormone therapy (File,
Heard, & Rymer, 2002). Declines on associate learning and logical memory performances were
found pre-operatively and post-operatively in women who had surgery before menopause compared
to women who did not (Farrag, et al., 2002). Another study found that associate learning and
paragraph recall performances improved or were no different compared to pre-operatively in
surgical menopausal women who had hormone therapy after surgery, whereas the performances
were no different or worsened in women with surgical menopause who received placebo (Phillips &
Sherwin, 1992). Smaller studies have found similar associations, including one that found that
surgically menopausal women on hormone therapy performed better on verbal memory and
constructional ability than those without therapy (Verghese, et al., 2000). These findings are
consistent with the abovementioned study that found estrogen levels positively associated with
verbal memory in healthy older women (Wolf & Kirschbaum, 2002) and the findings that fail to
support positive associations of estrogen use or levels with attention and executive function (Barrett-
Connor & Goodman-Gruen, 1999; Yaffe, Grady, Pressman, & Cummings, 1998).
Limitations of Existing Studies
Conclusions of existing research range from estrogen therapy and oophorectomy having
beneficial to harmful effects on cognitive performance and outcomes. Reasons for inconsistent
findings may include differences in the study designs (i.e., within or between group comparisons)
and sample characteristics including the ages of participants. A close examination of the existing
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 12
research reveals areas of improvements that could be made in future studies that might help
reconcile the various results. First, explicit identification of whether women had bilateral
oophorectomy, unilateral oophorectomy, and hysterectomy is needed. Cessation of menses after
surgery does not exclusively indicate surgical menopause, as ovarian disruption can also be caused by
hysterectomy if the ovarian branch of the uterine artery is ligated and the blood supply to the ovaries
is reduced. Women with hysterectomy or unilateral oophorectomy have been found to reach
menopause earlier than women with hysterectomy who retain both ovaries (Farquhar, Sadler,
Harvey, & Stewart, 2005). Furthermore, with respect to dementia, a dose effect with the number of
ovaries removed and the risk of dementia has been hypothesized (Phung, et al., 2010), hysterectomy
and hysterectomy with unilateral oophorectomy are not equivalent to bilateral oophorectomy or
surgical menopause, as they result in different levels of estrogen disruption. Despite this potentially
important distinction, studies do not specify these surgeries consistently (Phung, et al., 2010; Kritz-
Silverstein & Barrett-Connor, 2002). It is also important to consider that estrogen may affect
diagnostic outcomes such as dementia differently than cognitive function as a result of estrogen’s
vasoactive properties and other physiological mechanisms (Mortel & Meyer, 1995), so it is important
to consider both performance and disease outcomes.
Another critical limitation in studies of women who had bilateral oophorectomy is that
hormone therapy status, which alters estrogen exposure, is not consistently reported (McLay, Maki,
& Lyketsos, 2003; Nappi, et al., 2000; Farrag, et al., 2002). Some studies examined differences
among groups treated with hormone therapy versus placebo (Sherwin & Phillips, 1990). In other
studies, older women were not included (Szklo, et al., 1996), women were given hormone therapy
well after menopause (Espeland, et al., 2004), or previous hormone use was not reported (Nappi, et
al., 2000; Farrag, et al., 2002; Sherwin & Phillips, 1990).
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 13
In addition, it would be useful to investigate the group of women who had bilateral
oophorectomy before natural menopause separately from those who had the procedure after natural
menopause, but not all studies specify timing of surgery in relation to menstrual status (Kok, et al.,
2006). The timing of the surgery relative to natural menopause may also be important to consider.
One study of women who had oophorectomy found that the magnitude of the association between
oophorectomy and cognitive impairment or dementia increased with younger age at surgery (Rocca,
Grossardt, & Maraganore, 2008), and another found trends that prior hysterectomy with or without
bilateral oophorectomy (presumably pre-or peri-menopausal) was a risk factor for dementia with
onset before age 50 (Phung, et al., 2010). Another study found a significant association between
more time since surgical menopause and poorer performance in verbal memory (Nappi, et al., 2000).
These studies highlight that timing of oophorectomy, before or after menopause, may be a relevant
factor in understanding the relation between oophorectomy and cognitive effects.
In addition, some studies excluded participants who did not meet cognitive functioning
criteria (File, Heard, & Rymer, 2002; Verghese, et al., 2000; Szklo, et al., 1996), resulting in biased
samples where women showing cognitive impairment were not examined, potentially leading to
under-estimation of the negative cognitive effects of oophorectomy. Even with these differences in
individual studies, one meta-analysis (Kurita, Gatz, & Mack, unpublished) found oophorectomy to
be a risk factor for poorer performance in cognitive domains compared to no history of
oophorectomy.
The current study
This dissertation addresses some limitations and gaps in the existing literature. First, the
distinction between hysterectomy, unilateral oophorectomy, and bilateral oophorectomy is made.
Outcomes of both cognitive impairment and continuous measures of cognitive performance are
made. Adjustments are made for past hormone therapy, including for any history and where
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 14
available, its length. The study adjusts for both surgical menopause as distinct from bilateral
oophorectomy and age of surgery. Length of time from surgery to cognitive testing is also
considered as a relevant factor. The dissertation also includes participants regardless of level of
cognitive function.
Finally, other studies have used a mix of study designs such as within-subjects, repeated
measures designs, cross-over designs, and referent groups (Farrag, et al., 2002; Rocca, Grossardt, &
Maraganore, 2008; Polo-Kantola, et al., 1998). In this dissertation, using a population sample of
twins, a closely matched comparison group adjusts for genetic and familial confounds. This design is
ideal because twin controls are matched at least 50% for genetic factors and for some early
environmental factors. In twin studies, practice effects, inter-individual differences in cognitive
performance, and age differences at testing, which could influence results, can be minimized. This is
noteworthy as adjusting for genetic factors was considered important as they may play a role in
moderating an effect of estrogen in cognitive functioning (Yaffe, Haan, Byers, Tangen, & Kuller,
2000). To our knowledge, the effect of oophorectomy using twins as controls has not been
previously published.
Possible Mediating, Moderating, or Confounding Factors
Given the conflicting findings of the relation between oophorectomy and cognitive function
in existing studies, it is possible that there are factors that mediate or moderate the relation, as it may
be likely that the relation is more complex than main effects. Therefore, in this dissertation, the
possible role of hormone therapy as a moderator and the possible role of affective symptoms as a
mediator are examined. One benefit of exploring these factors is that they could provide a deeper
understanding of the relation. In addition, we consider health conditions as possible confounding
factors that may be additionally relevant.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 15
Affective symptoms. In this dissertation, depressive symptoms as a partial mediator are
examined. Research suggests that physiological mechanisms explain an association between low
estrogen levels and risk of depression (Rasgon, Shelton, & Halbreich, 2005; Walf & Frye, 2006). For
example, it is understood that estrogen has effects on the central nervous system such that
administration of estrogen decreases depressive behavior and administration of an estrogen receptor
antagonist increases depressive behavior in animals (Walf & Frye, 2006). In addition, depressive
symptoms may act as a psychological mediator because the decision to remove one’s ovaries may be
difficult, possibly following conditions that may accompany pain or bleeding. If ovaries are removed
before natural menopause, it marks a permanent end to reproductive life, which may lead to feelings
of loss and depressed mood. Depressive symptoms have already been examined in the literature.
Women who had bilateral oophorectomy, even 2 to 6 years after surgery, report more depressive
symptoms than those who had hysterectomy but with ovaries conserved (Nathorst-Böös, Von
Schoultz, & Carlström, 1993). Bilateral oophorectomy has been found to be associated with an
increased risk for affective symptoms including depression many years later compared to women
who did not have bilateral oophorectomy, even among women who never experienced these
symptoms before surgery (Rocca, et al., 2008). There was a continuing increase in the incidence of
depressive symptoms even after many years, with an increased long-term risk of depression among
women who had surgical menopause compared to women who had intact ovaries. Depressive
symptoms have been found to negatively affect cognitive domain performance, in memory and
learning areas, and subsequent cognitive decline (Rosenberg, Mielke, Xue, & Carlson, 2010;
Greendale, et al., 2010; Miller, Conney, Rasgon, Fairbanks, & Small, 2002). Thus, depressive
symptoms that persist for years after surgery may partially explain cognitive changes. Few studies
included depression as a factor in examining the association between cognitive function and
oophorectomy; those that did considered it as an adjustment variable or exclusion criterion (Rocca,
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 16
et al., 2007; File, Heard, & Rymer, 2002; Verghese, et al., 2000; Farquhar, et al., 2005). Studies that
adjusted for depressive symptoms still revealed that bilateral oophorectomy was a risk factor for
poor cognitive function. For example, women who had surgical menopause and used estrogen
therapy for 10 years performed worse on cognitive tests compared to women who did not have
surgical menopause matched for age, IQ, years of education, and occupation and adjusting for
depression (File, Heard, & Rymer, 2002). Because depressive symptoms are associated with both
oophorectomy and cognitive function, it is worth examining as a potential mediator.
Health conditions including malignancy, cardiovascular history, and adiposity. This
study considers medical and other health-related information in examining the relation between
cognitive functioning and oophorectomy. Although some studies have excluded women who had
oophorectomy as a result of malignant disease (Rocca, et al., 2007; Nappi, et al., 2000; Farrag, et al.,
2002; Phillips & Sherwin, 1992), it may be worth examining the extent to which cognitive outcomes
of surgery associated with cancer diagnosis is different from those associated with other indications
(Heflin, et al., 2005), as women who undergo oophorectomy because of malignant disease may be
characteristically different than those who undergo the surgery for other reasons.
Bilateral oophorectomy is a risk factor for cardiovascular disease (Lobo, 2007) and there is
some evidence that higher estrogen levels are protective against cardiovascular disease (Yang &
Reckelhoff, 2011). In addition, cardiovascular disease and history of coronary artery-bypass surgery
are risk factors for cognitive difficulties (Harman, Naftolin, Brinton, & Judelson, 2005; Newman, et
al., 2001; van Dijk, et al., 2000). Despite the potential explanatory factor of cardiovascular history, its
status has not been reported in some studies (Nappi, et al., 2000; Kritz-Silverstein & Barrett-Connor,
2002). Therefore, in this study, cardiovascular history is adjusted for statistically in the sample using
twins and as an exclusion criterion in another sample of healthy women.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 17
Adiposity, measured by body mass index (BMI), has been associated with increased levels or
impaired functioning of estrogen, as local estrogen biosynthesis is a primary source for estrogen in
postmenopausal women (Hughes, Borenstein, Schofield, Wu, & Larson, 2009; Parihar, 2003;
Simpson, et al., 2000). Finally, hormone therapy, as mentioned above, is examined as a possible
moderator, and use of contraceptives, which often contain estrogen and other sex hormones, is also
considered as a possible confound.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 18
Rationale and Hypotheses
This dissertation investigates cognitive functioning of older women who had bilateral
oophorectomy, its timing before versus after natural menopause, and hormone therapy use. The
analyses were performed on two existing data sets that allowed for well-matched comparison
groups, investigation into relevant factors such as timing of surgery and length of therapy, and
adjustments for potential demographic and medical confounds that have not been consistently
considered in current studies.
The sample in Study I included older-adult females in the Swedish Twin Registry, which is
population-based and therefore representative of the general population of Sweden (Lichtenstein, et
al., 2006). It was linked to the nation-wide, Inpatient Discharge Registry to identify oophorectomy
cases, a method that is only possible in a country with universal health care. The twin design was
used to explore the degree to which genetic factors and familial environments mediate associations.
Cognitive outcomes included dichotomous outcome of cognitively impaired versus not impaired
and a continuous score on a cognitive screening test. The sample in Study II uses data from 3
clinical trials that included healthy, postmenopausal women free of signs or symptoms of
cardiovascular disease and diabetes. Cognitive outcomes include domain-specific performance,
allowing for a fine examination of domains that may be differentially affected, as well as global
cognitive performance.
The two studies complemented each other. Having both datasets allows evaluation of
oophorectomy and cognitive functioning in two independent datasets, with different samples,
cognitive measures, and analytic designs. To the extent that results converge and replicate, despite
these design differences, the conclusions would be strengthened.
We expected that older women who had ovarian removal would be at an increased risk for
cognitive impairment compared to women without a similar history, that the risk would be greater
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 19
for women who had the procedure before natural menopause, and that differences in cognitive
function would be found mostly in verbal learning and logical memory domains. We studied one
moderator and one mediator. We predicted that hormone therapy would be more protective when
there was a history, compared to no history, of ovarian removal, and that depressive symptoms
would partially mediate the expected association between oophorectomy and impaired cognition.
Study I
Specific Aim 1. Examine the extent that bilateral oophorectomy, compared to intact ovaries, is
associated with cognitive impairment.
Hypothesis 1.1: Bilateral oophorectomy is a risk factor for cognitive impairment and
associated with more errors (lower score) on cognitive screening.
Specific Aim 2. Examine the extent that any relation between bilateral oophorectomy and
cognitive impairment is influenced by genetic or early environmental factors (familiality).
Hypothesis 2.1: Bilateral oophorectomy is a risk factor for cognitive impairment and
associated with more errors (lower score) on cognitive screening compared to co-twins with no
oophorectomy.
Hypothesis 2.2: Familiality explains a significant portion of any association detected in
Hypothesis 1.1.
Specific Aim 3. Examine the extent that earlier, compared to later, timing of bilateral
oophorectomy predicts cognitive impairment.
Hypothesis 3.1: Bilateral oophorectomy before natural menopause is associated with a
greater risk for cognitive impairment and associated with more errors (lower score) on cognitive
screening than the surgery after natural menopause.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 20
Hypothesis 3.2: Surgical menopause at a younger age is associated with a greater risk for
cognitive impairment and associated with more errors (lower score) on cognitive screening
compared to surgical menopause at a later age.
Specific Aim 4. Examine depressive symptoms as mediating the relation between oophorectomy
and cognitive impairment.
Hypothesis 4.1: Depressive symptoms partially explain the relation between bilateral
oophorectomy and increased risk for cognitive impairment.
Specific Aim 5. Examine hormone therapy as a protective factor in the relation between bilateral
oophorectomy and cognitive impairment.
Hypothesis 5.1: The associations in Hypotheses 1.1, 2.1, 3.1, 3.2, and 4.1 attenuate if
hormone therapy was used around the time of bilateral oophorectomy such that hormone therapy
would be more protective when there was a history, compared to no history, of ovarian removal.
Study II
Specific Aim 6. Examine the extent that bilateral oophorectomy compared to intact ovaries and its
earlier, compared to later, timing are associated with verbal learning, logical memory, and global
cognitive functioning.
Hypothesis 6.1: Bilateral oophorectomy predicts lower cognitive performance compared to
no history of bilateral oophorectomy.
Hypothesis 6.2: Surgical menopause predicts lower cognitive performance compared to
surgery after natural menopause.
Specific Aim 7. Examine depressive symptoms as mediating the relation between bilateral
oophorectomy and verbal learning, logical memory, and global cognitive functioning.
Hypothesis 7.1: Depressive symptoms partially explain the relation between bilateral
oophorectomy and poorer cognitive performance.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 21
Specific Aim 8. Examine the extent that bilateral oophorectomy compared to intact ovaries is
associated with verbal learning, logical memory, and global cognitive functioning over time.
Hypothesis 8.1: Bilateral oophorectomy predicts a decrease in cognitive performance
beyond the effects of age at cognitive testing.
Specific Aim 9. Examine hormone therapy as a protective factor in the relation between bilateral
oophorectomy and verbal learning, logical memory, and global cognitive functioning.
Hypothesis 9.1: The associations in Hypotheses 6.1, 6.2, 7.1, and 8.1 attenuate if hormones
or estrogen was used around the time of bilateral oophorectomy such that hormone therapy would
be more protective when there was a history, compared to no history, of ovarian removal.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 22
Study I
Method
Participants and Procedures
Study I used the Swedish Twin Registry, a unique database comprising in principle all twins
in Sweden (Lichtenstein, et al., 2002). Same-sex twin pairs were initially contacted in 1961(if born
before 1926) or in 1973 (if born after 1925) when study participants were between age 30 and 50
years old, to enroll them in the twin register, and to survey them for demographic, medical, and
lifestyle information. Twins in the Swedish Twin Registry who were born before 1936 and who
participated in data collection in 1998 to 2001, i.e., when the participants were aged 65 and older, are
part of the Study of Dementia in Swedish Twins. The Institutional Review Board at the University
of Southern California and the Ethics Committee at Karolinska Institutet approved the protocol of
Study of Dementia in Swedish Twins. Questions related to health were asked over the telephone
directly to the participant, and in some cases, an informant. As part of the health interview, a
cognitive screening test was also given.
For the sample used in this dissertation, twins who met the following criteria were included:
(a) are female; (b) are age 65 or older; (c) completed a telephone interview between 1998 and 2001
and whose screening for cognitive impairment was completed; and (d) had both ovaries intact or
both removed at the time of cognitive screening. Of the 14,126 women in the registry, after
excluding twins who had emigrated, had an unlisted telephone number, or died prior to being
contacted for the screening, 11,727 women were contacted. Of them, after excluding twins who
refused, were not reachable, were deceased prior to screening, or had a relative who could not be
contacted to be interviewed, 7,840 women completed the cognitive screening (see Figure 1).
Of the 7,840 female participants, 86 were excluded because they had a unilateral
oophorectomy. This left 7,754 women in the classic case-control analysis, with 154 women who had
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 23
bilateral oophorectomy or their second unilateral oophorectomy before cognitive screening and
7,600 women who had their ovaries intact at the time of cognitive screening (of whom 155 women
consequently had bilateral oophorectomy after cognitive screening but were considered in the study
as having intact ovaries). Sample characteristics are described in the top sections of Tables 1 and 2.
The co-twin control sample was created with the women who had bilateral oophorectomy in
the classic case-control analysis sample. Out of the 154 participants, 47 had male co-twins, 3 had co-
twins who also had bilateral oophorectomy (and therefore a total of 6 females were not included), 3
had co-twins who had unilateral oophorectomy, and 35 had co-twins who did not complete a
cognitive screening. This left 63 same-sex discordant pairs for co-twin control analysis. Sample
characteristics are found in the third sections of Tables 1 and 2.
In the classic case-control analyses, all individuals who met inclusion criteria were used, with
bilateral oophorectomy as the exposure condition. Because the individuals with bilateral
oophorectomy were not matched with their co-twin or with any other factor, the sample of twins is
considered to be representative of a larger female population of Sweden. The twins’ relatedness and
data dependency were statistically adjusted in these analyses.
In contrast, in the co-twin control analysis, the twin pairs are matched by design for age, sex,
and timing of cognitive screening, as screening was performed for both members of a twin pair at
approximately the same time. In co-twin control analyses, the informative pairs are those who are
discordant on both cognition and oophorectomy status. Thus, only complete female monozygotic
and same-sex female dizygotic twin pairs, where one twin had a bilateral oophorectomy and the
other twin had both ovaries intact are used.
Measures
Cognitive impairment. The telephone interview included a cognitive screening protocol named
TELE, constituting a brief telephone cognitive screening test administered to the twin participant.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 24
The TELE cognitive screening (see Figure A1 in the Appendix) has been previously described and is
a valid measure for identifying possible cases of dementia (Gatz, Reynolds, Nikolic, et al., 1995). The
TELE cognitive screening includes the 10-item Mental Status Questionnaire (MSQ; Kahn,
Goldfarb, Pollack, & Peck, 1960) and questions representing three other cognitive domains. The
TELE cognitive screening also included asking the respondent if she had consulted a physician or
retired or required help due to problems with memory or thinking. Scores on the TELE cognitive
screening can range from 0 to 19, based on number of items correct, where lower scores thus
indicate more errors. For this sample, 6,945 had TELE cognitive screening scores ranging from 1 to
19 (mean = 15.38, SD = 2.5).
If the participant could not respond to the screening questions or did poorly on them, an
interview was requested with an informant, who was given the Blessed Dementia Rating Scale
(BDRS; Blessed, Tomlinson, & Roth, 1968), consisting of questions about whether the twin
participant required help to function in daily life due to problems with memory or thinking. The
BDRS is a brief behavioral scale based on an interview with a close informant that has been
demonstrated to correlate well with neuropsychological testing and has been found to discriminate
between nondemented and demented individuals (Juva, et al., 1997).
An algorithm that combined the scores from the twin’s TELE cognitive screening and the
informant’s BDRS was used to create an indicator of cognitive impairment resulting in an ordinal
variable that ranged from 0 (“cognitively intact”) to 3 (“cognitive dysfunction”). This procedure
permitted classifying participants who were unable to complete the TELE and was shown to have
good sensitivity in identifying dementia cases (Gatz, et al., 2002). A “3” was assigned to a participant
if she made more than 3 MSQ errors (the cutoff previously identified by Kahn and colleagues
[Kahn, Goldfarb, Pollack, & Peck, 1960; Kahn, Pollack, & Goldfarb, 1961]), by being impaired
across multiple domains on the TELE and needing activities of daily living assistance due to
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 25
memory problems as evidenced by a score greater than 1.5 on the BDRS. Scores of “2”, “1” and “0”
corresponded to number of domains in which errors were made, in combination with an absence of
functional impairment. Those who received a score of “3” were regarded as having cognitive
impairment, while those with other scores were regarded as not impaired.
For the present study, both the continuous TELE cognitive screening score and the
dichotomous cognitive impairment variable were used in the analyses.
Reproductive surgery status. Both the Inpatient Discharge Registry (IDR), a national database of
hospital discharges in the country, and self-report data were examined to identify history of
oophorectomy. The IDR’s surgical procedure data used 4-digit procedure codes from the
Classification of Operations 6
th
Edition (“Klassifikation av Operationer Sjatte upplagan”) and the 5-
character Nordic Medico-Statistical Committee (NOMESCO) classification.
A participant was considered to have undergone bilateral oophorectomy if the IDR included
a record of one of the following procedures: LAE20, LAE21, LAF10, LAF11, LAF30, 7021, 7022,
7031, or 7032. In addition, a participant was considered to have undergone bilateral oophorectomy
if the IDR included two independent procedures, and not duplicate records, of one of the following
procedures indicating unilateral oophorectomy: LAE10, LAE11, LAF00, LAF01, LAF20, 7020, or
7030. A table of frequencies for each code is presented in Table A2 of the Appendix, which includes
records of oophorectomy for women who also did not have a cognitive screening. As shown in that
table, of the total of 14,126 women, there were 534 records of bilateral oophorectomy and 183
records of unilateral oophorectomy, which resulted in 545 women who had bilateral oophorectomy
(as 11 had 2 records of unilateral oophorectomy) and 161 women who had unilateral oophorectomy,
for a total of 706 women who had any oophorectomy, resulting in a rate for our sample of 5.00%
with any oophorectomy. Another 1,038 had a record in the IDR indicating a hysterectomy, resulting
in 1,418 who had either hysterectomy or any oophorectomy; thus, 10.04% had either hysterectomy
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 26
or oophorectomy. These numbers correspond with the range of frequencies of oophorectomy
reported in Swedish women (one study reported 10% had hysterectomy or oophorectomy and
another study reported 7% oophorectomy and 12% hysterectomy; Hammer, Berg, Fahraeus, &
Larsson-Cohn, 1984; Li, et al., 2003).
The IDR also included dates of surgeries. Age of surgery was determined by the difference
between birthdate and date of bilateral oophorectomy or the second date if two unilateral
oophorectomies were performed.
Hysterectomy and oophorectomy can be performed independently or together in the same
visit. In the IDR, hysterectomy was coded as a separate procedure from oophorectomy. Thus, we
captured hysterectomy as a variable for the study. A participant was considered to have undergone
hysterectomy if the IDR includes a record of “LCDnn” (total excision of the uterus) where “n” is a
number, such as LCD00 and LCD01, and 7220, 7221, 7222, 7223, or 7262.
Self-reports of gynecologic procedures were independently coded. In the telephone
interview, each participant was asked the following questions: “How old were you when you had
your last menstrual period?” “Why did your menstrual periods cease?” “Have you ever had a D&C,
laparoscopy, or removal of the uterus, ovaries, or fallopian tubes?” If a participant responded
positively to the latter question, she was asked “What type of operation?” followed by asking when
the procedure was performed. A participant was coded as having a self-report of oophorectomy if
she stated she had an ovariectomy. The interview did not ask whether her surgery was unilateral or
bilateral. Of those who responded to the self-report question, 10.2% reported oophorectomy.
As IDR records are considered to be more accurate, agreement between the oophorectomy
information from the IDR and self-report information was investigated, taking IDR as the “gold
standard.” Of the 7,754 participants, answers to the self-report of oophorectomy questions were
available for 4,745 (or 61.2%). Of these women, 4,267 (or 55.0%) provided responses that were
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 27
consistent with IDR (4,193 had answered that they did not have an operation; 74 answered that they
did). Of the remainder, 409 reported that they had ovariectomy when IDR records show none and
69 answered that they had no operation when IDR records show that they had bilateral
oophorectomy. For these data: Cohen’s kappa = 19.9%, sensitivity = 51.7%, specificity = 91.1%,
positive predictive value = 15.3%, and negative predictive value = 98.4%.
Of those in the false positive cell, 1 had unilateral oophorectomy in the IDR, 3 had correctly
reported that they had oophorectomy but within 6 months before or after of their cognitive screen
(and therefore were not considered to have bilateral oophorectomy in our analyses), and 129
reported that the age or year in which they had ovariectomy was before the last age or year of
menstrual period (which would be implausible if they had bilateral oophorectomy). Examination of
the remaining 276 participants’ records do not reveal obvious reasons for the false positive reports,
but some women had a history of gynecologic biopsies and resections or other lower abdominal
surgeries or procedures such as colonoscopies.
The 69 false negative participants included 33 who reported they did not have any
gynecologic procedure performed and others who reported gynecologic procedures other than
oophorectomy such as dilation and curettage. Table A3 of the Appendix shows the frequencies of
bilateral oophorectomy by date of surgery and by age of the women in the IDR as well as the
accurate self-reports. The table suggests that the distribution of recorded oophorectomies are mostly
consistent except for prior to 1970 and that accuracy of self-report seems to decline for women
older than 66. Researchers have observed that self-report of oophorectomy has not been as accurate
as self-report of other gynecologic conditions such as hysterectomy (Phipps & Buist, 2009). What
we conclude from the kappa or sensitivity reported above and from the detailed examination of the
women in the false positive and false negative cells is that the telephone health interview as
administered is not a satisfactory method of ascertaining bilateral oophorectomy. Therefore, the
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 28
IDR records were primarily used in the analyses.
Hormone therapy. Hormone use was provided by self-report as part of the telephone health
interview. Two variables were created for the main analyses. The first, hormone ever-use, was a
dichotomous variable based on whether the participant reported ever using hormones (“Do you
have or have you used hormones (e.g., pills, skin patch, vaginal suppositories, or cream)?”).
Participants who had used hormones were asked “What hormones do you use or have you
used?” For each hormone listed, they were asked “How long have you used [the name of the
hormone].” Participants either provided the length of time they were on the medication or the years
they started and ended the therapy. A second, categorical variable indicating hormone use around
the time of surgery was created to reflect the critical window hypothesis. The categories included no
self-reported hormone therapy, hormone therapy initiated before or during the year of the last
menstrual period, hormone therapy initiated within 3 years of last menstrual period, or hormone
therapy initiated 3 years after the last menstrual period or if the timing was unknown.
Because the comparison of self-report and records for ovarian removal did not result in
satisfactory agreement, the self-report responses of hormone therapy may be concerning. However,
the self-report of hormone therapy is used with more confidence. Compared to a one-time
procedure such as ovarian removal, hormone therapy typically requires more active participation by
women, such as taking a dose every day, and more maintenance, such as filling a prescription over
months or years. Due to participants’ greater engagement with hormone therapy, it is assumed that
they would have better recall of their history of it.
Depressive symptoms. The interview schedule included the 11-item Iowa short form (Kohout,
Berkman, Evans, & Cornoni-Huntley, 1993) from the Center for Epidemiologic Studies–Depression
(CES-D; Radloff, 1977) that asks participants to rate how often they experienced a symptom in the
past week. Each item was rated on a 4-point scale from 0 to 3 of never or almost never, seldom,
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 29
often, and always or almost always. The composite score was calculated as the sum of the 11 items,
with 2 of the items (“You were happy” and “You enjoyed life”) reversed-scored. Higher scores
indicate higher levels of depressive symptoms. If a participant was missing 2 or fewer of the items,
the mean of the remaining items was substituted for the missing items. If a participant was missing
more than 2 items, her composite score was not calculated. In this sample, the Cronbach coefficient
alpha was .80 and deleting any one item would not have improved the coefficient.
Demographic, medical, and other covariate information. Demographic information was extracted
from the telephone interview and from other Swedish Twin Registry sources. This information
included sex, years of education, zygosity, and the age when the participant finished the interview
including the cognitive screening. Other covariates included weight, height, and contraceptive use
(ever/never), all from self-report. Weight and height were used to calculate Body Mass Index (BMI)
using the metric formula (weight in kilograms)/(height in meters)
2
. BMI was then coded into a 4-
point ordinal variable based on standard BMI categories of underweight (below 18.5), normal
(between 18.5 and 24.9), overweight (between 25.0 and 29.9), and obese (above 30). Increasing
values of the BMI categorical variable corresponded to increasing BMI. Cancer diagnosis and
gynecologic cancer diagnosis as a reason for surgery were available in the IDR. Dichotomous
variables were created to capture this information. Cancer diagnosis was coded as to indicate
whether a malignant neoplasm of any site was diagnosed. Gynecologic cancer diagnosis was coded
to indicate whether a malignant neoplasm of female genital organs was diagnosed.
The telephone health interview also included questions about cardiovascular history. A
dichotomous variable, cardiovascular history, was created to indicate whether there was a history of
symptoms or diagnoses including at least one of the following conditions: angina pectoris, heart
attack, heart failure, high blood pressure, cerebral hemorrhage or clot in the brain (stroke), transient
ischemic attacks, irregular cardiac rhythm/atrial fibrillation, and history of coronary bypass.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 30
Statistical Analysis
Unmatched, classic case-control and regression analyses were performed on all eligible twins.
Analyses used generalized estimating equations (using PROC GENMOD in SAS), which are
conceptually equivalent to logistic regression for the analysis of classic case-control design, but
adjust for the dependency between twin pairs (the potential cluster-correlation within twin pairs)
(Mucci, Hsieh, Williams, Dickman, Bjorkman, & Pederson, 2004). Outcomes of both the
dichotomous variable of cognitive impairment and the continuous TELE cognitive screening score
were examined. To predict the dichotomous variable, PROC GENMOD with distribution =
multinomial and link = clogit were specified and to predict the screening score, distribution =
normal and link = identity were specified. For both, the correlation structure type = ind
(independent) and repeated subject = pairid, where pairid indicates the identification for each twin
pair. Parameter estimates of the regression models, including log odds for models that included the
dichotomous variable, were reported.
For matched, co-twin control and regression analyses, complete female twin pairs discordant
for history of bilateral oophorectomy were selected. Analyses and results included calculating
matched-pair odds ratios from conditional logistic regression (using PROC PHREG in SAS) and
examining the difference in the continuous cognitive screening score between twins in the pair using
a matched t-test.
In the regression analyses predicting cognitive outcomes, variables were entered in blocks,
where the first block included the covariates (Model I) that were significant (p < .05) in their
associations with cognitive function such as demographic and medical variables. Subsequent models
introduced additional blocks with main effect variables and other factors such as depressive
symptoms and hormone therapy.
In order to explore the question of timing of oophorectomy, both its timing before or after
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 31
natural menopause and age at last menstrual period were considered. Because the age at menopause
was self-reported and the age of oophorectomy was derived from the IDR, it was difficult to be
certain whether menopause was caused by the surgery. Therefore, if oophorectomy and date of last
menstrual period occurred within 1.5 years of one another, women were divided in groups where
one group had oophorectomy at or before the age of 50 and the other group had the surgery at 51
years of age or older. Given that 50 is the typical age at menopause in Sweden, it was more likely
that those with oophorectomy before age 50 had oophorectomy before natural menopause, or
surgical menopause; and more likely that those with oophorectomy at 51 or older had their
oophorectomy after natural menopause. There was a further group of women with oophorectomy
performed at least 1.5 years after date of last menstrual period, which affords some confidence that
surgery was after natural menopause. The mean age of last menstrual period for groups is shown in
Table 1, and for the whole sample it is 49.07 (SD = 5.10, lower quartile = 47, upper quartile = 52).
In addition, there was a group of women for whom timing of oophorectomy relative to menopause
was unknown or implausible (including those women whose date or age of last menses was
unknown), based on a date for menopause that was well after the year of bilateral oophorectomy. A
final category captured those who had no history of oophorectomy. These categories permitted an
exploration of the effect of timing of oophorectomy relative to menopause. In addition to the
covariates, Model IIa included history of oophorectomy and Model IIIa included indicators of the
timing of oophorectomy relative to the last menstrual period.
In addition to the covariates and history of oophorectomy, Model IIb included history of
whether hormone therapy was ever used and Model IIc included length of hormone therapy.
Similarly, in addition to the covariates and the timing of oophorectomy, Model IIIb included the
history of whether hormone therapy was ever used and Model IIIc included length of hormone
therapy. To investigate the possible effect of the length of hormone therapy on women who had
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 32
oophorectomy, an interaction variable, the product of oophorectomy status and length of hormone
treatment, was created. To address possible multicollinearity of the interaction and related terms,
centering the length of hormone treatment variable was considered. Examination of each of these
terms in the models with interactions did not reveal variance inflation or condition indices > 30.
Because of this reason, and the fact that the length of hormone treatment is interpretable at 0,
centering was not performed on the variable for length of hormone treatment. For models that
included the interaction variable, hierarchical regression analyses were conducted such that models
with covariates and the main effects of oophorectomy and length of hormone use were followed by
models with covariates, the main effects, and the interaction term (Baron & Kenny, 1986). These
successive models allowed us to examine the extent to which each block of the variables significantly
contributed to explaining the variance in cognitive performance. Model IId included the interaction
term and its main effects, oophorectomy status and length of hormone therapy.
To examine the extent to which depressive symptoms partially explained the relation
between bilateral oophorectomy and cognitive outcomes, a series of regression models was to be
performed (Baron & Kenny, 1986). With covariates in the models, depressive symptoms were
regressed on oophorectomy status, cognitive outcome was to be regressed on oophorectomy status,
and cognitive outcome was to be regressed on oophorectomy status and depressive symptoms. A
potential for mediation would have existed if a previously significant relation between
oophorectomy status and cognitive outcome lost significance after adjusting for the correlations
between depressive symptoms and cognitive outcomes and oophorectomy status. Depressive
symptoms were included in models with history of oophorectomy and timing of oophorectomy,
Models IV and V, respectively. These models were to include length of hormone therapy if they
were significantly associated with the cognitive outcomes, depending on the results of Model IIc and
IIIc.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 33
To summarize, the aims of Study I were examined by the following models. Hypothesis 1.1
was tested by examining whether history of bilateral oophorectomy was significant in Model IIa.
Hypotheses 2.1 and 2.2 were tested by examining whether history of bilateral oophorectomy was
significant in its association with cognitive outcomes in co-twin analysis. Hypotheses related to the
timing of oophorectomy, 3.1 and 3.2, were tested in Model IIIa. If oophorectomy around the time
of the last menstrual period prior to natural menopause is significantly associated with cognitive
outcomes, it suggests that there is evidence that bilateral oophorectomy before natural menopause is
a risk factor. If a significant relation existed, then age of surgery was entered into the model to test if
it is negatively associated. To examine the extent to which depressive symptoms partially mediate the
relations for Hypothesis 4.1, Baron and Kenny’s method (1986) was to be followed by examining
models that include depressive symptoms in Model IV. The hypothesis related to hormone therapy,
5.1, was examined in Models IIb through IId, IIIb and IIIc. If hormone therapy is protective in its
association with cognitive outcomes, then we expected a negative, significant association. These
models were run to examine both the outcome of cognitive impairment and the continuous
cognitive screening score.
With the exception of depressive symptoms, imputation was not performed on missing data,
so complete cases were used in the analysis. The results provide information about incomplete data
and comparisons of cases used in the analyses and those not included.
SAS Version 9.2 was used for data analysis. Statistical significance was set at two-tailed p <
.05 and associations are summarized as odds ratios with 95% confidence limits.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 34
Results
Age, years of education, CES-D score, length of time on hormone therapy, and age of last
period in participants who had both ovaries removed or both ovaries intact were examined for
group differences. Compared to women with intact ovaries, women with both ovaries removed were
younger at the time of telephone interview and had more years of education. Compared to women
without cognitive impairment, women with cognitive impairment were older, had fewer years of
education, more depressive symptoms, and shorter time on hormone therapy. Compared to their co-
twins with intact ovaries, women with both ovaries removed had more years of education.
Pearson product-moment, Spearman rank-order, point biserial, tetrachoric, and rank biserial
correlations between cognitive screening and other variables (Table 3) revealed that age at cognitive
screening and years of education correlated with the continuous cognitive screening score and thus
were included as demographic covariates. Medical variables of higher BMI category and lack of
cardiovascular history also correlated with better performance on the continuous cognitive screening
score and were also included as covariates in the models. Post-hoc power analysis revealed that for
the classic case-control analysis with n = 7,754, power = .59.
Bilateral oophorectomy and cognitive impairment
The results examining the main effect of bilateral oophorectomy on the dichotomous
outcome of cognitive impairment in the classic case-control analysis are shown in Table 4. The
covariates entered in Model I were all significant. Increased age, less education, lower BMI
categories, and cardiovascular history were associated with cognitive impairment.
The results reveal that oophorectomy generally was not significantly associated with
cognitive impairment (Model IIa). When separate variables representing the timing of the
oophorectomy relative to menopause were considered (Model IIIa), history of bilateral
oophorectomy with unknown or implausible timing with respect to cessation of menses was
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 35
significantly associated with cognitive impairment (Model IIIa), whether or not including use of
hormone therapy (Model IIIc) or depressive symptoms (Model V) in the models. The 27 women in
this group either had unknown timing of menopause (19) or had reported implausible dates of
menopause (8). This subgroup of women whose timing of oophorectomy was unknown or
implausible represented the majority (9 out of 13) of the women who had cognitive impairment. The
mean age of women in the unknown or implausible timing subgroup was 74.70 (SD = 5.81),
suggesting they were not significantly different in age than the rest of the women in the sample. Of
this subgroup, those who had cognitive impairment had a mean age of 79.44 (SD = 5.10), suggesting
that they were not significantly different in age than the women in the sample who were diagnosed
with cognitive impairment. T-tests of continuous measures did not reveal differences in this
subgroup and the rest of the sample, but nonparametric analysis revealed that the subgroup were
more likely to have had hysterectomy (p < .01), had lower BMI (p < .04), and were more likely to
have been diagnosed with cancer (p < .01) compared to the others in the sample. As shown in Table
1, only 33.9% of the women who had cognitive impairment reported age of last menses, compared
to 86.7% of those who did not have cognitive impairment. Separating the group of women who had
oophorectomy within 1.5 years of the date of their last menstrual period by age (T3a and T3b) did
not change the significance of the results.
The timing of bilateral oophorectomy was explored further as post-hoc analysis by running a
model with the covariates (as in Model I) and age of bilateral oophorectomy (not shown in tables).
When the dichotomous outcome of cognitive impairment was predicted, the association between
older age of surgery and cognitive impairment was not significant (parameter estimate = -.03; 95%
CI: -.08, .02; p = .22).
Because of the remarkable difference between the self-report and IDR records of
oophorectomy history, a model with oophorectomy status indicated by only the self-report and not
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 36
the IDR and a separate model with only the women who had accurately self-reported either history
of oophorectomy or no history were also performed. For the former model (which included the
covariates in Model I), the association between self-reported oophorectomy and cognitive
impairment was significant (parameter estimate = -.55; 95% CI: -1.06, -.05; p = .03), suggesting that
oophorectomy is protective. However, this result may be inaccurate given that women with
cognitive impairment were less likely to be asked the self-report question and would have been
included as not having had an oophorectomy. For the latter model, the association between self-
reported oophorectomy confirmed by IDR and cognitive impairment was not significant (parameter
estimate = -.16; 95% CI: -1.34, 1.02; p = .79).
Bilateral oophorectomy and cognitive screening score
The models predicting the continuous cognitive screening (TELE) score are described in
Table 5. Results of models testing the association between the continuous cognitive screening score
and oophorectomy were similar to those testing the association with the dichotomous cognitive
impairment variable. The covariates entered in Model I were all significant. Increased age, less
education, lower BMI categories, and cardiovascular history were associated with lower cognitive
screening scores. Bilateral oophorectomy was not significantly associated with cognitive screening
score (Table 5, Model IIa). However, a significant association was found between worse
performance on the cognitive screening score and the group where timing of oophorectomy was
unknown or implausible in Models IIIa through IIIc (including hormone therapy), Model V
(including depressive symptoms), and Model VIIb (including hormone therapy and age of last
menstrual period).
When the timing of bilateral oophorectomy was explored further as post-hoc analysis by
running a model with the covariates (as in Model I) and age of bilateral oophorectomy (not shown),
the association between older age of surgery and cognitive screening score was not significant
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 37
(parameter estimate = .03; 95% CI: -.005, .07; p = .09).
A model with oophorectomy status indicated by only the self-report and the covariates did
not result in a significant association between self-reported oophorectomy and cognitive screening
score (parameter estimate = .06; 95% CI: -.14, .25; p = .56). A model with only the women who had
accurately self-reported either history of oophorectomy or no history did not result in a significant
association between self-reported oophorectomy and cognitive screening score (parameter estimate
= -.02; 95% CI: -.42, .39; p = .93).
Depressive symptoms as a mediator
Because a significant association was not found between history of oophorectomy and
cognitive outcomes, the analysis described by Baron and Kenny (1986) was not performed.
Regardless, depressive symptoms were included in the models (Models IV and V in Tables 4 and 5)
to examine if they were predictive of cognitive outcomes. Results showed that the depressive
symptoms were significantly associated with cognitive impairment and with the continuous measure
of cognitive functioning, independent of the covariates, length of hormone therapy use, and ovarian
status. The coefficient estimates for the group whose timing of oophorectomy was unknown in both
Model V’s increased by 20% or more from the respective estimates in Model IIIc’s, indicating that
depressive symptoms may partially be associated with the relation between bilateral oophorectomy
and both cognitive impairment and the continuous measure of cognitive function. The increase in
estimates may suggest that depressive symptoms have a suppression effect in the association
between oophorectomy and cognitive outcomes.
It is likely that women with cognitive impairment were unable to complete the interview and
therefore would not have responded to the depressive symptoms items. There were group
differences between those for whom depressive symptoms were reported and for whom depressive
symptoms were not scored. The latter group was older (mean age = 79.93, SD = 7.94; compared to
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 38
mean age = 78.20, SD = 6.29; p < .01) and had fewer years of education (mean years = 7.27, SD =
2.10; compared to mean years = 8.42, SD = 2.82; p < .01), fewer years of hormone therapy (mean
years = .11, SD = 1.39; compared to mean years = 1.05, SD = 3.61; p < .01), lower BMI (p < .01),
hysterectomy performed less (p < .01), and fewer diagnoses of cancer (p < .01).
Age of last menstrual period
Age of last menstrual period, which considers the possibility of earlier depletion of estrogen
through natural menopause or surgery, was included in Models VIa through VIc, VIIa, and VIIb. It
was not found to be significantly associated with cognitive impairment or the continuous cognitive
screening score.
It is again likely that women with cognitive impairment were unable to complete the
interview and therefore would not have responded to the items related to age of last menstrual
period. There were group differences between those for whom age of last menstrual period was
reported and for whom it was not reported. The latter group was older (mean age = 78.90, SD =
7.85; compared to mean age = 72.98, SD = 6.18; p < .01) and had fewer years of education (mean
years = 7.50, SD = 2.33; compared to mean years = 8.44, SD = 2.83; p < .01), fewer years of
hormone therapy (mean years = .31, SD = 2.10; compared to mean years = 1.07, SD = 3.64; p <
.01), more depressive symptoms (mean = 5.01; SD = 4.59; compared to mean = 4.13, SD = 4.01; p
< .01), lower BMI (p < .01), hysterectomy performed less frequently (p < .01), and fewer diagnoses
of cancer (p < .01).
Hormone therapy as a protective factor
For Hypothesis 5.1, all models that included either an indication of hormone therapy or
length of hormone therapy (Models IIb through d, IIIb and IIIc, IV, V, VIb and VIc, and VIIb in
Tables 4 and 5) consistently show that history of hormone therapy and longer lengths of hormone
therapy were protective against cognitive impairment, using either the dichotomous indicator or the
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 39
continuous cognitive screening score. The interaction of the length of hormone therapy and
oophorectomy was not significant (Models IId and VIc), revealing that hormone therapy seems to
be a main protective factor against cognitive impairment and is not moderated by history of bilateral
oophorectomy.
The inclusion of the categorical variable representing hormone therapy around the time of
oophorectomy with Model IIa or with Model IIc did not significantly predict cognitive impairment
or the continuous cognitive score (not shown).
Again, it is likely that women with cognitive impairment were unable to complete the
interview and therefore would not have responded to the items related to hormone therapy. There
were group differences between those for whom history of hormone therapy use was reported and
for whom it was not reported. The latter group was older (mean age = 80.42, SD = 7.89; compared
to mean age = 73.24, SD = 6.31; p < .01) and had fewer years of education (mean years = 7.15, SD
= 2.00; compared to mean years = 8.41, SD = 2.82; p < .01), lower BMI (p < .01), hysterectomy
performed less frequently (p < .01), and fewer diagnoses of cancer (p < .01).
Bilateral oophorectomy using co-twin control analyses
In the test of Hypothesis 2.1, twins with oophorectomy and their co-twins who did not
experience oophorectomy were not found to have a different risk of cognitive impairment, resulting
in an odds ratio of 1.00, 95% CI .20, 4.95. In 3 out of the 63 pairs, the members of the twin pair
who had bilateral oophorectomy were later assessed with cognitive impairment and in 3 pairs the
members of the twin pair who did not have bilateral oophorectomy were later assessed with
cognitive impairment. None of the twins in the remaining 57 discordant pairs was assessed as having
cognitive impairment. A matched t-test comparing cognitive screening scores between twins with
oophorectomy and their co-twins without oophorectomy resulted in t = .43 and p = .67 for n = 60.
Matched t-tests of depressive symptoms and age of last menstrual period also did not reveal
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 40
significant differences in the twins and the co-twins. Wilcoxon signed-rank test for paired samples
revealed significant differences in cardiovascular history (p = .048), where 10 more twins with
oophorectomy had cardiovascular history than their co-twins. Wilcoxon signed-rank tests for paired
samples did not reveal significant differences in hormone use or BMI category.
The size of the co-twin sample precluded analyses of timing of oophorectomy relative to
menopause. With respect to overall tests of the association between cognition and oophorectomy,
classic case control and the matched co-twin control analyses gave consistent results, and no further
analyses were conducted to test Hypothesis 2.2.
Differential survival
Because an association between bilateral oophorectomy and cognitive impairment was only
found in the group whose oophorectomy relative to last menstrual period was unknown and there is
a lack of clear differences related to the timing of the oophorectomy in the woman’s reproductive
lifecycle, we conducted additional analyses to test for potential explanations, including whether
findings might have been affected by differential survival. Confounding by differential survival can
occur when a cross-sectional or prevalence design is used. In this regard, we considered the
possibility that more of the female twins who had bilateral oophorectomy died prior to cognitive
screening compared to their co-twins or to other females of the same age who did not have the
surgery. If this were true, the sample used in this study would be biased, underrepresenting the
cognitive outcomes of those who had survived oophorectomy. Particularly given that an average of
13.5 years passed from oophorectomy to cognitive screening (by examining the difference in age
when they occurred), there would be opportunity for differences to emerge. To examine this
possibility, mortality rates of those who had bilateral oophorectomy and those who did not were
compared in the entire sample of female twins in the Study of Dementia in Swedish Twins,
including those who had died prior to participating in the cognitive screening. Table 6 shows that
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 41
the rate of mortality was lower for those who had bilateral oophorectomy (13.60%) compared to
those who did not have the surgery (16.76%). These rates do not support the interpretation that
females who had oophorectomy were not sufficiently represented in the study due to differential
survival prior to cognitive screening. When person-time of follow-up was considered by using
estimates of mean age of each of the samples, the incidence rate difference = -.00041 and incidence
rate ratio = .83.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 42
Study II
Method
Participants
Study II used data from 3 clinical trials: B-Vitamin Atherosclerosis Intervention Trial
(BVAIT), the Women's Isoflavone Soy Health Trial (WISH), and the Early Versus Late Intervention
Trial with Estradiol (ELITE). The objectives of the trials were to test the effects of randomized
interventions (B-vitamins, soy, and oral estradiol, respectively) on progression of early or subclinical
atherosclerosis among healthy men and postmenopausal women. All three trials were randomized,
double-blind trials. A neuropsychological battery was given at baseline and at the end of the trials as
secondary trial outcomes. The participants in the trials were healthy as they (a) did not have diabetes
mellitus; (b) did not have clinical signs or symptoms of cardiovascular disease; (c) did not have
untreated thyroid disease; (d) did not have uncontrolled hypertension; and (e) had a life expectancy
of at least 5 years. Exclusion criteria did not include cognitive or psychiatric conditions. Detailed
information about the trials has been previously published (Hodis, et al., 2009; Henderson, et al.,
2012; Henderson, et al., 2013). Participants signed a written informed consent approved by the
Institutional Review Board at the University of Southern California. Study II includes participants in
the three trials who met inclusion criteria: (a) were postmenopausal women; (b) completed at least
the baseline wave of neuropsychological testing (testing scores for a second time point were
available for BVAIT and WISH participants); (c) had both ovaries intact or both ovaries removed at
the beginning of the trial. Table A4 of the Appendix describes the participants included in the
sample in more detail. In this study, a total of 1,149 women was available for analysis (see Figure 2).
Tables 7 and 8 list demographic and clinical characteristics. Participants are generally well-
educated with approximately one third holding graduate degrees and ethnically heterogeneous with
approximately one third with an ethnicity other than Caucasian. More than a quarter are obese
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 43
(BMI>30).
Procedures
Participants were drawn from the general population living in Los Angeles County
predominantly through English language media advertisement. Interventions took place on the
Health Sciences Campus of the University of Southern California, Los Angeles. Participants were
asked to complete questionnaires at the beginning of the trials.
BVAIT was conducted from November 2000 to June 2006. Both men and postmenopausal
women over 40 years old were recruited. Of the 5,309 men and women prescreened by telephone,
3,245 did not meet inclusion criteria and 1,558 declined to participate. Computer-generated random
numbers were used to assign participants to daily supplementation with folic acid with vitamins B12
and B6 or matching placebo. Clinic visits occurred every 3 months and treatment adherence was
assessed at each visit. The initial study period was 2.5 years (the treatment period was subsequently
extended).
WISH took place from April 2004 to March 2009. Of 1,063 women who were prescreened
by telephone, 660 did not meet inclusion criteria and 53 declined to participate. Computer-generated
random numbers were used to assign participants to a daily dose of isoflavone-rich soy protein or to
milk protein-matched placebo. Clinic visits occurred every month for the first 6 months and every 2
months afterwards, for a total of 2.5 years. Participants were instructed not to consume soy
supplements.
ELITE started in July 2004. Recruitment has completed but the study is still ongoing. A total
of 643 women initially participated in the trial. Of the 2,814 women were prescreened by telephone
for eligibility, 895 women were screened in person. Computer-generated randomization sequence
was prepared before trial initiation to assign participants to a 17β-estradiol or placebo. Clinic visits
occurred every month for the first 6 months and every 2 months afterwards.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 44
Measures
Cognitive functioning. Fourteen neuropyschological tests (see Table A5 in the Appendix) were
administered by one trained psychometrist to measure executive functioning, semantic memory,
logical memory, visual memory, and verbal learning. The tests were administered at baseline for all
of the three trials and for BVAIT and WISH, at the end of the trials, about 2.5 years later. End of
trial cognitive measurements for the ELITE trial were not completed at the time of the analysis of
this dissertation. The same neuropsychological battery was given both at baseline and at the end of
the trials.
A global functioning score and five domain-specific scores were calculated for each
participant. The 5 cognitive domain scores were based on the factor loadings generated from
principal components analysis performed on the cognitive tests, using consecutive uncorrelated
factors extracted for WISH and BVAIT which have been published previously (Gatto, et al., 2008;
Gatto, et al., 2009). The means and standard deviations calculated for each test within each trial were
used to calculate a Z-score for each administration and participant. Then, a composite score for each
cognitive domain was calculated by adding the Z-scores of each test identified as a factor in the
domain, except for the Z-score for Trails B, which was subtracted from the other scores. For
example, for logical memory, the Z-scores from the Immediate Recall and the Delayed Recall of the
Wechsler Memory Scale, 3rd edition (WMS-III) Logical Memory I Paragraph Recall test were totaled
for each participant. Table A5 in the Appendix lists the tests included in each composite score.
Global functioning was calculated as the sum of the Z-scores of all of the domains (this is different
from how the trial findings are reported). Change scores for global functioning and domain
composite scores were calculated by subtracting the baseline composite scores from the end-of-trial
scores for each participant. Higher Z-scores indicate better cognitive performance.
If a participant had a missing test score for a domain, the domain score was considered
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 45
missing. If a participant was missing any one of the domain scores, the global functioning score was
considered missing. Of the domains, the most data were missing from executive function, where
2.61% of the participants were missing scores. For global functioning, 3.74% were missing scores.
T-tests of continuous measures and nonparametric analyses for non-continuous measures were
performed to identify group differences. No differences in age, ethnicity, education category, BMI
category, and oophorectomy status were found between those for whom global functioning score
was available and those for whom it was missing.
Reproductive surgery status. At the beginning of the trials, participants completed a reproductive
history questionnaire that included history of oophorectomy and hysterectomy. If oophorectomy
was reported, they were asked whether one or both ovaries were removed, and an attempt was made
to verify the information with medical records. (Verification minimized false positives in this sample,
which was important given that the notable differences of self-report and records for ovarian
removal in Study I. Minimizing false positive was particularly important because a higher proportion
of women reported oophorectomy than were verified by IDR in Study I. ) Women who had both
ovaries removed were considered to have bilateral oophorectomy. Women who reported that only
one ovary was removed were excluded from the current study. None of the participants reported a
change in ovarian status during the trial. Age at surgery was calculated by the difference between
birthdate and reported date (or year, if only year was given) of surgery. Age at menopause was
calculated by the difference between birthdate and self-reported date (or year, if only year was given)
of last menstrual period. A set of categorical variables was created to indicate one of the following
conditions related to timing of oophorectomy with respect to menopause: a) no reported
hysterectomy or oophorectomy (resulting in natural menopause); b) oophorectomy and
hysterectomy at least 3 years after natural menopause, or if either date of surgery or menopause is
unknown; c) oophorectomy and hysterectomy within 3 years after natural menopause; and d)
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 46
oophorectomy and hysterectomy within 1 year of menopause (which resulted in surgical
menopause). Regarding the second subgroup, dates of surgery and menopause were unknown for 34
out of 50 women (for one woman, date of surgery was known but the date of menopause was
unknown). They were grouped together with those who had surgery after natural menopause as a
conservative assumption was made that the women had surgery after natural menopause. The last
category of women who experienced surgical menopause was subdivided into 2 groups close to the
mean of the age of menopause, with one group of women whose age of surgical menopause was at
45 years of age or younger and the second group of women whose age of surgical menopause was
older than 45 years of age.
Hormone therapy. Hormone use was also extracted by a questionnaire that the participants
completed at the beginning of the trial. Variables were created for the main analyses. Hormone ever-
use was a dichotomous variable that indicated whether the participant reported ever using hormones
for menopausal symptoms (“Have you ever taken hormones for menopause?”). Length of hormone
use was also provided by self-report (“If yes, how long did you take hormones?”). Finally, current
hormone use was a dichotomous variable indicating if the participant reported current use in the
BVAIT trial. Women who were currently on hormones in the WISH or ELITE trials stayed off for
a period of time to qualify for the trials.
Depressive symptoms. At the beginning of the trial, participants were asked to complete the full-
scale, 20-item CES-D (Radloff, 1977) by questionnaire. Participants completed each item on a 4-
point scale ranging from 0 (“never or almost never”) to 3 (“always or almost always”). The
composite score was calculated as the sum of all items, with 4 of the items, including “You were
happy” and “You enjoyed life,” reversed-scored. Higher scores indicate higher levels of depressive
symptoms. A total score for the CES-D was computed using all 20 items. If a participant did not
complete 3 or fewer of the items, the mean of the remaining items was used in place of the missing
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 47
items. If a participant did not complete more than 3 items, her composite score was not calculated.
In the BVAIT sample, the Cronbach coefficient alpha was .84 and in the WISH sample, the
coefficient was .82. Although deleting one item from each sample would have improved the
coefficient by .01, for completeness, all of the items were left in the composite score.
Demographic, medical, trial, and other covariate information. Demographic information was
extracted from the questionnaires at the beginning of the trial. This information included age and
information that was captured as dichotomous variables including ethnicity (Caucasian or other),
marital status (married or other), and oral contraceptive use (history of use or never). Years of
education were coded into 7 categories: 8th grade or less, some high school, high school graduate,
trade/business school, some college, Bachelor's degree, and graduate/professional degree.
Increasing values of the education categorical variable roughly corresponded to more years of
education.
Weight and height were directly measured at the time of the neuropsychological assessment
and were used to calculate BMI (703 x [weight in pounds])/[height in inches]
2
. BMI was then coded
into a 4-point ordinal variable based on standard BMI categories of underweight, normal,
overweight, and obese (under 18.5, greater or equal to 18.5 but less than 25, greater or equal to 25
but less than 30, and greater or equal to 30, respectively). Increasing values of the BMI categorical
variable corresponded to increasing BMI. Time since menopause was calculated by the difference of
age at baseline and age at menopause.
The trial from which each participant was recruited (e.g., BVAIT, WISH, ELITE) was
indicated by a categorical variable to statistically adjust for inclusion and exclusion criteria and
unique trial population as described in Table A4 in the Appendix. A dichotomous variable that
indicated whether the participant received treatment during the trial was also created. The variable
indicated whether women were on the active treatment arm of either WISH (who received
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 48
isoflavone, which is capable of weak, estrogen-like effects) or BVAIT (who received administration
of folic acid) or on the placebo arm of either WISH or BVAIT.
Statistical Analysis
The analyses examined the extent to which there are differences in cognitive performance in
women with a history of bilateral oophorectomy compared to those who have their ovaries intact
and how the trajectories of cognitive changes differ among these two groups of women. Linear
regression models were used to examine bilateral oophorectomy in cross-sectional cognitive
performance scores and changes in the scores over time. Separate models were analyzed for each
cognitive domain and global cognitive functioning. Each composite score was regressed first on a
block of only covariates (Model I), followed by blocks of predictor variables. Covariates included
demographic and medical variables that were significant in their associations with cognitive
performance domains. Together with the covariates, Model II included history of oophorectomy
and Model III included indicators of the timing of oophorectomy relative to natural menopause. In
addition, the analyses that examined cognitive performance changes included the indicator variable
of the trial from which each participant was recruited and the dichotomous variable indicating
treatment or placebo.
To examine the possible effect of prior hormone therapy on women who had
oophorectomy, an interaction variable, the product of oophorectomy status and length of hormone
treatment, was created. To address possible multicollinearity of the interaction and related terms,
centering the length of hormone treatment variable was considered. Examination of each of these
terms in the models did not reveal variance inflation or condition indices > 30 other than the trial
variable. For this reason, and the fact that the length of hormone treatment is interpretable at 0,
centering was not performed on the variable for length of hormone treatment. The interaction
variable was included in hierarchical regression analyses such that models with covariates and the
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 49
main effects of oophorectomy and length of hormone use were followed by models with covariates,
the main effects, and the interaction term (Baron & Kenny, 1986). These successive models allowed
us to examine the extent to which each block of the variables significantly contributed to explaining
the variance in cognitive performance.
To examine the extent to which depressive symptoms partially explained the relation
between bilateral oophorectomy and cognitive functioning, a series of regression models was to be
performed (Baron & Kenny, 1986). With covariates in the models, depressive symptoms were
regressed on oophorectomy status, cognitive performance was to be regressed on oophorectomy
status, and cognitive performance was to be regressed on oophorectomy status and depressive
symptoms. A potential for mediation would have existed if a previously significant relation between
oophorectomy status and cognitive performance lost significance after adjusting for the correlations
between depressive symptoms and cognitive performance and oophorectomy status. Depressive
symptoms were included in models with timing of oophorectomy.
The aims of Study II were examined by the following models. Hypothesis 6.1 was tested by
Model II as it included the variable for oophorectomy history and Hypothesis 6.2 was tested by
Model III as it included timing of oophorectomy. To examine the extent to which depressive
symptoms partially mediate the relations as described in Hypothesis 7.1, Baron and Kenny’s method
(1986) was followed by examining each of the relations, including adding depressive symptoms to
models including oophorectomy status and timing of oophorectomy. To examine cognitive
performance over time for Hypothesis 8.1, the change scores were regressed on a model that
included the variables in the model that predicted timing of oophorectomy, the baseline cognitive
score, treatment arm indicator, and trial indicator. To test Hypothesis 9.1, related to hormone
therapy, the interaction term was added to the models with the main effects of history of
oophorectomy and time on hormone therapy.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 50
SAS Version 9.2 was used for data analysis. Statistical significance was set at two-tailed p <
.05 and linear regression coefficients with 95% confidence limits were used to summarize model
results.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 51
Results
Table 9 describes the 152 women who had bilateral oophorectomy and the 997 women who
had both ovaries intact in the sample. Among age, CES-D score, length of time on hormone
therapy, age of menopause, and length of time since menopause, women who had bilateral
oophorectomy were using hormone therapy longer, were younger at menopause, and had more time
since menopause. Table 10 shows Pearson product-moment, Spearman rank-order, point biserial,
tetrachoric, and rank biserial correlations of the demographic and medical factors with cognitive
domain performance. Younger age, higher education level, Caucasian ethnicity, and lower BMI
category were included in all models as demographic covariate variables as they were associated with
better scores for at least 5 out of the 6 cognitive domains and global performance scores. These
covariates are similar to the ones included in Study I.
The extent to which there were differences in the cognitive performances among the
participants in the trials was examined. Instead of testing for group differences across the 3 trials
using the raw scores of the 14 individual tests, Z-scores of the cognitive domain performance were
compared. Mean Z-scores (Table 11) were calculated using means and standard deviations from the
combination of all of the scores from the trials for both test administrations so that general
comparisons could be made across trials and administrations. These Z-scores were calculated for the
purposes of comparing the scores across the trials but not used in the regression analyses. Raw score
differences were not tested. The results show that the participants in the BVAIT trial had generally
lower performances (with significantly lower education levels than the other two trials with p < .01;
not shown) than those of the other trials.
Post-hoc power analysis
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 52
Given n = 400, observed R
2
= .06, probability level = .05, and number of predictors = 10,
the observed statistical power for multiple regression = .95. Because most of the models had more
observations, larger R
2
, or fewer predictors, there is greater statistical power for most of the models.
Bilateral oophorectomy and cognitive performance
Table 12 describes the results of the series of linear regression models predicted baseline
cognitive performance of the 5 cognitive domains and global cognitive functioning with the first
model showing the results with only the covariates, a second model with with covariates and any
oophorectomy history, and a third model with covariates and oophorectomy given its timing before
or after natural menopause and age at menopause. In the third model, oophorectomy status was
represented by the 4 variables that indicate the timing of oophorectomy with respect to menopause.
Oophorectomy at age 45 or later that led to surgical menopause was negatively associated with
performance scores in verbal learning. This significance, however, no longer held with the inclusion
of length of time on hormone therapy in the model (model R
2
= .11, parameter estimate for h3b = -
.51, p = .075; not shown). Other cross-sectional analyses did not show oophorectomy or its timing
to be significantly associated with performance in domains when testing Hypotheses 6.1 and 6.2.
Because the education levels of BVAIT participants significantly differed compared to those in the
other trials, the second and third models were performed with BVAIT participants and the WISH
and ELITE participants separately. These models did not result in other significant associations
between oophorectomy and cognitive domain performance.
Depressive symptoms
Because oophorectomy was not found to be significantly associated with cognitive domain
functions in the cross-sectional analyses, the analysis to examine depressive symptoms, Hypothesis
7.1, was not examined. Instead, the extent to which the inclusion of depressive symptoms with
oophorectomy in the models predicting cognitive performance was examined (Table 13). We found
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 53
that with the exception of visual memory, greater depressive symptoms were significantly associated
with lower performance in all cognitive domains for models with either oophorectomy status or
timing of oophorectomy. History of oophorectomy after 45 years of age continued to be
significantly associated, with no parameter change, with lower verbal learning performance when
depressive symptoms were included in the model.
Cognitive performance over time
Table 14 describes the results of regressing cognitive performance on both oophorectomy
status and the timing of oophorectomy after adjusting for baseline cognitive performance and
demographic and trial-related (trial and treatment arm) factors to investigate Hypothesis 8.1. Figure
3 provides graphical representations of the cognitive domain performance. History of surgical
menopause after 45 years of age was significantly associated with lower performance in visual
memory (mean difference between the end-of-the-trial and baseline score for women with both
ovaries intact was .04 and SD = 1.30; for all women without ovaries, mean = -.26 and SD = 1.30;
for women who had surgical menopause after 45 years of age, mean = -1.17 and SD = 1.28). Other
domains did not appear to show that the timing of oophorectomy is significantly associated with
declines in other areas of cognitive performance.
Hormone therapy
Length of time on hormone therapy was examined with depressive symptoms (Table 15) and
with depressive symptoms and time since menopause (Table 16) in cross-sectional analysis.
Significant associations between bilateral oophorectomy and cognitive performance were not found
in these models. Length of hormone therapy was also not found to be significantly associated with
the cognitive performance.
When the main effect of length of time on hormone therapy was added into the models over
time to test Hypothesis 9.1 (Table 17), the results were consistent to Table 14 and revealed that
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 54
surgical menopause after the age of 45 was significantly associated with lower visual memory domain
performance over time. In a model where the interaction variable of length of time on hormone
therapy and oophorectomy (Table 18) was included with oophorectomy status, oophorectomy status
was associated with lower visual memory domain performance over time. However, the interaction
was not found to be significant in any of the models. Including baseline depressive symptoms in
both of the models did not change the significance of the results (not shown). With depressive
symptoms, oophorectomy after the age of 45 was still found to be significant for lower visual
memory performance over time (parameter estimate for h3b = -1.05 and p <.01) and oophorectomy
status was still found to be significant for lower visual memory performance over time (parameter
estimate for oophorectomy = -.48 and p =.03). In the latter model, it appears that with inclusion of
depressive symptoms, the parameter estimate for oophorectomy increases with depressive
symptoms by approximately 11%. In a model with the main effects of oophorectomy status and
length of time on hormone therapy without the interaction variable (a model similar to the one in
Table 14 but with length of time on hormone therapy), oophorectomy status is no longer significant
in being associated with a decline in visual memory (not shown; p = .09).
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 55
Discussion
This dissertation examined the extent to which bilateral oophorectomy, its timing relative to
natural menopause, and hormone treatment are associated with cognitive functioning using two
complementary studies. Study I used a twin sample to examine outcomes of a cognitive screening
test and cognitive impairment. Study II combined baseline assessments from 3 clinical trials and
follow-up assessments for 2 of the trials to examine outcomes of performance of global functioning
and 5 cognitive domains. In both studies, the possible mediating effect of depressive symptoms and
moderating effect of hormone therapy were examined. Although oophorectomy was not found to
be associated with general cognitive outcomes, some evidence of oophorectomy and its timing as a
risk factor for lower cognitive domain performance was found. Findings also suggested associations
with depressive symptoms and beneficial effects of hormone treatment.
In Study I, history of bilateral oophorectomy was not associated with cognitive impairment
in either the classical case control or co-twin control analyses. Similarly, no significant association
between history of bilateral oophorectomy and the continuous cognitive screening score was found.
The one exception is that for the women whose timing of menopause was reported as unknown or
implausible, the surgery was positively associated with both cognitive impairment and lower scores
on the cognitive screening score compared to those who did not have the surgery. An interpretation
regarding timing of oophorectomy with respect to menopause was not made as most women with
cognitive impairment did not report the age of last menses (9 of the 13 women with cognitive
impairment and oophorectomy). Because information about timing of oophorectomy, use of
hormone therapy, and depressive symptoms were all included in the telephone interview, these
variables were more likely to be missing for women with cognitive impairment. These women would
either have been unable to participate in the telephone interview or would have had the interview
terminated by design and an informant solicted. It may also be possible that women with cognitive
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 56
impairment were not able to provide accurate timing of last menses because it was perhaps confused
with the date of oophorectomy, difficult to identify because menses may become irregular for some
time before finally ceasing, or forgotten as its occurrence was approximately 20 years earlier. Because
oophorectomy was not found to be a significant risk factor in the other groups, with or without the
consideration with respect to menopause, evidence to support the study’s hypotheses of bilateral
oophorectomy as a risk factor for cognitive impairment or lower performance on a cognitive
screening test was not found.
Study II found evidence of oophorectomy status as a risk factor for cognitive domain
performance. Women who had oophorectomy surgery within one year of menopause after the age
of 45 showed a lower mean level of verbal learning performance compared to those with intact
ovaries. In analyses examining cognitive performance over time, the same subset of women had
lower scores in visual memory compared to those with intact ovaries. Although no other cognitive
domain showed significant relations with oophorectomy status regardless of timing, it may be
noteworthy that surgical menopause occurring later in life, even after 45 years of age, may be
associated with declines in some cognitive domains. Studies examining shorter-term outcomes have
reported declines in verbal memory associated with oophorectomy, but visual memory has not been
discussed as regularly (Verghese, et al., 2000; Nappi, et al., 2000; Costa, et al., 1997).
The results of Studies I and II were consistent in not finding oophorectomy status or its
timing to be significantly associated with broader measures of cognitive functioning, including
cognitive impairment, performance on a cognitive screening test, or the composite measure of
global functioning. This suggests that if oophorectomy is a risk factor for poorer performance in
specific cognitive domains, the declines may not be so broad or significant as to have an impact on
activities of daily living such that it entails cognitive impairment or globally poor performance. This
conclusion may be consistent with others’ observations that surgical menopause may be deleterious
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 57
for cognitive functioning, at least short-term (Henderson & Sherwin, 2007), with the effects waning
over time as other risk factors for cognitive impairment accumulate. This idea may be further
supported as women in Study II were approximately 14 years younger (and were tested 7 or more
years sooner after menopause) compared to women in Study I.
There are two points that may be worth considering in regard to the lack of significant
findings in Study I. First, it is worth noting that in Study I, women who had both ovaries removed
were younger in age at the time of cognitive screening and had more years of education compared to
women whose ovaries were intact. Although these factors were adjusted for as covariates in the
models, it is possible that there are factors related to age and more years of education that were
unaccounted for in this study that could be protective against cognitive impairment or poor
performance on a screening test, such as better maintenance of health and well-being. Therefore,
given that the women who had oophorectomy had the advantage of unaccounted protective factors,
it is possible that they would have been expected to cognitively perform no worse than their
comparison group. In addition, a prior study that examined discordant twins for cancer found that
women who had localized treatment for gynecologic cancers (most likely surgery which would have
affected reproductive organs) were at an increased risk for cognitive impairment (Kurita,
Meyerowitz, Hall, & Gatz, 2011).
The dissertation findings do not show that hormone use was a risk factor in any of the
cognitive outcomes as has been found by Shumaker and colleagues (2003, 2004). Estrogen is
thought to be associated with certain domains as it may have beneficial effects on neurotransmitter
systems and density of estrogen receptors in certain regions of the brain related to attention and
memory (McEwen & Woolley, 1997; McEwen, 2002). In Study I, the main effect of hormone use
was protective against cognitive impairment and poorer performance on cognitive screening score
regardless of ovarian status, suggesting that the use of hormone therapy may be protective against
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 58
cognitive impairment or decrease in cognitive function later in life. Not only was this association
found with hormone use history, but a dose-response relationship was found as the main effect of
length of hormone use was also found to be protective in both the dichotomous and continuous
measures of cognition. This is consistent with the findings of a study that used an overlapping
sample (Rasgon, et al., 2005) that found that hormone use was associated with a reduced risk of
cognitive impairment among Swedish twins. The interaction of length of hormone use and ovarian
status was not found to be significantly associated with cognitive impairment or worse cognitive
screening scores, so evidence of an association between length of hormone use and cognitive
impairment or worse performance on the cognitive screening test by oophorectomy status was not
supported.
In Study II, the main effect of length of hormone therapy was not associated with any
specific cognitive domain performance regardless of ovarian status in any of the models or
outcomes. When hormone use as a moderating factor was examined with the inclusion of the
interaction term of hormone therapy and ovarian status in the model, oophorectomy status was
found to be associated with lower performance scores in visual memory but no other cognitive
domains. In addition, hormone therapy use as a moderator was not supported in either study.
There are differences in hormone therapy usage that might account for significant findings
in Study I but not in Study II. Most of the Study II women had reported hormone therapy history
(70.41% had a history of hormone use) whereas most of Study I women did not (17.94% had a
history of hormone use). Among the women who were on hormone therapy, women in Study I
were using it for less time than those in Study II (for Study I mean = 5.34 years, SD = 6.62; for
Study II, mean = 6.99 years, SD = 6.92). Among women without oophorectomy, women in Study I
used hormone therapy less than those in Study II (for Study I, it was used by 21% for less than a
year; for Study II, it was used by 67% for approximately 4 years). Hormone therapies used by the
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 59
participants in the studies may have varied in the hormones included (i.e., estrogen only, estrogen
and progesterone, etc.), compounds used, administration method, and timing of administration.
These factors may have different associations with cognitive performance, as suggested by studies
that have looked at various hormones in cross-over study designs (Sherwin, 1988; Schiff, Bulpitt,
Wesnes, & Rajkumar, 2005).
Depressive symptoms were studied as a mediator in the relation between oophorectomy and
cognitive functioning in both Studies I and II. In Study I, depressive symptoms were positively
associated with cognitive impairment and lower cognitive screening scores, independent of
oophorectomy or hormone therapy, which is consistent with existing literature (Rocca, et al., 2008).
Additionally, in Study II, depressive symptoms were associated with poorer cognitive performance
in all domains except for visual memory. In both studies, complete tests for mediation were not
conducted due to the insufficient evidence of its effect as a mediator in the relation between
oophorectomy and cognitive function.
In both studies, the covariates were significantly associated with cognitive function. In Study
I, increased age, less education, and lower BMI was generally associated with cognitive impairment
and lower performance on the cognitive screening score, and in Study II, increased age, less
education, ethnicity other than Caucasian, and higher BMI was generally associated with lower
cognitive domain performance. Comparing the frequencies of the BMIs in the two studies seem to
show that the Study II sample was skewed to higher BMIs than Study I. The relation between BMI
and cognitive functioning is dependent on stage of life. Several studies have found that in mid-life
(representing the sample of Study II), higher BMI has been associated with lower cognitive ability
and poor cognitive function but that in later life (representing the sample of Study I), higher BMI
may be protective (Cournot, et al., 2006; Gustafson, et al., 2009; Nourhashemi, et al., 2003). This
was also found in a study with an associated Swedish twin sample (Dahl, et al., 2010).
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 60
Some of the findings in this dissertation are not consistent with the meta-analysis (Kurita,
Gatz, & Mack, unpublished) that found that women with bilateral oophorectomy had poorer overall
cognitive performance compared to those who did not have the surgery by a standardized mean
difference of 0.137 standard deviation units. The domains that were associated with oophorectomy
in the meta-analysis (global function, construction and reasoning, and logical memory) did not
include verbal memory, which was associated with oophorectomy in this dissertation. The meta-
analysis also found that in pre-post designs, women had poorer overall cognitive performance after
versus before bilateral oophorectomy by a standardized mean difference of 0.611 standard deviation
units with significant associations for global function, executive function, verbal learning, logical
memory, and visual episodic memory. The meta analysis found its smaller effects in between-group
design studies, which is most similar to the study designs in this dissertation. In this dissertation, the
domain that was associated over time with oophorectomy was visual memory. It may be interesting
to include the results from these studies in the meta-analysis.
There were differences in the sample characteristics in the two study samples. First, as
mentioned above, the samples of the two studies are different in that the women in Study II are
about a decade younger and healthier as determined by its study inclusion criteria, than the women
in Study I. In addition, also mentioned above, they varied in their use of hormones. The participants
in Study II also volunteered in a multi-year clinical trial that included multiple administrations of a
battery of neuropsychological assessments. Perhaps relatedly, they are made up of women who were
generally well-educated with approximately one third of them having a graduate or professional
degree, whereas the sample in Study I is a larger population-based sample of twins, excluding only
those who refused to complete the telephone screening. They are also of different cohorts, as based
on the mean ages. The Study I sample was born between 1891 to 1935 (most born from 1920 to
1930) and the Study II sample approximately 1937 to 1948. The two samples are from different
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 61
cultures where the practices of medicine may be different, so the indications for ovarian removal
and administration of hormone therapy may differ, leading to variances in the detection of cognitive
effects. The rates of oophorectomy in the two samples also differed. In Sample I, 1.99% of the
sample had bilateral oophorectomy as confirmed by medical records, and in Sample II, it was
13.23%. Neither rate of oophorectomy likely represents its respective population. This was especially
the case for Study I, where there were women who were considered to have intact ovaries at the
time of cognitive screening but who in fact had oophorectomy later.
Limitations and strengths
This study has a few limitations. For Study I, the data from the IDR were assumed to be
accurate, comprehensive, and inclusive and allowed all oophorectomy surgeries to be coded. The
earliest date recorded for any surgery is December 31, 1934, and for bilateral oophorectomy is July
9, 1964, with 4 oophorectomies being recorded prior to 1971 (and the youngest of women who had
recorded oophorectomy was in her 40s). If the youngest women were born in the mid-1930s, then
the earliest age any women could have been recorded with an oophorectomy was in her 30s (and the
earliest age of surgery recorded for this sample is after 40). There is a slight possibility that
oophorectomy in women earlier than in her 40s or before 1970s may have been missed in the study,
although this would only affect the few who were born earliest in the cohort. Furthermore, although
a few of the procedures recorded were performed as an outpatient (0 days of inpatient stays), it is
likely that many laparoscopic oophorectomies were performed as outpatient procedures, and
therefore not included in the IDR. There is some evidence to support this, including the remarkable
difference in self-report and IDR record rates. However, the lack of remarkable differences in the
number of surgeries by periods suggests that it may not be an issue of comprehensiveness of the
IDR during the time periods for which records exist. For Study II, although participants were
representative of an ethnically-diverse population, they self-selected to participate and were healthy,
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 62
which may not be representative of a general population. Also, with technology advances in
medicine, it is possible that more recent oophorectomy procedures are different than earlier ones in
terms of safety or invasiveness. Therefore there may be variability in the surgery such that the
indications and post-surgical experiences differ for the women who had oophorectomy earlier or
later, and more detailed health information of the women (other than cardiovascular disease and
cancer) prior to surgery may have been useful. The low number of women who had oophorectomy
may also have also been a limitation compared to other studies with larger samples (Rocca, et al.,
2007). Ideally, having more information or corroborating data regarding hormone therapy may have
also been helpful.
The strengths of this dissertation are that cognitive functioning and history of bilateral
oophorectomy in older women were closely examined while adjusting for important demographic
and medical variables, including BMI category and cardiovascular history, which have not been
consistently applied in other studies. Examination also included the factors that are hypothesized to
be relevant in cognitive performance including timing of the surgery, history of hormone therapy,
and length of hormone therapy. We examined some factors of time including the timing of
oophorectomy with respect to menopause stage, age at cognitive testing, age of last menses, and the
timing of hormone therapy. Both studies verified oophorectomy using medical records. Finally,
depressive symptoms as a mediating factor was also considered in the analyses. As a whole, the two
samples were complementary.
Clinical implications
The results from this dissertation continue to support the literature that there are
associations with ovarian removal, hormone therapy, and cognitive functioning, although
inconsistent. These findings are encouraged to be considered by providers and women considering
oophorectomy. Specifically, bilateral oophorectomy that leads to surgical menopause may be
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 63
associated with lower verbal learning domain performance and may eventually affect visual memory
over time, no matter the age. However, and perhaps fortunately, it may not directly result long-term
in cognitive impairment. Thus, these findings may be of some relief to women who decide to have
the surgery. Providers may encourage women to monitor their cognitive function after surgery and
report issues. Providers may want to consider the results that did not find that use of hormone
therapy was a risk factor for worse cognitive outcomes and consider the timing of its administration
as described by others (Rocca, Grossardt, & Shuster, 2011).
Future research
One reason oophorectomy was studied in this dissertation was that it provided a model that
could be used to understand how estrogen may be associated with cognitive function in older adults.
To continue this line of research, there are several areas that may be beneficial to explore further.
One area is to investigate further cognitive function and details of hormone therapy administration.
For example, examining the specific components of hormone therapy, which may include only
estrogen or estrogen and progesterone, patterns (continuous versus cyclical), dose, and routes of
administration may be beneficial, as some studies suggest that different compounds may be
associated with varying cognitive domain changes (Sherwin, 1988; Shumaker, et al., 2003; Shumaker,
et al., 2004). It may also be worthwhile to extend other factors to study to include reproductive
events, such as parity, which has been included in other studies examining the relation between
estrogen and cognitive function (Dunkin, et al., 2005). Exploring the role of depressive symptoms in
the relation between estrogen and cognitive function may be interesting. Identifying which of the
depressive factors, such as the ones identified for CES-D (affect, well-being, somatic, and
interpersonal; Hertzog, et al., 1990) may help describe the role of depressive symptoms in the
relation between estrogen and cognitive function. Given the increase of both depressive symptoms
and decline in cognitive function in older adults, it may be also interesting to study which and how,
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 64
if at all, depressive symptoms change longitudinally pre- and post-menopause. Finally, it may be
worth continuing to examine other factors (genetic or non-genetic) that may interact with
oophorectomy status and its association with cognitive function (Rocca, Grossardt, & Shuster,
2011). These studies will hopefully provide a better understanding of the modifiable factors that are
protective in cognitive function in older adults.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 65
References
American College of Obstetricians and Gynecologists. (2008). ACOG practice bulletin No. 89.
Elective and risk-reducing salpingo-oophorectomy. Obstetrics and Gynecology, 111, 231-241.
doi: 10.1097/01.AOG.0000291580.39618.cb
Amaducci, L. A., Fratiglioni, L., Rocca, W. A., Fieschi, C., Livrea, P., Pedone, D.,...Bino, G. (1986).
Risk factors for clinically diagnosed Alzheimer's disease: A case-control study of an Italian
population. Neurology, 36, 922-931. doi: 10.1097/00002093-198701020-00006
Asante, A., Whiteman, M. K., Kulkarni, A., Cox, S., Marchbanks, P. A., & Jamieson, D. J. (2010).
Elective oophorectomy in the united states: Trends and in-hospital complications, 1998–
2006. Obstetrics & Gynecology, 116, 1088-1095. doi: 10.1097/AOG.0b013e3181f5ec9d
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51, 1173-1182. doi: 10.1037/0022-3514.51.6.1173
Barrett-Connor, E., & Kritz-Silverstein, D. (1993). Estrogen replacement therapy and cognitive
function in older women. Journal of the American Medical Association, 269, 2637-2641. doi:
10.1001/jama.1993.03500200051032
Barrett-Connor, E., & Goodman-Gruen, D. (1999). Cognitive function and endogenous sex
hormones in older women. Journal of the American Geriatrics Society, 47, 1289-1293. Retrieved
from http://www.mdconsult.com.libproxy.usc.edu/das/article/body/229778328-
2/jorg=journal&source=&sp=11116109&sid=0/N/159430/1.html?issn=0002-8614
Blessed, G., Tomlinson, B. E., & Roth, M. (1968). The association between quantitative measures of
dementia and of senile changes in the cerebral grey matter of elderly subjects. British Journal of
Psychiatry, 114, 797-811. doi: 10.1192/bjp.114.512.797
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 66
Brenner, D. E., Kukull, W. A., Stergachis, A., Van Belle, G., Bowen, J. D., McCormick, W.
C.,...Larson, E. B. (1994). Postmenopausal estrogen replacement therapy and the risk of
Alzheimer's disease: A population-based case-control study. American Journal of Epidemiology,
140, 262-267. Retrieved from http://aje.oxfordjournals.org/content/140/3/262.abstract
Carroll, J. C., Rosario, E. R., Chang, L., Stanczyk, F. Z., Oddo, S., LaFerla, F. M., & Pike, C. J.
(2007). Progesterone and estrogen regulate Alzheimer-like neuropathology in female 3xTg-
AD mice. Journal of Neuroscience, 27, 13357-13365. doi: 10.1523/JNEUROSCI.2718-07.2007
Costa, A., Nappi, R. E., Sinforiani, E., Bono, G., Poma, A., & Nappi, G. (1997). Cognitive function
at menopause: Neuroendocrine implications for the study of the aging brain. Functional
Neurology, 12, 175 – 180. Retrieved from
http://www.researchgate.net/publication/14000540_Cognitive_function_at_menopause_ne
uroendocrine_implications_for_the_study_of_the_aging_brain
Cournot, M., Marquié, J. C., Ansiau, D., Martinaud, C., Fonds, H., Ferrières, J., et al. (2006). Relation
between body mass index and cognitive function in healthy middle-aged men and women.
Neurology, 67, 1208-1214.
Dahl, A., Hassing, L. B., Fransson, F., Berg, S., Gatz, M., Reynolds, C.A , et al. (2010). Being
overweight in midlife is associated with lower cognitive ability and steeper cognitive decline
in late life. Journal of Gerontology: Medical Sciences, 65, 57-62. doi:10.1093/gerona/glp035
Dunkin, J., Rasgon, N., Wagner-Steh, K., David, S., Altshuler, L, & Rapkin, A. (2005). Reproductive
events modify the effects of estrogen replacement therapy on cognition in healthy
postmenopausal women. Psychoneuroendocrinology, 30, 284 – 296. doi:
10.1016/j.psyneuen.2004.09.002
Espeland, M. A., Rapp, S. R., Shumaker, S. A., Brunner, R., Sherwin, B. B., Hsia, J., et al. (2004).
Conjugated Equine Estrogens and Global Cognitive Function in Postmenopausal Women:
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 67
Women’s Health Initiative Memory Study. Journal of the American Medical Association, 291, 2959
– 2968.
Farquhar, C. M., Sadler, L., Harvey, S. A., & Stewart, A. W. (2005). The association of hysterectomy
and menopause: A prospective cohort study. British Journal of Obstetrics and Gynaecology, 112,
956-962. doi: 10.1111/ j.1471-0528.2005.00696.x
Farrag, A. K., Khedr, E. M., Abdel-Aleem, H., & Rageh, T. A. (2002). Effect of surgical menopause
on cognitive functions. Dementia and Geriatric Cognitive Disorders, 13, 193-198. doi:
10.1159/000048652
File, S. E., Heard, J. E., & Rymer, J. (2002). Trough oestradiol levels associated with cognitive
impairment in post-menopausal women after 10 years of oestradiol implants.
Psychopharmacologia, 161(1), 107-112. doi: 10.1007/s00213-002-1032-3
Gatto, N. M., Henderson, V. W., St. John, J. A., McCleary, C., Hodis, H. N. , & Mack, W. J. (2008) .
Metabolic syndrome and cognitive function in healthy middle-aged and older adults without
diabetes. Aging, Neuropsychology, and Cognition, 15(5), 627-641. doi:
10.1080/13825580802036936
Gatto, N. M., Henderon, V. W., St. John, J. A., McCleary, C., Detrano, R., Hodis, H. N., & Mack,
W. J. (2009). Subclinical atherosclerosis is weakly associated with lower cognitive function in
healthy hyperhomocysteinemic adults without clinical cardiovascular disease. International
Journal of Geriatriatric Psychiatry, 24, 390–399. doi: 10.1002/gps.2134
Gatz, M., Reynolds, C., John, R., Johansson, B., Mortimer, J. A., & Pedersen, N. L. (2002).
Telephone screening to identify potential dementia cases in a population-based sample of
older adults. International Psychogeriatrics, 14, 273-289. doi: 10.1017/S1041610202008475
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 68
Gatz, M., Reynolds, C., Nikolic, J., Lowe, B., Karel, M., & Pedersen, N. (1995). An empirical test of
telephone screening to identify potential dementia cases. International Psychogeriatrics, 7, 429-
438. doi: 10.1017/S1041610295002171
Greendale, G. A., Wight, R. G., Huang, M.-H., Avis, N., Gold, E. B., Joffe, H, et al. (2010).
Menopause-associated symptoms and cognitive performance: results from the Study of
Women’s Health Across the Nation. American Journal of Epidemiology, 171, 1214-1224. doi:
10.1093/aje/kwq067
Gustafson, D. R., Backman, K., Waern, M., Östling, S., Guo, X., Zandi, P. et al. (2009). Adiposity
indicators and dementia over 32 years in Sweden. Neurology, 73, 1559-1556.
Hammar, M., Berg, G., Fahraeus, L., Larsson-Cohn. (1984). Climacteric symptoms in an unselected
sample of Swedish women. Maturitas, 6, 345 – 350.
Harman, S. M., Naftolin, F., Brinton, E. A., & Judelson, D. R. (2005). Is the estrogen controversy
over? deconstructing the women's health initiative study: A critical evaluation of the
evidence. Annals of the New York Academy of Sciences, 1052(1), 43-56. doi:
10.1196/annals.1347.004
Heflin, L. H., Meyerowitz, B. E., Hall, P., Lichtenstein, P., Johansson, B., Pedersen, N. L. et al.
(2005). Cancer as a risk factor for long-term cognitive deficits and dementia. Journal of the
National Cancer Institute, 97, 854-856. doi: 10.1093/jnci/dji137
Henderson, V. W., & Sherwin, B. B. (2007). Surgical versus natural menopause: cognitive issues.
Menopause, 14, 572- 579. doi: 10.1097/gme.0b013e31803df49c
Henderson, V. W., St. John, J. A., Hodis, H. N., Kono, N., McCleary, C. A., Franke, A. A., et al.
(2012). Long-term soy isoflavone supplementation and cognition in women: a randomized,
controlled trial. Neurology. 78(23), 1841-1848. doi: 10.1212/WNL.0b013e318258f822
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 69
Henderson, V. W., St. John, J. A., Hodis, H. N.,McCleary, C. A., Stanczyk, F. Z., Karim, R., et al.
(2013). Cognition, mood, and physiological concentrations of sex hormones in the early and
late postmenopause. Proceedings of the National Academy of Sciences of the United States of America,
110, 20290-20295. doi: 10.1073/pnas.1312353110
Hertzog, C., van Alstine, J., Usala, P.D., Hultsch, D. F., & Dixon, R. (1990). Measurement
properties of the Center for Epidemiological Studies Depression Scale (CES-D) in older
populations. Psychological Assessment, 2, 64 – 72.
Hodis, H., Mack, W. J., Dustin, L., Mahrer, P. R., Azen, S. P., Detrano, R., et al. (2009). High-Dose
B Vitamin supplementation and progression of subclinical atherosclerosis. A randomized
controlled trial. Stroke. doi: 10.1161/STROKEAHA.108.526798
Hogervorst, E., Williams, J., Budge, M., Riedel, W., & Jolles, J. (2000). The nature of the effect of
female gonadal hormone replacement therapy on cognitive function in post-menopausal
women: A meta-analysis. Neuroscience, 101, 485-512. doi: 10.1016/S0306-4522(00)00410-3
Hogervorst, E., & Bandelow, S. (2010). Sex steroids to maintain cognitive function in women after
the menopause: A meta-analyses of treatment trials. Maturitas, 66, 56-71. doi:
10.1016/j.maturitas.2010.02.005
Hughes, T., Borenstein, A., Schofield, E., Wu, Y., & Larson, E. (2009). Association between late-life
body mass index and dementia: The kame project. Neurology, 72, 1741-1746. doi:
10.1212/WNL.0b013e3181a60a58
Juva, K., Makela, M., Erkinjuntii, T., Sulkava, R., Yukoski, R., Valvanne, J. et al. (1997). Functional
assessment scales in detecting dementia. Age and Aging, 26, 393 - 400. doi:
10.1093/ageing/26.5.393
Kahn, R. L., Goldfarb, A. I., Pollack, M., & Peck, A. (1960). Brief objective measures for the
determination of mental status in the aged. The American Journal of Psychiatry, 117, 326 - 328.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 70
Retrieved from
http://ajp.psychiatryonline.org/article.aspx?Volume=117&page=326&journalID=13
Kahn, R. L., Pollack, M., & Goldfarb, A. I. (1961). Factors related to individual differences in mental
status of institutionalized aged. Psychopathology of aging. New York: Grune & Stratton, 104-13.
Keshavarz, H., Hillis, S. D., Kieke, B. A., & Marchbanks, P. A. (2002). Hysterectomy surveillance—
United states, 1994–1999. Morbidity and Mortality Weekly Report, 51(5505), 1-8. Retrieved from
http://www.cdc.gov/mmwr/preview/mmwrhtml/ss5105a1.htm
Kohout, F. J., Berkman, L. F., Evans, D. A., & Cornoni-Huntley, J. (1993). Two shorter forms of
the CES-D depression symptoms index. Journal of Aging and Health, 5, 179-193. doi:
10.1177/089826439300500202
Kok, H. S., Kuh, D., Cooper, R., van der Schouw, Y. T., Grobbee, D. E., Wadsworth, M. E. J., &
Richards, M. (2006). Cognitive function across the life course and the menopausal transition
in a British birth cohort. Menopause, 13(1), 19-27. doi: 10.1097/01.gme.0000196592.36711.a0
Korse, C. M., Bonfrer, J. M. G., van Beurden, M., Verheijen, R. H. M., & Rookus, M. A. (2009).
Estradiol and testosterone levels are lower after oophorectomy than after natural
menopause. Tumor Biology, 30(1), 37-42. doi: 10.1159/000199449
Kritz ‐Silverstein, D., & Barrett ‐Connor, E. (2002). Hysterectomy, oophorectomy, and cognitive
function in older women. Journal of the American Geriatrics Society, 50(1), 55-61. doi:
10.1046/j.1532-5415.2002.50008.x
Kurita, K., Gatz, M., & Mack, W. (2014, unpublished). Cognitive functioning in women who had
oophorectomy: A meta-analytic review. Manuscript in preparation.
Kurita, K., Meyerowitz, B. E., Hall, P., & Gatz, M. (2011). Long-term cognitive impairment in older
adult twins discordant for gynecologic cancer treatment. Journal of Gerontology: Medical Sciences,
6A, 1343–1349. doi: 10.1093/gerona/glr140
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 71
Li, C., Samsioe, G., Borgfeldt, C., Lidfeldt, J., Agardh, C. D., & Nerbrand, C. (2003). Menopause-
related symptoms: What are the background factors? A prospective population-based cohort
study of Swedish women (The Women’s Health in Lund Area study). American Journal of
Obstetrics and Gynecology, 189, 1646 – 1653. doi:10.1016/S0002-9378(03)00872-X
Lichtenstein, P., de Faire, U., Floderus, B., Svartengren, M., Svedberg, P., & Pederson, N. L. (2002).
The Swedish Twin Registry: a unique resource for clinical, epidemiological and genetic
studies. Journal of Internal Medicine, 252, 184-205. doi: 10.1046/j.1365-2796.2002.01032.x\
Lichtenstein, P., Sullivan, P. F., Cnattingius, S., Gatz, M., Johansson, S., Carlström, E., Björk, C.,
Svartengren, M., Wolk, A., Klareskog, L., de Faire, U., Schalling, M., Palmgren, J., &
Pedersen, N. L. (2006). The Swedish Twin Registry in the Third Millennium: An Update.
Twin Research and Human Genetics, 9(6), 875–882. doi: dx.doi.org/10.1375/twin.9.6.875
Lobo, R. A. (2007). Surgical menopause and cardiovascular risks. Menopause, 14, 562 – 566. doi:
10.1097/gme.0b013e318038d333
Low, L-F. & Anstey, K. J. (2006). Hormone replacement therapy and cognitive performance in
postmenopausal women—a review by cognitive domain. Neuroscience and Biobehavioral
Reviews, 30, 66-84. doi:10.1016/j.neubiorev.2005.05.003
Maki, P. M. (2005). A systematic review of clinical trials of hormone therapy on cognitive function:
Effects of age at initiation and progestin use. Annals of the New York Academy of Sciences, 1052,
182-197. doi: 10.1196/annals.1347.012
Margolis, K. L., Mucci, L., Braaten, T., Kumle, M., Lagerros, Y. T., Adami, H-O., et al. (2005).
Physical activity in different periods of life and the risk of breast cancer: The Norwegian-
Swedish Women's Lifestyle and Health Cohort Study. Cancer Epidemiology, Biomarkers &
Prevention, 14, 27-32.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 72
McEwen, B. (2002). Estrogen actions throughout the brain. Recent Progress in Hormone
Research, 57, 357-384. Retrieved from
http://rphr.endojournals.org/cgi/content/short/57/1/357
McEwen, B. S. & Woolley, C. S. (1994). Estrodial and progesterone regulate neuronal structure and
synaptic connectivity in adult as well as developing brain. Experimental Gerontology, 29, 431 –
436.
McLay, R. N., Maki, P. M., & Lyketsos, C. G. (2003). Nulliparity and late menopause are associated
with decreased cognitive decline. Journal of Neuropsychiatry and Clinical Neurosciences, 15, 161-
167. doi: 10.1176/appi.neuropsych.15.2.161
McTiernan, A., Wu, L. L., Barnabei, V. M., Chen, C., Hendrix, S., Modugno, F., et al. (2008).
Relation of demographic factors, menstrual history, reproduction and medication use to sex
hormone levels in postmenopausal women. Breast Cancer Research and Treatment, 108, 217-231.
doi 10.1007/s10549-007-9588-6
Miller, K. J., Conney, J. C., Rasgon, N. L., Fairbanks, L.A., & Small, G. W. (2002). Mood symptoms
and cognitive performance in women estrogen users and nonusers and men. Journal of the
American Geriatric Society, 50, 1826-1830. doi: 10.1046/j.1532-5415.2002.50511.x
Mortel, K. F., & Meyer, J. S. (1995). Lack of postmenopausal estrogen replacement therapy and the
risk of dementia. Journal of Neuropsychiatry and Clinical Neurosciences, 7, 334 -337.
Mucci, L. A., Hsieh, C-C., Williams, P. L., Dickman, P. W., Björkman, L., & Pedersen, N. L. (2004).
Birth order, sibship size, and housing density in relation to tooth loss and periodontal
disease: A cohort study among Swedish twins. American Journal of Epidemiology, 159, 499-506.
doi: 10.1093/aje/kwh063
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 73
Nappi, R., Sinforiani, E., Mauri, M., Bono, G., Polatti, F., & Nappi, G. (2000). Memory functioning
at menopause: Impact of age in ovariectomized women. Gynecologic and Obstetric Investigation,
47(1), 29-36. doi: 10.1159/000010058
Nathorst-Böös, J., Von Schoultz, B., & Carlström, K. (1993). Elective ovarian removal and estrogen
replacement therapy—effects on sexual life, psychological well-being and androgen status.
Journal of Psychosomatic Obstetrics & Gynecology, 14, 283-293. doi: 10.3109/01674829309084451
Newman, M. F., Kirchner, J. L., Phillips-Bute, B., Gaver, V., Grocott, H., Jones, R. H., et al. (2001).
Longitudinal assessment of neurocognitive function after coronary-artery bypass surgery.
New England Journal of Medicine, 344, 395-402. doi: 10.1056/NEJM200102083440601
Nourhashemi, F., Deschamps, V., Larrieu, S, Letenneur, L., Dartigues, J.-F., & Barberger-Gateau, P.
(2003). Body mass index and incidence of dementia: The PAQUID study. Neurology, 60, 117-
119.
Paganini-Hill, A., & Henderson, V. W. (1994). Estrogen deficiency and risk of Alzheimer’s disease
in women. American Journal of Epidemiology, 140, 256-261. Retrieved from
http://aje.oxfordjournals.org/content/140/3/256.abstract
Parazzini, F., Negri, E., La Vecchia, C., Luchini, L., & Mezzopane, R. (1993). Hysterectomy,
oophorectomy, and subsequent ovarian cancer. Obstetrics & Gynecology, 81, 363-366.
Retrieved from
http://journals.lww.com/greenjournal/Abstract/1993/03000/Hysterectomy,_Oophorecto
my,_and_Subsequent_Ovarian.10.aspx
Parihar, M. (2003). Obesity and infertility. Reviews in Gynaecological Practice, 3(3), 120-126. doi:
10.1016/S1471-7697(03)00061-3
Parker, W. H., Jacoby, V., Shoupe, D., & Rocca, W. (2009). Effect of bilateral oophorectomy on
women’s long-term health. Women’s Health, 5, 565-576. doi: 10.2217/whe.09.42
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 74
Phillips, S. M., & Sherwin, B. B. (1992). Effects of estrogen on memory function in surgically
menopausal women. Psychoneuroendocrinology, 17, 485-495. doi: 10.1016/0306-4530(92)90007-
T
Phipps, A. I. & Buist, D. S. M. (2009). Validation of self-reported history of hysterectomy and
oophorectomy among women in an integrated group practice setting. Menopause, 16, 576-581.
Phung, T. K. T., Waltoft, B. L., Laursen, T. M., Settnes, A., Kessing, L. V., Mortensen, P. B., &
Waldemar, G. (2010). Hysterectomy, oophorectomy and risk of dementia: A nationwide
historical cohort study. Dementia and Geriatric Cognitive Disorders, 30(1), 43-50. doi:
10.1159/000314681
Polo-Kantola, P., Portin, R., Polo, O., Helenius, H., Irjala, K., & Erkkola, R. (1998). The effect of
short-term estrogen replacement therapy on cognition: A randomized, double-blind, cross-
over trial in postmenopausal women. Obstetrics & Gynecology, 91, 459-466. Retrieved from
http://journals.lww.com/greenjournal/Abstract/1998/03000/The_Effect_of_Short_Term_
Estrogen_Replacement.26.aspx
Preacher K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects in
simple mediation models. Behavior Research Methods, Instruments, & Computers, 36(4), 717-731.
Retrieved from http://www.mendeley.com/research/spss-sas-procedures-estimating-
indirect-effects-simple-mediation-models-5/
Radloff, L. S. (1977). The CES-D scale: A self report depression scale for research in the general.
Applied Psychological Measurement, 1, 385-401. doi: 10.1177/014662167700100306
Rasgon, N. L., Magnusson, C., Johansson, A. L. V., Pedersen, N. L., Elmand, S., & Gatz, M. (2005).
Endogenous and exogenous hormone exposure and risk of cognitive impairment in Swedish
twins: A preliminary study. Psychoneuroendocrinology, 30, 558–567. doi:
10.1016/j.psyneuen.2005.01.004
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 75
Rasgon, N. L., Shelton, S., & Halbreich, U. (2005). Perimenopausal mental disorders: epidemiology
and phenomenology. CNS Spectrum, 10, 471-478. Retrieved from
http://www.cnsspectrums.com/aspx/articledetail.aspx?articleid=354
Rebbeck, T.R., Levin, A. M., Eisen, A., Snyder, C., Watson, P., Cannon-Albright, L., Isaacs, C.,
Olopade, O., Garber, J. E.,. Godwin, A. K., Daly, M. B., Narod, S. A., Neuhausen, S. L.,
Lynch, H. T., & Weber, B. L. (1999). Breast cancer risk after bilateral prophylactic
oophorectomy in BRCA1 mutation carriers. Journal of the National Cancer Institute, 91(17),
1475-1479. doi: 10.1093/jnci/91.17.1475
Rice, M. M. Graves, A. B., PhD; McCurry, S. M., Gibbons, L. E., Bowen, J.D., McCormick, W. C.,
& Larson, E. B. (2000). Postmenopausal estrogen and estrogen-progestin use and 2-year rate
of cognitive change in a cohort of older Japanese American women: The Kame Project.
Archives of Internal Medicine, 160, 1641 – 1649. Retrieved from http://archinte.ama-
assn.org/cgi/content/abstract/160/11/1641
Rocca, W., Bower, J., Maraganore, D., Ahlskog, J., Grossardt, B., De Andrade, M., & Melton III, L.
(2007). Increased risk of cognitive impairment or dementia in women who underwent
oophorectomy before menopause. Neurology, 69, 1074-1083. doi:
10.1212/01.wnl.0000276984.19542.e6
Rocca, W. A., Grossardt, B. R., Geda, Y. E., Gostout, B. S., Bower, J. H., Maraganore, D.
M.,...Melton III, L. J. (2008). Long-term risk of depressive and anxiety symptoms after early
bilateral oophorectomy. Menopause, 15, 1050-1059. doi: 10.1097/gme.0b013e318174f155
Rocca, W. A., Grossardt, B. R., & Maraganore, D. M. (2008). The long-term effects of
oophorectomy on cognitive and motor aging are age dependent. Neurodegenerative Diseases, 5,
257-260. doi: 10.1159/000113718
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 76
Rocca, W. A., Grossardt, B. R., & Shuster, L. T. (2011). Oophorectomy, menopause, estrogen
treatment, and cognitive aging: Clinical evidence for a window of opportunity. Brain Research,
1379, 188-198. doi:10.1016/j.brainres.2010.10.031
Rosenberg, P. B., Mielke, M. M., Xue, Q.-L., & Carlson, M. C. (2010). Depressive symptoms predict
incident cognitive impairment in cognitive healthy older women. American Journal of
Geriatric Psychiatry, 18, 204-211. doi:10.1097/JGP.0b013e3181c53487
Schiff, R., Bulpitt, C. J., Wesnes, K. A., & Rajkumar, C. (2005). Short-term transdermal estradiol
therapy, cognition and depressive symptoms in healthy older women. A randomised placebo
controlled pilot cross-over study. Psychoneuroendocrinology, 30, 309 – 315.
doi:10.1016/j.psyneuen.2004.08.007
Sherwin, B. B. (1988). Estrogen and/or androgen replacement therapy and cognitive functioning in
surgically menopausal women. Psychoendocrinology, 13, 345-357.
Sherwin, B. B. (2005). Estrogen and memory in women: How can we reconcile the findings?
Hormones and Behavior, 47, 371-375. doi: 10.1016/j.yhbeh.2004.12.002
Sherwin, B. B. (2009). Estrogen therapy: Is time of initiation critical for neuroprotection? Nature
Reviews Endocrinology, 5, 620-627. doi: 10.1038/nrendo.2009.193
Sherwin, B. B., & Phillips, S. (1990). Estrogen and cognitive functioning in surgically menopausal
women. Annals of the New York Academy of Sciences, 592, 474-475. doi: 10.1111/j.1749-
6632.1990.tb30379.x
Shoupe, D., Parker, W. H., Broder, M. S., Liu, Z., Farquhar, C., & Berek, J. S. (2007). Elective
oophorectomy for benign gynecological disorders. Menopause, 14, 580-585. doi:
10.1097/gme.0b013e31803c56a4
Shumaker, S. A., Legault, C., Kuller, L., Rapp, S. R., Thal, L., Lane, D. S., et al. (2004). Conjugated
equine estrogens and incidence of probable dementia and mild cognitive impairment in
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 77
postmenopausal women: Women's health initiative memory study. Journal of the American
Medical Association, 291, 2947-2958. doi: 10.1001/jama.291.24.2947
Shumaker, S. A., Legault, C., Rapp, S. R., Thal, L., Wallace, R. B., Ockene, J. K., et al. (2003).
Estrogen plus progestin and the incidence of dementia and mild cognitive impairment in
postmenopausal women: The women's health initiative memory study: A randomized
controlled trial. Journal of the American Medical Association, 289, 2651-2662. doi:
10.1001/jama.289.20.2651
Simpson, E., Rubin, G., Clyne, C., Robertson, K., O'Donnell, L., Jones, M., & Davis, S. (2000). The
role of local estrogen biosynthesis in males and females. Trends in Endocrinology and Metabolism,
11, 184-188. doi: 10.1016/S1043-2760(00)00254-X
Sohrabji, F. (2005). Estrogen: A neuroprotective or proinflammatory hormone? Emerging evidence
from reproductive aging models. Annals of the New York Academy of Sciences, 1052(1), 75-90.
doi: 10.1196/annals.1347.006
Sohrabji, F., & Bake, S. (2006). Age-related changes in neuroprotection. Endocrine, 29, 191-197. doi:
10.1385/ENDO:29:2:191
Szklo, M., Cerhan, J., Diez-Roux, A. V., Chambless, L., Cooper, L., Folsom, A. R.,...Nieto, F. J.
(1996). Estrogen replacement therapy and cognitive functioning in the atherosclerosis risk in
communities (ARIC) study. American Journal of Epidemiology, 144, 1048-1057. Retrieved from
http://aje.oxfordjournals.org/content/144/11/1048.abstract
van Dijk, D., Keizer, A., Diephuis, J. C., Durand, C., Vos, L. J., & Hijman, R. (2000).
Neurocognitive dysfunction after coronary artery bypass surgery: A systematic review. The
Journal of Thoracic and Cardiovascular Surgery, 120, 632-639. doi: 10.1067/mtc.2000.108901
Vearncombe, K. J., & Pachana, N. A. (2009). Is cognitive functioning detrimentally affected after
early, induced menopause? Menopause, 16, 188-198. doi: 10.1097/gme.0b013e3181775eb4
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 78
Verghese, J., Kuslansky, G., Katz, M. J., Sliwinski, M., Crystal, H. A., Buschke, H., & Lipton, R. B.
(2000). Cognitive performance in surgically menopausal women on estrogen. Neurology, 55,
872-874. Retrieved from http://www.neurology.org/content/55/6/872.abstract
Walf, A. A., & Frye, C. A. (2006). A review and update of mechanisms of estrogen in the
hippocampus and amygdala for anxiety and depression behavior. Neuropsychopharmacology, 31,
1097-1111. doi:10.1038/sj.npp.1301067
Weiderpass, E., Baron, J.A., Adami, H.O., Magnusson, C., Lindgren, A., Bergström, R., Correia, N.,
& Persson, I. (1999). Low-potency oestrogen and risk of endometrial cancer: a case-control
study. Lancet, 353, 1824-1828.
Whitmer, R.A., Quesenberry, C. P., Zhou, J., & Yaffe, K. (2011). Timing of hormone therapy and
dementia: The Critical Window Theory revisited. Annals of Neurology, 69, 163-169. doi:
10.1002/ana.22239
Wolf, O. T., & Kirschbaum, C. (2002). Endogenous estradiol and testosterone levels are associated
with cognitive performance in older women and men. Hormones and Behavior, 41, 259-266.
doi: 10.1006/hbeh.2002.1770
Yaffe, K., Haan, M., Byers, A., Tangen, C., & Kuller, L. (2000). Estrogen use, APOE, and cognitive
decline: Evidence of gene-environment interaction. Neurology, 54, 1949-1953. Retrieved from
http://www.neurology.org/content/54/10/1949.full
Yaffe, K., Grady, D., Pressman, A., & Cummings, S. (1998). Serum estrogen levels, cognitive
performance, and risk of cognitive decline in older community women. Journal of the American
Geriatrics Society, 46, 816-821. Retrieved from
http://www.mdconsult.com.libproxy.usc.edu/das/article/body/229778328-
2/jorg=journal&source=&sp=10322676&sid=0/N/115725/1.html?issn=0002-8614
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 79
Yang, X-P, & Reckelhoff, J. F. (2011). Estrogen, hormonal replacement therapy and cardiovascular
disease. Current Opinion in Nephrology and Hypertension, 20(2), 133–138. doi:
10.1097/MNH.0b013e3283431921
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 80
STUDY 1 RESULTS TABLES
Table 1
Descriptives of the Sample
Group Means (Standard Deviations); n
Women with Both
Ovaries Removed
Women with Both
Ovaries Intact
Age at Cognitive Screening, in years** 72.45 (5.58); 154 74.23 (7.00); 7,600
Years of Education* 8.88 (3.26); 154 8.24 (2.75); 7,574
Depressive Symptoms (short-form CES-D Score) 4.08 (3.90); 138 4.20 (4.55); 6,465
Age of Last Menstrual Period 49.39 (4.67); 135 49.06 (5.11); 6,023
Years on Hormones, for those on therapy 5.17 (6.13); 46 5.35 (6.64); 1,276
Age of Bilateral Oophorectomy 58.94 (8.55); 154 not applicable
Cognitive Screening Score (TELE Score) 15.62 (2.06); 146 15.38 (2.51); 6,799
Women with Cognitive
Impairment
Women without
Cognitive Impairment
Age at Cognitive Screening, in years** 80.55 (7.41); 1,072 73.18 (6.35); 6,682
Years of Education** 7.21 (1.99); 1,054 8.41 (2.83); 6,674
Depressive Symptoms (short-form CES-D Score)** 5.57 (5.23); 471 4.10 (4.46); 6,132
Age of Last Menstrual Period 48.79 (7.04); 363 49.08 (4.95); 5,795
Years on Hormones, for those on therapy** 4.21 (6.17); 39 5.38 (6.63); 1,283
Age of Bilateral Oophorectomy 62.01 (10.35); 13 58.65 (8.35); 141
Cognitive Screening Score (TELE Score) 10.55 (2.86); 551 15.80 (1.98); 6,394
Twins with Both
Ovaries Removed
Co-twins with Both
Ovaries Intact
Age at Cognitive Screening, in years 70.81 (4.84); 63 70.80 (4.85); 63
Years of Education* 9.36 (3.42); 63 8.56 (2.55); 63
Depressive Symptoms (short-form CES-D Score) 3.42 (3.66); 57 3.06 (3.91); 58
Age of Last Menstrual Period 48.97 (4.22); 58 49.96 (5.82); 55
Years on Hormones, for those on therapy 4.29 (6.21); 21 3.57 (6.16); 21
Cognitive Screening Score (TELE Score) 16.22 (1.91); 62 16.06 (1.82); 61
Note. Significant differences in groups with *p < .05, **p < .001
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 81
Table 2
Descriptives of the Sample
Group Frequencies (%)
Women with Both
Ovaries Removed
Women with Both
Ovaries Intact
Timing of oophorectomy
T0: no oophorectomy
T1: timing of oophorectomy relative to last menstrual period
unknown
a
T2: oophorectomy after last menstrual period
T3a: oophorectomy within 1.5 years of last menstrual period,
<= age 50
T3b: oophorectomy within 1.5 years of last menstrual period,
> age 50
27 (17.53%)
97 (62.99%)
15 (9.74%)
15 (9.74%)
7,600 (100.00%)
Cognitive impairment
not cognitively impaired
cognitively impaired
141 (91.56%)
13 (8.44%)
6,541 (86.07%)
1,059 (13.93%)
BMI***
1. underweight (<18.5)
2. normal (>=18.5, <25)
3. overweight (>=25, <30)
4. obese (>=30)
21 (13.64%)
57 (37.01%)
58 (37.66%)
18 (11.69%)
1,697 (22.33%)
3,173 (41.75%)
2,158 (28.39%)
572 (7.53%)
Zygosity of twins
monozygotic
dizygotic
unknown
opposite sex dizygotic
34 (22.08%)
68 (44.16%)
5 (3.25%)
47 (30.52%)
1,888 (24.84%)
3,279 (43.14%)
127 (1.67%)
2,306 (30.34%)
Hormone use, ever***
at least once
never
unknown
47 (30.52%)
82 (53.25%)
25 (16.23%)
1,344 (17.68%)
4,925 (64.80%)
1,331 (17.51%)
Hormone use
first year is year of last menstrual period or earlier
first year is within 3 years of last menstrual period
first year unknown or 3 years after last menstrual period
no hormone history/unknown
6 (3.90%)
4 (2.60%)
37 (24.03%)
107 (69.48%)
85 (1.12%)
65 (.86%)
1,194 (15.71%)
6,256 (82.32%)
Contraceptive pill use
at least once
never
unknown
1 (.65%)
55 (35.71%)
98 (63.64%)
85 (1.12%)
1,471 (19.36%)
6,044 (79.53%)
Hysterectomy***
history of surgery
no history of surgery
49 (31.82%)
105 (68.18%)
512 (6.74%)
7,088 (93.26%)
Cardiovascular issues
self-reported history
no self-reported history
unknown
77 (50.00%)
66 (42.86%)
11 (7.14%)
3,329 (43.80%)
3,404 (44.79%)
867 (11.41%)
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 82
Women with
Cognitive
Impairment
Women without
Cognitive
Impairment
Timing of oophorectomy
T0: no oophorectomy
T1: timing of oophorectomy relative to last menstrual period
unknown
a
**
T2: oophorectomy after last menstrual period**
T3a: oophorectomy within 1.5 years of last menstrual period,
<= age 50
T3b: oophorectomy within 1.5 years of last menstrual period,
> age 50
1,059 (98.79%)
9 (.84%)
3 (.28%)
1 (<.01%)
0 (0.00%)
6,541 (97.89%)
18 (.27%)
94 (1.41%)
14 (.21%)
15 (.22%)
BMI***
1. underweight (<18.5)
2. normal (>=18.5, <25)
3. overweight (>=25, <30)
4. obese (>=30)
729 (68.00%)
196 (18.28%)
108 (10.07%)
39 (3.64%)
989 (14.80%)
3,034 (45.41%)
2,108 (31.55%)
551 (8.25%)
Zygosity of twins***
monozygotic
dizygotic
unknown
opposite sex dizygotic
291 (27.15%)
519 (48.41%)
30 (2.80%)
232 (21.64%)
1,631 (24.41%)
2,828 (42.32%)
102 (1.53%)
2,121 (31.74%)
Hormone use, ever***
at least once
never
unknown
71 (6.62%)
421 (39.27%)
580 (54.10%)
1,860 (27.84%)
4,365 (65.32%)
457 (6.84%)
Hormone use
first year is year of last menstrual period or earlier
first year is within 3 years of last menstrual period
first year unknown or 3 years after last menstrual period
no hormone history/unknown
2 (.19%)
0 (0.00%)
46 (4.29%)
1,024 (95.52%)
89 (1.33%)
69 (1.03%)
1,185 (17.73%)
5,339 (79.9%)
Contraceptive pill use
at least once
never
unknown
3 (.28%)
54 (5.04%)
1,015 (94.68%)
83 (1.24%)
1,472 (22.03%)
5,127 (76.73%)
Hysterectomy***
history of surgery
no history of surgery
41 (3.82%)
1,041 (97.11%)
520 (7.78%)
6,162 (92.22%)
Cardiovascular issues***
self-reported history
no self-reported history
unknown
320 (29.85%)
231 (21.55%)
521 (48.60%)
3,086 (46.18%)
3,239 (48.47%)
357 (5.34%)
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 83
Twins with Both
Ovaries Removed
Co-twins with
Both Ovaries
Intact
BMI
1. underweight (<18.5)
2. normal (>=18.5, <25)
3. overweight (>=25, <30)
4. obese (>=30)
9 (14.29%)
22 (34.92%)
25 (39.68%)
7 (11.11%)
5 (7.94%)
27 (42.86%)
26 (4.27%)
5 (7.94%)
Zygosity of twins
monozygotic
dizygotic
unknown
24 (38.1%)
37 (58.73%)
2 (3.17%)
24 (38.1%)
37 (58.73%)
2 (3.17%)
Hormone use, ever
at least once
never
unknown
28 (44.44%)
31 (49.21%)
4 (6.35%)
21 (33.33%)
38 (60.32%)
4 (6.35%)
Hormone use
first year is year of last menstrual period or earlier
first year is within 3 years of last menstrual period
first year unknown or 3 years after last menstrual period
no hormone history/unknown
1 (1.59%)
1 (1.59%)
1 (1.59%)
60 (95.24%)
0 (0.00%)
0 (0.00%)
7 (11.11%)
56 (88.89%)
Contraceptive pill use
never
unknown
23 (36.51%)
40 (63.49%)
17 (26.98%)
46 (73.02%)
Hysterectomy
history of surgery
unknown
22 (34.92%)
41 (65.08%)
6 (9.52%)
57 (90.48%)
Cardiovascular issues*
self-reported history
no self-reported history
unknown
33 (52.38%)
26 (41.27%)
4 (6.35%)
23 (36.51%)
36 (57.14%)
4 (6.35%)
Note.
a
Includes 8 women who reported that their last menstrual period were after the recorded date of
bilateral oophorectomy.
Significant differences in groups with *p < .05, **p < .01, ***p < .001
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 84
Table 3
Correlations Among Cognitive Function and Potential Covariates for All Participants
r, p, n
1 2 3 4 5 6. 7 8 9 10 11
1. Cognitive — -.31 .23 .11 -.07 .09 .02 .14 <-.01 -.09 .05
screening score <.01 <.01 <.01 <.01 <.01 .10 <.01 .95 <.01 <.01
6,945 6,943 6,945 6,813 6,945 6,158 6,398 1,612 6,603 6,945
2. Age at — -.22 -.31 .15 -.11 -.02 -.18 -.11 .11 -.12
screening, in <.01 <.01 <.01 <.01 .09 <.01 <.01 <.01 <.01
years 7,728 7,754 6,876 7,754 6,158 6,398 1,612 6,603 7,754
3. Years of — .09 -.09 .14 .05 .17 .02 -.03 .03
education <.01 <.01 <.01 .01 <.01 .53 .02 <.01
7,728 6,874 7,728 6,158 6,398 1,612 6,603 7,728
4. BMI — .08 .13 .01 .05 .06 -.02 .07
category <.01 <.01 .56 <.01 .02 .09 <.01
6,876 7,754 6,158 6,398 1,612 6,603 7,754
5. Cardiovascular — -.02 -.04 -.03 -.12 .15 -.03
history .11 <.01 <.01
6,876 6,158 6,395 1,612 6,601 6,876
6. Years on — .03 .57 -.12 .02 .06
hormone .04 <.01 <.01 .12 <.01
therapy 6,158 6,398 1,612 6,603 7,754
7. Age of last — .04 .01 -.02 -.01
menstrual <.01 .81 .18 .39
period 5, 845 1,519 6,047 6,158
8. Use of — -.76 .04 .19
hormones <.01
1,612 6,271 6,398
9. Use of — -.01 -.10
contraceptives .73
1,588 1,612
10. Depressive — -.02
symptoms .05
(short-form 6,603
CES-D score)
11. Hysterectomy —
history
Note. Cardiovascular issues, use of hormones, use of contraceptives, and hysterectomy are dichotomous
variables with values of 0 if there is no history and 1 for history.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 85
Pearson product-moment correlations are presented where 2 variables are continuous, Spearman rank-order
correlations are presented where at least one variable is ordinal (e.g., education and BMI category) and the
other variable is not dichotomous, point biserial correlations are presented where one variable is dichotomous
and the other is continuous, tetrachoric correlations are presented where both variables are dichotomous, and
rank biserial are presented where one variable is ordinal and the other is dichotomous.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 86
Table 4
Models Examining Oophorectomy on Cognitive Impairment (Dichotomous Outcome)
Parameter
Estimate
Standard
Error
95%
Confidence
Limits
p n QIC
Model I: covariates only 6,874 3404.04
intercept
age, in years
years of education
BMI category
cardiovascular history
-7.54
.09
-.12
-.50
.21
.69
.01
.02
.07
.09
-8.88, -6.19
.08, .11
-.16, -.07
-.65, -.36
.02, .39
<.01
<.01
<.01
<.01
.03
Model IIa: Model I + oophorectomy 6,874 3406.07
intercept
age, in years
years of education
BMI category
cardiovascular history
oophorectomy
-7.53
.09
-.12
-.50
.21
-.03
.69
.01
.02
.07
.09
.38
-8.88, -6.19
.08, .11
-.16, -.07
-.65, -.36
.02, .39
-.77, .71
<.01
<.01
<.01
<.01
.03
.93
Model IIb: Model IIa + hormone therapy 6,395 3059.25
intercept
age, in years
years of education
BMI category
cardiovascular history
hormone therapy, ever
oophorectomy
-7.19
.09
-.11
-.42
.28
-.54
.15
.72
.01
.02
.08
.10
.16
.38
-8.61, -5.77
.07, .10
-.15, -.06
-.57, -.27
.08, .48
-.85, -.23
-.60, .89
<.01
<.01
<.01
<.01
<.01
<.01
.70
Model IIc: Model IIa + years on hormone therapy 6,874 3394.39
intercept
age, in years
years of education
BMI category
cardiovascular history
years on hormone therapy
oophorectomy
-7.38
.09
-.11
-.50
.21
-.09
-.01
.69
.01
.02
.07
.09
.04
.38
-8.74, -6.01
.08, .11
-.15, -.07
-.64, -.36
.02, .39
-.17, -.01
.78, .71
<.01
<.01
<.01
<.01
.03
.02
.99
Model IId: Model IIa + years on hormone therapy + interaction 6,874 3395.63
intercept
age, in years
years of education
BMI category
cardiovascular history
years on hormone therapy
years x oophorectomy
oophorectomy
-7.38
.09
-.11
-.50
.21
-.10
.06
-.05
.70
.01
.02
.07
.09
.04
.10
.39
-8.74, -6.01
.08, .11
-.15, -.07
-.64, -.36
.02, .39
-.18, -.01
-.14, .27
-.82, .72
<.01
<.01
<.01
<.01
.03
.02
.54
.89
Model IIIa: Model I + timing of oophorectomy 6,874 3401.87
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 87
Parameter
Estimate
Standard
Error
95%
Confidence
Limits
p n QIC
intercept
age, in years
years of education
BMI category
cardiovascular history
T1
T2
T3
-7.55
.09
-.12
-.50
.21
1.83
-.75
-.31
.69
.01
.02
.07
.09
.64
.62
1.01
-8.91, -6.20
.08, .11
-.16, -.08
-.64, -.36
.02, .39
.58, 3.08
-.20, .46
-2.29, 1.66
<.01
<.01
<.01
<.01
.03
<.01
.22
.76
Model IIIb: Model IIIa + hormone therapy 6,395 3055.77
intercept
age, in years
years of education
BMI category
cardiovascular history
hormone therapy, ever
T1
T2
T3
-7.22
.09
-.11
-.42
.28
-.54
1.90
-.58
-.04
.73
.01
.02
.08
.10
.16
.65
.61
1.03
-8.65, -5.80
.07, .10
-.15, -.06
-.57, -.27
.08, .48
-.85, -.23
.63, 3.17
-1.78, .62
-2.07, 1.99
<.01
<.01
<.01
<.01
.01
<.01
<.01
.34
.97
Model IIIc: Model IIIa + years on hormone therapy 6,874 3389.97
intercept
age, in years
years of education
BMI category
cardiovascular history
years on hormone therapy
T1
T2
T3
-7.40
.09
-.11
-.50
.21
-.09
1.92
-.75
-.21
.70
.01
.02
.07
.09
.04
.66
.61
1.01
-8.77, -6,03
.08, .11
-.15, -.07
-.64, -.35
.02, .39
-.17, -.02
.62, 3.21
-1.95, .46
-2.21, 1.78
<.01
<.01
<.01
<.01
.03
.02
<.01
.23
.83
Model IV: Model IIc + depressive symptoms 6,601 3046.14
intercept
age, in years
years of education
BMI category
cardiovascular history
years on hormone therapy
oophorectomy
depressive symptoms
-7.47
.09
-.11
-.42
.22
-.10
.10
.04
.74
.01
.02
.08
.10
.04
.38
.01
-8.92, -6.02
.07, .10
-.15, -.06
-.57, -.27
.02, .42
-.17, -.01
-.65, .85
.03, .06
<.01
<.01
<.01
<.01
.03
.02
.79
<.01
Model V: Model IIIc + depressive symptoms 6,601 3039.76
intercept
age, in years
years of education
BMI category
cardiovascular history
years on hormone therapy
T1
T2
T3
depressive symptoms
-7.47
.09
-.11
-.42
.23
-.09
2.30
-.63
-.24
.04
.75
.01
.02
.08
.10
.04
.62
.62
1.03
.01
-8.93, -6.01
.07, .10
-.15, -.06
-.57, -.27
.03, .43
-.17, -.01
1.08, 3.52
-1.84, .57
-2.26, 1.79
.03, .06
<.01
<.01
<.01
<.01
.03
.02
<.01
.30
.82
<.01
Model VIa: Model IIa + age of last menstrual period 6,158 2587.91
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 88
Parameter
Estimate
Standard
Error
95%
Confidence
Limits
p n QIC
intercept
age, in years
years of education
BMI category
cardiovascular history
age of last menstrual period
oophorectomy
-7.13
.08
-.11
-.28
.33
-.01
-.09
1.12
.01
.02
.09
.11
.01
.43
-9.33, -4.93
.06, .10
-.16, -.06
-.45, -.11
.10, .55
-.03, .02
-.93, .76
<.01
<.01
<.01
<.01
<.01
.68
.85
Model VIb: Model VIa + years on hormone therapy 6,158 2579.96
intercept
age, in years
years of education
BMI category
cardiovascular history
age of last menstrual period
years on hormone therapy
oophorectomy
-6.98
.08
-.10
-.28
.33
-.01
-.10
-.06
1.13
.01
.02
.09
.11
.01
.05
.43
-9.19, -4.78
.06, .10
-.15, -.05
-.45, -.12
.10, .55
-.03, .02
-.18,- .01
-.90, .78
<.01
<.01
<.01
<.01
<.01
.71
.04
.89
Model VIc: Model VIa + years on hormone therapy + interaction 6,158 2580.97
intercept
age, in years
years of education
BMI category
cardiovascular history
age of last menstrual period
years on hormone therapy
years x oophorectomy
oophorectomy
-6.99
.08
-.10
-.28
.33
-.01
-.10
.09
-.14
1.13
.01
.02
.09
.11
.01
.05
.10
.44
-9.20, -4.78
.06, .10
-.15, -.05
-.45, -.11
.10, .55
-.03, .02
-.19,- .01
-.11, .30
-1.00, .73
<.01
<.01
<.01
<.01
<.01
.71
.045
.37
.76
Model VIIa: Model IIIa + age of last menstrual period 6,158 2585.08
intercept
age, in years
years of education
BMI category
cardiovascular history
age of last menstrual period
T1
T2
T3
-7.05
.08
-.11
-.28
.33
-.01
2.44
-.55
-.22
1.12
.01
.02
.09
.11
.01
.08
.60
1.02
-9.24, -4.85
.06, .10
-.16, -.06
-.45, -.11
.10, .55
-.04, .02
.93, 3.95
-1.72, .62
-2.22, 1.79
<.01
<.01
<.01
<.01
<.01
.60
<.01
.35
.83
Model VIIb: Model VIIa + years on hormone therapy 6,158 2577.48
intercept
age, in years
years of education
BMI category
cardiovascular history
age of last menstrual period
years on hormone therapy
T1
T2
T3
-6.90
.08
-.10
-.28
.33
-.01
-.09
2.43
-.54
-.13
1.13
.01
.02
.09
.11
.01
.04
.83
.59
1.03
-9.11, -4.69
.06, .10
-.15, -.06
-.45, -.12
.10, .56
-.03, .02
-.18, -.01
.80, 4.06
-1.71, .62
-2.15, 1.89
<.01
<.01
<.01
<.01
<.01
.63
.04
<.01
.36
.90
Note. The criterion variable, cognitive impairment = 0 if cognitively intact and = 1 if impaired.
years x oophorectomy = years on hormone therapy x oophorectomy status.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 89
T1 = oophorectomy relative to last menstrual period unknown; T2 = oophorectomy after last menstrual
period; T3a = oophorectomy and last menstrual period within 1.5 years, <= age 50; T3b = oophorectomy
and last menstrual period within 1.5 years, > age 50.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 90
Table 5
Models Examining Oophorectomy on the Continuous Cognitive Scores (TELE score)
Parameter
Estimate
Standard
Error
95%
Confidence
Limits
p n QIC
Model I: covariates only 6,811 6,817.31
intercept
age, in years
years of education
BMI category
cardiovascular history
19.89
-.09
.16
.42
-.13
.43
.01
.01
.04
.06
19.05, 20.73
-.10, -.08
.15, .18
.34, .50
-.25, -.02
<.01
<.01
<.01
<.01
.02
Model IIa: Model I + oophorectomy 6,811 6,817.63
intercept
age, in years
years of education
BMI category
cardiovascular history
oophorectomy
19.90
-.09
.16
.42
-.13
-.11
.43
.01
.01
.04
.06
.16
19.05, 20.74
-.10, -.08
.15, .18
.34, .50
-.25, -.02
-.42, .20
<.01
<.01
<.01
<.01
.02
.48
Model IIb: Model IIa + hormone therapy 6,395 6,402.14
intercept
age, in years
years of education
BMI category
cardiovascular history
hormone therapy, ever
oophorectomy
19.60
-.08
.15
.30
-.16
.35
-.19
.42
.01
.01
.04
.06
.06
.17
18.77, 20.43
-.09, -.07
.13, .17
.22, .38
-.27, -.05
.23, .47
-.53, .14
<.01
<.01
<.01
<.01
<.01
<.01
.25
Model IIc: Model IIa + years on hormone therapy 6,811 6818.04
intercept
age, in years
years of education
BMI category
cardiovascular history
years on hormone therapy
oophorectomy
19.82
-.09
.16
.42
-.13
.03
-.13
.43
.01
.01
.04
.06
.01
.16
18.98, 20.67
-.1, -.08
.14, .18
.34, .50
-.25, -.02
.02, .04
-.44, .18
<.01
<.01
<.01
<.01
.02
<.01
.43
Model IId: Model IIa + years on hormone therapy + interaction 6,811 6,818.44
intercept
age, in years
years of education
BMI category
cardiovascular history
years on hormone therapy
years x oophorectomy
oophorectomy
19.83
-.09
.16
.42
-.13
.03
-.02
-.09
.43
.01
.01
.04
.06
.01
.04
.17
18.98, 20.67
-.10, -.08
.14, .18
.34, .50
-.25, -.02
.02, .04
-.10, .05
-.42, .25
<.01
<.01
<.01
<.01
.02
<.01
.54
.61
Model IIIa: Model I + timing of oophorectomy 6,811 6,817.88
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 91
Parameter
Estimate
Standard
Error
95%
Confidence
Limits
p n QIC
intercept
age, in years
years of education
BMI category
cardiovascular history
T1
T2
T3
19.89
-.09
.16
.42
-.13
-1.05
.06
-.17
.43
.01
.01
.04
.06
.46
.20
.27
19.05, 20.74
-.10, -.08
.15, .18
.34, .50
-.25, -.02
-1.95, -.15
-.33, .45
-.71, .37
<.01
<.01
<.01
<.01
.02
.02
.76
.54
Model IIIb: Model IIIa + hormone therapy
intercept
age, in years
years of education
BMI category
cardiovascular history
hormone therapy, ever
T1
T2
T3
19.60
-.09
.15
.30
-.16
.35
-1.21
-.02
-.20
.42
.01
.01
.04
.06
.06
.49
.21
.30
18.78, 20.43
-.09, -.07
.13, .17
.22, .38
-.27, -.05
.23, .47
-2.17, -.25
-.43, .39
-.80, .39
<.01
<.01
<.01
<.01
.01
<.01
.01
.92
.50
6,395 6,402.65
Model IIIc: Model IIIa + years on hormone therapy 6,811 6,818.34
intercept
age, in years
years of education
BMI category
cardiovascular history
years on hormone therapy
T1
T2
T3
19.82
-.09
.16
.42
-.13
.03
-1.07
.05
-.21
.43
.01
.01
.04
.06
.01
.48
.20
.27
18.98, 20.67
-.10, -.08
.14, .18
.34, .50
-.25, -.02
.02, .04
-2.01, -.13
-.33, .45
-.74, .31
<.01
<.01
<.01
<.01
.02
<.01
.03
.77
.43
Model IV: Model IIc + depressive symptoms 6,601 6,609.21
intercept
age, in years
years of education
BMI category
cardiovascular history
years on hormone therapy
oophorectomy
depressive symptoms
19.79
-.08
.15
.29
-.13
.03
-.15
-.03
.42
.01
.01
.04
.06
.01
.16
.01
18.97, 20.60
-.09, -.07
.13, .17
.22, .37
-.24, -.02
.01, .04
-.46, .16
-.04, -.01
<.01
<.01
<.01
<.01
.02
<.01
.33
<.01
Model V: Model IIIc + depressive symptoms 6,601 6,609.49
intercept
age, in years
years of education
BMI category
cardiovascular history
years on hormone therapy
T1
T2
T3
depressive symptoms
19.78
-.08
.15
.29
-.13
.03
-1.41
.06
-.23
-.03
.42
.01
.01
.04
.06
.01
.47
.20
.27
.01
18.97, 20.60
-.09, -.07
.13, .17
.22, .37
-.24, -.02
.01, .04
-2.34, -.49
-.33, .45
-.75, .29
-.04, -.01
<.01
<.01
<.01
<.01
.02
<.01
<.01
.77
.38
<.01
Model VIa: Model IIa + age of last menstrual period 6,158 6,166.22
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 92
Parameter
Estimate
Standard
Error
95%
Confidence
Limits
p n QIC
intercept
age, in years
years of education
BMI category
cardiovascular history
age of last menstrual period
oophorectomy
19.28
-.07
.14
.21
-.16
.01
-.13
.53
.01
.01
.04
.06
.01
.16
18.25, 20.32
-.08, -.06
.13, .16
.13, .29
-.27, -.05
-.01, .01
-.45, .19
<.01
<.01
<.01
<.01
<.01
.69
.42
Model VIb: Model VIa + years on hormone therapy 6,158 6,166.74
intercept
age, in years
years of education
BMI category
cardiovascular history
age of last menstrual period
years on hormone therapy
oophorectomy
19.24
-.07
.14
.21
-.16
.01
.03
-.14
.53
.01
.01
.04
.06
.01
.01
.16
18.20, 20.27
-.08, -.06
.12, .16
.13, .29
-.27, -.05
-.01, .01
.01, .04
-.46, .18
<.01
<.01
<.01
<.01
<.01
.74
<.01
.39
Model VIc: Model VIa + years on hormone therapy + interaction 6,158 6,166.96
intercept
age, in years
years of education
BMI category
cardiovascular history
age of last menstrual period
years on hormone therapy
years x oophorectomy
oophorectomy
19.23
-.07
.14
.21
-.16
.01
.03
.01
-.15
.53
.01
.01
.04
.06
.01
.01
.04
.17
18.20, 20.27
-.08, -.06
.12, .16
.13, .29
-.27, -.05
-.01, .01
.01, .04
-.06, .08
-.49, .19
<.01
<.01
<.01
<.01
<.01
.74
<.01
.87
.39
Model VIIa: Model IIIa + age of last menstrual period 6,158 6,167.37
intercept
age, in years
years of education
BMI category
cardiovascular history
age of last menstrual period
T1
T2
T3
19.27
-.07
.14
.21
-.16
.01
-.96
-.03
-.24
.53
.01
.01
.04
.06
.01
.76
.20
.26
18.23, 20.30
-.08, -.06
.13, .16
.13, .29
-.27, -.05
-.01, .01
-2.46, .53
-.42, .36
-.76, .29
<.01
<.01
<.01
<.01
<.01
.67
.21
.88
.38
Model VIIb: Model VIIa + years on hormone therapy 6,158 6,167.97
intercept
age, in years
years of education
BMI category
cardiovascular history
age of last menstrual period
years on hormone therapy
T1
T2
T3
19.22
-.07
.14
.21
-.16
.01
.03
-.94
-.03
-.27
.53
.01
.01
.04
.06
.01
.01
.78
.20
.26
18.19, 20.26
-.08, -.06
.12, .16
.13, .29
-.27, -.05
-.01, .01
.01, .04
-2.48, .59
-.42, .35
-.79, .24
<.01
<.01
<.01
<.01
<.01
.71
<.01
.23
.86
.30
Note. Higher scores of the criterion variable, cognitive functioning, indicate better performance.
years x oophorectomy = years on hormone therapy x oophorectomy status.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 93
T1 = oophorectomy relative to last menstrual period unknown; T2 = oophorectomy after last menstrual
period; T3a = oophorectomy and last menstrual period within 1.5 years, <= age 50; T3b = oophorectomy
and last menstrual period within 1.5 years, > age 50.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 94
Table 6
Comparing Mortality for Those With and Without Oophorectomy
Frequencies (%)
Deceased before
cognitive screening
(% of deceased)
Alive at time of
cognitive screening
(% of those alive)
% deceased before
cognitive screening
Without oophorectomy 2,245 (95.94%) 11,152 (95.10%) 16.76%
With unilateral oophorectomy 24 (1.03%) 124 (1.06%) 16.22%
With bilateral oophorectomy or two
unilateral oophorectomies
71 (3.03%) 451 (3.85%) 13.60%
Total 2,340 (100%) 11,727 (100%) 16.63%
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 95
STUDY II RESULTS TABLES
Table 7
Descriptives of the Sample
Group Means (Standard Deviations); n
BVAIT WISH ELITE Combined
Age at Test Administration
1, in years
62.70 (7.89); 184 60.79 (7.11); 337 59.92 (6.93); 628 60.63 (7.21); 1,149
Age at Test Administration
2, in years
65.75 (7.66); 162 63.47 (7.03); 301 not applicable 64.27 (7.33); 463
Depressive symptoms
(CES-D score)
7.56 (7.59); 180 7.28 (6.79); 336 8.35 (8.65); 628 7.91 (7.99); 1,144
Years on hormones, for
those on therapy
8.21 (7.86); 130 6.43 (6.48); 237 6.94 (6.94); 442 6.99 (6.92); 809
Age at menopause 45.36 (9.28); 32 49.05 (5.64); 337 49.43 (5.45); 626 49.18 (5.71); 995
Years since menopause 16.45 (11.84); 32 11.74 (8.93); 337 10.50 (8.39); 626 11.11 (8.77); 995
Test Administration I Test Administration II
Women without
both ovaries
Women with
ovaries intact
Women without
both ovaries
Women with
ovaries intact
Symbol Digit 51.89 (8.56); 152 52.92 (9.14); 996 51.76 (9.72); 66 53.00 (9.97); 395
WMS-III, Letter-Number
Sequencing
9.38 (2.35); 151 9.67 (2.42); 991 9.11 (2.41); 66 9.41 (2.44); 393
Judgment of Line
Orientation
23.42 (5.13); 152 23.40 (4.86); 996 22.36 (5.16); 66 22.83 (5.24); 394
Trail Making Test Part B 87.60 (52.81);
149
81.73 (41.95);
986
90.27 (55.12); 66 83.66 (47.83); 391
Block Design 12.14 (3.23); 152 11.84 (2.96); 992 12.24 (3.62); 62 12.18 (3.01); 387
Shipley 13.46 (4.44); 151 14.22 (3.91); 993 12.98 (4.72); 63 14.07 (4.00); 386
CVLT-II, Immediate Recall 27.43 (6.19); 152 27.79 (6.09); 993 27.19 (5.45); 64 28.12 (6.43); 389
CVLT-II, Delayed Recall 9.10 (3.45); 152 9.52 (3.12); 989 9.27 (2.89); 64 9.51 (3.40); 386
WMS-III, Logical Memory
I, Immediate Recall
4.68 (1.02); 151 4.73 (1.12); 997 4.70 (.93); 66 4.65 (1.18); 397
WMS-III, Logical Memory
II, Delayed Recall
4.48 (1.09); 151 4.59 (1.15); 994 4.48 (.95); 66 4.511 (1.18); 397
WMS-III, Faces I 35.40 (4.59); 152 35.81 (4.64); 997 35.77 (4.77); 66 36.69 (4.94); 394
WMS-III Faces II 36.89 (4.26); 152 37.13 (4.41); 993 36.67 (4.25); 66 37.61 (4.59); 394
Boston Naming Test 26.47 (3.10); 151 26.58 (3.02); 990 26.20 (3.68); 66 26.77 (2.85); 394
Category Fluency 29.03 (9.62); 152 29.91 (9.44); 993 29.30 (9.68); 64 30.25 (9.85); 391
Note. CVLT-II = California Verbal Learning Test, 2nd edition. WMS-III = Wechsler Memory Scale, 3rd
edition.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 96
Table 8
Descriptives of the Sample
Group Frequencies (%)
BVAIT WISH ELITE Combined
Oophorectomy status
bilateral
both intact
36 (19.57%)
148 (80.43%)
35 (10.39%)
302 (89.61%)
81 (12.90%)
547 (87.10%)
152 (13.23%)
997 (86.77%)
Education
1. 8th grade or less
2. some high school
3. high school graduate
4. trade/business school
5. some college
6. bachelor's degree
7. graduate/professional
8. other, unknown
1 (.54%)
2 (1.09%)
17 (9.24%)
8 (4.35%)
65 (35.33%)
50 (27.17%)
40 (21.74%)
1 (.54%)
1 (.30%)
4 (1.19%)
13 (3.86%)
16 (4.75%)
101 (29.97%)
84 (24.93%)
118 (35.01%)
1 (.16%)
2 (.32%)
19 (3.03%)
16 (2.55%)
175 (27.87%)
167 (26.59%)
248 (39.49%)
3 (.26%)
8 (.70%)
49 (4.26%)
40 (3.48%)
341 (29.68%)
301 (26.20%)
406 (35.34%)
1 (.09%)
Ethnicity
Caucasian
Other
118 (64.13%)
66 (35.87%)
215 (63.80%)
122 (36.20%)
427 (67.99%)
201 (32.01%)
760 (66.14%)
389 (33.86%)
BMI
1. underweight (<18.5)
2. normal (>=18.5, <25)
3. overweight (>=25, <30)
4. obese (>=30)
2 (1.09%)
54 (29.35%)
62 (33.70%)
66 (35.87%)
5 (1.48%)
145 (43.03%)
112 (33.23%)
75 (22.26%)
8 (1.27%)
223 (35.51%)
232 (36.94%)
165 (26.27%)
15 (1.31%)
422 (36.73%)
406 (35.34%)
306 (26.63%)
Hysterectomy status
hysterectomy
intact uterus
unknown
66 (32.61%)
124 (67.39%)
63 (18.69%)
274 (81.31%)
104 (16.56%)
522 (83.12%)
2 (.32%)
227 (19.76%)
920 (80.07%)
2 (.17%)
Hormone use, current
current use
no current use
unknown
77 (41.85%)
101 (54.89%)
6 (3.26%)
1 (.30%)
319 (94.66%)
17 (5.04%)
3 (.48%)
622 (99.04%)
3 (.48%)
81 (7.05%)
1,042 (90.69%)
26 (2.26%)
Hormone use, ever
at least once
never
unknown
130 (70.65%)
54 (29.35%)
237 (71%)
100 (28%)
442 (70.38%)
184 (29.30%)
2 (.32%)
809 (70.41%)
338 (29.42%)
2 (.17%)
Contraceptive pills
at least once
never
unknown
184 (100.00%)
251 (74.48%)
86 (25.52%)
499 (79.46%)
123 (19.59%)
6 (.96%)
750 (65.27%)
209 (18.19%)
190 (16.54%)
Treatment
active
placebo
unknown/not disclosed
92 (50.00%)
92 (50.00%)
170 (50.45%)
167 (49.55%)
628 (100.00%)
262 (22.80%)
259 (22.54%)
628 (54.66%)
Test Administration
Test 1
Test 2
184
162
337
301
628
not applicable
1,149
463
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 97
Table 9
Characteristics of Women Who Had Oophorectomy and Women Who Did Not
Group Means (Standard Deviations)
Women who had bilateral
oophorectomy (n = 152)
Women who have both
ovaries intact (n = 997)
Age at Test Administration 1, mean
(SD)
61.66 (6.97) 60.47 (7.23)
CES-D score at Test Administration 1,
mean (SD)
8.92 (7.96) 7.76 (7.99)
Years on hormones, mean (SD)*** 10.08 (8.90) 4.16 (5.84)
Age at menopause, mean (SD)*** 45.04 (5.77) 49.73 (5.47)
Years since menopause, mean (SD)*** 16.11 (8.84) 10.44 (8.55)
BMI category, frequency
1 (underweight)
2 (normal)
3 (overweight)
4 (obese)
51 (33.55%)
51 (33.55%)
50 (32.89%)
15 (1.5%)
371 (37.21%)
355 (35.61%)
256 (25.68%)
Timing of surgery, frequency
h1
h2
h3a
h3b
50 (32.89%)
20 (13.16%)
39 (25.66%)
42 (27.63%)
Hormone use, frequency
current
not current
unknown
23 (15.13%)
127 (83.55%)
2 (1.32%)
58 (5.82%)
915 (91.78%)
24 (2.41%)
Hormone history, frequency
at least once
never
unknown
139 (91.45%)
11 (7.24%)
2 (1.32%)
670 (67.20%)
327 (32.80%)
Mean change in Z-score
global functioning
executive functioning
semantic memory
logical memory
visual memory
verbal learning
-.58
-.03
-.16
.07
-.26
-.24
-.37
-.22
-.08
-.04
.04
-.07
Note. h1 = no oophorectomy but hyst, oophorectomy and hyst at least 3 years after natural menopause, or
oophorectomy but dates of oophorectomy or menopause unknown;
h2 = oophorectomy within 3 years after natural menopause;
h3 = oophorectomy within 1 year of menopause;
h3a = oophorectomy within 1 year of menopause, where age of oophorectomy <= 45;
h3b = oophorectomy within 1 year of menopause, where age of oophorectomy > 45;
Mean change in cognitive composite Z-scores = Z-scores from Test Administration 2 – Z-scores from Test
Administration 1)
Significant differences in groups with ***p < .0001
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 98
Table 10
Correlations of Potential Covariates and Cognitive Performance
r, p, n
1 2 3 4 5
1. Age, in years — -.09 .11 -.11 .01
p < .01 p < .01 p < .01 p = .84
n = 1,148 n = 1,149 n = 1,148 n = 1,106
2. Education — .14 .10 -.09
category p < .01
n = 1,148 n = 1,147 n = 1,148
3. Caucasian — .04 -.16
n = 1,148 n = 1,149
4. Married — -.09
n = 1,148
5. BMI category —
Global -.22 .31 .36 .09 -.13
cognitive p < .01 p < .01 p < .01 p < .01 p < .01
functioning n = 1,106 n = 1,105 n = 1,106 n = 1,105 n = 1,106
Executive -.20 .30 .37 .07 -.13
functioning p < .01 p < .01 p < .01 p = .02 p <.01
n = 1,119 n = 1,118 n = 1,119 n = 1,118 n = 1,119
Semantic -.11 .30 .39 .05 -.12
memory p = .02 p < .01 p < .01 p = .12 p < .01
n = 1,138 n = 1,137 n = 1,138 n = 1,137 n = 1,138
Logical -.06 .11 .24 .02 -.06
memory p = .06 p < .01 p <.01 p = .61 p = .05
n = 1,145 n = 1,144 n = 1,145 n = 1,144 n = 1,145
Visual -.19 .14 .04 .07 -.01
memory p < .01 p < .01 p = .21 p = .02 p = .91
n = 1,145 n = 1,144 n = 1,145 n = 1,144 n = 1,145
Verbal -.20 .21 .15 .09 -.09
learning p < .01 p < .01 p < .01 p < .01 p < .01
n = 1,141 n = 1,140 n = 1,141 n = 1,140 n = 1,141
Note. Pearson product-moment correlations are presented where 2 variables are continuous, Spearman rank-
order correlations are presented where at least one variable is ordinal (e.g., education and BMI category) and
the other variable is not dichotomous, point biserial correlations are presented where one variable is
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 99
dichotomous and the other is continuous, tetrachoric correlations are presented where both variables are
dichotomous, and rank biserial are presented where one variable is ordinal and the other is dichotomous.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 100
Table 11
Cognitive Performance Composite Z-scores Across All Trials and Administrations for Women in Study II
Group Means (Standard Deviation); n
BVAIT WISH ELITE
All participants
Test 1 global cognitive functioning
executive functioning
semantic memory
logical memory
visual memory
verbal learning
-2.22 (9.15); 175
-.99 (4.93); 179
-.02 (1.60); 182
-.35 (2.08); 183
-.56 (1.76); 181
-.52 (1.83); 181
.06 (8.56); 317
-.03 (4.29); 325
.02 (1.92); 329
.04 (1.90); 335
-.17 (1.79); 336
-.11 (1.77); 333
1.28 (7.71); 614
.67 (3.97); 615
.05 (1.61); 627
.15 (1.77); 627
.17 (1.75); 628
.22 (1.88); 627
Test 2 global cognitive functioning
executive functioning
semantic memory
logical memory
visual memory
verbal learning
-1.01 (8.98); 149
-.81 (4.73); 152
.21 (1.72); 156
-.34 (1.78); 162
-.09 (1.82); 159
-.15 (2.05); 155
1.30 (8.81); 284
.49 (4.48); 288
.06 (1.74); 297
.05 (1.95); 301
.38 (1.90); 300
.15 (1.86); 295
Women without both ovaries
Test 1 global cognitive functioning
executive functioning
semantic memory
logical memory
visual memory
verbal learning
-2.48 (7.38); 35
-1.23 (4.55); 35
.01 (1.37); 36
-.08 (1.79); 36
-.84 (1.44); 36
-.59 (1.63); 36
-1.26 (10.74); 32
-.99 (5.61); 33
-.34 (2.36); 34
-.28 (1.93); 34
.10 (1.65); 35
.09 (2.09); 35
.49 (7.93); 79
.59 (4.20); 79
-.03 (1.55); 81
-.01 (1.57); 81
.02 (1.90); 81
-.08 (2.04); 81
Test 2 global cognitive functioning
executive functioning
semantic memory
logical memory
visual memory
verbal learning
-1.28 (7.72); 30
-.61 (4.79); 30
.17 (1.74); 32
-.10 (1.45); 34
-.58 (1.70); 34
-.39 (1.48); 32
-.76 (9.70); 32
-.67 (5.68); 32
-.48 (2.15); 32
-.03 (1.59); 32
.33 (1.68); 32
.10 (1.85); 32
Women with ovaries intact
Test 1 global cognitive functioning
executive functioning
semantic memory
logical memory
visual memory
verbal learning
-2.16 (9.57); 140
-.93 (5.03); 144
-.02 (1.66); 146
-.41 (2.14); 147
-.49 (1.83); 145
-.51 (1.88); 145
.21 (8.29); 285
.14 (4.11); 292
.06 (1.86); 295
.08 (1.90); 301
-.20 (1.80); 301
-.13 (1.73); 298
1.40 (7.68); 535
.68 (3.94); 536
.06 (1.62); 546
.17 (1.80); 546
.19 (1.73); 547
.26 (1.85); 546
Test 2 global cognitive functioning
executive functioning
semantic memory
logical memory
visual memory
verbal learning
-.94 (9.30); 119
-.86 (4.73); 122
.22 (1.73); 124
-.40 (1.84); 125
.05 (1.84); 125
-.09 (2.17); 123
1.56 (8.68); 252
.64 (4.30); 256
.12 (1.68); 265
.06 (1.99); 269
.38 (1.92); 268
.16 (1.87); 263
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 101
Table 12
Models Examining Oophorectomy on Cognitive Performance
Global functioning
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Model I
intercept
age, in years
education category
Caucasian
BMI category
BVAIT
WISH
1,105
.28 6
2.70
-.28
2.03
6.11
-.65
1.79
.57
2.38
.03
.19
.46
.26
.62
.49
1.14
-9.16
10.65
13.24
-2.46
2.90
1.15
.26
<.01
<.01
<.01
.01
<.01
.25
Model II
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
BVAIT
WISH
1,105 .28 7
2.70
-.28
2.03
6.11
-.64
-.20
1.80
.56
2.38
.03
.19
.46
.26
.63
.62
.49
1.13
-9.13
10.63
13.23
-2.44
-.31
2.91
1.14
.26
<.01
<.01
<.01
.01
.75
<.01
.25
Model III:
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
BVAIT
WISH
1,105 .28 10
2.54
-.27
2.04
6.09
-.66
.86
.38
-.83
-1.38
1.57
.56
2.38
.03
.19
.46
.26
1.09
1.69
1.17
1.15
.65
.50
1.07
-9.02
10.71
13.17
-2.50
.79
.22
-.71
-1.20
2.43
1.13
.29
<.01
<.01
<.01
.01
.43
.82
.48
.23
.02
.26
Executive functioning
Model n R
2
df Parameter
estimate
Standard
error
t-value P
Model I:
intercept
age, in years
education category
Caucasian
BMI category
BVAIT
WISH
1,118
.27 6
1.23
-.13
1.00
3.16
-.37
.86
.16
1.24
.02
.10
.24
.14
.32
.25
1.00
-8.48
10.08
13.17
-2.72
2.68
.64
.32
<.01
<.01
<.01
<.01
.01
.52
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 102
Model n R
2
df Parameter
estimate
Standard
error
t-value P
Model II:
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
BVAIT
WISH
1,118 .27 7
1.23
-.13
1.00
3.16
-.37
-.04
.86
.16
1.23
.02
.10
.24
.14
.33
.32
.26
1.00
-8.46
10.07
13.17
-2.71
-.11
2.68
.63
.32
<.01
<.01
<.01
.01
.91
.01
.53
Model III:
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
BVAIT
WISH
1,118 .27 10
1.14
-.13
1.00
3.15
-.38
.36
.78
-.49
-.46
.78
.16
1.24
.02
.10
.24
.14
.57
.86
.61
.60
.34
.26
.92
-8.33
10.12
13.13
-2.78
.64
.91
-.79
-.76
2.32
.61
.36
<.01
<.01
<.01
.01
.52
.37
.43
.45
.02
.54
Semantic memory
Model n R
2
df Parameter
estimate
Standard
error
t-value P
Model I:
intercept
age, in years
education category
Caucasian
BMI category
BVAIT
WISH
1,137
.23 6
-.77
-.03
.35
1.33
-.08
.26
.08
.50
.01
.04
.10
.06
.13
.10
-1.54
-5.04
8.84
13.77
-1.52
2.02
.79
.12
<.01
<.01
<.01
.13
.04
.43
Model II:
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
BVAIT
WISH
1,137 .23 7
-.77
-.03
.35
1.33
-.08
-.04
.26
.08
.50
.01
.04
.10
.06
.13
.13
.10
-1.54
-5.02
8.82
13.77
-1.50
-.27
2.03
.78
.12
<.01
<.01
<.01
.13
.78
.04
.43
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 103
Model n R
2
df Parameter
estimate
Standard
error
t-value P
Model III:
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
BVAIT
WISH
1,137 .23 10
-.78
-.03
.35
1.33
-.08
.07
<-.01
-.18
-.24
.08
.09
.50
.01
.04
.10
.06
.23
.35
.26
.24
.14
.10
-1.55
-4.98
8.83
13.73
-1.52
.28
-.01
-.74
-.35
1.76
.75
.12
<.01
<.01
<.01
.13
.78
.99
.46
.73
.08
.45
Logical memory
Model n R
2
df Parameter
estimate
Standard
error
t-value P
Model I:
intercept
age, in years
education category
Caucasian
BMI category
BVAIT
WISH
1,144
.07 6
-.32
-.02
.17
.91
-.03
.12
.06
.60
.01
.05
.11
.07
.15
.12
-.54
-2.59
3.49
7.91
-.48
.75
.53
.59
.01
<.01
<.01
.63
.45
.60
Model II:
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
BVAIT
WISH
1,144 .07 7
-.32
-.02
.17
.91
-.03
-.06
.12
.06
.60
.01
.05
.12
.07
.16
.16
.12
-.54
-2.57
3.48
7.91
-.46
-.37
.77
.51
.59
.01
<.01
<.01
.65
.71
.44
.61
Model III:
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
BVAIT
WISH
1,144 .08 8
-.33
-.02
.17
.91
-.03
.22
-.08
-.36
-.12
.06
.06
.60
.01
.05
.12
.07
.27
.41
.30
.29
.16
.12
-.55
-2.56
3.55
7.88
-.48
.79
-.18
-1.22
-.41
.37
.45
.58
.01
<.01
<.01
.63
.43
.85
.22
.68
.71
.65
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 104
Visual memory
Model n R
2
df Parameter
estimate
Standard
error
t-value P
Model I:
intercept
age, in years
education category
Caucasian
BMI category
BVAIT
WISH
1,144
.06 6
1.53
-.05
.21
.17
.01
.22
.05
.58
.01
.05
.11
.06
.15
.12
2.62
-6.43
4.42
1.49
.08
1.42
.46
.01
<.01
<.01
.14
.94
.16
.65
Model II:
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
BVAIT
WI SH
1,144 .06 7
1.53
-.05
.21
.17
.01
-.03
.22
.05
.58
.01
.05
.11
.06
.15
.15
.12
2.62
-6.41
4.41
1.49
.09
-.19
1.43
.45
.01
<.01
<.01
.14
.93
.85
.15
.65
Model III:
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
BVAIT
WISH
1,144 .06 10
1.53
-.05
.21
.17
.01
-.20
-.23
.37
-.22
.25
.06
.59
.01
.05
.11
.06
.27
.40
.29
.28
.16
.12
2.61
-6.38
4.37
1.49
.12
-.73
-.59
1.30
-.80
1.58
.53
<.01
<.01
<.01
.14
.91
.47
.56
.19
.42
.11
.60
Verbal learning
Model n R
2
df Parameter
estimate
Standard
error
t-value P
Model I:
intercept
age, in years
education category
Caucasian
BMI category
BVAIT
WISH
1,140
.11 6
1.34
-.05
.31
.59
-.12
.25
.06
.59
.01
.05
.11
.07
.15
.12
2.25
-7.03
6.47
5.15
-1.89
1.60
.51
.02
<.01
<.01
<.01
.06
.11
.61
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 105
Model n R
2
df Parameter
estimate
Standard
error
t-value P
Model II:
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
BVAIT
WISH
1,140 .11 7
1.34
-.05
.31
.59
-.12
-.03
.25
.06
.59
.01
.05
.11
.07
.16
.15
.12
2.26
-7.01
6.46
5.15
-1.88
-.18
1.61
.50
.02
<.01
<.01
<.01
.06
.85
.11
.61
Model III:
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
BVAIT
WISH
1,140 .11 10
1.26
-.05
.31
.58
-.13
.36
.26
-.11
-.61
.16
.06
.59
.01
.05
.11
.07
.27
.40
.29
.28
.16
.12
2.14
-6.85
6.64
5.09
-2.02
1.35
.64
-.38
-2.21
1.03
.54
.03
<.01
<.01
<.01
.04
.18
.52
.71
.03
.30
.60
Note. Model I includes covariates only, Model II includes Model I + oophorectomy, and Model III + timing
of oophorectomy, where h1 = no oophorectomy but hyst, oophorectomy and hyst at least 3 years after
natural meno, or oophorectomy but dates unknown; h2 = oophorectomy within 3 years of natural meno; h3a
= oophorectomy within 1 year of meno, before or at 45 years of age; h3b = oophorectomy within 1 year of
meno, after 45 years of age.
ETRIAL = intercept - BVAIT – WISH
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 106
Table 13
Models Examining Depressive Symptoms on Cognitive Performance
Models with oophorectomy status
Model n R
2
df Parameter
estimate
Standard
error
t-value P
Global functioning
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
depressive sxs
BVAIT
WISH
1,102 .29 8
4.87
-.29
1.93
6.06
-.58
-.07
-.11
1.66
.43
2.42
.03
.19
.46
.26
.63
.03
.62
.49
2.01
-9.58
10.12
13.19
-2.22
-.12
-4.24
2.68
.87
.04
<.01
<.01
<.01
.03
.91
<.01
.01
.39
Executive function
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
depressive sxs
BVAIT
WISH
1,113 .27 8
2.27
-.14
.95
3.11
-.35
.02
-.05
.79
.11
1.26
.02
.10
.24
.14
.33
.01
.32
.25
1.80
-8.85
9.56
13.03
-2.56
.06
-3.72
2.45
.42
.07
<.01
<.01
<.01
.01
.95
<.01
.01
.68
Semantic memory
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
depressive sxs
BVAIT
WISH
1,132 .24 8
-.47
-.03
.34
1.33
-.08
-.02
.02
.24
.06
.51
.01
.04
.10
.06
.13
.01
.13
.10
-.91
-5.30
8.45
13.75
-1.36
-.14
-3.06
1.83
.55
.36
<.01
<.01
<.01
.17
.89
<.01
.07
.58
Logical memory
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
depressive sxs
BVAIT
WISH
1,139 .08 8
.18
-.02
.14
.90
-.02
-.03
-.03
.08
.04
.61
.01
.05
.12
.07
.16
.01
.16
.12
.29
-2.98
2.98
7.76
-.23
-.21
-3.81
.54
.31
.77
<.01
<.01
<.01
.82
.83
<.01
.59
.76
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 107
Model n R
2
df Parameter
estimate
Standard
error
t-value P
Visual memory
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
depressive sxs
BVAIT
WISH
1140 .06 8
1.48
-.05
.21
.17
.01
-.03
<.01
.22
.06
.60
.01
.05
.11
.07
.16
.01
.15
.12
2.47
-6.33
4.36
1.51
.13
-.19
.53
1.42
.52
.01
<.01
<.01
.13
.89
.85
.59
.15
.60
Verbal learning
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
depressive sxs
BVAIT
WISH
1,136 .12 8
1.79
-.06
.29
.58
-.11
-.02
.22
.04
.61
.01
.05
.11
.07
.15
.15
.12
2.96
-7.37
6.03
5.09
-1.74
-.04
-3.47
1.43
.29
<.01
<.01
<.01
<.01
.08
.97
<.01
.15
.77
Models with timing of oophorectomy
Model n R
2
df Parameter
estimate
Standard
error
t-value P
Global functioning
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
depressive sxs
BVAIT
WISH
1,102 .29 11
4.71
-.29
1.95
6.04
-.60
.92
.51
-.52
-1.29
-.11
1.45
.43
2.43
.03
.19
.46
.26
1.10
1.68
1.17
1.14
.03
.65
.49
1.95
-9.47
10.19
13.14
-2.27
.84
.30
-.45
-1.13
-4.20
2.25
.87
.05
<.01
<.01
<.01
.02
.40
.76
.65
.26
<.01
.02
.39
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 108
Model n R
2
df Parameter
estimate
Standard
error
t-value P
Executive function
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
depressive sxs
BVAIT
WISH
1,113 .28 11
2.17
-.14
.96
3.11
-.36
.40
.82
-.35
-.42
-.05
.72
.10
1.27
.02
.10
.24
.14
.57
.86
.61
.60
.01
.34
.25
1.72
-8.71
9.61
13.00
-2.63
.69
.96
-.57
-.70
-3.68
2.14
.40
.09
<.01
<.01
<.01
<.01
.49
.34
.57
.48
<.01
.03
.69
Semantic memory
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
depressive sxs
BVAIT
WISH
1,132 .24 11
-.48
-.03
.34
1.33
-.08
.06
.03
-.13
-.07
-.02
.22
.05
.51
.01
.04
.10
.06
.23
.35
.25
.24
.01
.14
.10
-.93
-5.25
8.45
13.71
-1.37
.25
.08
-.24
-.30
-3.02
1.63
.53
.35
<.01
<.01
<.01
.17
.80
.93
.60
.76
<.01
.10
.60
Logical memory
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
depressive sxs
BVAIT
WISH
1,139 .09 11
.17
-.02
.15
.90
-.01
.23
-.04
-.30
-.11
-.03
.03
.03
.61
.01
.05
.11
.07
.28
.41
.30
.28
.01
.16
.12
.27
-2.96
3.05
7.76
-.25
.82
-.10
-1.00
-.37
-3.75
.18
.26
.78
<.01
<.01
<.01
.80
.41
.92
.32
.71
<.01
.85
.80
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 109
Model n R
2
df Parameter
estimate
Standard
error
t-value P
Visual memory
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
depressive sxs
BVAIT
WISH
1140 .06 11
1.49
-.05
.21
.17
.01
-.19
-.24
.36
-.23
<.01
.25
.07
.60
.01
.05
.11
.07
.27
.40
.29
.28
.01
.16
.12
2.47
-6.31
4.32
1.50
.16
-.68
-.61
1.26
-.81
.46
1.55
.60
.01
<.01
<.01
.13
.87
.50
.54
.21
.42
.65
.12
.55
Verbal learning
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
depressive sxs
BVAIT
WISH
1,136 .12 11
1.72
-.05
.30
.60
-.12
.37
.29
-.04
-.61
-.02
.15
.04
.61
.01
.05
.11
.07
.27
.40
.29
.28
.01
.16
.12
2.85
-7.22
6.19
5.03
-1.87
1.34
.73
-.13
-2.17
-3.46
.91
.34
<.01
<.01
<.01
<.01
.06
.18
.47
.89
.03
<.01
.36
.74
Note. ETRIAL = intercept - BVAIT – WISH
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 110
Table 14
Models Examining Cognitive Performance Over Time
Models with oophorectomy status
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Global functioning
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
baseline
treatment
BVAIT
416 .09 8
1.99
-.11
.50
.50
.40
-.21
-.13
.42
.20
2.10
.03
.17
.44
.23
.54
.03
.38
.41
.95
-4.01
2.88
1.14
1.74
-.38
-4.72
1.12
.47
.34
<.01
<.01
.26
.09
.70
<.01
.26
.64
Executive function
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
baseline
treatment
BVAIT
432 .08 8
.96
-.05
.21
.19
.14
.20
-.09
-.07
-.10
.90
.01
.07
.19
.10
.24
.02
.16
.17
1.07
-3.85
2.87
1.03
1.45
.83
-4.28
-.44
-.54
.29
<.01
<.01
.30
.15
.41
<.01
.66
.59
Semantic memory
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
baseline
treatment
BVAIT
446 .07 8
-.38
<-.01
.05
.20
.01
-.10
-.16
.04
-.01
.48
.01
.04
.10
.05
.12
.03
.09
.10
-.78
-.41
1.36
1.96
.17
-.78
-5.45
.49
.10
.44
.68
.17
.05
.86
.44
<.01
.62
.92
Logical memory
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
baseline
treatment
BVAIT
459 .33 8
.01
-.03
.19
.08
.19
.13
-.63
.29
.09
.90
.01
.07
.18
.10
.23
.05
.16
.17
.02
-2.75
2.83
.45
1.89
.57
-14.00
1.80
.52
.99
<.01
<.01
.66
.06
.57
<.01
.07
.60
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 111
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Visual memory
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
baseline
treatment
BVAIT
454 .19 8
1.52
-.04
.10
.17
.07
-.25
-.31
.15
.14
.62
.01
.05
.12
.07
.16
.03
.11
.12
2.44
-4.96
2.05
1.44
1.44
1.01
-9.65
1.34
1.19
.01
<.01
.04
.15
.31
.11
<.01
.18
.24
Verbal learning
intercept
age, in years
education category
Caucasian
BMI category
oophorectomy
baseline
treatment
BVAIT
446 .17 8
.30
-.02
.15
.37
-.05
-.07
-.35
.03
.08
.70
.01
.06
.14
.08
.18
.04
.12
.14
.43
-2.41
2.64
2.73
-.67
-.40
-9.31
.25
.57
.67
.02
<.01
<.01
.51
.69
<.01
.80
.57
Models with timing of oophorectomy
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Global functioning
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
baseline
treatment
BVAIT
416 .09 11
1.83
-.11
.50
.50
.38
-.27
.63
1.61
-1.29
-.13
.38
.20
2.10
.03
.18
.44
.23
.75
1.92
1.37
1.02
.03
.38
.43
.87
-3.86
2.88
1.14
1.64
-.36
.33
1.17
-1.27
-4.83
1.02
.48
.38
<.01
<.01
.25
.10
.72
.74
.24
.21
<.01
.31
.63
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 112
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Executive function
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
baseline
treatment
BVAIT
432 .08 11
.94
-.05
.21
.19
.14
.20
-.03
.68
<-.01
-.10
-.08
-.10
.91
.01
.07
.19
.10
.33
.76
.60
.45
.02
.16
.18
1.04
-3.78
2.85
1.02
1.42
.62
-.04
1.13
-.01
-4.29
-.49
-.55
.30
<.01
<.01
.31
.16
.53
.97
.26
.99
<.01
.63
.59
Semantic memory
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
baseline
treatment
BVAIT
446 .07 11
-.40
<-.01
.06
.20
.01
.06
-.10
-.01
-.44
-.17
.04
-.03
.48
.01
.04
.10
.05
.17
.46
.31
.23
.03
.09
.10
-.83
-.36
1.52
1.92
.09
.35
-.22
-.04
-1.91
-5.54
.43
-.31
.40
.72
.13
.06
.93
.73
.83
.97
.06
<.01
.67
.76
Logical memory
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
baseline
treatment
BVAIT
459 .33 11
-.02
-.03
.19
.08
.19
.14
-.25
.77
-.08
-.63
.28
.08
.90
.01
.07
.18
.10
.32
.78
.62
.44
.05
.16
.18
-.02
-2.67
2.81
.43
1.87
.43
-.32
1.26
-.19
-13.97
1.72
.46
.99
.01
.01
.67
.06
.67
.75
.21
.85
<.01
.09
.65
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 113
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Visual memory
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
baseline
treatment
BVAIT
454 .21 11
1.43
-.04
.10
.16
.05
-.09
.56
.08
-1.04
-.30
.14
.10
.62
.01
.05
.12
.07
.22
.53
.39
.30
.03
.11
.13
2.31
-4.71
2.17
1.34
.77
-.42
1.06
.21
-3.45
-9.49
1.26
.80
.02
<.01
.03
.18
.44
.68
.29
.83
<.01
<.01
.21
.43
Verbal learning
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
baseline
treatment
BVAIT
446 .18 11
.29
-.02
.14
.38
-.05
-.40
.23
.36
.21
-.34
.03
.16
.70
.01
.06
.14
.08
.25
.60
.45
.34
.04
.12
.14
.41
-2.34
2.43
2.79
-.63
-1.58
.38
.81
.61
-9.13
.25
1.10
.68
.02
.02
.01
.53
.11
.71
.42
.54
<.01
.80
.27
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 114
Table 15
Models Examining Years of Hormone Therapy and Oophorectomy on Cognitive Performance
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Global functioning
intercept
age, in years
education category
Caucasian
BMI category
depressive sxs
ooph x horm
hormone, in years
oophorectomy
BVAIT
WISH
1,099 .29 10
4.70
-.28
1.93
6.07
-.58
-.11
.10
-.03
-.83
1.66
.45
2.54
.03
.19
.46
.26
.03
.08
.04
.95
.62
.49
1.85
-8.56
9.99
13.09
-2.20
-4.27
1.22
-.82
-.88
2.67
.91
.06
<.01
<.01
<.01
.03
<.01
.22
.42
.38
.01
.36
Executive function
intercept
age, in years
education category
Caucasian
BMI category
depressive sxs
ooph x horm
hormone, in years
oophorectomy
BVAIT
WISH
1,110 .27 10
2.50
-.14
.94
3.10
-.34
-.05
.05
<-.01
-.39
.78
.12
1.32
.02
.11
.24
.14
.01
.04
.02
.49
.32
.26
1.89
-.21
9.33
12.84
-2.47
-3.79
1.12
-.16
-.80
2.42
.46
.06
<.01
<.01
<.01
.01
<.02
.27
.87
.43
.02
.64
Semantic memory
intercept
age, in years
education category
Caucasian
BMI category
depressive sxs
ooph x horm
hormone, in years
oophorectomy
BVAIT
WISH
1,129 .24 10
-.48
-.03
.34
1.32
-.08
-.02
<.01
<-.01
-.05
.24
.06
.54
.01
.04
.10
.06
.01
.02
.01
.20
.13
.10
-.89
-4.78
8.41
13.58
-1.35
-3.07
.23
-.22
-.26
1.83
.54
.37
.02
<.01
<.01
.18
<.01
.81
.83
.79
.07
.60
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 115
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Logical memory
intercept
age, in years
education category
Caucasian
BMI category
depressive sxs
ooph x horm
hormone, in years
oophorectomy
BVAIT
WISH
1,136 .09 10
-.02
-.02
.15
.92
-.02
-.03
-.01
<-.01
.09
.09
.03
.64
.01
.05
.12
.07
.01
.02
.01
.24
.16
.12
-.04
-2.36
3.16
7.88
-.36
-3.74
-.48
-.58
.38
.57
.26
.97
.02
<.01
<.01
.72
<.01
.63
.63
.56
.56
.79
Visual memory
intercept
age, in years
education category
Caucasian
BMI category
depressive sxs
ooph x horm
hormone, in years
oophorectomy
BVAIT
WISH
1,137 .06 10
1.61
-.05
.20
.16
.01
<.01
.02
<-.01
-.17
.21
.06
.63
.01
.05
.11
.07
.01
.02
.01
.23
.15
.12
.57
-5.95
4.19
1.38
.21
.44
.81
-.01
-.74
1.37
.53
.01
<.01
<.01
.17
.84
.66
.42
.99
.46
.17
.60
Verbal learning
intercept
age, in years
education category
Caucasian
BMI category
depressive sxs
ooph x horm
hormone, in years
oophorectomy
BVAIT
WISH
1,133 .12 10
1.51
-.05
.29
.60
-.11
-.02
.02
-.02
-.12
.23
.04
.63
.01
.05
.11
.07
.01
.02
.01
.23
.15
.12
2.39
-6.06
6.02
5.27
-1.80
-3.40
1.04
-1.64
-.54
1.49
.37
.02
<.01
<.01
<.01
.07
<.01
.30
.10
.59
.14
.71
Note. ooph x horm = oophorectomy status (dichotomous) x years on hormone treatment (continuous
variable).
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 116
Table 16
Models Examining Time Since Menopause on Cognitive Performance
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Global functioning
intercept
age, in years
education category
Caucasian
BMI category
depressive sxs
ooph x horm
hormone, in years
oophorectomy
menopause time
BVAIT
WISH
960 .27 11
5.61
-.28
1.77
5.92
-.58
-.12
.11
-.03
-1.25
-.01
2.57
.41
3.25
.05
.21
.51
.29
.03
.09
.05
1.10
.05
1.31
.50
1.72
-5.38
8.40
11.71
-2.04
-4.22
1.22
-.67
-1.14
-.17
1.97
.82
.08
<01
<.01
<.01
.04
<.01
.22
.50
.25
.87
.049
.41
Executive function
intercept
age, in years
education category
Caucasian
BMI category
depressive sxs
ooph x horm
hormone, in years
oophorectomy
menopause time
BVAIT
WISH
968 .25 11
2.36
-.12
.84
2.99
-.38
-.06
.06
-.01
-.51
-.01
1.62
.08
1.69
.03
.11
.26
.15
.02
.05
.03
.57
.02
.68
.26
1.40
-4.58
7.67
11.41
-2.58
-3.78
1.19
-.21
-.90
-.52
2.39
.34
.16
<.01
<.01
<.01
.01
<.01
.23
.83
.37
.60
.02
.74
Semantic memory
intercept
age, in years
education category
Caucasian
BMI category
depressive sxs
ooph x horm
hormone, in years
oophorectomy
menopause time
BVAIT
WISH
984 .22 11
-.61
-.03
.31
1.28
-.09
-.02
.02
<-.01
-.19
-.01
.79
.06
.69
.01
.04
.11
.06
.01
.02
.01
.23
.01
.28
.10
-.89
-2.31
6.96
11.69
-1.36
-3.16
.86
-.12
-.81
-1.01
2.82
.53
.37
.02
<.01
<.01
.18
<.01
.39
.91
.42
.31
.01
.60
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 117
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Logical memory
intercept
age, in years
education category
Caucasian
BMI category
depressive sxs
ooph x horm
hormone, in years
oophorectomy
menopause time
BVAIT
WISH
990 .09 11
.49
-.03
.14
.94
.02
-.03
-.01
<-.01
-.06
.01
.32
.04
.82
.01
.05
.13
.07
.01
.02
.01
.27
.01
.33
.12
.60
-2.34
2.70
7.47
.24
-3.48
-.46
-.18
-.23
.45
.96
.29
.55
.02
.01
<.01
.81
<.01
.64
.86
.82
.65
.34
.77
Visual memory
intercept
age, in years
education category
Caucasian
BMI category
depressive sxs
ooph x horm
hormone, in years
oophorectomy
menopause time
BVAIT
WISH
992 .05 11
1.48
-.04
.19
.12
.01
<.01
.01
<.01
-.08
<.01
.05
.05
.81
.01
.05
.12
.07
.01
.02
.01
.27
.01
.33
.12
1.84
-3.56
3.66
.98
.16
.39
.39
.10
-.31
.13
.16
.44
.07
<.01
<.01
.33
.87
.69
.70
.92
.76
.89
.87
.66
Verbal learning
intercept
age, in years
education category
Caucasian
BMI category
depressive sxs
ooph x horm
hormone, in years
oophorectomy
menopause time
BVAIT
WISH
989 .12 11
1.91
-.06
.29
.60
-.12
-.03
.04
-.02
-.31
<.01
-.14
.05
.80
.01
.05
.12
.07
.01
.02
.01
.27
.01
.33
.12
2.38
-4.40
5.49
4.81
-1.65
-3.51
1.58
-1.73
-1.16
.23
-.42
.38
.02
<.01
<.01
<.01
.10
<.01
.11
.08
.25
.82
.67
.71
Note. ooph x horm = oophorectomy status (dichotomous) x years on hormone treatment (continuous
variable).
ETRIAL = intercept - BVAIT – WISH
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 118
Table 17
Models Examining Hormones on Cognitive Performance Over Time
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Global functioning
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
hormone, in years
baseline
treatment
BVAIT
416 .10 12
2.49
-.12
.50
.38
.42
-.52
.41
1.20
-1.35
.05
-.13
.33
.18
2.13
.03
.17
.44
.23
.76
1.92
1.39
1.02
.03
.03
.38
.43
1.17
-4.18
2.83
.86
1.81
-.69
.22
.87
-1.33
1.66
-4.84
.87
.41
.24
<.01
<.01
.39
.07
.49
.83
.39
.18
.10
<.01
.38
.68
Executive function
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
hormone, in years
baseline
treatment
BVAIT
432 .08 12
1.02
-.05
.21
.17
.15
.16
-.06
.62
-.01
.01
-.10
-.09
-.10
.92
.01
.07
.19
.10
.33
.76
.61
.45
.01
.02
.16
.18
1.11
-3.78
2.83
.93
1.47
.49
-.08
1.02
-.03
.54
-4.29
-.53
-.56
.27
<.01
<.01
.36
.14
.62
.94
.31
.98
.59
<.01
.59
.57
Semantic memory
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
hormone, in years
baseline
treatment
BVAIT
446 .08 12
-.24
-.01
.06
.17
.01
<.01
-.15
-.12
-.46
.01
-.17
.02
-.03
.49
.01
.04
.10
.05
.18
.46
.31
.23
.01
.03
.09
.10
-.49
-.86
1.46
1.67
.28
.00
-.33
-.38
-1.96
1.69
-5.58
.27
-.36
.63
.39
.15
.10
.78
1.00
.74
.70
.051
.09
<.01
.79
.72
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 119
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Logical memory
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
hormone, in years
baseline
treatment
BVAIT
459 .33 12
.11
-.03
.19
.05
.19
.08
-.28
.70
-.09
.01
-.63
.27
.08
.91
.01
.07
.18
.10
.33
.78
.63
.44
.01
.05
.16
.18
.12
-2.77
2.76
.30
1.93
.24
-.36
1.12
-.21
.72
-13.96
1.65
.44
.90
<.01
<.01
.76
.054
.81
.72
.26
.83
.47
<.01
.10
.66
Visual memory
intercept
age, in years
education
Caucasian
BMI category
h1
h2
h3a
h3b
hormone, in years
baseline
treatment
BVAIT
454
.21
12
1.43
-.04
.10
.16
.05
-.09
.56
.08
-1.04
<.01
-.30
.14
.10
.63
.01
.05
.12
.07
.22
.53
.40
.30
.01
.03
.11
.13
2.26
-4.49
2.17
1.32
.77
-.41
1.05
.21
-3.44
.01
-9.48
1.25
.79
.02
<.01
.04
.19
.44
.68
.29
.83
<.01
.99
<.01
.21
.43
Verbal learning
intercept
age, in years
education category
Caucasian
BMI category
h1
h2
h3a
h3b
hormone, in years
baseline
treatment
BVAIT
446 .18 12
.41
-.02
.14
.36
-.04
-.44
.19
.27
.20
.01
-.34
.02
.15
.72
.01
.06
.14
.08
.25
.60
.46
.34
.01
.04
.13
.14
.58
-2.50
2.38
2.60
-.52
-1.73
.31
.61
.58
.89
-9.12
.16
1.07
.56
.01
.02
.01
.60
.08
.76
.54
.56
.38
<.01
.87
.29
Note. ETRIAL = intercept - BVAIT – WISH
Treatment = 1 for those in the treatment group of WISH or BVAIT. Treatment = 0 for those in the Placebo
group of WISH or BVAIT.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 120
Table 18
Models Examining Years of Hormone Therapy and Oophorectomy on Cognitive Performance Over Time
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Global functioning
intercept
age, in years
education category
Caucasian
BMI category
ooph x horm
hormone, in years
oophorectomy
baseline
treatment
BVAIT
416 .09 10
2.73
-.13
.49
.35
.45
-.04
.07
-.18
-.13
.34
.16
2.13
.03
.17
.44
.23
.06
.04
.74
.03
.38
.41
1.28
-4.39
2.79
.79
1.94
-.57
1.89
-.25
-4.68
.92
.39
.20
<.01
<.01
.43
.053
.57
.06
.81
<.01
.36
.69
Executive function
intercept
age, in years
education category
Caucasian
BMI category
ooph x horm
hormone, in years
oophorectomy
baseline
treatment
BVAIT
432 .08 10
1.07
-.05
.21
.16
.15
-.02
.01
.32
-.09
-.08
-.09
.92
.01
.07
.19
.10
.03
.02
.32
.02
.16
.18
1.17
-3.90
2.80
.88
1.52
-.76
.94
.98
-4.24
-.52
-.54
.24
<.01
<.01
.38
.13
.45
.35
.33
<.01
.60
.59
Semantic memory
intercept
age, in years
education category
Caucasian
BMI category
ooph x horm
hormone, in years
oophorectomy
baseline
treatment
BVAIT
446 .07 10
-.20
-.01
.05
.17
.02
<-.01
.01
-.13
-.16
.03
<.01
.49
.01
.04
.10
.05
.01
.01
.17
.03
.09
.09
-.40
-.96
1.28
1.66
.38
-.24
1.75
-.76
-5.48
.30
.03
.69
.34
.20
.10
.71
.81
.08
.45
<.01
.77
.98
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 121
Model n R
2
df Parameter
estimate
Standard
error
t-value p
Logical memory
intercept
age, in years
education category
Caucasian
BMI category
ooph x horm
hormone, in years
oophorectomy
baseline
treatment
BVAIT
459 .33 10
.17
-.03
.19
.05
.20
-.01
.01
.15
-.63
.27
.09
.91
.01
.07
.18
.10
.03
.02
.32
.05
.16
.18
.18
-2.89
2.76
.29
1.97
-.32
.89
.46
-13.99
1.70
.49
.85
<.01
.01
.77
.049
.75
.37
.65
<.01
.09
.62
Visual memory
intercept
age, in years
education category
Caucasian
BMI category
ooph x horm
hormone, in years
oophorectomy
baseline
treatment
BVAIT
454 .20 10
1.56
-.04
.10
.17
.07
.02
<-.01
-.43
-.31
.15
.14
.64
.01
.05
.12
.07
.02
.01
.22
.03
.11
.12
2.44
-4.80
2.10
1.38
1.01
1.12
-.25
-1.99
-9.68
1.37
1.12
.02
<.01
.04
.17
.31
.26
.80
.047
<.01
.17
.26
Verbal learning
intercept
age, in years
education category
Caucasian
BMI category
ooph x horm
hormone, in years
oophorectomy
baseline
treatment
BVAIT
446 .17 10
.42
-.02
.15
.35
-.04
-.01
.01
-.07
-.34
.02
.07
.72
.01
.06
.14
.08
.02
.01
.25
.04
.13
.14
.59
-2.576
2.58
2.53
-.56
-.28
.88
-.27
-9.26
.15
.53
.56
.01
.01
.01
.58
.78
.38
.79
<.01
.88
.60
Note. ETRIAL = intercept - BVAIT – WISH
ooph x horm = oophorectomy status (dichotomous) x years on hormone treatment (continuous variable).
Treatment = 1 for those in the treatment group of WISH or BVAIT. Treatment = 0 for those in the Placebo
group of WISH or BVAIT.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 122
STUDY I AND II FIGURES
Figure 1. Sample for Study I.
Note. “Twins age 65+ in STR” represents the twins in the Swedish Twin Registry who were alive and
65 years or older when they were contacted during 1998 – 2001.
Shaded boxes indicate ineligible participants.
* Not included because twins were deceased, emigrated, or “classified,” or had an unlisted telephone
number.
** Not available because twins refused, were not reachable, were deceased prior to screening, had a
relative who could not be contacted to be interviewed, etc.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 123
women
in BVAIT
n = 197
women in ELITE
n = 643
women in WISH
n = 350
no cognitive
baseline
n = 1
uni ooph or
ovarian status
unknown n = 12
no cognitive
baseline
n = 1
uni ooph or
ovarian status
unknown n = 12
women
from BVAIT
n = 184
women
from WISH
n = 337
women
from ELITE
n = 638
no cognitive
baseline
n = 1
uni ooph or
ovarian status
unknown n = 14
total women for
cross sectional
analyses
n = 1,149
no cognitive
follow-up
n = 22
women
from BVAIT
n = 162
no cognitive
follow-up
n = 36
women
from WISH
n = 301
total women for
change score
analyses
n = 463
Figure 2. Participants in Study II
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 124
change in Z score at Test 2
-30
-20
-10
0
10
20
30
Z score at Test 1
-30 -20 -10 0 10 20 30
General cognitive functioning, for women without both ovaries
Z score at Test 2
-30
-20
-10
0
10
20
30
Z score at Test 1
-30 -20 -10 0 10 20 30
General cognitive functioning, for women with ovaries intact
change in Z score at Test 2
-30
-20
-10
0
10
20
30
Z score at Test 1
-30 -20 -10 0 10 20 30
Executive functioning, for women without both ovaries
Z score at Test 2
-30
-20
-10
0
10
20
30
Z score at Test 1
-30 -20 -10 0 10 20 30
Executive functioning, for women with ovaries intact
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 125
change in Z score at Test 2
-30
-20
-10
0
10
20
30
Z score at Test 1
-30 -20 -10 0 10 20 30
Semantic memory, for women without both ovaries
Z score at Test 2
-30
-20
-10
0
10
20
30
Z score at Test 1
-30 -20 -10 0 10 20 30
Semantic memory, for women with ovaries intact
change in Z score at Test 2
-30
-20
-10
0
10
20
30
Z score at Test 1
-30 -20 -10 0 10 20 30
Logical memory, for women without both ovaries
Z score at Test 2
-30
-20
-10
0
10
20
30
Z score at Test 1
-30 -20 -10 0 10 20 30
Logical memory, for women with ovaries intact
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 126
change in Z score at Test 2
-30
-20
-10
0
10
20
30
Z score at Test 1
-30 -20 -10 0 10 20 30
Visual memory, for women without both ovaries
Z score at Test 2
-30
-20
-10
0
10
20
30
Z score at Test 1
-30 -20 -10 0 10 20 30
Visual memory, for women with ovaries intact
change in Z score at Test 2
-30
-20
-10
0
10
20
30
Z score at Test 1
-30 -20 -10 0 10 20 30
Verbal learning, for women without both ovaries
Z score at Test 2
-30
-20
-10
0
10
20
30
Z score at Test 1
-30 -20 -10 0 10 20 30
Verbal learning, for women with ovaries intact
Figure 3. Z-Scores of domain performance in Study II
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 127
APPENDIX
Table A1
Incidence
age
group
A. num of
hyst, 1994 -
1999
B. est num
of hyst per
year
(=A 6)
C. % of
hyst
with bil
ooph
D. num of
bil ooph
with hyst
per year
(=B C)
E. % of hyst
with bil
ooph that
are benign
F. num of
bil ooph
with hyst
per year
that are
benign
(=D E)
G. num of
bil ooph
per year
that are
benign
15-44 1,845,693
1
307,616
3
40%
1
123,046
3,4
71%
3
87,363
3,6
45-54 998,476
1
166,413
3
75%
1
124,810
3,5
71%
3
142,071
3,6
>55 681,068
1
113,511
3
75,291
3
71%
3
all 3,525,237
1
587,540
3
55%
1
323,147
3
71%
1
229,434
3
all 600,000
1
55%
1
330,000
3
71%
1
234,300
3,7
all 300,000
2,7
Note. num = number, est = estimated, hyst = hysterectomy, bil ooph = bilateral oophorectomy.
1
data provided from Keshavarz, Hillis, Kieke, & Marchbanks, 2002
2
data provided from Parker, Jacoby, Shoupe, & Rocca, 2009
3
calculated or assumed from data provided
4
undergone surgical menopause, assuming natural menopause would have occurred over age 44
5
a portion of these women would have undergone surgical menopause, assuming natural menopause
would have occurred between ages 45 and 54
6
surgical menopause was a prophylactic decision for some or all
7
data may not be equal as women may have undergone bilateral oophorectomy without concomitant
hysterectomy
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 128
Table A2
Frequency of Oophorectomy by IDR Codes
n
earliest date
of surgery
latest date
of surgery
mean
number of
days
inpatient
range of
days
inpatient
bilateral oophorectomy
LAE20 5 1998-11-20 2004-07-25 10.20 6 - 14
LAE21 1 1999-01-28 1.00 1
LAF10 253 1997-01-12 2010-12-15 8.19 0 - 36
LAF11 16 1998-01-28 2010-01-20 3.44 1 - 6
LAF30
7021 32 1964-05-05 1995-07-10 11.50 4 - 25
7022 1 1977-08-08 8.00 8
7031 223 1966-04-25 1997-10-15 12.26 1 - 68
7032 3 1975-05-26 1980-11-09 10.00 9 - 12
total
534
unilateral oophorectomy
LAE10 2 2000-10-09 2000-09-30 11.00 7 - 15
LAE11 1 1998-04-30 1.00 1
LAF00 33 1998-03-18 2009-03-16 11.70 4 - 60
LAF01 7 1999-06-30 2009-04-20 1.71 1 - 3
LAF20 2 2000-07-17 2004-01-25 6.50 5 - 8
7020 25 1967-05-08 1994-04-08 11.92 0 - 39
7030 113 1964-04-21 1997-12-07 10.88 1 - 48
total 183
Note. This table includes all records of oophorectomy in the IDR for the 14,126 women in the
registry, including those not included in the analysis such as those who had surgeries after cognitive
screening.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 129
Table A3
Frequency of Bilateral Oophorectomy Surgeries by Date and Age Range According to IDR
n
number who
also self-
reported
accurately
date of surgery
through December 31, 1970 4 3
January 1, 1971 through December 31, 1975 26 11
January 1, 1976 through December 31, 1980 24 12
January 1, 1981 through December 31, 1985 26 17
January 1, 1986 through December 31, 1990 23 10
January 1, 1991 through December 31, 1995 26 12
after January 1, 1996 25 9
total
154 74
age at surgery
until 40 years old 0 0
41 through 45 years old 4 2
46 through 50 years old 28 16
51 through 55 years old 34 20
56 through 60 years old 24 13
61 through 65 years old 32 16
66 through 70 years old 19 5
71 through 75 years old 12 2
76 years old and over 1 0
total 154 74
Note. This data includes the true positives and the false negatives from the self-report.
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 130
Table A4
Three Trials for Study II
General Information
BVAIT WISH ELITE
Randomized, double-blind trial
Placebo-controlled
Study dates November 2000 –
February 2007
March 2004 – June
2009
July 2004 – ongoing
Purpose of the trial To examine
whether vitamin B
supplementation
will reduce the
progression of early
atherosclerosis in
individuals over 40
years old and
without clinical
evidence of
cardiovascular
disease.
To determine
whether soy
supplements can
reduce hardening of
the arteries in
postmenopausal
women. This study
will also determine
the effects of soy
supplements on
mental processes,
bone mineral
density, and breast
tissue density.
To examine the
effects of oral 17B-
estradiol (estrogen)
on the progression of
early (subclinical)
atherosclerosis and
cognitive decline in
healthy
postmenopausal
women.
Total enrollment, men and women;
women only
506; 127 350; 350 643; 643
Inclusion Criteria
BVAIT WISH ELITE
Age
40 years and older
30 years and older
none
Menopause status postmenopausal postmenopausal (no
vaginal bleeding for
at least 1 year and
having a serum
estradiol level higher
than 20 pg/ml)
postmenopausal less
than 6 years, OR 10
years or longer (no
periods for 6 months
or more and a serum
estradiol level 25
pg/ml or less)
Fasting plasma homocysteine 8.5
micromoles per liter (µmol/L) or
greater
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 131
Exclusion Criteria
BVAIT WISH ELITE
Clinical signs or symptoms of
cardiovascular disease
Diabetes mellitus or fasting serum
glucose 140 mg/dL or greater
Serum creatinine
(greater than 1.6
mg/dL)
(greater than 2.0
mg/dL)
(greater than 2.0
mg/dL)
Untreated thyroid disease
Uncontrolled hypertension (systolic blood
pressure 160
mmHg or greater
and/or diastolic
blood pressure 100
mmHg or greater)
(diastolic blood
pressure 110 mmHg
or greater)
Triglyceride levels 500mg/dL or
greater
Life expectancy less than 5 years
Unable to confirm that both ovaries
were removed
Current hormone replacement
therapy (HRT)
Alcohol intake greater than 5 drinks
per day or substance abuse
Soy- or nut-related food allergies
Follow a vegan diet
Those who have had a hysterectomy
only and no oophorectomy (since
time from menopause cannot be
determined)
Cirrhosis or liver disease
Deep vein thrombosis or pulmonary
embolism
History of breast cancer
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 132
Table A5
Neuropsychological Tests Administered in Study II
Executive
functioning
Verbal
learning
Logical
memory
Visual
memory
Semantic
memory
Global
functioning
Symbol Digit Modalities Test
X
X
Wechsler Memory Scale, 3rd
edition (WMS-III) Logical
Memory I Paragraph Recall,
Immediate Recall
X X
WMS-III Faces I
X X
WMS-III Letter-Number
Sequencing
X X
Judgment of Line Orientation
X X
Boston Naming Test (30-
items)
X X
Trail Making Test Part B
X X
WMS-III, Logical Memory II
Paragraph Recall, Delayed
Recall
X X
WMS-III Faces II
X X
California Verbal Learning
Test, 2nd edition (CVLT–
II) Immediate Recall
X X
Wechsler Adult Intelligence
Scale, 3rd edition Block
Design
X X
Category Fluency (animal
naming, 60 seconds)
X X
Shipley Institute of Living
Scale Abstraction Subset
Part II
X X
CVLT – II Delayed Recall
X X
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 133
COGNITIVE FUNCTIONING FOLLOWING OVARIAN REMOVAL 134
Figure A1. TELE
Abstract (if available)
Abstract
Estrogen has been associated with cognitive function, and cognitive difficulties become more prevalent in older age when estrogen is reduced. However, the long‐term cognitive function of older women who had reductions in estrogen through having their ovaries removed is not well understood. This dissertation examined the extent to which bilateral oophorectomy and its timing relative to natural menopause are associated with cognitive functioning through two complementary studies. Depressive symptoms as a mediator and hormone therapy as a protective factor were also examined. In Study I, twins over 65 years of age in the Swedish Twin Registry with linkages to the national Inpatient Discharge Registry were used in classic, case‐control (154 women with bilateral oophorectomy and 7,600 women with ovaries intact) and discordant co‐twin (n = 63 pairs) analyses to examine associations between oophorectomy and both cognitive impairment and scores on a cognitive screening test. In Study II, oophorectomy status and performance on global cognitive function and five cognitive domains among healthy postmenopausal women in 3 clinical trials conducted in Los Angeles County were analyzed in linear regression models both cross‐sectionally (n = 1,149) and on repeat cognitive assessment 2.5 years after baseline (n = 463). In both studies, covariates included age, education, and health factors. In Study I, neither oophorectomy nor its timing was associated with cognitive impairment. In Study II, oophorectomy after age 45 compared to no history of oophorectomy was a risk factor for poorer performance in verbal learning (p = .03) in the cross‐sectional design and with a decrease in visual memory (p < .01) performance over time. Oophorectomy status was associated with a decrease in visual memory (p < .05) over time. Neither study showed a significant interaction of oophorectomy and hormone treatment. However, in Study I, hormone treatment was protective against cognitive impairment (p < .02 for length of time on hormone treatment, p < .01 for history of any hormone treatment). In both studies, depressive symptoms were not found to be a mediator but were found to be associated with poorer cognitive function. The dissertation findings suggest that ovarian removal may be associated with poorer performance in specific cognitive domains but not with longer‐term cognitive impairment.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Personality and cognitive aging: are positive traits protective?
PDF
Neuropsychological functioning in older adult long-term cancer survivors
PDF
Cardiovascular disease risk factors and cognitive function
PDF
Relationship of blood pressure and antihypertensive medications to cognitive change in the BVAIT, WISH, and ELITE clinical trials
PDF
Association between endogenous sex hormone levels and cognition: the Women’s Isoflavone Soy Health (WISH) trial
PDF
Longitudinal relationships between bereavement and physical health‐related quality of life in middle‐ to older‐aged women
PDF
Estrogen and progesterone-based hormone therapy and the development of Alzheimer's disease
PDF
Association of carotid artery stiffness with measures of cognition in older adults
PDF
Subclinical carotid atherosclerosis, psychosocial measures, and cognitive function in middle- to older-aged adults
PDF
Cross-sectional association of blood pressure, antihypertensive medications, MRI volumetric measures and cognitive function scores in an aging population
PDF
Health-related quality of life correlates of the treatment decision making process of newly diagnosed prostate cancer patients
PDF
Effects of estradiol on cortisol response, working memory, and emotional memory during stress in young and post-menopausal women
PDF
The influences of anxiety, coping, and social support on physical functioning among heart failure patients
PDF
Psychosocial adjustment among patients with metastatic colorectal cancer
PDF
Effects of AT-1 receptor blockers on cognitive decline and Alzheimer's disease
PDF
Psychosexual adjustment among low-income Latinas with cervical cancer
PDF
Contributions of dynamic cerebrovascular function to cognitive decline and dementia: development and validation of a novel neuroimaging approach
PDF
Age-related androgen depletion and the development of Alzheimer's disease
PDF
Indexing cerebrovascular resistance in cognitive decline and Alzheimer's disease
PDF
Relationships between lifetime chronic stress exposure, vascular risk, cognition, and brain structure in HIV
Asset Metadata
Creator
Kurita, Keiko
(author)
Core Title
Cognitive functioning following ovarian removal before or after natural menopause
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Psychology / Public Health
Publication Date
08/10/2015
Defense Date
05/21/2014
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
cognitive function,Menopause,OAI-PMH Harvest,oophorectomy
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Meyerowitz, Beth E. (
committee chair
), Gatz, Margaret (
committee member
), Mack, Wendy Jean (
committee member
), Pike, Christian J. (
committee member
), Wood, Wendy (
committee member
)
Creator Email
kkurita@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-459253
Unique identifier
UC11287109
Identifier
etd-KuritaKeik-2804.pdf (filename),usctheses-c3-459253 (legacy record id)
Legacy Identifier
etd-KuritaKeik-2804.pdf
Dmrecord
459253
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Kurita, Keiko
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
cognitive function
oophorectomy