Neuropsychology Review, Vol. 9, No.

4, 1999

Review Article
Hormones and Cognition: Current Concepts and Issues in
Neuropsychology

David M. Erlanger,1'2 Kenneth C. Kutner,3 and Alan R. Jacobs3

This article provides an extensive and comprehensive review of the effects of hormones on cognition.
Studies detailing specific neurocognitive functions affected by variation in hormone levels across the
life span are presented. Dysregulation of hormone levels is considered from models of both normal
and diseased functioning. Patterns of cognitive dysfunction are described for a range of syndromes
involving the neuroendocrine system, and evidence of specific neurophysiological mechanisms that
can account for these findings is outlined. This review includes discussion of treatment outcomes and
the permanency of endocrine-related cognitive dysfunction. The authors present a set of guidelines for
clinical neuropsychologists to use for assessment of patients with neuroendocrine system dysfunction.
Clinical and methodological issues in research and treatment settings are discussed.
KEY WORDS: Neuropsychology; neuroendocrinology; hormones; cognition; review article.

INTRODUCTION The review is organized as follows: First, the en-

docrine system functions are briefly delineated. Second,
Human behavior is governed by two interdependent the neuroanatomical features and neurophysiological ac-
systems: the nervous system and the endocrine system. tions of the principal components of the neuroendocrine
Clinical neuropsychology and neuropsychological re- system are summarized. Third, neuroendocrine syndromes
search have been concerned primarily with the former. marked by dysfunction of multiple hormone systems are
Nevertheless, neuropsychologists frequently encounter noted, and principal pathological processes are described.
patients with endocrine system dysfunction in conditions Fourth, the literature demonstrating the discrete effects of
ranging from diabetes to Graves' disease. In pediatric, hormones on cognition is reviewed. Attention is paid to
adolescent, and adult patients, research has consistently neuropsychological and neurocognitive features, mecha-
shown profound effects of hormonal dysregulation on be- nisms of action, and assessment methods. The review con-
havior, affect, and cognition. These effects may be due cludes with a set of guidelines for assessment of patients
to severe syndromes or even to subclinical endocrine dys- with neuroendocrine dysfunction in clinical and research
function. Researchers and clinicians are increasingly settings and a rationale for providing interventions, in-
studying hormones for their potential in diagnosing and cluding cognitive remediation.
treating cognitive problems associated with a variety of
conditions that affect the central nervous system. This re-
FUNDAMENTALS OF ENDOCRINE FUNCTION
view of research is intended to update neuropsychologists
in clinical and research settings on the neuropsychology
The endocrine system produces its effects on bod-
of endocrine functions.
ily and mental functions by means of various hormones.
A hormone is a chemical substance that is secreted by
1
Department of Neurosciences, Columbia University, New York, New a specific endocrine gland and exerts physiological con-
York. trol on other cells. Unlike neurotransmitters, which act lo-
2
All correspondence should be directed to the author at 3 East 65th Street,
Suite #5B, New York, New York 10021.
cally, hormones may act on cells located at a distance. In
3
Department of Neurology, Weill Medical College of Cornell University, this way, hormones perform a range of functions, such as
New York, New York. maintaining metabolism, regulating the rates of chemical

175
l040-7308/99/l200-0175$16.00/0 c 1999 Plenum Publishing Corporation
176 Erlanger, Kutner, and Jacobs

Table I. Principal Endocrine Glands and Hormones

Anterior pituitary gland

Growth hormone causes growth of most bodily tissues and cells.
Prolactin promotes breast development and milk secretion.
Follicle-stimulating hormone is involved in reproductive and gonadal functioning.
Luteinizing hormone causes ovulation in females and stimulates the gonads in both sexes.
Adrenocorticotropin stimulates the adrenal cortex to secrete adrenocortical hormones.
Thyroid-stimulating hormone causes the release of thyroid hormones by the thyroid gland.
Posterior pituitary gland
Vasopressin causes the kidneys to retain water and regulates blood vessels.
Oxytocin facilitates childbirth and lactation.
Thyroid gland
Thyroxine and triiodothyronine regulate the chemical processes in cells.
Calcitonin causes the deposit of calcium in bones.
Adrenal cortex
Cortisol contributes to the control of proteins, carbohydrates, and fats.
Aldosterone regulates sodium and potassium levels.
Pancreas
Insulin promotes the storage of glucose into cells.
Glucagon promotes the release of glucose into the bloodstream.
Testes
Testosterone causes growth of male sex organs and secondary sex characteristics.
Ovaries
Estrogen causes growth of female sex organs and secondary sex characteristics.
Progesterone nurtures the developing fetus and causes development of milk apparatus.
Pineal gland
Melatonin regulates circadian rhythms and aspects of sexual behavior.
Parathyroid gland
Parathormone controls calcium ion concentrations in extracellular fluid.

reactions in cells, and promoting growth. Through their As noted, individual hormones may perform a num-
actions on the central nervous system, hormones influ- ber of roles and have a multiplicity of receptors. Not all
ence affect, sexual and social behavior, and cognition. the hormones included in Table I exert direct or even
Certain hormones can affect virtually all cells in the body, secondary effects on cognitive processes. As is the case
whereas others have specific receptors on various organs, with any chemical substance, a given hormone's effects
including the brain. Metabolic control of hormone lev- on behavior are due not only to the identity, dose, and
els is maintained by feedback loops of associated hor- duration of the hormone, but also to the brain substrate
mones; these loops are referred to as axes. For example, the upon which it acts. In this way, cognition may be affected
hypothalamic-pituitary-thyroid axis, or HPT axis, com- as a result of (a) abnormal amounts of hormones acting
prises TRH (a hypothalamic releasing hormone), TSH (a on a normal brain (e.g., Cushing's syndrome), (b) nor-
pituitary trophic hormone), and T3 and T4 (thyroid hor- mal amounts of hormones acting on an abnormal brain
mones), which provide feedback to the hypothalamus and substrate (e.g., premenstrual syndrome), or (c) abnormal
thus regulate stable thyroid hormone levels. amounts of hormones acting on an abnormal brain (e.g.,
Hormones are secreted at different rates (a) accord- congenital adrenal hyperplasia). Following is a review of
ing to metabolic needs, (b) according to periodic cycles the literature pertinent to the hormones known to influence
ranging from daily to monthly to seasonally, and (c) in cognition.
response to intrapsychic and environmental stimuli. In ad-
dition, levels of certain hormones vary widely during key
growth phases across the life span: in the fetal and peri- THE NEUROENDOCRINE SYSTEM: BASIC
natal environments, during adolescence, and as a natural NEUROANATOMY AND NEUROPHYSIOLOGY
consequence of aging.
The body creates more than 50 basic hormones. How- The Hypothalamus and Its Connections
ever, the essential elements of the endocrine system can be
summarized as including the major endocrine glands and The hypothalamus is well known to neuropsycholo-
their respective principal hormones as outlined in Table I. gists as the principal autonomic center of the brain, acting
Hormones and Cognition 177

on the sympathetic and parasympathetic nervous systems modulated by means of feedback loops and various neuro-
to maintain homeostasis and to prepare the body to deal transmitter mechanisms. Known as the master endocrine
with emergencies. No less important is the role of the hy- gland, the pituitary affects the body's overall metabolism,
pothalamus as a component of the limbic system; in this the nervous system, and many aspects of behavior both
role, it mediates drives for hunger and thirst, sexual activ- through the direct effect of its hormones and secondarily
ity, and aggression. In addition, the hypothalamus directly through its effects on other endocrine glands.
regulates the majority of endocrine functions, largely by Anatomically, the gland consists of two distinct com-
means of its regulatory effect on the pituitary gland. Be- ponents, the anterior pituitary (adenohypophysis) and the
haviorally, all these components may act in concert. For posterior pituitary (neurohypophysis), which are con-
example, an aggressive response will raise a person's heart nected by a relatively avascular region known as the pars
rate, induce anger, increase glucose concentrations, and intermedia. Stimulation of the anterior pituitary results
elevate serum cortisol levels. from the influence of specialized releasing and inhibitory
The hypothalamus has been referred to as the head hormones generated within the hypothalamus. The hor-
ganglion of the internal milieu. It is composed of a single mones in turn are secreted through small blood vessels in
grouping of small nuclei, weighing approximately 4 g and the hypothalamic-hypophysial portal system. The anterior
lying ventral to the thalamic nuclei. The inferior surface is pituitary secretes a number of hormones vital to the body's
exposed to the subarachnoid space and is bounded by the metabolic function and capacity to deal with stress. Known
optic chiasm, the optic tracts, and the mamillary bodies. as trophic hormones, these substances act to stimulate their
Although complex, the connections of the hypothalamus target organs at distances from the pituitary. Chief among
can be broken down into three principal divisions: these are growth hormone (GH), adrenocorticotropic hor-
mone (ACTH), thyroid-stimulating hormone (TSH), the
1. Reciprocal pathways connect the hypothalamus to
gonadotropins known as follicle-stimulating hormone
the brain stem and spinal cord, interconnecting var-
(FSH) and luteinizing hormone (LH), and prolactin.
ious visceral and somatic nuclei, both motor and
Stimulation of the posterior pituitary is accomplished
sensory.
by means of nerve tracts that originate in the supraoptic
2. Reciprocal pathways connect the hypothalamus with
and paraventricular nuclei of the hypothalamus and pass
the limbic system, including the septal area, hip-
through the pituitary stalk. These nerve signals cause the
pocampus, and amygdala. Additional afferents arise
secretion of the posterior hormones from secretory gran-
in the orbital cortex of the frontal lobes and optic
ules that lie on the surface of the capillaries. Chief among
tracts.
these hormones are vasopressin (VP), also known as an-
3. Efferent pathways connect to the pituitary gland.
tidiuretic hormone (ADH), and oxytocin.
These rich interconnections illustrate not only the wide-
spread influence of the hypothalamus, but also how ac-
tivation can be affected by autonomic processes as well COGNITION AND GENERALIZED
as through psychic and environmental stimuli. They also DYSFUNCTION OF THE NEUROENDOCRINE
underscore the interdependence of the discrete systems, SYSTEM
shedding light on why, for example, patients with tempo-
ral lobe epilepsy may experience thyroid dysfunction or Dysfunction of the hypothalamus, whether due to
why persons who are diabetic may experience memory trauma, vascular disease, primary hypothalamic neoplasm,
problems. or other etiology, typically manifests a combination of
symptoms entailing both autonomic functions and behav-
ioral responses. Symptoms include temperature dysreg-
The Pituitary Gland ulation, diabetes insipidus, obesity, aphagia and emaci-
ation, hypersomnia, and sexual dystrophy. Episodic fear
Immediately juxtaposed to the hypothalamus, within and rage reactions may also result. Severe compromise of
the sella turcica at the base of the brain, the pituitary hypothalamic functions may lead to stupor and coma.
gland, or hypophysis, measures approximately 1 cm in Commonly, hypothalamic dysfunction may result
diameter and weighs approximately 1 g. It is connected from compression secondary to a pituitary adenoma. Pitu-
to the hypothalamus by the pituitary stalk, which con- itary adenomas are age-linked, increasing in incidence into
tains the hypothalamic-hypophysial portal system. Pitu- the eighth decade. Macroadenomas (>10 mm) frequently
itary hormones are secreted primarily under stimulation come to medical attention as a result of their prominent
by hormones generated in the hypothalamus, and they are mass effect, producing symptoms such as headache, due to
178 Erlanger, Kutner, and Jacobs

traction of the meninges, and visual abnormalities, due to DISCRETE EFFECTS OF HORMONES
extension of the tumor into the suprasellar region, which ON COGNITION
results in compression of the optic chiasm or optic tract.
These problems are accompanied by generalized pituitary Growth Hormone
dysfunction manifesting as combinations of symptoms
such as those noted previously, due both to hypersecretion Growth hormone (GH), also known as somatotropic
of a hormone as a result of the tumor and to hyposecretion hormone or somatotropin, is necessary for promoting
of other hormones secondary to the tumor's mass effect growth and plays a role in other aspects of metabolic
on the pituitary gland and hypothalamus. Hormonal de- processes. These include suppressing glucose utilization,
clines due to mass effect follow a pattern: Hyposecretion increasing the rate of protein synthesis in cells, and in-
of GH precedes that of LH and FSH, which precedes that creasing the utilization of fatty acids for energy. Normal
of TSH, which precedes that of ACTH. Cognitive deficits serum levels are < 10 ng/ml for children and <5 ng/ml for
including amnesia and executive dysfunction may persist adults.4 It is secreted by the anterior pituitary under stim-
following treatment (Grattan-Smith et al., 1992; Peace ulation by the hypothalamic growth factor-releasing hor-
et al., 1997). However, the etiologies of such deficits ap- mone (GFRH) and itself stimulates production of insulin-
pear to be multifactorial. Possible factors include surgical like growth factor-I (IGF-I) in the liver.
sequelae, pre- and postsurgical hormonal imbalance, tu- GH variation in childhood is typically associated with
mor size, secondary hydrocephalus, and radiotherapy (see more widespread hypopituitary symptoms resulting in
McCord et al., 1997, regarding long-term outcome and dwarfism and permanent sexual immaturity, although for
sequelae). approximately one-third of this population there is a defi-
All adults experience age-related changes in multiple ciency in GH alone. Among this latter group, early treat-
hormone systems, including reduced production of GH, ment with GH therapy can completely cure the abnor-
thyroid hormones, sex hormones, certain adrenal andro- mality. The incidence of GH deficiency is approximately
gens, and pancreatic hormones (Lamberts et al., 1997). 1:4,000, and there is a higher incidence in boys than in
Moreover, adults may experience a generalized hypopitu- girls by a 2.5:1 ratio. Adults may acquire GH deficiency
itarism syndrome characterized by chronic pituitary insuf- as part of a generalized pituitary syndrome due either to
ficiency. Patients experience a range of symptoms such as tumor or to thrombosis of pituitary blood vessels. Symp-
decreased energy, loss of libido, depressed mood, and/or toms associated with these conditions are due to depressed
increased irritability. Women may experience irregular pituitary, thyroid, adrenal, and/or gonadal functioning and
menses or amenorrhea, and men may experience impo- include lethargy, decreased libido, and weight gain.
tence and fertility problems. GH may also be produced in excess as a result of
Pituitary microadenomas (<10 mm), as well as vas- dysfunction or tumor. If this occurs before adolescence,
cular lesions, inflammation, trauma, immune system dys- gigantism will occur. When the excess is caused by tumor,
function, and other etiologies, may also produce discrete the pituitary gland is typically damaged either by mass ef-
hypo- or hypersecretion of individual hormones. For sim- fect or by local invasion, which results in hyposecretion
plicity in compiling evidence regarding the cognitive ef- of other hormones. Hyperglycemia is also common as a
fects of individual hormones, these syndromes—such as result of the effect of GH on the pancreatic beta cells
acromegaly and Cushing's disease—are reviewed in the that regulate insulin production. If occurring after adoles-
next section with regard to the specific hormone or hor- cence, excessive GH manifests as acromegaly, a condition
monal axis involved, and not necessarily the gland of ori- in which the bones increase in thickness, especially those
gin. That is, the gonadotropic hormones FSH and LH are in the hands and feet, but also the membranous bones of
discussed in the context of the reproductive hormones be- the face, forehead, and jaw. Hypersecretion of GH is also
cause of their role in stimulating production of andro- associated with heart failure.
gens and ovarian hormones by the gonads. The hypothala- A number of studies have attempted to determine
mic releasing hormones corticotropin-releasing hormone the possible effects of GH on the brain and cognition. In
(CRH) and thyrotropin-releasing hormone (TRH), as well their study of 29 children with idiopathic GH deficiency,
as the pituitary hormones ACTH and TSH, can also affect Meyer-Bahlburg et al. (1978) found average IQ and no dif-
cognition. They do so, however, primarily through their in- ference in IQ between children with GH deficiency treated
fluence on the production of cortisol by the adrenal glands
and of thyroid hormones by the thyroid gland, respectively. 4
Throughout the text, hormone levels noted are approximate. Also, units
Their actions are therefore discussed in the next section of measurement vary. For simplicity, levels are supplied in the more
with regard to the glands they affect. commonly used unit formats.
Hormones and Cognition 179

with GH replacement therapy and those not treated. Two pituitary tumor, a head injury, or another disease process,
additional studies of short-stature children revealed simi- as well as of pituitary surgery, most typically that affecting
lar findings (M. Gordon et al., 1982; Stabler et al., 1994). the hypothalamus or upper pituitary stalk.
One double-blind, placebo-controlled study reported in- Although early animal studies indicated that learning
creased attentional functioning in children with GH de- could be facilitated by VP (e.g., De Wied, 1965), attempts
ficiency who received replacement therapy (M. Q. Smith to demonstrate this in humans have met with only limited
et al., 1985). success thus far. Short-term administrations of VP to older
Among adult populations, however, there is evidence subjects showed improvements that could be attributed
associating congenital GH deficiency with memory weak- to modestly enhanced memory or attentional processes
nesses. In a study comparing men with GH deficiency (Jennings et al., 1986; Nebes et al., 1984; Weingartner
with men with multiple hormone deficiencies (includ- et al., 1981) or to increased cortical arousal (Fehm-
ing GH deficiency) and healthy control subjects, Deijen Wolfsdorf and Born, 1991). Although one study suggested
et al. (1996) identified significant weaknesses on list- possible memory benefits for both older subjects and pa-
learning (Bouma and Lindeboom, 1988) and paired asso- tients with dementia (Legros and Gilot, 1979), two studies
ciates (Emmen et al., 1988) tasks as the factors associated found that neither short-term nor long-term VP therapy
with both patient groups, but not with the normal subjects. improved memory and other cognitive deficits in patients
Further, two studies demonstrated statistically significant with Alzheimer's disease (AD) (Peabody et al., 1985;
improvement in mnestic functioning, independent of de- Wolters et al., 1990).
pression and other psychological factors, in adults with In a double-blind, placebo-controlled study of older
GH deficiency who received GH therapy (Clopper, 1990; healthy volunteers administered long-term VP therapy,
Deijen et al., 1998). Perras et al. (1997) found an increased primacy effect on
Although no mechanism has yet been identified to the Rey Auditory Verbal Learning Test (Rey, 1964; Taylor,
account for these memory weaknesses, Deijen et al. (1996) 1959) for the subjects receiving VP therapy vs. the con-
noted the existence of GH receptors in the hippocampus trols (m = 42.6 ±3.1% vs. ,n = 37.2 ±3.1%, p < .05). A
(Lai et al., 1991), as well as the possible effects of GH on trend was also noted in the reduction of proactive interfer-
dopamine activity at that site. However, GH deficiency in ence effects on the distractor list. Although no improve-
Deijen's subjects was present at least since birth and likely ment in learning ability was evident, the results were con-
during the early fetal environment as well. It is not clear, sistent with increased attention and semantic encoding,
therefore, whether the assessed mnestic dysfunction was and possible decreased distractibility. The authors inter-
due to decreased circulating GH, chronic GH deficiency, preted these results as consistent with enhanced "mnes-
possible effects of GH deficiency on the developing brain, tic preactivation" because of the effect of VP on general
or some combination of these factors. arousal.

Vasopressin Introduction to the Reproductive Hormones

Vasopressin (VP), also known as antidiuretic hor- Virtually every discipline in the neurosciences, psy-
mone (ADH), is secreted by the posterior pituitary. It chology, sociology, and anthropology has conducted ex-
causes retention of water by the kidneys to regulate serum tensive investigations regarding the origin and nature of
osmololality (i.e., the relative concentration of water to gender differences in humans and other animals. After ge-
electrolytes in blood) and can also elevate blood pres- netic/chromosomal differences, which determine gender,
sure by the constriction of blood vessels throughout the hormonal differences have long been looked to as a leading
body. The normal plasma reference range is 2-12 pg/ml if factor for explanations of behavioral differences: gender
serum osmolality >290 mosm/kg, or <2 pg/ml if serum identity, sexual preference, emotional and social behavior,
osmolality is <290 mosm/kg. VP receptors are present in and cognition. Certainly, there is much evidence that hor-
numerous central nervous system locations, including the mones play an important role in gender differences. How-
hippocampi (Van Wimersma Greidanus et al., 1986). ever, regarding cognitive factors, caution is required in
Diabetes insipidus is caused by a lack of VP and attributing gender-based behavioral differences to repro-
is characterized by polyuria (passage of large quantities ductive hormone levels. Central to this caveat is the follow-
of dilute urine) and polydipsia (increased water intake), ing observation: Reproductive hormones are not gender
which together may lead to dehydration, causing stupor, specific. Furthermore, although androgens are present at
coma, and death. Diabetes insipidus may be the result of a higher levels in males, as are ovarian hormones in females,
180 Erlanger, Kutner, and Jacobs

small variations may produce larger effects in the "oppo- (absence of menses with persistent milk secretion) that is
site" sex as a result of the relative size of change from due to excessive prolactin secretion, which inhibits go-
baseline levels. Finally, because of synergistic effects of nadotropin production, no reports of cognitive sequelae
hormones on the same axis, a rise in one hormone may unique to this syndrome have yet been described.
cause a decrease in another, which makes causal attribu- Nevertheless, there is evidence that these hormones
tions difficult. may influence cognition. There is one study of LH and
As outlined in many of the following discussions of FSH levels and cognitive functioning in a normal pop-
the literature, in studies of group differences, women gen- ulation. H. W. Gordon and colleagues (H. W. Gordon
erally outperform men on certain verbal tasks, and men et al., 1986; H. W. Gordon and Lee, 1986) assayed FSH,
surpass women on certain visuospatial tasks (Halpern, LH, and gonadal hormones in healthy men and women
1992; Maccoby and Jacklin, 1974). However, the differ- who were administered a number of visuospatial and ver-
ences are small and there is considerable overlap in the bal/sequential tests from the Cognitive Laterality Battery
range of performance across groups. Existing evidence (H. W. Gordon, 1986). For males, FSH correlated neg-
suggests a relatively minor role for circulating levels of re- atively with the visuospatial tests, with high FSH levels
productive hormones in accounting for these differences. associated with poorer performance. In contrast, LH had
Variability in performance due to circulating hormone lev- positive, but weaker, correlations with both the visuospa-
els in healthy normal subjects has been identified primarily tial and the verbal/sequential tests. For women, FSH was
by utilizing very sensitive tests designed to detect sub- also negatively correlated with the visuospatial tasks, but
tle changes in the laboratory setting. For this reason, for less strongly and only after the effects of ovarian hormones
a better understanding of the role of hormones in cog- were partialled out. FSH and LH were both positively cor-
nition, researchers have largely turned their attention to related with the verbal fluency task. These results suggest
individuals with developmental abnormalities affecting that FSH and LH have independent properties that affect
hormone levels (e.g., congenital adrenal hyperplasia), per- cognition. Unfortunately, for populations in which signif-
sons with naturally occurring or surgically induced de- icant variation in gonadotropin levels occurs as a result
clines in hormone levels (e.g., menopause), or individuals of genetic and/or developmental abnormalities (e.g., id-
with changes in hormone levels due to disease processes iopathic hypogonadotropic hypogonadism, Turner's syn-
(e.g., AD). drome, Klinefelter's syndrome), it is not possible to parse
out the unique effects of the individual gonadotropins from
one another or from the effects of variation in gonadal hor-
Gonadotropins: Follicle-Stimulating Hormone mones. Also, genetic factors likely play a key role in the
and Luteinizing Hormone cognitive problems associated with these conditions. (For
discussions of these conditions, see Collaer and Hines,
Follicle-stimulating hormone (FSH) and luteinizing 1995, and H. W. Gordon et al., 1988).
hormone (LH), referred to jointly as gonadotropic hor-
mones, are secreted by the anterior pituitary gland and
stimulate the production of gonadal hormones: andro- Androgens
gens, estrogens, and progestins. However, their functions
within the sexes differ. In the male testes, FSH promotes In males, androgens are secreted principally by the
spermatogenesis and LH promotes testosterone produc- testes. By far, the most abundant of these is testosterone.
tion. Normal adult male serum levels are 2.25-20 IU/L Almost nonexistent during most of childhood, testosterone
and 3.6-17.1 IU/L, respectively. In the female ovary, FSH is nevertheless produced in large quantities in the newborn
causes follicle growth and estrogen production, and LH in- male in the first few months, as well as following the onset
duces ovulation and promotes androgen production. Nor- of puberty. Normal levels of total serum testosterone are
mal serum levels in adult premenopausal women are 300-1,000 ng/dl for males. Women also produce endoge-
5-24 IU/L and 4.5-24.3 IU/L, respectively, and are con- nous testosterone, with normal levels of approximately
siderably higher during the midcycle peak. Normal FSH 30-70 ng/dl.
and LH levels in postmenopausal females are 50-300 IU/L The term androgen refers to any steroid hormone
and 29-189 IU/L, respectively. with masculinizing effects. Although 95% of these are
The gonadotropins' role in cognition is not clear, produced in the testes, androgens are produced at other
in part because receptors have not yet been identified in sites as well, including the adrenal glands. Because of
the brain. Although pituitary adenomas are frequently the the low rate of production, adrenal androgens do not typ-
source of a syndrome known as amenorrhea/galactorrhea ically cause significant masculinizing activity. However,
Hormones and Cognition 181

excessive masculinization can occur secondary to over- measured by means of route-learning tasks and mental
production of these androgens due to dysfunction, such as rotation tasks, although many neuropsychological stud-
occurs in congenital adrenal hyperplasia (CAH). ies examine performances on visuospatial tasks such as
The principal functions of the androgens include their the Wechsler Adult Intelligence Scale—Revised (WAIS-
role in spermatogenesis and masculinization of the fetus R: Wechsler, 1981) Block Design subtest or across a range
and the development of male secondary sex characteris- of related tasks, such as those composing the WAIS-R Per-
tics. Regulation of production is maintained by a feedback formance domain subtests.
loop, which begins with the production of gonadotropin- There is support from the animal literature for en-
releasing hormone (GnRH) by the hypothalamus. This in hanced spatial ability in males. In a number of nonhuman
turn stimulates the secretion of the two anterior pituitary species, spatial-navigational ability is sexually differen-
gland gonadotropic hormones: LH, the main stimulus for tiated, with males learning to use routes more efficiently
testosterone production by the testes, and FSH, which (Beatty, 1984). Further, relative to controls, male mice cas-
principally stimulates spermatogenesis. Gonadotropin trated at birth exhibit reduced spatial efficiency in adult-
secretion can be affected by environmental and intrapsy- hood (Williams et al., 1990).
chic stimuli as well, particularly those involving the lim-
bic system. For instance, the amygdalae give large af-
ferent projections that directly modulate the function of Congenital Adrenal Hyperplasia
the hypothalamic neurosecretory cells. A feedback loop
is provided by the action of testosterone on the hypo- Excessive levels of androgens are associated with
thalamus. congenital adrenal hyperplasia (CAH), an endocrine dis-
There is reason to believe that the presence of andro- order beginning in the early prenatal environment. The
gens underlies morphological differences between the two condition is typically detected at birth, and androgen lev-
genders in the human brain. Differences in the size of the els are normalized. Although early studies found increases
anterior hypothalamic nuclei (Allen et al., 1989; Swaab in Full Scale IQ in patients with CAH, these findings were
and Fliers, 1985) and a number of midline structures, apparently largely due to selection bias (Nass and Baker,
such as the splenium of the corpus callosum and the ante- 1991a). Nevertheless, because of their high androgen lev-
rior commissure, have been identified (Allen and Gorski, els in utero, patients with CAH may be expected to show
1986; Holloway and de LaCoste, 1986). Geschwind and enhanced spatial ability. Researchers have focused largely
colleagues (Geschwind and Behan, 1982; Geschwind and on females with CAH because identifying enhanced spa-
Galaburda, 1985a,b) hypothesized that perinatal exposure tial ability in males with CAH may be more difficult.
to testosterone differentially affects development of the Research by Resnick et al. (1986) utilizing a sample of
right cerebral hemisphere, citing enhanced spatial abil- adolescent females with CAH revealed selectively better
ity, increased frequency of left-handedness, and greater performances on the Mental Rotations Test (Vandenberg
incidence of verbal learning disabilities in males. Behav- and Kuse, 1978), the Card Rotation Test (Ekstrom et al.,
iorally, androgens have been associated with sexual drive 1976), and a hidden figures test, compared with the perfor-
and increased aggression in both genders (see Hines and mance of unaffected siblings. Strengthening the case for
Green, 1991). However, because aggressive encounters al- a prenatal organizational effect of androgens on the de-
ter testosterone levels, no clear cause-effect relationship veloping brain, Hampson et al. (1998) found a similar ad-
between the two variables has been established. Males re- vantage (= 1 SD) in spatial capacity among preadolescent
ceiving testosterone replacement therapy report enhanced females with CAH (compared with controls), by utilizing
well-being and increased energy. However, testosterone in the Spatial Relations Test from the Primary Mental Abil-
males is converted, or aromatized, into estrogens, which ities battery (Thurstone and Thurstone, 1963). Although
obscures the interpretation of these reports. other researchers have not found strength on certain visuo-
Research on a possible role for androgens in cogni- spatial tasks in females with CAH (Baker and Ehrhardt,
tion has focused on spatial abilities because statistically 1974; McGuire et al., 1975; Perlman, 1973), it is likely
significant differences favoring males are consistently re- that discrepant findings may be attributed to differences
ported (see Linn and Petersen, 1985). Moreover, these dif- in test sensitivity (Collaer and Hines, 1995). In addition,
ferences become apparent following the onset of puberty Hampson et al. (1998) found that boys with CAH scored
(Maccoby and Jacklin, 1974) and have been identified significantly lower (>1 SD) than did control boys on the
across various groups for many decades, which suggests Spatial Relations Test, which suggests that overexposure
that differences are not likely due to psychosocial factors to androgens may have a "demasculinizing" effect on spa-
(Gladue and Baily, 1995). These capacities are typically tial ability in males.
182 Erlanger, Kutner, and Jacobs

Significant Verbal IQ-Performance IQ (VIQ-PIQ) levels (Gouchie and Kimura, 1991; Shute et al., 1983).
discrepancies favoring performance domain abilities have Recently, Moffat and Hampson (1996) reported a curvi-
been reported among female patients with CAH, relative linear relationship between testosterone and a composite
to the abilities of unaffected sibling controls (Nass and spatial abilities score obtained from the Mental Rotations
Baker, 1991b). Similarly, using controls matched for IQ Test (Vandenberg and Kuse, 1978) and the Paper Folding
in their study of adult females with CAH, Helladay et al. Test (Ekstrom et al., 1976) that was evident only in right-
(1994) found greater discrepancies on verbal vs. visuospa- handed subjects. As in the earlier studies, testosterone was
tial/nonverbal reasoning tasks favoring the latter domain positively correlated with a spatial task for right-handed
among the patient sample. In this study, however, the females but negatively correlated with spatial abilities for
observed differences appeared to be due largely to su- right-handed males. Notably, this finding is consistent with
perior logical inductive reasoning scores, not to spatial the previously described evidence of the masculinizing ef-
abilities. The authors postulated that these inconsistent fects of excessive androgen exposure in girls with CAH
findings could be attributed to the presence of low andro- and the demasculinizing effects of excessive androgen ex-
gen levels in the CAH group at the time of assessment, and posure in boys with CAH (Hampson et al., 1998).
suggested that this finding supported other research indi- Two additional lines of inquiry further suggest that
cating that variation in circulating androgen levels may optimal, not extreme, levels of testosterone may be asso-
influence spatial abilities in humans. ciated with enhanced spatial ability in normal subjects.
Janowsky et al. (1994) administered testosterone supple-
ments to men undergoing normal age-associated declines
Androgen Insensitivity in androgen levels. Subjects' performance on the WAIS-
R (Wechsler, 1981) Block Design subtest was significantly
Patients with androgen insensitivity (AI) are genetic (p<.04) improved (/* = 27.96±7.56 to /x = 30.17±
males who produce androgens but manifest partial to to- 6.78) relative to that of controls (/ti = 28.72 ± 8.61 to m =
tal insensitivity of androgen receptors. Depending on the 27.90 ± 8.57). In contrast, performances on the California
degree of insensitivity, they either are born with external Verbal Learning Test (CVLT: Delis et al., 1987), Wechsler
female genitalia (and no female reproductive organs)— Memory Scale—Revised (WMS-R: Wechsler, 1987) Vi-
total AI—and are raised as girls, or are born with ambigu- sual Reproduction subtest, Grooved Pegboard test (K10ve,
ous genitalia—partial AI—and are raised as either girls or 1963), and Trail Making Test (Reitan and Davison, 1974)
boys (Grumbach and Conte, 1992). Patients with AI typi- were unchanged for both experimental and control groups.
cally demonstrate a VIQ-PIQ discrepancy with decreased Additional support for the influence of circulating andro-
PIQ (Imperato-McGinley et al., 1991;Masica et al., 1969; gens on visuospatial functions has been found in mea-
Perlman, 1973). In the Imperato-McGinley et al. (1991) sures of natural fluctuation in testosterone. Moffat and
study, patients with AI were compared with both male and Hampson's study (1996, discussed previously) also mea-
female controls, which suggests that the differences were sured subjects at a time interval that allowed assessment
not due to psychosocial factors. However, it was ambigu- according to the diurnal drop in testosterone levels. In
ous whether the lower PIQ and individual subtest scores keeping with other findings, for men, higher levels of
were due to visuospatial deficits or attentional factors on testosterone were associated with more errors on spatial
speeded tests. tests, and for women, errors correlated with lower levels.
Finally, studies of transsexuals undergoing cross-sex
hormone treatment prior to surgery (Slabbekoorn et al.,
Androgens in Normal Subjects 1999; Van Goozen, 1994; Van Goozen et al., 1995) sup-
port an activating role for testosterone. In these studies,
Studies of androgens in normal populations suggest female-to-male patients undergoing testosterone supple-
a possible inverted U-shaped curve regarding the effects mentation demonstrated improved spatial performance,
of androgens on spatial tasks. Using somatic indices of an- and male-to-female patients receiving antiandrogens and
drogenization such as muscle-fat distribution, genital and estrogen showed declines in spatial ability as measured by
breast size, body shape, and amount of pubic hair, Petersen the Card Rotation Test (Ekstrom et al., 1976) and the Ro-
(1976) found that more-masculine women outperformed tated Figures, Three-Dimensional (Vandenburg and Kuse,
less-masculine women on spatial tasks. More-masculine 1978). In the most recent of the studies, the effects did not
men, however, performed worse than less-masculine men disappear after termination of cross-sex hormone therapy
on the same tasks. These results have been tentatively for a period of 5 weeks. Instead, the differences continued
supported by studies that directly measured testosterone to increase slightly.
Hormones and Cognition 183

An alternative hypothesis exists for certain of the pre- occurring on average at age 51, estradiol and progesterone
ceding findings, because testosterone acts to lower estro- production declines sharply.
gen levels in females. It is therefore possible that estrogen Of the estrogens, yS-estradiol is far more potent than
has a negative effect on spatial ability for the tests involv- either estrone or estriol. Their principal functions include
ing female subjects. stimulating the development of adult female sex organs
and secondary sex characteristics. Serum levels range
Summary from 20-100 pg/ml in the early follicular phase to 100-
350 pg/ml in the preovulatory and luteal phases. In adult
Despite the evidence outlined associating androgens males, serum estradiol typically ranges from 20-50 pg/ml.
with visuospatial ability, it is unclear how circulating an- Of the progestins, progesterone is generally considered to
drogens influence cognition. Evidence exists that pre- and be the single most important hormone. In addition to its
perinatal exposure to androgens may lay the neural sub- role in reproductive capacity, progesterone aids in nutri-
strate for later visuospatial ability. There is also support for tion of the fertilized ovum. Serum levels range in females
a role for testosterone in producing a transient activation from 0.3-0.8 ng/ml in the follicular phase to 4-20 ng/ml in
of spatial ability (either through a direct effect or through the luteal phase. In males, serum levels range from 0.12-
its influence on estrogen levels) according to an inverted 0.3 ng/ml.
U-shaped curve. However, it is unclear to what degree cir- Affective features associated with normal variation
culating androgen effects may depend on or interact with in ovarian hormone levels have been the subject of nu-
the neural substrate. The finding of stronger correlations merous investigations. Premenstrual syndrome (PMS) is
of circulating testosterone levels and visuospatial skills a cyclical disorder with depressive/mood-related and so-
for right-handed subjects suggests a contribution of each matic symptoms occurring during the luteal phase of the
of these two factors. menstrual cycle and clearing with the onset of menstru-
ation. A recent study by P. J. Schmidt et al. (1998) has
demonstrated that PMS symptoms are not due to increased
Ovarian Hormones levels of reproductive hormones, but to an abnormal re-
sponse of the brain substrate to normal changes in hor-
The female reproductive hormones (estrogens and mone levels. Menopause has been researched because of
progestins) are primarily secreted by the ovaries. As in its association with depressive disorders. Although there
the male, these hormones are secreted in response to the is evidence that major depression is more common in post-
two anterior pituitary gland gonadotropic hormones: LH menopausal women (Weissman, 1996), estrogen replace-
and FSH. These are in turn regulated by the hypothala- ment therapy (ERT) alone does not appear to alleviate
mic hormone GnRH. However, the female hormonal sys- depressive symptoms (M. A. Schneider et al., 1977). The
tem differs markedly from the male hormonal system re- utility of ERT for the prevention of major depression and
garding hormonal function and the regulation of hormonal as an adjunct to antidepressant therapy is currently under
cycles. investigation. In contrast to the depressive features asso-
Female reproductive hormones are thought to be ciated with declines in estrogen, there is evidence from
largely responsible for the feminization of the brain dur- research on epileptic disorders that progesterone is a po-
ing fetal stages (Gorski, 1991). With the onset of puberty tent anxiolytic by means of its role as a y-aminobutyric
and lasting until menopause, these hormones are secreted acid (GABA) agonist (Herzog, 1991). Indeed, studies have
at vastly different rates according to the female menstrual demonstrated that progesterone has significant anticonvul-
cycle. Approximately every 28 days, LH and FSH cause sant effects (Herzog, 1995).
the growth of new follicles within the ovaries. During the Cognitively, female hormones are thought to account
first 2 weeks, estrogen levels rise steadily to a midcycle in part for sexually dimorphic characteristics in brain struc-
peak and then dip just after ovulation. Following ovu- ture and function, particularly with regard to verbal abili-
lation, progesterone production begins in the corpus lu- ties in general (Linn and Petersen, 1985) and verbal mem-
teum and is maintained at a high level for approximately ory in particular (Sherwin, 1994). Morphologically, Hines
2 weeks. In the absence of fertilization, the corpus luteum et al. (1992) pointed to the relatively larger size of the
then ceases progesterone and estrogen production. These female corpus callosum as well as greater language lat-
declines mark the onset of menstruation and a new ovarian eralization in males as underlying this female verbal ad-
cycle. Beginning when a woman reaches approximately vantage. An ongoing role for estrogen in the regulation
40 years of age, the ovaries produce decreasing amounts of verbal abilities is suggested by the emergence of this
of estradiol, the principal estrogen. Following menopause, advantage following the onset of puberty (Maccoby and
184 Erlanger, Kutner, and Jacobs

Jacklin, 1974). Sherwin (1998) has proposed that the speci- in the placebo condition demonstrated a significant decline
ficity of estrogen's effect (i.e., on verbal memory) suggests on the verbal paired associates task. Similarly, paragraph
that estrogen serves to activate neural pathways estab- recall was found to be significantly decreased (%1 SD)
lished under the influence of this hormone during prenatal in a group of premenopausal women with benign uterine
life. neoplasms who were administered an agent causing sup-
pression of estrogen production. This decrement reversed
Estrogen Levels and Cognition with "add-back" ERT (Sherwin and Tulandi, 1996).
Animal studies support the existence of a role for es-
Evidence of a role for estrogen in accounting for trogen in enhanced memory processes during the estrous
enhanced verbal skills in females comes from research (i.e., menstrual) cycle. Estrogen promotes the formation
on women during their menstrual cycle. An early study of dendritic spines and synapses (Wooley and McEwen,
compared preovulatory women (in whom estrogen lev- 1993) and enhances long-term potentiation in the rat hip-
els were high) with women taking birth control pills (in pocampal CA1 region (Warren et al., 1995). Ovariectomy
whom estrogen levels were low) and with men. The pre- has been shown to decrease hippocampal synaptic den-
ovulatory women performed better than the control groups sity, a process that can be reversed with the addition of
on tests of color naming and color reading (Komnenich estradiol (Wooley and McEwen, 1993). Two behavioral
et al., 1978). More recently, Hampson (1990a,b) found learning paradigms utilizing ovariectomized rats have as-
modestly improved performance on speeded tests of ver- sociated a lack of estrogen with deficits on learning tasks
bal articulation (Mateer and Kimura, 1977), color naming and estradiol replacement with the maintenance of learn-
(Uhlmann, 1962), and several fine motor tasks during the ing (O'Neal et al., 1996; Singh et al., 1994). Estrogen is
high-estrogen phase of the menstrual cycle. Keenan et al. also known to enhance activation of the cholinergic sys-
(1992) replicated earlier findings of improved color nam- tem (Toran-Allerand et al., 1992), which could account in
ing (Stroop Test: Stroop, 1935) and found greater mental part for the aforementioned findings associated with pro-
flexibility (Trails B: Reitan and Davison, 1974) as well. cessing by the basal forebrain and frontal lobes in general
Also, Phillips and Sherwin (1992) found an association (i.e., color naming and mental flexibility).
between the high-estrogen phase of the menstrual cycle One study provides direct evidence that the hormonal
and the Paired Associates subtest of the Wechsler Memory milieu modulates the prefrontal cortex in humans. Berman
Scale (Wechsler, 1974). However, using the Cognitive Lat- et al. (1997) found increased cerebral blood flow in phar-
erality Battery (H. W. Gordon, 1986), H. W. Gordon and macologically controlled conditions in young women by
colleagues (H. W. Gordon et al., 1986; H. W. Gordon and using positron emission tomography (PET) scans obtained
Lee, 1993) twice failed to find relationships between ovar- during performance of the Wisconsin Card Sorting Test
ian hormone levels and gender-based verbal/sequential (Berg, 1948; Grant and Berg, 1948). However, although
ability according to natural hormonal variations in women. patterns of enhanced activation were found for both estro-
Although the authors described a number of possible con- gen and progesterone, there was no change in task perfor-
founding factors, their results point to the need for further mance parameters.
research in this area.
Another set of investigations has examined the re- Postmenopausal Estrogen Replacement Therapy
sponse to ERT in premenopausal women undergoing med-
ical procedures that disrupted estrogen production. In a ERT studies provide another format for consideration
series of well-controlled studies (Phillips and Sherwin, of the effects of estrogen on cognition in human females.
1992; Sherwin, 1988; Sherwin and Phillips, 1990), cogni- Following menopause, the ovaries virtually stop produc-
tive performances were examined in premenopausal ing estradiol, and a weaker hormone produced by the
women before and after surgically induced menopause adrenal glands, estrone, becomes the predominant estro-
(e.g., hysterectomy) under ERT and placebo conditions. gen. ERT has been used for many years because of its vaso-
Subjects were also compared with women who had un- active properties that reduce the risk of stroke (Paganini-
dergone hysterectomy but whose ovaries were not re- Hill et al., 1988), heart disease (Barrett-Connor and Bush,
sected. Using an adaptation of the Wechsler Memory Scale 1991), and vascular dementia (Funk et al., 1991). ERT is
with alternate forms (Russell, 1975; Stone et al., 1946; also known to prevent osteoporosis (Lindsay et al., 1976).
Wechsler, 1945), the investigators found that paragraph- Despite these potential benefits, a controversy exists re-
recall capacity was maintained at presurgical levels for all garding the utility of ERT for women with a family his-
women whose estrogen levels remained constant, whether tory of breast cancer (see Colditz et al., 1995; Grady and
as a result of natural secretion or of ERT. Moreover, women Ernster, 1991).
Hormones and Cognition 185

Memory problems are a frequent complaint of post- declined. Along these lines, in a recent study of older men
menopausal women (Anderson et al., 1987). Because and women (/x age = 72.1), Carlson and Sherwin (1998)
memory functions that require the encoding of new in- found significantly better performances on Digit Span To-
formation are increasingly compromised with age in both tal and Digit Span Forward subtests of the WAIS-R for
genders (Craik, 1984), these memory problems are not ERT users and men than for nonusers. Women ERT users
likely due entirely to low circulating levels of estrogen. also had higher Digit Span Backward scores than those of
Nevertheless, studies of ERT in both normal and patient nonusers. Women ERT users scored higher than but did
populations suggest a role for estrogen in maintenance not differ significantly from nonuser and male subjects on
of cognitive functions, and perhaps verbal memory in the Logical Memory and Paired Associates subtests of the
particular. WMS-R (Wechsler, 1987) and on a selective reminding
The first such study examined cognitive perform- task (Buschke and Fuld, 1974).
ances of two groups of 75-year-old females who were In contrast, in a prospective/cross-sectional study of
administered either estradiol or a placebo for 12 months. a large population of older, educated women, Barrett-
Those in the treatment group demonstrated increases in Connor and Kritz-Silverstein (1993) found no verbal mem-
VIQ and memory, whereas those in the placebo group ex- ory or other cognitive advantages for women who had un-
perienced declines in both domains. After discontinuation dergone ERT for 20 or more years, compared with women
of estrogen treatment, the cognitive abilities of those in the who had discontinued ERT or who had never been placed
experimental group declined to below pretreatment levels on it. D. M. Jacobs et al. (1998) noted numerous demo-
(Caldwell, 1954; Caldwell and Watson, 1952). However, graphic and ERT history differences between these two
the possible inclusion of a number of subjects with mild study populations as factors that could account for the
dementia may have confounded these findings. discrepant findings. Supportive of the Barrett-Connor and
A number of subsequent studies yielded inconsis- Kritz-Silverstein findings, Yaffe, Sawaya, et al. (1998)
tent results: Three (Feydor-Freybergh, 1977; Hackman concluded in their meta-analysis that ERT benefits cog-
and Galbraith, 1977; R. Schmidt et al., 1996) reported en- nition only in symptomatic, recently menopausal women.
hanced general cognitive functioning with ERT, and three Unfortunately, their meta-analysis did not include the stud-
(Polo-Kantola et al., 1998;Rauramo et al., 1975; Vanhulle ies of D. M. Jacobs etal. (1998) and Carlson and Sherwin
and Demol, 1976) reported no effect. Differences among (1998), which were published at about the same time.
the subject populations, experimental designs, functions
assessed, and measures make it difficult to reconcile these
disparate findings. However, several of these studies may Estrogen and Alzheimer's Disease
be reconciled by the hypothesis that short-term ERT is
not as effective as long-term ERT in enhancing general A potential role for female reproductive hormones
cognitive functioning in symptomatic women. regarding Alzheimer's disease (AD) is suggested by the
Recently, Kampen and Sherwin (1994) found that disease's greater prevalence among women than men, even
women on ERT performed better on the Story Memory test after differences in life expectancy have been adjusted for
from the Wechsler Memory Scale (Russell, 1975; Stone (Jorm et al., 1987). Likewise, several studies have found
et al., 1946; Wechsler, 1945) than did woman never on that women with AD perform worse than men with AD
ERT who had been matched for age, education, and socioe- on various verbal tasks (Henderson and Buckwalter, 1994;
conomic status. Similarly, in a community-based epidemi- Ripich et al., 1995), despite the women's premorbid ad-
ological study of 727 women, D. M. Jacobs et al. (1998) vantage on such tasks. A number of other clinical and
found that women who had used estrogen replacement experimental findings, including body weight, neurophys-
scored higher on measures of verbal memory (Buschke iological processes, and genetic mechanisms, further sup-
Selective Reminding Test: Buschke and Fuld, 1974), nam- port the relevance of estrogen to AD (see Henderson, 1997,
ing (Boston Naming Test: Kaplan et al., 1983), and ab- for a discussion).
stract reasoning (WAIS-R, Similarities: Wechsler, 1981) Clinically, ERT has been used to treat women with
than did nonusers, regardless of duration of treatment, de- AD for many years. In a number of studies, ERT has
mographic factors, or Apolipoprotein (APOE) genotype. been shown to be useful in improving cognition in gen-
All tests were significant for p < .05, with differences eral (Fillit et al., 1986; Honjo et al., 1989, 1993; Ohkura
averaging about 0.5 SD. Follow-up data were available et al., 1994a,b). Some limitations noted in these studies
for a subset of subjects. Women with histories of ERT include small sample size and the use of general cognitive
showed improved verbal memory relative to their own screening instruments, not norm-referenced neuropsycho-
baseline performances, whereas nonusers' performances logical tests. Also, certain of these studies were criticized
186 Erlanger, Kutner, and Jacobs

in part for not parsing out cognitive improvements due trend of decreasing odds ratios with increasing duration
to mood enhancement. In a 9-month longitudinal study of ERT use.
of nondepressed women with mild AD, Birge (1997) re- Prospective epidemiological studies also suggest a
ported improvement in general cognitive functioning in 8 significant reduction in risk for AD in women receiv-
of 10 subjects receiving ERT and either no change or de- ing ERT. Henderson, Paganini-Hill, and colleagues
cline in 10 control subjects. Results consistent with these (Henderson et al, 1994; Paganini-Hill and Henderson,
findings were reported for women receiving ERT as part 1994) found an approximatly 30% lower risk for ERT-
of a large study of the effects of tacrine (Knapp et al., using females than that for matched controls in their study
1994; L. S. Schneider et al., 1996). A recent double- of dementia at time of death in upper-middle-class women
blind, placebo-controlled, parallel group study examined at the Leisure World retirement community. Researchers
the response to transdermal estrogen therapy in post- for the Baltimore Longitudinal Study of Aging (Kawas
menopausal women with AD (Asthana et al., 1999). Pa- et al., 1997; Morrison et al., 1996) found a reduced risk of
tients self-corrected mistakes on the Stroop (Stroop, 1935) more than 50% for women who had ERT, compared with
interference trial more frequently (p < .03) and recalled that for women who never had ERT. Notably, these investi-
more words (p < .02) on the delayed cued recall condi- gators did not find an effect for duration of ERT. Similarly,
tion of the Buschke Selective Reminding Task (Buschke, Tang et al. (1996) found a 60% reduced risk and a later
1973). Performances declined when ERT was dis- age of AD onset for women with a history of ERT. They
continued. found that this was true even for women with other signif-
Henderson et al. (1996) designed a study to identify icant AD risk factors such as possession of the APOE s4
specific cognitive functions enhanced by ERT in an AD genotype. Only one study, by Brenner etal (1994), found
population. Men and women with similar levels of AD no reduced risk, and this discrepancy may be partly due to
symptoms were matched to women with AD who were differences in ERT delivery mechanisms (i.e., oral vs. oral
receiving ERT. The women in the treatment condition per- plus injection or cream vs. injection or cream only) consid-
formed significantly better than the women in the control ered in that study. In their meta-analysis, Yaffe, Sawaya,
group on the Boston Naming Test (Kaplan et al., 1983; et al. (1998) concluded that there is a 29% reduced risk of
H = 34.8 ± 13.1 vs. /i = 19.1 ± 12.2, p < .003), the Digit AD for estrogen users.
Span Forward (/z = 5.8±0.7 vs./x, = 4.2±2.1,p < .008)
and Digit Span Backward (/z = 3.8±2.0 vs. /z= 1.9 ±
1.7, p < .01) subtests of the WAIS-R (Wechsler, 1981), Summary
and a modified Clock Drawing task (Henderson et al.,
1989, after Goodglass and Kaplan, 1983; //= 11.1 ± 3.6 The data are inconsistent regarding circulating lev-
vs. fi = 6.4 ±5.2, p<.0l). Women on ERT also per- els of estrogen and cognitive abilities for premenopausal
formed better than control group men, but the differences women. The most consistent finding is a strength in color
were not statistically significant. The authors interpreted naming, although some researchers have found an effect
their results as consistent with findings that females with on mental flexibility, fine motor dexterity, and verbal skills,
AD may have gender-associated differences that reflect a including verbal memory. The data are also mixed re-
state of relative acquired estrogen deficiency. garding the effect of ERT on verbal memory for healthy
Several retrospective studies of clinical registries post-menopausal women. Although controlled studies of
have found at least a 45% lower relative risk factor for de- younger women who undergo surgical menopause consis-
mentia in women receiving ERT (Birge, 1994; Henderson tently demonstrate enhanced ability on paragraph-recall
et al., 1994; Mortel and Meyer, 1995). However, demo- tasks, for postmenopausal women, ERT immediate bene-
graphic factors and possible treatment variables in clinics fits may be limited to symptomatic women in the imme-
make it difficult to generalize these findings to the popula- diate postmenopausal period. The issues regarding how
tion at large. Also, a risk reduction for vascular dementia affective features associated with menopause may interact
has been linked to ERT, and not all these studies discrimi- with attention and memory have not yet been teased apart.
nated women with AD from those with vascular dementia. Also, distinctions between the effect of ERT on symp-
Nevertheless, in a retrospective study that matched pa- tomatic vs. asymptomatic postmenopausal women remain
tients with AD to controls on variables of age at menarche to be clarified. Nevertheless, ERT does appear to im-
and age at menopause, Waring et al. (1999) reported an prove verbal memory and general cognitive functioning in
odds ratio of 0.42 for ERT of at least 6 months in control women with AD. Of even greater consequence are several
subjects vs. those with AD, which remained significant studies that have found that ERT appears to lower women's
after education was adjusted for. There was a significant risk of developing AD, perhaps by as much as 50%.
Hormones and Cognition 187

Thyroid Hormones electroconvulsive therapy may be mitigated by adminis-

tration of TRH (Khan et al., 1994; Stern et al., 1991).
The thyroid gland is located just below the larynx
and anterior to the trachea, opposite the fifth, sixth, and Hyperthyroidism
seventh cervical vertebrae and the first thoracic vertebra. It
consists of two oblong, shield-shaped lobes, each approx- In most patients with hyperthyroidism, the thyroid
imately 2.5 cm wide by 4.5 cm long and 20 g in weight, gland undergoes a two- to three-fold increase in size. When
connected by a narrow portion termed the isthmus. This associated with diffuse goiter, ophthalmopathy, and der-
gland consists of densely packed follicles containing ap- mopathy due to an autoimmune disorder (i.e., the pres-
proximately 100 days' average output of thyroglobulin, ence of antibodies to TSH receptors), hyperthyroidism
a protein that contains the thyroid hormones within its is referred to as Graves' disease. Notable symptoms in-
molecule. The main function of the thyroid gland is the clude excitability, intolerance to heat, increased sweating,
synthesis of the principal thyroid hormones triiodothyro- weight loss, muscle weakness, fatigue with dyssomnia,
nine and thyroxine—commonly referred to as T3 and T4, anxiety, and tremor. Increased metabolic rates may be ac-
respectively—which are largely responsible for regulat- companied by abnormal cardiovascular rates, myopathy,
ing the metabolic rate of the body. The functions of T3 and autonomic tremor. These symptoms are mediated by
and T4 are the same, but they differ in speed and inten- increased serum levels of T3 and T4 and their action on the
sity of action. T3 is approximately 10 times more potent catecholaminergic system. The incidence of this disorder
than T4 but is present in smaller quantities and persists is approximately 0.3:1,000 in the United States. The con-
for shorter periods. The thyroid hormones result from dition peaks during the third and fourth decades of life and
iodine processed in the presence of TSH, a product of is 7 to 10 times more common in women than in men. The
the anterior pituitary gland. This in turn is regulated by explanation for this gender difference and for the under-
TRH, which is synthesized in the hypothalamus and is lying cause of the disease is unknown. However, immune
subject to the influence of a number of neurotransmit- dysfunction is more common in women in general. Among
ters as well as exogenous factors such as temperature and older patients, symptoms of hyperthyroidism may be more
stress. Normal serum levels for TSH, total T3, and total subtle and include apathy, limb myopathy, and cardiovas-
T4 are 0.4-5.0 /LtU/ml, 88-160 ng/dl, and 5.0-12.0 /ng/dl, cular disease. Also, there is an association between hyper-
respectively. thyroidism and myasthenia gravis (MG): Approximately
Although the classic action of thyroid hormones is 5% of patients with MG manifest hyperthyroidism.
the stimulation of heat production, their influence on meta- Prominent psychiatric symptoms have complicated
bolic processes throughout the organs and tissues is sig- the differential diagnosis since the earliest descriptions by
nificant. Accordingly, variability of thyroid functioning Parry (1825) and Graves (1835), both of whom attributed
occurs both as a result of primary dysfunction of the thy- elevated thyroid activity to psychogenic etiology. Mobius
roid gland and as a secondary result of systemic illnesses. (1886) was the first to identify these symptoms as sec-
The hormones and their receptors are widely distributed ondary to hyperfunction of the thyroid gland. Psychiatric
throughout the brain. Their regulatory action on the central symptoms such as irritability, anxiety, insomnia, weight
nervous system appears to include the effects on adrener- loss, change in appetite, and hyperkinesis, paired with rel-
gic function, striatal dopaminergic activity, and levels of atively subtle cognitive problems in aspects of attention
substance P and serotonin. and mental control, can continue to lead patients and health
Administration of TRH in normal, neurological, and professionals to form an initial impression of an anxi-
psychiatric populations appears to increase levels of moti- ety or a mood disorder. Additional diagnostic ambiguity
vation, relaxation, and adaptation in a non-disease-specific may result from the presentation of typical somatic symp-
manner (Loosen, 1992). Among patients who are depres- toms of anxiety such as tremor and tachycardia, which
sed, subclinical levels of hypothyroidism may be present. are consistent with peripheral sympathetic activation. En-
Additionally, administration of tricyclic antidepressants hanced sympathetic tone in hyperthyroidism is not due to
(TCA) appears to be augmented by T3 and can convert increased catecholamine levels, however, but is thought to
response failure with TCAs to success in up to 66% of pa- be due to increased sensitivity of certain receptors to nore-
tients (Joffe and Singer, 1990). TRH is known to directly pinephrine, secondary to increased levels of thyroid hor-
affect the central nervous system (Bennett et al., 1989) mones. Classic clinical features of hyperthyroidism that
and has been investigated for use as a potential thera- may clarify the differential diagnosis include increased
peutic agent in AD (Kelly, 1995). Also, there is evidence heat sensitivity, prominence of the eyes, diplopia, and
that postictal cognitive dysfunction in patients undergoing goiter.
188 Erlanger, Kutner, and Jacobs

According to a survey by Stern et al. (1996), it is continued to demonstrate weakness on the paired asso-
likely that many health professionals initially interpret the ciates task. Alvarez et al. (1983) did not find a correlation
presenting symptoms of Graves' disease as a primary psy- between T4 and attention but did find that patients with hy-
chiatric disorder. Patients' psychiatric symptom patterns perthyroidism performed significantly worse on a test of
in Graves' disease vary; many individuals meet criteria attention (Tolouse-Pieron Concentration Attention Test:
for multiple diagnoses including major depression, gener- Szekely, 1966) than did healthy controls.
alized anxiety disorder, panic disorder, agoraphobia, hy- In a study comparing 26 patients with hyperthyro-
pomania, obsessive-compulsive disorder, and obsessive- idism and a control group with nontoxic goiter, 2 years and
compulsive personality disorder. Nevertheless, a common 10 years following treatment, Perrild et al. (1986) found
symptom across a number of studies appears to be the evidence of ongoing neurocognitive dysfunction due to
presence of anxiety, which is common in both frequency hyperthyroidism. Fifty-four percent of the treated indi-
and severity of report. viduals manifested cognitive dysfunction at 10 years post
Studies investigating cognitive problems in Graves' treatment, and in half of these persons, the dysfunction
disease have focused on identifying the nature of cognitive was described as marked to severe in range. Weaknesses
dysfunction in the hyperthyroid state and on determining were identified in the areas of attention, concentration,
the degree of resolution upon the patient's returning to a conceptual reasoning, visuospatial processing, and long-
euthyroid state following treatment (Beckwith and Tucker, term memory. However, 31% of the controls scored in
1988). The first such study, by Artunkal and Togrol (1964), the mildly impaired range on certain of the measures.
compared patients with hyperthyroidism with matched The authors did not detail psychiatric or other factors
normal controls on tests of motor performance and reac- that might have accounted for these intra- and intergroup
tion time on visual and auditory discrimination tasks. The findings.
patient group performed significantly worse than the con- More recently, Trzepacz, McCue, Klein, Greenhouse,
trols did on a number of these measures, which produced et al. (1988) and Trzepacz, McCue, Klein, Levey, et al.
a clinical picture of fatigued individuals. When retested in (1988) examined the performances of 10 subjects with
a euthyroid state, the patients showed no significant im- hyperthyroidism on measures of psychiatric symptoms
provement in visual discrimination, reaction time, or fine and attention at three stages: baseline (hyperthyroidal) as-
motor ability. sessment, following treatment with propranolol (an anx-
Whybrow et al. (1969) compared patients with hyper- iolytic) for 2 weeks, and following 6 months of antithy-
thyroidism and patients with hypothyroidism on a num- roid treatment. Although psychiatric symptoms improved
ber of psychiatric and neuropsychological measures. Al- following both treatment phases, significant improvement
though both subject groups had reduced scores initially, in attention as measured by the Stroop Test color (^ =
patients in the hyperthyroidal group performed better on 77.2 ±8.4 to A/, = 85.8 ±5.8, p < .03) and color-word
the Trail Making (Reitan and Davison, 1974) and Porteus (/i = 39.2 ± 10.8 to /i = 52.0 ± 8.9, p < .01) conditions
Maze (Porteus, 1959) tests following treatment than did (Stroop, 1935) was observed only following the third stage
either the treated or the untreated patients with hypothy- of treatment with an antithyroid agent. This was in con-
roidism. Although these findings are suggestive of a read- trast to the earlier studies, which found only subtle dif-
ier recovery of cognitive functioning following treatment ferences or no improvement between hyperthyroid and
for hyperthyroidism, without the use of a control group of euthyroid states. Also in contrast to earlier studies, the au-
healthy matched subjects, one does not know whether the thors found a. positive correlation between levels of T4 and
results were due in part to differential retest characteris- tests of attention during the hyperthyroid state. The authors
tics of the study groups. Another study, comparing women suggested that these findings were consistent with a nora-
with hyperthyroidism and healthy matched controls on drenergic effect on attention related to dose in an inverted
measures of attention (Stroop Test: Stroop, 1935), paired U-shaped curve. Other neuropsychological tests were ad-
associates learning (Wechsler Memory Scale: Wechsler, ministered at treatment stages 1 and 3 only. As in earlier
1945), motor speed (Tapping Test: Reitan, 1955), and studies, reductions in fine motor speed and simple atten-
speed of information processing (The Spokes Test: Reitan, tion did not improve significantly with time. Significant
1955), revealed no statistically significant differences be- improvements on measures of IQ (WAIS-R: Wechsler,
tween the groups (MacCrimmon et al., 1979; Wallace 1981), conceptual reasoning (Category Test: Halstead,
et al., 1980). Within the patient group, however, T4 lev- 1947; Reitan and Davison, 1974), visual-spatial process-
els were associated with weaknesses in concentration and ing (Tactual Performance Test: Reitan and Davison, 1974),
memory. This correlation disappeared following treat- and memory (Wechsler Memory Scale: Wechsler, 1945)
ment, but patients with initial elevated levels of T4 were attributed to practice effects and/or a lessening of
Hormones and Cognition 189

psychiatric symptoms, not to structural or neurophysio- (Mendorla et al., 1988). Similarly, rearing in an environ-
logical changes. Supporting the hypothesis that thyroid ment deficient in iodine results in cognitive weaknesses,
hormones may correlate with attention according to an in- particularly with respect to the development of psychomo-
verted U-shaped curve, Schlote et al. (1992) also found tor skills. The severity of the cognitive dysfunction appears
reduced psychomotor speed in overt but not subclinical to be linked to the extent of the iodine deficiency (Aghini-
cases. Lombardi et al., 1995; Fenzi et al., 1990; Sankar et al.,
Despite the inconclusive nature of the aforemen- 1994; Vitti et al., 1992).
tioned investigations regarding the relationship of hyper- Adult hypothyroidism, typically associated with an
function of the thyroid and neurocognitive test perfor- autoimmune mechanism, is the result of decreased serum
mance, these findings are generally consistent with patient levels of thyroid hormones. The incidence of primary hy-
reports of initial dysfunction in both psychiatric and cog- pothyroidism is approximately one-eighth that of hyper-
nitive realms followed by a partial return to baseline func- thyroidism, being four to seven times more common in
tioning. Moreover, Bommer et al. (1990) found that a less females than in males. Hypothyroidism is also commonly
complete neuropsychological recovery was characteristic associated with hypothalamic-pituitary disease. The con-
of patients who relapsed within 2.5 years of initial treat- dition can also result from Hashimoto's thyroiditis, a con-
ment. In Stern et al.'s (1996) survey, 33.1% of respon- dition roughly comparable to Graves' disease in incidence,
dents reported being prescribed psychotropic medication in which the thyroid gland at first enlarges and subse-
following diagnosis of Graves' disease, even though only quently atrophies. Another important cause is iatrogenic
7.7% of the sample reported such treatments as part of (e.g., after treatment for Graves' disease). The number of
their premorbid history. Furthermore, 24% reported ongo- older adults at risk for hypothyroidism steadily increases
ing problems in cognitive functioning, especially slowed with age (Osterweil et al., 1992).
mental processing and memory problems. Overall, pa- The clinical symptoms of hypothyroidism reflect the
tients reported significant differences between premorbid numerous physiological systems subject to thyroid dys-
and current (euthyroid) levels of functioning in the areas function: intolerance to cold; puffy face; coarse, dry skin
of memory, attention, planning, and productivity, which and hair; fatigue and somnolence; muscular sluggishness;
suggests that cognitive problems persist following stabi- decreased heart rate and cardiac output; weight gain; de-
lization of thyroid functioning. However, no tests for inter- pressed growth of hair; development of a husky voice; and
actions between ongoing psychiatric and cognitive symp- edematous body tissue. Because a lack of thyroid hormone
toms were reported, so it is unclear whether these problem increases levels of serum cholesterol, hypothyroidism is
areas are independent or related. also associated with arteriosclerosis. Psychiatric symp-
toms include depression, paranoia, and psychotic ideation,
and neurological symptoms range from myopathy and
Hypothyroidism ataxia to cognitive confusion and dementia. Diagnostic
findings include low circulating T3 and T4 levels, elevated
Children born with congenital hypothyroidism (CH) TSH, low radioiodine uptake by the thyroid gland, and an
are typically identified at birth. In the United States, the increase in cerebrospinal fluid protein.
incidence is 1:5,000 in the White population and 1:32,000 Cognitively, severe hypothyroidism is considered a
in the Black population. Early intervention prevents the "reversible" dementia with treatment by thyroid replace-
development of severe motor and sensory impairments ment therapy. However, variables mediating the extent of
as well as mental retardation. Nevertheless, meta-analysis recovery following treatment are still being investigated.
demonstrates a relatively small but significant trend to- Mennemeier et al. (1993) pointed out that although thyroid
ward lower IQ (=6.3 points) and reduced motor skills in replacement therapy is associated with remission of psy-
treated children with CH (Derksen-Lubsen, 1996). This chiatric symptoms, the link with improved cognitive func-
finding has been replicated in a study of identical twins, tioning is less well established. Although both patients
only one of whom was affected by CH, secondary to thy- with hyperthyroidism and those with hypothyroidism re-
roid dysgenesis. Although the affected twin performed port a range of psychiatric symptoms that have a great
within normal limits at 8 years of age, her IQ was 7 points deal in common, hypothyroidism most typically presents
lower and she consistently performed worse than her sister with depression in combination with motor retardation,
on tasks of writing, reading, and verbal memory (Bargagna not anxiety and agitation (Whybrow et al., 1969).
et al., 1997). Absence or delay in treatment leads to se- Impairment in cognition has been noted in patients
vere cognitive problems: In one study, 45% of patients ob- with hypothyroidism since Gull's (1873) initial descrip-
tained IQ scores below 70 and only 28% scored above 85 tion of "a cretinous state supervening in adult life in
190 Erlanger, Kutner, and Jacobs

women" due to hypothyroidism. Similarly, some of the A subset of these patients was available for retesting
first modern reports noted impairment in intellectual ca- following 5 months of thyroid replacement therapy. Im-
pacities (Crown, 1949; Reitan, 1953). Moreover, of six provement was detected on the Symbol Digit Modalities
studies utilizing unselected patient populations—Crown, Test (A. Smith, 1982; ^ = 33.6 to /i = 37.9, p < .016),
1949; Denicoff et al., 1990; Jain, 1972; Reitan, 1953; Trail Making A (Reitan and Davison, 1974; n = 67.9 to
Schon et al., 1961; and Whybrow et al., 1969—cognitive /u, = 56.4, p < .03), and Inglis Paired Associates medium
impairment, not psychiatric disturbance, was noted to be association items (Inglis, 1959; //, = 0.86 to /u. = 0.93%
the primary problem in four—Crown, 1949; Denicoff correct, p < .0006), which suggests that attention, psy-
et al., 1990; Reitan, 1953; and Schon et al., 1961. Although chomotor speed, and learning were the principal functions
several of the six studies attempted to assess response to that responded to treatment. A lack of significant improve-
treatment, results have been inconclusive. Initial reports ment on a number of other tests, including animal naming,
(Crown, 1949; Schon et al., 1961) indicated improve- Digits Forward and Digits Backward (WAIS-R; Wechsler,
ment in general cognitive functioning following thyroid 1981), and Trail Making B (Reitan and Davison, 1974),
replacement therapy. However, design and measurement suggested that some dysfunction in aspects of effortful at-
problems may have distorted these findings. Whybrow tention may persist for certain patients despite treatment.
et al. (1969) reported subjective complaints of memory Finally, the authors reported evidence that age may in-
problems, observer reports of poor general cognition, and crease vulnerability to the effects of hypothyroidism on
weaknesses in concentration (Trail Making: Reitan and cognitive functioning.
Davison, 1974) and problem solving (Porteus Maze Test: Thyroid replacement therapy such as that used in the
Porteus, 1959), with no improvement following return to aforementioned studies typically consists of administra-
a euthyroid condition. Interpretation of these findings is tion of thyroxine (T4) alone. A recent study (Bunevicius
complicated by use of a comparison group comprising et al., 1999) compared cognitive improvement in patients
patients with hyperthyroidism, not healthy normal sub- with hypothyroidism treated with T4 alone vs. T4 plus tri-
jects. Denicoff et al. (1990), using a mental status exam iodothyronine (T3). Each patient received a 5-week treat-
and the WAIS-R Digit Symbol subtest (Wechsler, 1981), ment of each protocol, and the order was randomized.
also reported no change in cognitive functioning following Subjects receiving the combination therapy performed sig-
therapeutic intervention. nificantly better on Digits Backward (Wechsler, 1981)
One recent large-scale study by Osterweil et al. and on incidental learning recall of Digit Symbol pairs
(1992) examined hypothyroidism and the effect of thyroid (Wechsler, 1981). Although the results were statistically
replacement among older adults without dementia. Fifty- significant, their clinical significance is unclear. For Dig-
four men and women with hypothyroidism ranging from its Backward, raw score means of 5.5 ± 1.6 and 6.0 ±1.3
minimal to overt and untreated for varying lengths of time (p < .04) were obtained for monotherapy and combination
were compared with 30 euthyroid controls. Hypothyroidal therapy, respectively. Similarly, for Digit Symbol pairs,
patients in general performed significantly worse than raw score means of 5.5 ±2.3 and 6.3 ±2.1 (p < .05) were
controls did (p < .01) on tests of general cognitive func- obtained for monotherapy and combination therapy, re-
tion (Mini-Mental State Exam; /x = 26.1 vs. /i = 28.7), spectively. The results of other cognitive measures did
cube drawing, paired associate learning (Inglis, 1959), an- not reach significance, including those for Digits For-
imal naming, and psychomotor speed (Trail Making A: ward, speed indices on Digit Symbol, and a test of visual
Reitan and Davison, 1974; Symbol Digit Modalities Test: scanning.
A. Smith, 1982). However, among the patients with min- Researchers of minimal and subclinical hypothy-
imal hypothyroidism, no performances were significantly roidism have found that a complete return to baseline
reduced relative to those of controls. Moreover, differ- following treatment may be anticipated. Monzani et al.
ences in cognitive functioning were apparent between the (1993) obtained pretreatment Memory Quotient scores
minimal and overt hypothyroidal groups on all indices ex- on the WMS-R (Wechsler, 1974) of 89.1 for a group
cept Digits Forward (Wechsler, 1981) and a set of Yes/No of subclinical patients with hypothyroidism. Following
questions and Sequential Commands adapted from the treatment and return to a euthyroid state, patients' scores
Boston Diagnostic Aphasia Examination (Goodglass and rose to 99.9. Similarly, Baldini et al. (1997) found im-
Kaplan, 1983). Also, the control subjects and the patients proved memory performance on the WMS-R in a subclin-
with overt hypothyroidism whose duration of thyroid dys- ical hypothyroid population following return to a euthy-
function was brief (10-20 days, secondary to discontinua- roidal state.
tion of thyroid hormone replacement following thyroidec- Nevertheless, recent case studies provide evidence of
tomy) displayed no statistically significant differences on persistent cognitive deficits despite thyroid replacement
neuropsychological tests. therapy in patients with severe hypothyroidism. Leentjens
Hormones and Cognition 191

and Kappers (1995) reported the case of a 43-year-old than for severe hypothyroidism, in which thyroid replace-
female administrator who was severely incapacitated by ment therapy may serve to arrest a progressive process.
problems with concentration and memory following ex- Moreover, in both hyper- and hypothyroidism, impairment
tensive treatment for hypothyroidism. Reduced perfor- and recovery of cognitive functioning appears to be rela-
mances on a continuous performance test, Digit Span and tively independent of resolution of psychiatric symptoms.
Digit Symbol subtests of the WAIS-R (Wechsler, 1981),
recall of the Rey Complex Figure (Rey, 1941) and Benton
Visual Retention Test (Benton, 1974; Sivan, 1992), and the Adrenal Hormones I: Cortisol
Rivermead Behavioral Memory Test (Wilson et al., 1985)
supported the patient's complaints of difficulty reading The adrenal glands weigh approximately 4 g each
books and magazines, watching television, following con- and are located at the superior poles of the kidneys. Each
versations, and driving, despite the absence of depressive gland is divided into two distinct portions: the adrenal
symptoms. medulla, which secretes the hormones epinephrine and
In another single case study, with the benefit of mul- norepinephrine and is functionally related to the sympa-
tiple control subjects, Mennemeier et al. (1993) reported thetic nervous system, and the adrenal cortex, which se-
initial impairment on a number of verbal and visual mem- cretes the hormones known as corticosteroids, of which
ory indices. Improvement in a range of cognitive and cortisol is of principal interest in regard to cognitive func-
psychiatric symptoms following diagnosis but preceding tions. Another adrenal hormone, dehydroepiandrosterone
treatment suggested the influence of nonspecific treatment (DHEA), has recently been studied for its possible ef-
effects. Repeated measures obtained during 7 months of fects on mood and memory and is discussed subsequently.
thyroid replacement therapy revealed significant ongo- The adrenal glands also produce small amounts of certain
ing memory dysfunction relative to the function of con- androgens, which are similar in function to testosterone.
trols, despite modest improvement. Measures included These are discussed in the context of CAH in the preceding
the Wechsler Memory Scale (Wechsler, 1974) and the section on male sex hormones.
Buschke Selective Reminding Test (Buschke and Fuld, The hypothalamic-pituitary-adrenal axis is the en-
1974) with various alternate forms. The authors suggested docrine system most integral to the body's reaction to
that although memory may not have been significantly physiological stressors. It prepares the organism to re-
helped by thyroid replacement therapy in this patient, such spond to environmental stimuli and maintains homeosta-
treatment may have arrested a progressive deterioration of sis. Cortisol is integral to these preparations through its
mnestic functioning due to hypothyroidism. role in increasing blood glucose concentrations, which in
Recent animal studies support the hypothesis of ir- turn mobilizes available energy stores. The axis is con-
reversible cognitive dysfunction due to hypothyroidism. trolled by a feedback loop as follows: Corticotropin-releas-
Selective cell destruction in the hippocampus was demon- ing hormone (CRH) is produced by hypothalamic neurons
strated in rats that underwent thyroidectomy or were made both according to a circadian pattern and in response to
hypothyroidal by injection with propylthiouracil (Madeira physiological stress. CRH regulates the release of adreno-
et al., 1992). Return to a euthyroid condition did not cor- corticotropic hormone (ACTH) from the pituitary. ACTH
rect for the cell loss. stimulates the adrenal glands, which in turn produce cor-
tisol. Cortisol completes the feedback loop by its effect
on the hypothalamus and other structures. Further mod-
Summary ulation of the hypothalamus and pituitary are provided
through the amygdala, the hippocampus, and other path-
Despite the high frequency of disorders of thyroid ways. The normal plasma ACTH level is 10-52 pg/ml.
function and the prominence of the associated cognitive Normal serum cortisol levels range from 3-20 mg/dl in
symptoms, research on neuropsychological patterns of the morning to 2.5-10.0 /^g/dl in the afternoon.
impairment and recovery is sparse. The preceding findings In normal subjects, short- and long-term adminis-
suggest that neurocognitive dysfunction results from both tration of exogenous corticosteroids has been shown to
increased and decreased levels of thyroid hormones. The produce mild dysfunction of verbal mnestic processes,
former is associated with impairment or reduction in fine although it is unclear whether this is due to disruption
motor speed, attention, and memory, and the latter with of attentional factors, memory processes, or both (Naber
general cognition, attention, learning, and psychomotor et al., 1996; Wolkowitz, 1994; Wolkowitz et al., 1990). On
speed. Recovery following treatment appears to be more a verbal learning test, higher rates of intrusion errors were
complete for hyperthyroidism and mild hypothyroidism associated with dexamethasone, and poor discrimination
192 Erlanger, Kutner, and Jacobs

on a recognition paradigm was associated with prednisone et al., 1981). Also, because the tumor may exert a mass
(Wolkowitz, 1994). McEwen (1982) has suggested that effect on other portions of the pituitary gland, other en-
deficits such as these may be due to impaired filtering docrine complications such as diabetes insipidus, hyper-
of stimuli by the hippocampus. In contrast, Naber et al. tension, and amenorrhea or impotence may be comorbid.
(1996) reported improved performances on Trails A (Re- Initial studies of neuropsychological function in pa-
itan and Davison, 1974) and verbal fluency measures, tients with Cushing's syndrome revealed a pattern of dif-
which they attributed to a hyperactivating effect or "sen- fuse bilateral frontal dysfunction manifesting as poor con-
sory sharpness" due to cortisol. centration, comprehension, and orientation, together with
Elevated cortisol levels are also evident among many impairments in visual memory, nonverbal reasoning, and
patients with major depression, in whom memory and at- spatial/constructional ability (Whelan et al., 1980). In a
tention problems have been frequently observed (Carroll, subsequent study, however, weaknesses in visuospatial
1982). However, there is not evidence that elevations in functions were found to correlate with affective and veg-
serum cortisol levels can account for the range of cognitive etative symptoms (Starkman et al., 1986), which suggests
problems associated with depression. Also, there is evi- that depressive symptoms had confounded the earlier find-
dence that hypersecretion of the hypothalamic-pituitary- ings. The authors noted, however, that they did not find
adrenal axis releasing hormone CRH, not cortisol, plays significant correlations between depressive symptoms and
the more important role in mediating cognitive problems memory problems.
in patients who are depressed (Muglia et al., 1995). Neuroimaging has suggested that the hippocampus
may play a role in mnestic dysfunction of patients with
Hypercortisolism Cushing's disease. Starkman et al. (1992) reported signif-
icant positive correlations between hippocampal volume
Cushing's syndrome refers to the clinical manifes- and measures of verbal memory, together with significant
tation of increased concentrations of cortisol and gener- negative correlations between hippocampal volume and
alized catabolism: truncal obesity, plethoric (full) facies, plasma cortisol levels. However, improvement in memory
hirsutism and baldness, osteoporosis, and generalized following normalization of cortisol levels indicated that
muscular weakness. Psychiatric symptoms may include corticosteroids can cause cognitive deficits independent of
depression, anxiety, and psychosis (i.e., "steroid hippocampal neuron loss. Such a mechanism is suggested
psychosis"). Cognitive functions may be affected, includ- by the animal literature, in which corticosterone has been
ing disruptions in orientation, concentration, memory, and shown to suppress hippocampal excitability and long-term
comprehension (Starkman and Schteingart, 1981; Whelan potentiation (Bliss and Lomo, 1973; Gufstafsson and
et al., 1980). The condition may be the result of a pituitary Wigstrom, 1988) in a concentration-dependent manner
adenoma, a primary adrenal tumor, ectopic production of (M. C. Bennett et al., 1991). However, generalization of
ACTH by a carcinoma of the lung, or, frequently, the long- these findings is uncertain because corticosterone and cor-
term treatment of a variety of diseases with exogenous tisol differ in their behavioral effects in humans.
cortisol such as cortisone or prednisone. Ectopic ACTH A recent study compared the performance of 25 pa-
syndrome due to a carcinoma is three times more preva- tients with Cushing's disease without psychosis and/or se-
lent in men than in women and has its highest incidence vere affective symptoms as determined by both clinical in-
between ages 40 and 60. terview and objective self-report (Minnesota Multiphasic
When due to the secondary effects of a pituitary ade- Personality Inventory [MMPI]; Hathaway and McKinley,
noma, the condition is known as Cushing's disease. In 1943) with the performance of a control group on an exten-
this case, a hypothalamic-pituitary defect, most typically sive neuropsychological battery (Mauri et al., 1993). Mod-
a pituitary tumor, causes hypersecretion of ACTH, which erate dysfunction was identified relative to the function
leads to increased cortisol secretion. Onset is typically of controls on the Logical Memory (I: /Li = 6.3 ± 2.2 vs.
between the ages of 20 and 40, but this disease has been /n = 8.3 ±2.4, p<.01; II: m=7.1±3.1 vs. /Li = 10.0±
reported in infants and in patients older than 70 as well. 2.5, p < .001) and Visual Reproductions (I: /Li = 8.8 ± 3.7
Unlike Cushing's syndrome, Cushing's disease develops vs. /z= 10.9 ±2.2, p<.05; II: /u. = 6.8 ±3.9 vs. /u,=
in females eight times more often than in males. There 10.0 ±2.2, p < .001) subtests of the Wechsler Memory
is some evidence that psychiatric disturbances may be Scale (Wechsler, 1945). In addition, weaknesses on the
more frequent in cases of pituitary hyperfunction (i.e., Digits Backward (/Li = 3.5 ±0.8 vs. p=4.1 ±0.8, p <
Cushing's disease) than in cases of adrenal adenomas .01) and the Digit Symbol (/u. = 36.8 ± 14.1 vs. n = 47.3 ±
(i.e., Cushing's syndrome) because of the higher levels 11.2,p<.01) subtests of the WAIS-R (Wechsler,
of ACTH associated with the former condition (Starkman 1981) indicated problems with attention/concentration
Hormones and Cognition 193

and, possibly, visuomotor functioning. No correlations be- is more common in females than in males (2.6:1) and is
tween ACTH/cortisol levels and cognitive performances usually diagnosed in the third to fifth decades of life.
were found for the patients with Cushing's disease. How- Despite reports of depression, apathy, and confu-
ever, in 8 of these subjects retested 6 months following sion, investigations of neurocognitive functioning of pa-
surgical ablation of the neoplasia, normalization of ACTH tients with hypocortisolism are few. One case of primary
and cortisol levels was accompanied by a significant ame- adrenal insufficiency due to a traumatic brain injury pro-
lioration of memory dysfunction. vides some insight into the nature of the condition. In their
Among older persons, increases in cortisol levels case study, Webster and Bell (1997) reported confusion
with time are predictive of weaknesses in explicit memory and severe problems with short-term memory and atten-
and selective attention (Lupien et al., 1994). Significantly, tion in a 31-year-old male 6 weeks after severe traumatic
increases in cortisol levels have been associated with hip- brain injury. The patient also manifested obsessive and
pocampal atrophy in patients with AD, independent of de- paranoid thoughts, anxiety, and depressive features. Di-
pressive symptoms (Davis et al., 1986;Dodt et al., 1991). agnosed at that point with primary adrenal insufficiency,
Also, hippocampal atrophy has been found in patients with the patient was treated with adrenal replacement therapy.
post-traumatic stress disorder, in whom elevated cortisol After 2 weeks, the patient was participating in 3 hours of
levels are thought to be associated with a traumatic emo- therapy per day, including a 45-min speech therapy ses-
tional stressor (Bremner et al., 1995). Further, although sion, which indicated significant improvement in atten-
corticoid receptors have been shown to be present through- tional processes. The patient's history of depressed skull
out the brain, a concentration of receptors with a high affin- fractures, bifrontal epidural hematomas, and frontal contu-
ity for corticosterone is present in the hippocampus (Ruel sions indicate that traumatic etiology, not adrenal insuffi-
and DeKloet, 1985; Sarrieau et al., 1988). These findings ciency, likely accounted for a significant degree of the cog-
support the hypothesis that increased adrenal activity can nitive dysfunction observed. However, the patient's pos-
account in part for age-related hippocampal pathology and itive response to the hormonal intervention underscores
memory dysfunction due to elevations in cortisol, ACTH, the importance of the adrenal system in general metabolic
and/or CRH. Moreover, there is evidence from the animal functioning. Notably, the patient was eventually able to
literature that the hippocampus plays a major role in the return to work, resume driving, and participate in leisure
regulation of pituitary-adrenal activity (for a discussion, activities.
see Jacobson and Sapolsky, 1991).
Summary

Chronic Hypocortisolism In studies of hypercortisolism due to exogenous and

endogenous factors, there is a clear association with im-
Adrenal cortical insufficiency, a primary disease of paired hippocampal functioning. Although improvement
the adrenals also known as Addison's disease, may result in mnestic capacity may result following normalization
in conditions ranging from mild generalized weakness to of cortisol levels, there is evidence supportive of perma-
sudden-onset vascular collapse. Insufficient production of nent neurotoxic effects on the hippocampal structures due
cortisol is accompanied by increased CRH and ACTH to hypercortisolism. Age-related increases in cortisol are
levels due to decreased feedback to the hypothalamus and similarly associated with memory dysfunction, indepen-
anterior pituitary. This disease is characterized by pigmen- dent of mood. Although little researched, hypocortisolism
tation of the skin and mucous membranes, nausea, vomit- appears to be associated with attentional and motivational
ing, weight loss, muscle weakness, fatigue, and dizziness. factors as they influence cognition. Finally, the different
Psychiatric symptoms include depression, confusion, ap- roles that CRH, ACTH, and cortisol play in cognition and
athy, psychosis, paranoia, schizophrenic behaviors, and behavior are only beginning to be understood, although
self-mutilation (Johnstone et al., 1990). The etiology is there is evidence that they may vary considerably.
typically an autoimmune adrenalitis due to tuberculosis,
malignancy, sarcoidosis, or infection but may also result
from bilateral adrenal hemorrhage after sepsis, trauma, Adrenal Hormones II: Dehydroepiandrosterone
surgery, or burns (Claussen et al., 1992; Murphy et al.,
1993; Szalados and Vukmir, 1994). Secondary adrenal in- The biological roles of dehydroepiandrosterone
sufficiency may also result from disease of the pituitary or (DHEA) and its sulfate (DHEA-S) have recently been the
hypothalamic lesions. Primary Addison's disease is rare, subject of much investigations because of the decrease of
with a prevalence of approximately 39 per 1 million. It these hormones with normal aging and their correlation
194 Erlanger, Kutner, and Jacobs

with age-related immune system decline (Thoman and DHEA-S levels did not predict cognitive performance or
Weigle, 1989). Epidemiological data have demonstrated decline in older women on the Trail Making B (Reitan
an association between low circulating DHEA levels and and Davison, 1974) or Digit Symbol subtest (Wechsler,
cardiovascular morbidity in males (Barrett-Connor et al., 1981). Elevation in DHEA-S has been identified as a pos-
1986) and breast cancer in females (Helzlsouer et al., sible cause of neurobehavioral disturbances in certain in-
1992). DHEA is also thought to affect behavior and cog- dividuals with anomalous brain substrates. A. R. Jacobs
nition. Although the precise mechanism of action has not et al. (1995) diagnosed adult-onset CAH in a series of 12
yet been identified, DHEA has been found in the brain and patients referred for refractory psychiatric disturbances.
may mediate certain interactions with GABA receptors A majority of anxiety-related disturbances underscored
(Robel and Baulieu, 1994). DHEA-S has been identified the action of DHEA-S as a GABA antagonist. Each pa-
as a potent GABA antagonist (Deutsch et al., 1992). The tient had evidence of an anomalous brain substrate such
reference range for serum DHEA is 160-800 ng/dl for as neuropsychological deficits and/or EEG abnormali-
adult males and adult premenopausal females. The Post- ties. Treatment with adrenal suppressive therapy produced
menopausal female range is 30-450 ng/dl. For DHEA-S, amelioration of psychiatric symptoms in all cases in which
normal serum levels are 110-690 /ctg/dl for adult males endocrine correction was obtained.
and 80-340 mg/dl for adult premenopausal females. For
males older than 50 years and postmenopausal females,
normal serum levels are 40-330 mg/dl and 17-77 mg/dl, Pancreatic Hormones and Insulin
respectively.
In healthy subjects, the effects of DHEA appear to be Along with its role in digestion, the pancreas se-
on general well-being. Morales et al. (1994) administered cretes a number of hormones that regulate glucose levels
replacement doses, sufficient to raise levels comparable in the bloodstream. In the pancreas, the islets of Langer-
to those of 20-year-olds, to a group of men and women hans contain three principal cell types—alpha, beta, and
aged 40-70. Self-report indices gathered during the 6- delta—which secrete glucagon, insulin, and somatostatin,
month duration of the double-blind, placebo-controlled respectively. Together, these hormones are regulated by
study indicated an effect of DHEA on a range of vari- means of feedback loops as well as through their direct
ables, including increased energy, deeper sleep, improved actions on one another within the islets of Langerhans.
mood, greater relaxation, and better stress-handling capac- Insulin is secreted in response to energy-giving foods,
ity. However, despite a number of animal studies demon- such as carbohydrates and proteins, and causes them to be
strating antiamnestic properties (Flood et al., 1988; Frye stored as glycogens in the liver and muscles, and as fats
and Sturgis, 1995; Melchior and Ritzmann, 1996), DHEA in other tissues, in order to meet future energy needs. It
administration had no benefit on memory tests for either inhibits the release of glucose into the bloodstream and
young subjects receiving a single replacement treatment causes enhanced glucose uptake from the blood by the
(Wolf, Koster, et al., 1997) or for older subjects receiving liver. It is important in promoting protein formation and
a 2-week DHEA replacement protocol (Wolf, Neumann, in preventing protein degeneration, thus promoting growth
et al., 1997). in a synergistic relationship with GH. The normal serum
Specific cognitive benefits were found, however, in level is 5-25 mU/ml (0.2-1 ng/ml) for fasting adults.
pilot studies of patients with major depression receiv- Glucagon is secreted in response to low blood glu-
ing DHEA supplements (Wolkowitz et al., 1994, 1997). cose concentrations and thus plays a role diametrically
Using standardized psychiatric rating scales and verbal opposed to insulin. That is, glucagon is responsible for
tasks measuring semantic encoding, free recall, recog- breaking down liver glycogens and for gluconeogenesis,
nition memory, and categorical fluency, the researchers the process that releases glucose into the bloodstream.
identified significant mood enhancement but no improve- Somatostatin acts directly to inhibit both insulin and
ment in cognitive performance following 6 months of re- glucagon within the islets of Langerhans.
placement therapy. For a single subject, however, a 63% Unlike other body tissues, which in addition to us-
improvement in categorical fluency was accompanied by ing blood glucose can use fats and proteins for energy,
substantial improvement on mood indices. These gains de- the brain depends on a steady glucose supply. Normal
clined to baseline levels when her treatment ended. Mood serum glucose levels are =90 mg/dl, and maintenance
and cognition were significantly related to plasma levels above a critical level is necessary for normal function-
of DHEA and DHEA-S for this patient. ing. Even mild transient hypoglycemia has been found
DHEA-S has also been the subject of study. Yaffe, to affect healthy subjects' performance on psychomotor
Ettinger, et al. (1998) found that during a 4-year period, tasks(Mellman et al., 1994). When glucose concentrations
Hormones and Cognition 195

fall too low, hypoglycemic shock develops, characterized sensation. Cognitively, the increased risk of stroke results
by nervousness, sweating, irritability, and fainting, which in increased rates of vascular dementia in persons with
leads to clonic seizures and, if untreated, coma. This hypo- diabetes (McCall, 1992).
glycemic reaction may develop as a consequence Cognitive dysfunction due to diabetes per se has been
of excessive administration of insulin by patients with the subject of a number of investigations. For adolescent
diabetes, causing a syndrome known as insulin shock. patients with Type I diabetes, the age of onset has been
Prolonged hypoglycemia typically results in permanent shown to be related to cognitive deficits (Rovet et al., 1987;
damage to the central nervous system secondary to both Ryan et al., 1985). Impairments in visuospatial tasks were
reduced cerebral oxygen consumption and decreased brain evident in patients afflicted before the age of 4 years, com-
glucose metabolism (McCall, 1992). pared with those diagnosed at later ages and with nondi-
abetic controls. Ryan (1988) pointed to higher rates of
hypoglycemic seizures and greater sensitivity of the pedi-
Diabetes Mellitus atric brain to metabolic or physiological insult as possible
explanations for these findings. Noting the deleterious ef-
Diabetes mellitus is a term that applies to a group fects of hypoglycemia on medial temporal lobe structures,
of disorders that produce chronic elevations in blood glu- Hershey et al. (1997) identified significantly lowered per-
cose levels, or hyperglycemia. In the United States, ap- formances on a delayed verbal recall task (Story Recall:
proximately 500,000 persons have been diagnosed Squire et al., 1987) in a group of patients with Type I dia-
with juvenile-onset insulin-dependent diabetes mellitus betes who had experienced severe hypoglycemic episodes
(IDDM), also known as Type I diabetes. The age of onset (IJ, = 3.3), compared with the performance of patients with
peaks at 10-14 years (Carter Center, 1985). Although the Type I diabetes but no history of severe hypoglycemia,
etiology is unclear, environmental factors are thought to and with that of controls. The same subjects performed
cause an autoimmune reaction in genetically vulnerable less well than the other two groups on the delayed re-
individuals that destroys the beta cells within the pan- call trial of the CVLT (Delis et al., 1987) and a Word
creas, which thus curtails the body's insulin supply. Treat- Stem Priming test (Squire et al, 1987), but differed sig-
ment is by means of intramuscular injection of insulin, nificantly only from the control group. Both IDDM groups
with the primary goal of therapy being the maintenance manifested significantly reduced performances on a ver-
of metabolic control by avoiding both hyperglycemic and bal fluency test (Benton, 1968; Milner, 1964) and a picture
hypoglycemic states. priming task. No between-group differences were found
Maturity-onset non-insulin-dependent diabetes on tests of attention and visuospatial functioning, or on a
(NIDDM), or Type II diabetes, is characterized by an in- serial reaction time test.
sidious onset of symptoms due to hyposecretion of in- There are conflicting findings regarding the relative
sulin and/or insulin resistance. The prevalence increases vulnerability of persons with Type I diabetes according
with age so that by 65 years, approximately 20% of the to gender: Although Rovet et al. (1987) initially found
population may be affected (Winograd et al., 1990). Esti- greater visuospatial skill dysfunction in young females, a
mates that include undiagnosed cases of impaired glucose recent study found that young males with diabetes may
tolerance more than double that figure (Minaker, 1990). perform more poorly on certain measures (Holmes et al.,
Reports of gender differences vary, with higher rates re- 1992). Complicating the picture, Ryan and colleagues
ported both for women (Carter Center, 1985) and for men (Ryan & Williams, 1993; Ryan et al., 1992), found main
(Winograd et al., 1990). Differences in diagnostic crite- effects for both group (diabetic vs. nondiabetic) and gen-
ria likely account for these conflicting data. For many der (male vs. female) but no interaction on both the Digit
individuals diagnosed with NIDDM, metabolic control is Vigilance (Lewis and Rennick, 1979) and Grooved
frequently achieved through a combination of diet and/or Pegboard (K10ve, 1963) tests, which illustrated a likely
oral hypoglycemic agents. selection problem.
Medical complications due to macro- and microvas- Specific deficits in adults with Type I diabetes appear
cular damage are common, and patients with diabetes are to be related to the individual's history of metabolic con-
at increased risk for stroke, heart attack, retinopathy, and trol. Long histories of poorly controlled glucose levels are
end-stage renal disease. Symptoms of peripheral neuropa- associated with the vascular complications noted previ-
thy may develop, including paresthesias and/or painful ously, the peripheral neuropathies being most apparent on
sensations in the distal limbs. Autonomic neuropathies neuropsychological tests (Ryan et al., 1992; Young et al.,
may also occur, including impotence and loss of bladder 1983). Skenazy and Bigler (1984) found decreased PIQ on
196 Erlanger, Kutner, and Jacobs

the WAIS-R (Wechsler, 1981) and reduced mental flexi- and found improvements in attention/concentration (Trail
bility (Trails B: Reitan and Davison, 1974) and concep- Making A: Reitan and Davison, 1974; ^ = 47.4 ±3.4 to
tual reasoning (Category Test: Halstead, 1947; Reitan and 1*, = 40.4±3.3, p<.05), reading speed (Stroop, Word
Wolfson, 1993) in a group of patients with Type I diabetes Naming: Stroop, 1935; y, = 31.9 ± 2.4 to /z = 28.5 ± 2.3,
compared with the performance of nondiabetic controls. p < .01), fine motor speed (Grooved Pegboard: K10ve,
When grouped according to history of metabolic control, 1963;/z = 98.6±5.1 to /* = 91.0±3.9, p < .05), learn-
those with more problematic histories had significantly ing (Buschke Cued Recall: Buschke and Fuld, 1974; ^ =
greater dysfunction on tasks involving a motor component, 31.9±2.4 to /i = 30.5±1.5, p < .05), and conceptual
including Trails B. Franceschi et al. (1984) found similar thinking (WAIS-R Picture Arrangement: Wechsler, 1981;
dysfunction in a group of patients with histories of poor /i = 8.4 ± 1.0 to /a = 10.1 ± 1.2, p < .05). After adjusting
control, but also reported lower overall Memory Quotient for Type I errors and possible practice effects, the authors
scores on the WMS-R (Wechsler, 1987). Poor metabolic concluded that the improvements in attention/concentra-
control has also been associated with decreased learning tion and fine motor speed were statistically significant.
efficiency (Bale, 1973; Lichty and Klachko, 1985). More These results are consistent with earlier findings that blood
recently, however, Ryan and Williams (1993), in a well- glucose levels at the time of testing may have an impact on
controlled study, found no impairment on measures of attention/concentration mechanisms (Holmes et al., 1983).
learning and memory. Significant differences were iden- The differential vulnerability of learning and mem-
tified on the WAIS-R PIQ (Wechsler, 1981; /z = 99.4 vs. ory processes in patients with Type II diabetes compared
m = 105.8, p < .001), an embedded figure test, a cancel- with patients with Type I diabetes has been discussed
lation task (Lewis and Rennick, 1979), and the Grooved by Ryan and Williams (1993), who suggested that mnes-
Pegboard test (K10ve, 1963). Greater impairment on a psy- tic functions are not likely due to chronic hyperglycemia
chomotor speed index was associated with a greater degree alone, because patients with Type I diabetes do not man-
of chronic hyperglycemia. Similarly, Ryan et al. (1992) re- ifest memory dysfunction. These authors also noted that
ported a strong association between psychomotor slowing patients with Type I have a more severe form of the disease,
and distal symmetrical polyneuropathy and a weaker as- marked by longer histories, more medical complications,
sociation of psychomotor slowing with a history of poor and less easily controlled glucose levels. An initial inves-
glycemic control. tigation also indicates that the mnestic dysfunction of pa-
Neurocognitive dysfunction has also been identified tients with Type II is not likely due to higher rates of cardio-
in patients with Type II diabetes. Meuter et al. (1980) vascular disease (Reaven et al., 1990). Instead, Ryan and
compared patients with Type I and Type II diabetes and Williams (1993) hypothesized an Age x Hyperglycemia
found slowed reaction time and weaknesses in learning interaction to account for the findings. The aging brain
and memory for patients with Type II. Slowed psychomo- has been found to be increasingly vulnerable to mnes-
tor speed and memory dysfunction were subsequently tic dysfunction in other patient groups (e.g., patients ad-
confirmed by a number of investigators (Mooradian et al., dicted to alcohol; Ryan and Butters, 1984). Animal models
1988; Perlmuter et al., 1984, 1987; Reaven et al., 1990; show evidence of neuronal loss in chronic hyperglycemia
Tun et al., 1987; U'Ren et al., 1990), with mnestic dys- (Jakobsen et al., 1987). Neurochemically, hypoglycemia
function correlating with history of metabolic control. is associated with a partial cerebral energy failure, by
Consistent with these findings, one recent prospective means of brain fuel starvation or neuroglycopenia (McCall,
study recruited only patients with Type II diabetics histo- 1992). Animal models have demonstrated the sensitivity
ries of good metabolic control; on a series of cognitive tests of the hippocampal neurons to fluctuations in glucose uti-
(Corsi Blocks: Milner, 1971; Raven Progressive Matrices: lization levels (Yashino et al., 1991), which further sup-
Raven, 1938; Rey Auditory Verbal Learning Test: Rey, ports the hypothesis that poorly controlled hyperglycemia
1964; Taylor, 1959; Digit Span: Wechsler, 1945), no dif- could account in part for chronic mnestic dysfunction.
ferences were found between the patients' test results and (See McCall, 1992, for an extended discussion of the im-
those of nondiabetic controls (Assisi et al., 1996). pact of diabetes on the central nervous system.) Other
Two studies have examined the effect of improved possible factors suggest a complex explanation. In a criti-
metabolic control on cognitive functions in patients with cal review of 19 published studies, Strachan et al. (1997)
Type II diabetes. Gradman et al. (1991) reported improved detailed many of the methodological problems in study-
learning (Buschke Cued Recall: Buschke and Fuld, 1974), ing Type II diabetes and discussed nine potential factors
complex psychomotor speed (visual choice reaction time), associated with cognitive dysfunction in Type II diabetes,
and attention following 2 months of treatment. Meneilly including depression, hyperlipidemia, impaired vision, re-
et al. (1993) measured subjects after 6 months of treatment nal failure, and others noted previously.
Hormones and Cognition 197

Summary under the influence of both exogenous melatonin—to

increase daytime levels (Deacon et al., 1994; Lieberman
Patients with Type I and Type II diabetes face the et al., 1984; Wynn and Arendt, 1988)—and melatonin-
risk of developing a mild chronic encephalopathy. Poorly suppression treatments—to decrease nighttime levels
controlled glucose levels are associated with psychomo- (Badia et al., 1990; Dollins et al., 1994; Myers and Badia,
tor slowing and possible decreases in concentration and 1993). In an elegantly designed single-case, double-blind,
mental flexibility for both groups of patients. Among pa- placebo-controlled study, Slotten and Krekling (1996)
tients with Type I, an early onset, a history of severe hy- found evidence that circulating melatonin has no direct
poglycemia, or both is associated with greater degrees of effect on cognitive capacities. Reduced speed and accu-
dysfunction. Patients with Type II diabetes with histories racy during peak serum melatonin levels did not differ
of poor glucose control also experience mnestic dysfunc- from the control condition. However, during the temper-
tion, perhaps due to the greater vulnerability of the aging ature trough that occurs approximately 2 hours after peak
brain in this patient group. Cognitive dysfunction in both serum levels, decrements in speed and accuracy on a com-
patient groups appears to be due to multiple etiologies, puterized battery of tests of logical reasoning, sustained
including metabolic crisis, vascular disease, course of ill- attention, visuospatial orientation, and reaction time (Per-
ness, and neurochemical disturbances. Patients, especially formance Assessment Battery: Thorne et al., 1985) were
those with Type II, may have no dysfunction if glucose lev- evident on all tasks. The authors hypothesized that the
els are well controlled. Also, cognition may normalize to cognitive weaknesses were due to reduced speed of infor-
a large extent with improved metabolic control. mation processing secondary to melatonin's hypothermic
properties.

The Pineal Gland and Melatonin

CONCLUSION
The pineal gland, along with the habenula and asso-
ciated structures, compose the epithaiamus. It is known Hormones interact with the brain and can produce
by many as Descartes' hypothetical "seat of the soul" and profound effects on behavior and cognition. As the pre-
for its phyloanatomical history as a remnant of a "third vious discussions illustrate, the actions of the neuroen-
eye" in the posterior portion of the head in lower ani-
docrine system are complex. Nevertheless, neuropsychol-
mals. It is located within the cranium, between the third
ogists should acquire a basic working understanding of
and fourth ventricles, at approximately the level of the
the principal disorders. It is suggested that clinicians use
midbrain. In addition to its possible role in the seasonal
the preceding review as a reference regarding the specific
regulation of human sexual behavior, the pineal gland has cognitive and neurobehavioral sequelae of their individual
received scrutiny because of its synthesis of the hormone
patients' neuroendocrine disorders. Also, Table II summa-
melatonin. rizes the principal characteristics and cognitive findings
Melatonin is secreted cyclically under the regula-
relevant to neuropsychological assessment and the neu-
tion of the suprachiasmatic nucleus of the hypothalamus
roendocrine system.
(Reiter, 1990). It serves as the "dark" signal to the brain,
The following guidelines will assist the clinician in
with low levels associated with daylight and increases conceptualization of neuroendocrine disorders, history
peaking toward midnight and then gradually returning to
taking, and, ultimately, protocol interpretation. Here, clin-
baseline by morning. Exogenous melatonin is currently icians should use these concepts, while referring to the
enjoying popularity as a treatment for jet lag because of
particular characteristics of the hormone involved.
its ability to reset the body's circadian rhythms (Lewy
et al., 1992) and to induce sleepiness (Lieberman et al., • Endocrine disorders can be analyzed according
1984). Its other principal effect is to suppress core body to principles familiar to neuropsychologists.
temperature (Deacon et al., 1994). Although the "clock- Patient symptoms can be "localized" to discrete
setting" capacities of melatonin are most likely controlled neuroendocrine systems according to the pattern
directly by afferents from the suprachiasmatic nucleus, of effects on behavior and cognition. Knowledge
its temperature-suppression function is governed by more of the underlying pathophysiology will aid the
complex hypothalamic thermoregulation factors (Saarela clinician in providing effective interventions and
and Reiter, 1994). in formulating recommendations.
Several studies have found significant variation in • Cognitive dysfunction may be due to any of three
cognitive functioning, particularly in reaction time tasks, conditions: (1) Abnormal hormonal levels
198 Erlanger, Kutner, and Jacobs
Hormones and Cognition 199

present in a normal brain, (2) normal hormonal the patient's compliance with his or her medi-
levels present in an abnormal brain, or (3) ab- cation regimen. Careful attention should be paid
normal hormonal levels present in an abnormal to recent treatment adjustments. When feasible,
brain. patients should be tested only when under ad-
• Pituitary adenomas produce deficit patterns that equate metabolic control. At times, neuropsy-
are due to multiple factors. These include presur- chological testing may need to be deferred for
gical hypo- or hypersecretion of one or more hor- a number of weeks to allow for accurate assess-
mones, deficits due to mass effect, sequelae of ment.
resection, and postsurgical hormone replacement
In the research setting, knowledge of cognitive func-
therapy. History taking should carefully track the
tions determined by neuropsychologists is increasingly
chronology of symptom onset and development
important to scientific endeavors. Recent advances regard-
along with the sequence of therapeutic proce-
ing the role of ovarian hormones in AD represent an ex-
dures. This will aid in more accurate attribution
citing opportunity to utilize complex knowledge of neu-
of symptoms and may be pertinent to referral
roendocrine functioning to treat a central nervous system
questions regarding employment and rehabilita-
disorder. Other hormone-based treatments are currently
tion.
being investigated to treat cognitive problems associated
• Histories of metabolic control may be complex.
with aging and disease processes. Most recently, DHEA
Psychiatric symptoms frequently mask an insid-
has generated considerable interest. It is hoped that the
ious onset of symptoms of hormone dysregula-
previous discussions will serve to facilitate future research
tion and delay diagnosis. Once such dysregula-
efforts by neuropsychologists.
tion is diagnosed, stabilization through hormone
The preceding discussions note many of the method-
replacement therapy may be an ongoing process.
ological problems inherent in studying the effects of hor-
Often, treatment of hyperfunction may result in
monal dysfunction on cognition. Indeed, many of the
hypofunction of an endocrine gland and, occa-
conflicting findings are due to selection problems, mea-
sionally, vice versa.
surement difficulties, and confounding factors. Still, a
• Special notice should be taken of episodes associ-
number of the studies reviewed demonstrate that careful
ated with extreme hormonal imbalance and acute
planning can mitigate these problems successfully. The
metabolic crises because even isolated events may
following guidelines are suggested:
produce persistent cognitive sequelae. Exam-
ples include significant traumatic experiences • Full neuropsychological batteries should be used
such as shock due to Addison's disease, hypo- whenever feasible. When this is impractical, spe-
glycemic shock and its associated seizure or cialized investigations into target domains should
coma, and post-traumatic stress disorder. utilize multiple measures within each domain
• Age of onset is an important factor. In general, (e.g., sustained visual attention vs. sustained au-
perinatal and early childhood events increase the ditory attention, list learning vs. incidental learn-
likelihood of neurocognitive deficits because of ing, etc.) to add specificity to the interpretation of
the greater vulnerability of the child brain and/or findings. This would also allow for more measures
greater chronicity. Also, aging produces lower to be duplicated between studies, which would
basal levels of endocrine function, and the aging enable reconciliation of disparate findings.
brain may be more vulnerable to certain types of • Robust IQ data are crucial for interpreting
deficits, particularly those entailing memory. For between-group differences on domain measures
instance, patients with Type II diabetes appear to because relatively small differences will likely be
experience memory dysfunction more frequently of interest.
than patients with Type I do. • Mood/symptom inventories are important in the
• Metabolic control at time of testing may affect study of neuroendocrine disorders because of
neuropsychological test performance. For ins- their common comorbidity and potential to af-
tance, glucose levels may directly affect a range fect neuropsychological test performance.
of cognitive capacities. Also, anxiety or depres- • A careful history of metabolic control should be
sive symptoms due to variation in thyroid or cor- obtained for each subject because individuals
tisol levels may affect attention in particular and with the same disorder may manifest different
other functions. Clinicians should obtain infor- patterns of neuropsychological weakness based
mation regarding current metabolic control and on this factor.
200 Erlanger, Kutner, and Jacobs

• Control subjects should be recruited with care, Badia, P., Culpepper, J., Myers, B., Boecker, M., and Harsh, J. (1990).
Several studies have clarified discrepancies in the Psychophysiological and behavioral effects of bright and dim light.
Sleep Research 10:387.
literature by utilizing siblings or matched con- Baker, S. W., and Ehrhardt, A. A. (1974). Prenatal androgen, intelligence
trols, or by using subjects as their own control. and cognitive sex differences. In Friedman, R. C., Richart, R. M.,
and Vande Wiele, R. L. (eds.), Sex Differences in Behavior, Wiley,
Finally, it is worth noting that neuropsychologists New York, pp. 53-76.
do not yet routinely provide cognitive remediation and Baldini, I. M., Wita, A., Mauri, M. C., et al. (1997). Psychopathological
and cognitive features in subclinical hypothyroidism. Progress in
other treatments to individuals with cognitive dysfunction Neum-Psychopliarmacology and Biological Psychiatry 21: 925-
secondary to hormonal dysregulation. Although the med- 935.
ical condition of these patients frequently improves with Bale, R. N. (1973). Brain damage in diabetes mellitus. British Journal
of Psychiatry 122: 22-39.
hormone replacement therapy, cognitive dysfunction per- Bargagna, S., Chiovato, L., Dinetti, D., et al. (1997). Neuropsycholog-
sists in many cases. Cognitive rehabilitation techniques ical development in a child with early-treated congenital hypothy-
are likely to be useful to patients in developing compen- roidism as compared with her unaffected identical twin. European
Journal of Endocrinology 136: 100-104.
satory resources. Because neuroendocrine disorders ap- Barrett-Connor, E., and Bush, T. L. (1991). Estrogen and coronary heart
pear in many instances to produce cognitive dysfunction disease in women. Journal of the American Medical Association
independent of neuronal degeneration, application of cog- 265: 1861-1867.
Barrett-Connor, E., Khaw, K., and Yen, S. S. C. (1986). A prospective
nitive remediation techniques may produce outcomes that study of DS, mortality and cardiovascular disease. New England
differ from those of patients with injuries that produce Journal of Medicine 315: 1519-1524.
necrosis, such as traumatic brain injury, stroke, and so Barrett-Connor, E., and Kritz-Silverstein, D. (1993). Estrogen replace-
ment therapy and cognitive function in older women. Journal of the
forth. However, there is no systematic research at present American Medical Association 269: 2637-2641.
in this area. Also, psychological issues regarding affective Beatty, W. W. (1984). Hormonal organization of sex differences in play
symptoms and adjustment to illness are familiar to most fighting and spatial behavior. Progress in Brain Research 61: 315-
329.
neuropsychologists, as are the significant family issues Beckwith, B. E., and Tucker, D. M. (1988). Thyroid disorders. In Tarter,
that pertain to dealing with chronic cognitive dysfunction. R. E., van Thiel, D. H., and Edwards, K. L. (eds.), Medical Neu-
Given the benefits of a therapy that integrates these diverse ropsychology: The Impact of Disease on Behavior, Plenum Press,
New York, pp. 197-221.
areas, neuropsychologists appear extremely well qualified Bennett, G. W., Marsden, C. A., Fone, K. C. F, Johnson, J. V., and Heal,
to play a major role in the multidisciplinary treatment of D. J. (1989). TRH-catecholamine interactions in brain and spinal
these patients. cord. Annals of the New York Academy of Sciences 553: 106-120.
Bennett, M. C., Diamond, D. M., Fleshner, M., and Rose, G. M. (1991).
Serum corticosterone level predicts the magnitude of hippocampal
primed burst potentiation and depression in urethane-anesthetized
REFERENCES rats. Psychobiology 19: 301-307.
Benton, A. L. (1968). Differential behavioral effects in frontal lobe dis-
Aghini-Lombardi, F. A., Pinchera, A., Anonangeli, L., et al. (1995). Mild ease. Neuropsychologia 6: 53-60.
iodine deficiency during fetal/neonatal life and neuropsychological Benton, A. L. (1974). Revised Visual Retention Test (4th ed.), The Psy-
impairment in Tuscany. Journal of Endocrinological Investigation chological Corporation, San Antonio, TX.
18:57-62. Berg, E. A. (1948). A simple objective technique for measuring flexibility
Allen, L. S.,andGorski,R. A. (1986). Sexual dimorphism of the human in thinking. Journal of General Psychology 39: 15-22.
anterior commissure. Anatomical Record 214: 3A. Berman, F. B., Schmidt, P. J., Rubinow, D. R., et al. (1997). Modulation of
Allen, L. S., Hines, M., Shryne, J. E., and Gorski, R. A. (1989). Two sex- cognition-specific cortical activity by gonadal steroids: A positron-
ually dimorphic nuclei in the human brain. Journal of Neuroscience emission tomography study in women. Proceedings of the National
9: 497-506. Academy of Sciences, USA 93: 8836-8841.
Alvarez, M. A., Gomez, A., Alavez, E., and Navarro, D. (1983). Attention Birge, S. J. (1994). The role of estrogen deficiency in the aging cen-
disturbance in Graves' disease. Psychoneumendocrinology 8:451- tral nervous system. In Lobo, R. A. (ed.), Treatment of the Post-
454. menopausal Woman: Basic and Clinical Aspects, Raven Press, New
Anderson, F, Hamburger, S., Liu, J. H., and Rebar, R. W. (1987). Charac- York, pp. 153-157.
teristics of menopausal women seeking assistance. American Jour- Birge, S. J. (1997). The role of estrogen in the treatment of Alzheimer's
nal of Obstetrics and Cynecology 156: 428-433. disease. Neurology 48 (Suppl. 7): S36-S41.
Artunkal, S., and Togrol, B. (1964). Psychological studies in hyper- Bliss, T. V. P., and Lomo, T. (1973). Long-lasting potentiation of synaptic
thyroidism. In Cameron, M.R., and O'Connor, M. (eds.), Brain- transmission in the dentate area of the anaesthetized rabbit following
Thyroid Relationships, Little Brown, Boston, pp. 92-113. stimulation of the perforant path. Journal of Physiology (London)
Assisi, A., Alimenti, M., Maceli, F, Di Pietro, S., Lalloni, G., and 232:331-356.
Montera, P. (1996). Diabetes and cognitive function: Preliminary Bommer, M., Eversmann, T., Pickardt, R., Leohnardt, A., and Naber, D.
studies. Archives of Gerontology and Geriatrics (Suppl. 5): 229- (1990). Psychopathological and neuropsychological symptoms in
232. patients with subclinical and remitted hyperthyroidism. Klinische
Asthana, S., Craft, S., Baker, L. D., Raskind, M. A., Birnbaum, R. S., Wochenschrift 6S: 552-558.
Lofgreen, C. P., Veith, R. C., and Plymate, S. R. (1999). Cogni- Bouma, L., and Lindeboom, J. (1988). Klinisch neumpsychologische As-
tive and neuroendocrine response to transdermal estrogen in post- sessment, een Handlekding voor de Praktijk. Internal report, Vrije
menopausal women with Alzheimer's disease: Results of a placebo- Universiteit, Amsterdam.
controlled, double-blind, pilot study. Psychoneumendocrinology Bremner, J. D., Randall, P., Scott, T. M., et al. (1995). MRI-based mea-
24: 657-677. surement of hippocampal volume in patients with combat-related
Hormones and Cognition 201

posttraumatic stress disorder. American Journal of Psychiatry 152: Derksen-Lubsen, G. (1996). Neuropsychologic development in early
973-981. treated congenital hypothyroidism: Analysis of literature data. Pe-
Brenner, D. E., Kukull, W. A., Stergachis, A., et al. (1994). Post- diatric Research 39: 561-566.
menopausal estrogen replacement therapy and the risk of Deutsch, S. I., Mastropaolo, J., and Hitri, A. (1992). GABA-active
Alzheimer's disease: A population-based case-control study. Amer- steroids: Endogenous modulators of GABA-gated chloride ion con-
ican Journal of Epidemiology 140: 262-267. ductance. Clinical Neuropharmacology 15: 352-364.
Bunevicius, R., Kazanavicius, G., Zalinkevicius, R., and Prange, A. J. De Wied, D. (1965). The influence of the posterior and intermediate
(1999). Effects of thyroxine as compared with thyroxine plus tri- lobe of the pituitary and pituitary peptides on the maintenance of
iodothyronine in patients with hypothyroidism. New England Jour- a conditioned avoidance response in rats. International Journal of
nal of Medicine 340: 424-429. Neuropharmacology 4: 157-167.
Buschke, H. (1973). Selective reminding for analysis of memory and Dodt, C., Dittmann, J., Hruby, J., et al. (1991). Different regulation
learning. Journal of Verbal Learning and Verbal Behavior 12: 543- of adrenocorticotropin and cortisol secretion in young, mentally
550. healthy elderly and patients with senile dementia of Alzheimer's
Buschke, H., and Fuld, P. A. (1974). Evaluation of storage, retention, type. Journal of Clinical Endocrinology and Metabolism 72: 272-
and retrieval in disordered memory and learning. Neurology 11: 276.
1019-1025. Dollins, A. B., Lynch, H. J., Wurtman, R. J., and Deng, M. H. (1994).
Caldwell, B. M. (1954). An evaluation of psychological effects of sex Effects of illumination on human nocturnal serum melatonin levels
hormone administration in aged women. Journal of Gerontology 9: and performance. Physiology and Behavior 53: 153-160.
168-174. Ekstrom, R. B., French, J. W., Harman, H. H., and Derman, D. (1976).
Caldwell, B. M., and Watson, R. I. (1952). An evaluation of psycho- Manual for Referenced Cognitive Tests, Educational Testing Ser-
logic effects of sex hormone administration in aged women. Re- vice, Princeton, NJ.
sults of therapy after six months. Journal of Gerontology 1: 228- Emmen, H., Hogendijk, E., Hooisma, J., Orlib eke, J., and Uijtdehaage,
244. S. (1988). Adaptation of two standardized international test bat-
Carlson, L. E., and Sherwin, B. B. (1998). Steroid hormones, memory and teries for use in the Netherlands for detection of exposure to neu-
mood in a healthy elderly population. Psychoneuroendocrinology rotonic compounds. Internal report, Medisch Biologisch Laborato-
23: 583-603. rium TNO, Rijswijk, The Netherlands.
Carroll, B. J. (1982). The dexamethasone suppression test for melancho- Fehm-Wolfsdorf, G., and Born, J. (1991). Behavioral effects of neuro-
lia. British Journal of Psychiatry 140: 292-304. hypophyseal peptides in healthy volunteers: 10 years of research.
Carter Center. (1985). Closing the gap: The problem of diabetes mellitus Peptides 12: 1399-1406.
in the United States. Diabetes Care 8: 391-406. Fenzi, G. F., Giusti, L. F, Agnini-Lombardi, F, et al. (1990). Neuropsy-
Claussen, M. S., Landercasper, J., and Cogbill, T. H. (1992). Acute chological assessment in schoolchildren from an area of moderate
adrenal insufficiency presenting as shock after trauma and surgery: iodine deficiency. Journal of Endocrinological Investigation 13:
Three cases and review of the literature. Journal of Trauma 32: 427-431.
94-100. Feydor-Freybergh, P. (1977). The influence of estrogen on the well-
Clopper, R. R. (1990). Assessing the effect of replacement hor- being and mental performance in climacteric and postmenopausal
mone treatment on psychosocial and psychosexual behavior in women. Acta Obstetricia et Gynecologica Scandinavica 64:
growth hormone deficient individuals. In Holme, C. S. (ed.), 5-69.
Psychoneuroendocrinology, Springer-Verlag, New York, pp. 56- Fillit, H., Weinreb, H.,Cholst, l.,et al. (1986). Observations in a prelimi-
78. nary open trial of estradiol therapy for senile dementia-Alzheimer's
Colditz, G. A., Hankinson, S. E., Hunter, D. J., et al. (1995). The use type. Psychoneuroendocrinology 11: 337-345.
of estrogens, and progestins and the risk of breast cancer in post- Flood, J. F., Smith, G. E., and Roberts, E. (1988). Dehydroepiandros-
menopausal women. New England Journal of Medicine 332: 1589- terone and its sulfate enhance memory retention in mice. Brain
1593. Research 447: 269-278.
Collaer, M. L.,and Hines, M. (1995). Human behavioral sex differences: Franceschi, M., Ceccetto, R., Minicucci, F., Smizne, S., Baio, G., and
A role for gonadal hormones during early development? Psycho- Canal, N. (1984). Cognitive processes in insulin-dependent dia-
logical Bulletin 118: 55-107. betes. Diabetes Care 7: 228-231.
Craik, R I. M. (1984). Age differences in remembering. In Squire, L. R., Frye, C. A., and Sturgis, J. D. (1995). Neurosteroids affect spa-
and Butters, N. (eds.), Neuropsychology of Memory, Guilford Press, tial/reference, working and long-term memory of female rats. Neu-
New York, pp. 3-12. robiology oj Learning and Memory 64: 83-96.
Crown, S. (1949). Notes on experimental study of intellectual deteriora- Funk, J. L., Mortel, K. F., and Meyer, J. S. (1991). Effects of estrogen
tion. British Medical Journal 2: 684-685. replacement therapy on cerebral perfusion and cognition among
Davis, K. L, Davis, B. M., Greenwald, B. S., etal. (1986). Cortisol and postmenopausal women. Dementia 2: 268-272.
Alzheimer's disease. I. Basal studies. American Journal of Psychi- Geschwind, N., and Behan, P. (1982). Left-handedness: Association with
atry 143: 300-305. immune disease, migraine, and developmental learning disorder.
Deacon, S., English, J., and Arendt, J. (1994). Acute phase-shifting ef- Proceedings of the National Academy of Sciences, USA 79: 5097-
fects of melatonin associated with suppression of body core tem- 5100.
perature in humans. Neuroscience Letters 178: 153-160. Geschwind, N., and Galaburda, A. M. (1985a). Cerebral lateralization.
Deijen, J. B., de Boer, H., Blok, G. J., and van der Veen, E. A. (1996). Part I. Archives of Neurology 42: 428-459.
Cognitive impairments and mood disturbances in growth hormone Geschwind, N., and Galaburda, A. M. (1985b). Cerebral lateralization.
deficient men. Psychoneuroendocrinology 21:313-322. Part II. Archives of Neurology 42: 521-552.
Deijen, J. B., de Boer, H., and van der Veen, E. A. (1998). Cognitive Gladue, B. S., and Baily, J. M. (1995). Spatial ability, handedness, and
changes during growth hormone replacement in adult men. Psy- human sexual orientation. Psychoneuroendocrinology 20:487-497.
choneuroendocrinology 23: 45-55. Goodglass, J., and Kaplan, E. (1983). The Assessment of Aphasia and
Delis, D. C., Kramer, H. H., Kaplan, E., and Ober, B. A. (1987). Califor- Related Disorders (2nd ed.), Lea & Febiger, Philadelphia.
nia Verbal Learning Test: Adult Version, The Psychological Corpo- Gordon, H. W. (1986). The Cognitive Laterality Battery: Tests of spe-
ration, San Antonio, TX. cialized cognitive function. International Journal of Neuroscience
Denicoff, K. D., Joffe, R. T, Lakshmanan, M. C., Robbins, J., and 29: 223-244.
Rubinow, D. R. (1990). Neuropsychiatric manifestations of al- Gordon, H. W., Corbin, E. D., and Lee, P. A. (1986). Changes in spe-
tered thyroid state. American Journal of Psychiatry 147: 94- cialized cognitive function following changes in hormone levels.
99. Cortex 22: 399-4115.
202 Erlanger, Kutner, and Jacobs

Gordon, H. W., and Lee, P. A. (1986). A relationship between go- Henderson, V. W. (1997). Epidemiology of estrogen replacement therapy
nadotropins and visuospatial function. Neuropsychologia 24: 563- and Alzheimer's disease. Neurology 48 (Suppl. 7): S27-S35.
576. Henderson, V. W., and Buckwalter, J. G. (1994). Cognitive deficits of
Gordon, H. W., and Lee, P. A. (1993). No differences in cognitive per- men and women with Alzheimer's disease. Neurology 44: 90-96.
formance between phases of the menstrual cycle. Psyclwneuroen- Henderson, V. W., Mack, W., and Williams, B. W. (1989). Spatial dis-
docrinology 18:521-531. orientation in Alzheimer's disease. Archives of Neurology 46: 391-
Gordon, H. W., Lee, P. A., and Tamres, L. K. (1988). The pituitary 394.
axis: Behavioral correlates. In Tarter, R. E., van Thiel, D. H., and Henderson, V. W., Paganini-Hill, A., Emanuel, C. K., et al. (1994).
Edwards, K. L. (eds.), Medical Neuropsychology: The Impact of Estrogen replacement therapy in older women: Comparisons be-
Disease on Behavior, Plenum Press, New York, pp. 159-196. tween Alzheimer's disease cases and non-demented control sub-
Gordon, M., Crouthamel, D., Post, E. M., and Richman, R. A. (1982). jects. Archives of Neurology 51: 896-900.
Psychosocial aspects of constitutional short stature: Social com- Henderson, V. W., Watt, L., and Buckwalter, J. G. (1996). Cognitive skills
petence, behavior problems, self esteem, and family functioning. associated with estrogen replacement in women with Alzheimer's
Journal of Pediatrics 101: 477-480. disease. Psychoneuroendocrinology 21: 421-430.
Gorski, R. A. (1991). Sexual differentiation of the endocrine brain and Hershey, T., Craft, S., Bhargava, N., and White, N. (1997). Memory
its control. In Motta, M. (ed.), Brain Endocrinology, Raven Press, and insulin dependent diabetes mellitus (IDDM): Effects of child-
New York, pp. 71-104. hood onset and severe hypoglycemia. Journal of the International
Gouchie, C., and Kimura, D. (1991). The relationship between Neuropsychological Society 3: 509-520.
testosterone levels and cognitive ability patterns. Psychoneuroen- Herzog, A. G. (1991). Reproductive endocrine considerations and hor-
docrinology 16: 323-334. monal therapy for women with epilepsy. Epilepsia 21 (Suppl. 6):
Gradman, T., Taus, A., Thompson, L., and Reaven, G. M. (1991). Effects S27-S33.
of glycemic control on cognitive function in older subjects with Herzog, A. G. (1995). Progesterone therapy in women with complex
NIDDM. Clinical Research 39: 14A. partial and secondary generalized seizures. Neurology 45: 1660-
Grady, D., and Ernster, V. (1991). Does postmenopausal hormone ther- 1662.
apy cause breast cancer? American Journal of Epidemiology 134: Hines, M., Chiu, L., McAdams, L. A., Beutler, P. M., and Lipcamon,
1396-1400. J. (1992). Cognition and the corpus callosum: Verbal fluency, vi-
Grant, D. A., and Berg, E. A. (1948). A behavioral analysis of degree suospatial ability, and language lateralization related to midsagillal
of impairment and ease of shifting to new responses in a Weigl- surface areas of callosal subregions. Behavioral Neuroscience 106:
type card sorting problem. Journal of Experimental Psychology 39: 3-14.
404-411. Hines, M., and Green, R. (1991). Human hormonal and neural correlates
Grattan-Smith, P. J., Morris, J. G. L., Shores, E. A., Bachelor, J., and of sex-typed behaviors. Review of Psychiatry 10: 536-555.
Sparks, R. S. (1992). Neuropsychological abnormalities in patients Holloway, R. L., and de LaCoste, M. C. (1986). Sexual dimorphism in
with pituitary tumors. Acta Neurologica Scandinavica 86:626-631. the human corpus callosum: An extension and replication study.
Graves, R. G. (1835). Newly observed affection of the thyroid gland in Human Neurobiology 5: 87-91.
females. Medicine and Surgery 7: 516. Holmes, C. S., Dunlap, W. P., Chen, R. S., and Cornwell, J. M. (1992).
Grumbach, M. M., and Conte, F. A. (1992). Disorders of sex differenti- Gender differences in cognitive performance in diabetes. Journal
ation. In Wilson, J. D., and Foster, D. W. (eds.), Williams Textbook of Consulting and Clinical Psychology 60: 698-704.
of Endocrinology, Saunders, Philadelphia, pp. 853-951. Holmes, C. S., Hayford, J. T., Gonzalez, J. L., and Weydert, J. A. (1983).
Gull, W. W. (1874). On a cretinoid state supervening in adult life in A survey of cognitive functioning at different glucose levels in
women. Transactions of the Medical Society of London 21: 298- diabetic persons. Diabetes Care 6: 180-185.
300. Honjo, H., Ogino, Y, Naitoh, K., et al. (1989). In vivo effects by
Gustafsson, B., and Wigstrom, T. (1988). Physiological mechanisms estrone sulfate on the central nervous system—senile dementia-
underlying long-term potentiation. Trends in Neuroscience 11:156- Alzheimer's type. Journal of Steroid Biochemistry 34: 521-
162. 525.
Hackman, B. W., and Galbraith, D. (1977). Six month study of oestrogen Honjo, H., Ogino, Y.,Tanaka, K., etal. (1993). An effect of conjugated
therapy with piperazine oestronesulphate and its effects on memory. estrogen to cognitive impairment in women with senile dementia-
Current Medical Research and Opinion 4: 21-27. Alzheimer's type: A placebo-controlled double-blind study. Journal
Halpern, D. F. (1992). Sex Differences in Cognitive Abilities (2nd ed.), of the Japanese Menopause Society 1: 167-171.
Erlbaum, Hillsdale, NJ. Imperato-McGinley, J., Pichardo, M., Gautier, T., Voyer, D., and Bryden,
Halstead, W. C. (1947). Brain and Intelligence, University of Chicago M. P. (1991). Cognitive abilities in androgen-insensitive subjects:
Press, Chicago. Comparison with control males and females from the same kindred.
Hampson, E. (1990a). Estrogen-related variations in human spatial and Clinical Endocrinology 34: 341-347.
articulatory-motor skills. Psychoneuroendocrinology 15: 97-111. Inglis, J. (1959). An experimental study of learning and "memory func-
Hampson, E. (1990b). Variations in sex-related cognitive abilities across tion" in elderly psychiatric elders. Journal of Mental Science 103:
the menstrual cycle. Brain and Cognition 14: 26-43. 796-803.
Hampson, E., Rovet, J. F., and Altmann, D. (1998). Spatial reasoning in Jacobs, A. R., Coleman, A. E., and Herzog, A. G. (1995). Late onset
children with congenital adrenal hyperplasia due to 21 -hydroxylase congenital adrenal hyperplasia: A treatable cause of anxiety and
deficiency. Developmental Neuropsychology 14: 299-320. maladaptive behavioral disorders. Paper presented at the American
Hathaway, S. R., and McKinley, J. C. (1943). Booklet for the Minnesota Neuropsychiatric Association, October, Pittsburgh, PA.
Multiphasic Personality Inventory, The Psychological Corporation, Jacobs, D. M., Tang, M. X., Stern, Y, et al. (1998). Cognitive function
New York. in nondemented older women who took estrogen after menopause.
Helladay, J., Bartfai, A., Martin Ritzen, E., and Forsman, M. (1994). Neurology 50: 368-373.
General intelligence and cognitive profile in women with congenital Jacobson, L., and Sapolsky, R. (1991). The role of the hippocampus in
adrenal hyperplasia (CAM). Psychoneuroendocrinology 19: 343- feedback regulation of the hypothalamic-pituitary-adrenocortical
356. axis. Endocrine Review 12: 118-134.
Helzlsouer, K. J., Gordon, G. B., Alberg, A., Bush, T. L., and Comstock, Jain, V. K. (1973). A psychiatric study of hypothyroidism. Psychiatric
G. W. (1992). Relationship of prediagnostic serum levels of DHEA Clinician5: 121-130.
and DS to the risk of developing premenopausal breast cancer. Can- Jakobsen, J., Sidenius, P., Gendersen, H., and Osterby, R. (1987). Quan-
cer Research 52: 1-4. titative changes of cerebral neocortical structure in insulin-treated
Hormones and Cognition 203

long-term streptozocin-induced diabetes in rats. Diabetes 36: 597- Linn, M. C., and Petersen, A. C. (1985). Emergence and characteriza-
601. tion of sex differences in spatial ability: A meta-analysis. Child
Janowsky, J. S., Oviatt, S. K., and Orwoll, E. S. (1994). Testosterone Development 56: 1479-1498.
influences spatial cognition in older men. Behavioral Neuroscience Loosen, P. T. (1992). Effects of thyroid hormones on central nervous
108:325-332. system in aging. Psychoneuroendocrinology 17: 355-374.
Jennings, J. R., Nebes, R. D., and Reynolds, C. R, III (1986). Vasopressin Lupien, S., Lecours, A. R., Lussier, I., el al. (1994). Basal cortisol levels
peptide (DDAVP) may narrow the focus of attention in normal el- and cognitive deficits in human aging. Journal of Neuroscience 14:
derly. Psychiatry Research 17: 31-39. 2893-2903.
Joffe, R. T., and Singer, W. (1990). A comparison of triiodothyronine and Maccoby, E. E., and Jacklin, C. N. (1974). The Psychology of Sex Dif-
thyroxine in the potentiation of tricyclic antidepressants. Psychiatry ferences, Stanford University Press, Stanford, CA.
Research 32: 241-251. MacCrimmon, D. J., Wallace, J. E., Goldberg, W. M., and Streiner, D. L.
Johnstone, P. A., Randell, J. R., and Esposito, M. (1990). Mental status (1979). Emotional disturbance and cognitive deficits in hyperthy-
changes of Addison's disease. Psychosomatic* 31: 103-107. roidism. Psychosomatic Medicine 41: 331.
Jorm, A. R, Korten, A. E., and Henderson, A. S. (1987). The preva- Madeira, M. D., Sousa, N., Lima-Andrade, M. T., Calheiros, P., Cadete-
lence of dementia: A quantitative integration of the literature. Acta Leite, A., and Paula-Barbosa, M. M. (1992). Selective vulnerability
Psychiatrica Scandinavica 76: 475-479. of the hippocampal pyramidal neurons to hypothyroidism in male
Kampen, D., and Sherwin, B. B. (1994). Estrogen use and verbal memory and female rats. Journal of Comparative Neurology 322: 501-518.
in healthy postmenopausal women. Obstetrics and Gynecology 83: Masica, D. N., Money, J., Ehrhardt, A. A., and Lewis, V. G. (1969). IQ,
979-983. fetal sex hormones and cognitive patterns: Studies in the testicu-
Kaplan, E., Goodglass, H., and Weintraub, S. (1983). Boston Naming lar feminizing syndrome of androgen insensitivity. Johns Hopkins
Test, Lea & Pebiger, Philadelphia. Medical Journal 124: 34—43.
Kawas, C., Resnick, S., Morrison, A., et al. (1997). A prospective Mateer, C., and Kimura, D. (1977). Impairment of nonverbal oral move-
study of estrogen replacement therapy and the risk of developing ments in aphasia. Brain and Language 4: 262-276.
Alzheimer's disease: The Baltimore Longitudinal Study of Aging. Mauri, M., Sinforiani, E., Bono, G., et al. (1993). Memory impairment
Neurology 48: 1517-1521. in Cushing's disease. Acta Neurologica Scandinavica 87: 52-55.
Keenan, P. A., Stern, R. A., Janowsky, D. S., and Pedersen, C. A. (1992). McCall, A. L. (1992). The impact of diabetes on the CNS. Diabetes 41:
Psychological aspects of premenstrual syndrome. I: Cognition and 557-570.
memory. Psychoneuroendocrinology 17: 179-187. McCord, M. W., Buatti, J., Fennell, E., et al. (1997). Radiotherapy for
Kelly, J. A. (1995). Thyrotrophin-releasing hormone: Basis and potential pituitary adenoma: Long-term outcome and sequelae. International
for its therapeutic use. Essays in Biochemistry 30: 133-149. Journal of Radiation Oncology and Biological Physiology 39: 437-
Khan, A., Mirolo, M. H., Claypoole, K., Bhang, J., Cox, 0., Horita, 444.
A., and Tucker, G. (1994). Effects of low-dose TRH on cognitive McEwen, J. W. (1982). Glucocorticoids and hippocampus: Receptors in
deficits in the ECT postictal state. American Journal of Psychiatry search of a function. In Ganten, D., and Pfaff, D. (eds.), Adrenal
151: 1694-1696. Actions on Brain, Springer-Verlag, New York, pp. 1-22.
K10ve, H. (1963). Clinical neuropsychology. In Forster, F. M. (ed.), The McGuire, L. S., Ryan, K. O., and Omen, G. S. (1975). Congenital adrenal
Medical Clinics of North America, Saunders, New York. hyperplasia. II. Cognitive and behavioral studies. Behavioral Ge-
Knapp, M. J., Knopman, D. S., Solomon, P. R., et al. (1994). A 30- netics 5: 175-188.
week randomized controlled trial of high-dose tacrine in patients Melchior, C. L., and Ritzmann, R. F. (1996). Neurosteroids block the
with Alzheimer's disease: The Tacrine Study Group. Journal of the memory-impairing effects of ethanol in mice. Pharmacology, Bio-
American Medical Association 271: 85-91. chemistry and Behavior 53: 51-56.
Komnenich, P., Lane, D. M., Dickey, R. P., and Stone, S. C. (1978). Mellman, M. J., Davis, M. R., Brisman, M., and Shamoon, H. (1994).
Gonadal hormones and cognitive performance. Physiological Psy- Effect of antecedent hypoglycemia on cognitive function and on
chology d: 115-120. glycemic thresholds for counterregulatory hormone secretion in
Lai, Z., Emtner, M., Roos, P., and Nyberg, R. (1991). Characterization of healthy humans. Diabetes Care 17: 183-188.
putative growth hormone receptors in human choroid plexus. Brain Mendorla.G., Sava, L., Calaciura, R, Lisi, E., Castorina, S., and Vigneri,
Research 46: 222-226. R. (1988). Personality traits and mental prognosis in patients with
Lamberts, S. W. J., van den Beld, A. W., and van der Lely, A. (1997). congenital hypothyroidism not treated from early life. Journal of
The endocrinology of aging. Science 278: 419-424. Endocrinological Investigation 11: 289-295.
Leentjens, A. F. G., and Kappers, E. J. (1995). Persistent cognitive de- Meneilly, G. S., Cheung, E., Tessier, D., Yakura, C., and Tuokko, H.
fects after corrected hypothyroidism. Psychopathology 28: 235- (1993). The effect of improved glycemic control on cognitive func-
237. tions in the elderly patient with diabetes. Journal of Gerontology
Legros, J., and Gilot, P. (1979). Vasopressin and memory in the human. 48: M117-M121.
In Krieger, D. T., Brownstein, M. J., and Martin, J. B. (eds.), Brain Mennemeier, M., Garner, R. D., and Heilman, K. M. (1993). Memory,
Peptides, Wiley, New York, pp. 347-364. mood and measurement in hypothyroidism. Journal of Clinical and
Lewis, R., and Rennick, P. (1979). Manual for the Repeatable Cognitive- Experimental Neuropsychology 15: 822-831.
Perceptual-Motor Battery, Axon, Grosse Point Park, MI. Meuter, R, Thomas, W., GrUneklee, D., Gries, F., and Lohmann,
Lewy, A. J., Ahmed, S., Latham Jackson, J. M., and Sack, R. (1992). R. (1980). Psychometric evaluation of performance in diabetes
Melatonin shifts circadian rhythms according to a phase-response mellitus. Hormone and Metabolic Research (Supplement} 9: 9-
curve. Chronobiology International 9: 380-392. 17.
Lichty, W., and Klachko, D. (1985). Memory in Type I diabetics Meyer-Bahlburg, H. F. L., Feinman, J. A., MacGillivray, M. H., and
[Abstract]. Diabetes 34 (Suppl. 1): 19. Acetor, T, Jr. (1978). Growth hormone deficiency, brain develop-
Lieberman, H. R., Waldhauser, R, Garfield, G., Lynch, H. J., and ment, and intelligence. American Journal of Diseases of Children
Wurtman, R. J. (1984). Effects of melatonin on human mood and 132:565-572.
performance. Brain Research 323: 201-207. Milner, B. (1964). Some effects of frontal lobectomy in man. In Warren,
Lindsay, R., Hart, D. M., Aitken, J. M., MacDonald, E. B., Anderson, J. M., and Akert, K. (eds.), The Frontal Granular Cortex and Be-
J. B., and Clarke, A. C. (1976). Long-term prevention of post- havior, McGraw-Hill, New York, pp. 313-331.
menopausal osteoporosis by oestrogen: Evidence for an increased Milner, B. (1971). Interhemispheric differences in the localization of
bone mass after delayed onset of oestrogen treatment. Lancet psychological processes in man. British Medical Bulletin 27: 272-
1: 1038-1040. 277.
204 Erlanger, Kutner, and Jacobs

Minaker, K. L. (1990). What diabetologists should know about elderly Parry, C. H. (1825). Cited in Beckwith, B. E., and Tucker, D. M. (1988).
patients. Diabetes Care 13 (Suppl. 2): 34. Thyroid disorders. In Tarter, R. E., van Thiel, D. H., and Edwards,
M5bius, P. J. (1886). Cited in Beckwith, B. E., and Tucker, D. M. (1988). K. L. (eds.), Medical Neumpsychology: The Impact of Disease on
Thyroid disorders. In Tarter, R. E., van Thiel, D. H., and Edwards, Behavior, Plenum Press, New York, pp. 197-221.
K. L. (eds.), Medical Neumpsychology: The Impact of Disease on Peabody, C. A., Theimann, S., Pigache, R., Miller, T. P., Berger, P. A.,
Behavior, Plenum Press, New York, pp. 197-221. Yesavage, J., and Tinklenberg, J. R. (1985). Desglyciamide-9-
Moffat, S. D., and Hampson, E. (1996). A curvilinear relationship be- arginine-8-vasopressin (DGAVP, Org 5667) in patients with de-
tween testosterone and spatial cognition in humans: Possible in- mentia. Neurobiology of Aging 6: 95-100.
fluence of hand preference. Psychoneuroendocrinology 21: 323- Peace, K. A., Orme, S. M., Thompson, A. R., Padayatt, S., Ellis, A. W.,
337. and Belchetz, P. E. (1997). Cognitive dysfunction in patients treated
Monzani, R., Del Guerra, P., Caraccio, N., et al. (1993). Subclinical for pituitary tumours. Journal of Clinical and Experimental Neu-
hypothyroidism: Neurobehavioral features and beneficial effect of mpsychology 19: 1-6.
L-thyroxine treatment. Clinical Investigator 11: 367-371. Perlman, S. M. (1973). Cognitive abilities of children with hormone
Mooradian, A., Ferryman, K., Fitten, J., Kavonian, G., and Morley, J. abnormalities: Screening by psychoeducational tests. Journal of
(1988). Cortical function in elderly non-insulin dependent diabetic Learning Disabilities 6: 21-29.
patients. Archives of Internal Medicine 148: 2369-2372. Perlmuter, L. C., Hakami, M., Hodgson-Harrington, C., Ginsberg, J.,
Morales, A. J., Nolan, J. J., Nelson, J. C., and Yen, S. C. C. (1994). Katz, J., Singer, D. E., and Nathan, D. M. (1984). Decreased cog-
Effects of replacement dose of dehydroepiandrosterone in men and nitive function in aging non-insulin-dependent diabetic patients.
women of advancing age. Journal of Clinical Endocrinology and American Journal of Medicine 77: 1043-1048.
Metabolism 78: 1360-1367. Perlmuter, L. C., Tun, P., Sizer, N., McGlinchey, R. E., and Nathan,
Morrison, A., Resnick, S., Corrada, M., Zonderman, A., and Kawas, D. M. (1987). Age- and diabetes-related changes in verbal fluency.
C. (1996). A prospective study of estrogen replacement therapy Experimental Aging Research 13: 9-14.
and the risk of developing Alzheimer's disease in the Baltimore Perras, B., Droste, C., Born, J., Fehm, H. L., and Pietrowsky, R. (1997).
Longitudinal Study of Aging. Neurology 46: A435-A436. Verbal memory after three months of intranasal vasopressin in
Mortel, K. E, and Meyer, J. S. (1995). Lack of postmenopausal estrogen healthy old humans. Psychoneuroendocrinology 22: 387-396.
replacement therapy and the risk of dementia. Journal ofNeuropsy- Perrild, H., Hansen, J. M., Arnung, K., Olsen, P. Z., and Danielsen,
chiatry and Clinical Neuroscience 1: 334—337. U. (1986). Intellectual impairment after hyperthyroidism. Acta En-
Muglia, L., Jacobson, L., Dikkes, P., and Majzoub, J. A. (1995). docrinologica 112: 185-191.
Corticotropin-releasing hormone deficiency reveals major fetal but Petersen, A, C. (1976). Physical androgyny and cognitive functioning in
not adult glucocorticoid need. Nature 373: 427-432. adolescence. Developmental Psychology 18: 77-94.
Murphy, J. E, Purdue, G. E, and Hunt, J. L. (1993). Acute adrenal in- Phillips, S. M., and Sherwin, B. B. (1992). Variations in memory function
sufficiency in the patient with bums. Journal of Burn Care and and sex steroid hormones across the menstrual cycle. Psychoneu-
Rehabilitation 14: 155-157. roendocrinology 17: 497-506.
Myers, B. L., and Badia, P. (1993). Immediate effects of different light Polo-Kantola, P., Portin, R., Polo, O., Helenius, H., Irjala, K., and
intensities on body temperature and alertness. Physiology and Be- Erkkola, R. (1998). The effect of short-term estrogen replacement
havior 52: 199-202. therapy on cognition: A randomized, double-blind, cross-over trial
Naber, D., Sand, P., and Heigl (1996). Psychopathological and neuropsy- in postmenopausal women. Obstetrics and Gynecology 91: 459-
chological effects of 8-days' corticosteroid treatment. A prospective 466.
study. Psychoneuroendocrinology 21: 25-31. Porteus, S. D. (1959). The Maze Test and Clinical Psychology, Pacific
Nass, R., and Baker, S. (199la). Androgen effects on cognition: Congen- Books, Palo Alto, CA.
ital adrenal hyperplasia. Psychoneuroendocrinology 16: 189-201. Rauramo, L., Lagerspetz, K., Engblom, P., and Punnonen, R. (1975).
Nass, R., and Baker, S. (1991b). Learning disabilities in children with The effect of castration and peroral estrogen therapy on some psy-
congenital adrenal hyperplasia. Journal of Child Neurology 6: 306- chological functions. Frontiers of Hormone Research 3: 94-104.
312. Raven, L. C. (1938). Progressive Matrices: A Perceptual Test of Intelli-
Nebes, R. D., Reynolds, C. E, and Horn, L. C. (1984). The effect of gence: Individual Form, H. K. Lewis, London.
vasopressin on memory in the healthy elderly. Psychiatry Research Reaven, G. M., Thompson, W. W.,Nahm, D., and Haskins, E. (1990). Re-
11:49-59. lationship between hyperglycemia and cognitive function in older
Ohkura, T., Isse, K., Akazawa, K., et al. (1994a). Evaluation of estrogen NIDDM patients. Diabetes Care 13: 16-21.
treatment in female patients with dementia of the Alzheimer type. Reitan, R. M. (1953). Intellectual function in myxedema. AMA Archives
Endocrine Journal 41: 361 -371. of Neurology and Psychiatry 69: 436-449.
Ohkura, T., Isse, K., Akazawa, K., et al. (1994b). Low-dose estrogen Reitan, R. M. (1955). Investigation of the validity of Halstead's measures
replacement therapy for Alzheimer disease in women. Menopause of biological intelligence. Archives of Neurological Psychiatry 73:
1: 125-130. 28.
O'Neal, M. E, Means, L. W., Poole, M. C., and Hamm, R. J. (1996). Es- Reitan, R. M., and Davison, L. A. (1974). Clinical Neumpsychology:
trogen affects performance of ovariectomized rats in a two-choice Current Status and Applications, Hemisphere, New York.
water-escape working memory task. Psvchoneuroendocrinology Reitan, R. M., and Wolfson, D. (1993). The Halstead-Reitan Neuropsy-
21:51-65. chological Test Battery: Theory and Clinical Interpretation, Neu-
Osterweil, D., Syndulko, K., Cohen, S. N., Pettier-Jennings, P. D., ropsychology Press, Tucson, AZ.
Hershman, J. M., Cummings, J. L., Tourtellotte, W. W., and Reiter, R. J. (1990). Pineal rhythmicity: Neural, behavioral, and en-
Solomon, D. H. (1992). Cognitive function in non-demented older docrine consequences. In Shafii, M., and Shafii, S. L. (eds.), Bi-
adults with hypothyroidism. Journal of the American Geriatric So- ological Rhythms, Mood Disorders, Light Therapy, and the Pineal
ciety 40: 325-335. Gland, American Psychiatric Press, Washington, DC, pp. 41-66.
Paganini-Hill, A., and Henderson, V. W. (1994). Estrogen deficiency Resnick, S. M., Berenbaum, S. A., Gottesman, I. I., and Bouchard,
and risk of Alzheimer's disease in women. American Journal of T. J., Jr. (1986). Early hormonal influences on cognitive functioning
Epidemiology 140: 256-261. in congenital adrenal hyperplasia. Developmental Psychology, 22,
Paganini-Hill, A., Ross, R. K., and Henderson, B. E. (1988). Post- 191-198.
menopausal oestrogen treatment and stroke: A prospective study. Rey, A. (1941). Psychological examination of traumatic encephalopa-
British Medical Journal 297: 519-522. thy. Archive de Psychologie 28: 286-340. (Sections translated by
Hormones and Cognition 205

J. Corwin and F. W. Bylsma, The Clinical Neuropsychologist, 1993, Sherwin, B. B. (1994). Estrogenic effects on memory in women. Annals
4-9.) of the New York Academy of Sciences 743: 756-762.
Rey, A. (1964). L'Examin Clinique en Psychologic, Presses Universi- Sherwin, B. B. (1998). Estrogen and cognitive functioning in women.
taires de France, Paris. Proceedings of the Society far Experimental Biology and Medicine
Ripich, D. N., Petril, S. A., Whitehouse, P. J., and Ziol, E. W. (1995). Gen- 217: 17-22.
der differences in language of AD patients: A longitudinal study. Sherwin, B. B., and Phillips, S. (1990). Estrogen and cognitive function
Neurology 45: 299-302. in surgically menopausal women. Annals of the New York Academy
Rebel, P., and Baulieu, E. E. (1994). Neurosteroids: Biosynthesis and of Sciences 592: 474-475.
function. Trends in Endocrinology and Metabolism 5: 1-8. Sherwin, B. B., and Tulandi, T. (1996). "Add-back" estrogen reverses
Rovet, J. F, Erlich, R. M., and Hoppe, M. G. (1987). Specific intellectual cognitive deficits induced by a gonadotropin releasing-hormone
deficits associated with the early onset of insulin-dependent diabetes agonist in women with leiomyomata uteri. Journal of Clinical En-
mellitus in children. Diabetes Care 10: 510-515. docrinology and Metabolism 81: 2545-2549.
Ruel, J. M.,and Dekloet, E. R. (1985). Two receptor systems forcorticos- Shute, V. J., Pellegrino, J. W., Hubert, L., and Reynolds, R. W. (1983). The
terone in rat brain: Microdistribution and differential occupation. relationship between androgen levels and human spatial abilities.
Endocrinology 117: 2505-2511. Bulletin of the Psychonomic Society 21: 465-468.
Russell, E. W. (1975). A multiple scoring method for the assessment Singh.M., Meyer, E.M.,Millard,W.J.,andSimpkins,J.W. (1994). Ovar-
of complex memory functions. Journal of Consulting and Clinical ian steroid deprivation results in a reversible learning impairment
Psychology 43: 800-809. and compromised cholinergic function in female Sprague-Dawley
Ryan, C. M. (1988). Neurobehavioral disturbances associated with disor- rats. Brain Research 644: 305-312.
ders of the pancreas. In Tarter, R. E., van Thiel, D. H., and Edwards, Sivan, A. B. (1992). Benton Visual Retention Test (5th ed.), The Psycho-
K. L. (eds.), Medical Neuropsychology: The Impact of Disease on logical Corporation, San Antonio, TX.
Behavior, Plenum Press, New York, pp. 121-158. Skenazy, J., and Bigler, E. (1984). Neuropsychological findings in dia-
Ryan, C. M., and Butters, N. (1984). Learning and memory impairments betes mellitus. Journal of Clinical Psychology 40: 246-258.
in young and old alcoholics: Evidence for the premature-aging hy- Slabbekoorn, D., van Goozen, S. H. M., Megens, J., Gooren, L. J. G.,
pothesis. Alcoholism: Clinical and Experimental Research 4: 288- and Cohen-Kettenis, P. T. (1999). Activating effects of cross-sex
293. hormones on cognitive functioning: A study of short-term and long-
Ryan, C. M., Vega, A., and Drash, A. (1985). Cognitive deficits in ado- term hormone effects in transsexuals. Psychoneuroendocrinology
lescents who developed diabetes early in life. Pediatrics 75: 921- 24: 423^47.
927. Slotten, H. A., and Krekling, S. (1996). Does melatonin have an effect on
Ryan, C. M., and Williams, T. M. (1993). Effects of insulin-dependent cognitive performance? Psychoneuroendocrinology 21: 673-680.
diabetes on learning and memory in adults. Journal of Clinical and Smith, A. (1982). Symbol Digit Modalities Test, Western Psychological
Experimental Neuropsychology 15: 685-700. Services, Los Angeles.
Ryan, C. M., Williams, T. M., Orchard, T, and Feingold, D. (1992). Psy- Smith, M. O., Shaywitz, S. E., Shaywitz, B. A., Gernter, J. M., Raskin,
chomotor slowing is associated with distal symmetrical polyneu- L. A., and Gelwan, E. M. (1985). Exogenous GH levels predict
ropathy in adults with diabetes mellitus. Diabetes 41: 107-113. attentional performance: A preliminary report. Journal of Develop-
Saarela, S., and Reiter, R. J. (1994). Function of melatonin in thermoreg- mental and Behavioral Pediatrics 6: 273-278.
ulatory processes. Life Science 54: 295-311. Squire, L. R., Shimamura, A. P., and Graf, P. (1987). Strength and du-
Sankar, R., Rai, B., Pulger, T, et al. (1994). Intellectual and motor ration of priming effects in normal subjects and amnestic patients.
functions in schoolchildren from severely iodine deficient region Neumpsychologia 25: 195-210.
in Sikkim. Indian Journal of Pediatrics 61: 231-236. Stabler, B., Clopper, R. R., Siegel, P. T, Stoppani, C., Compton, P. G.,
Sarrieau, A., Sharma, S., and Meaney, M. J. (1988). Postnatal develop- and Underwood, L. E. (1994). Academic achievement and psycho-
ment and environmental regulation of hippocampal glucocorticoid logical adjustment in short children. Journal of Developmental and
and mineralocorticoid receptors in the rat. Developmental Brain Behavioral Pediatrics 15: 1-6.
Research 43: 158-162. Starkman, M. N., Gebarski, S. S., Berent, S., et al. (1992). Hippocam-
Schlote, B., Nowotny, B., Schaaf, L., et al. (1992). Subclinical hyperthy- pal formation volume, memory dysfunction and cortisol levels in
roidism: Physical and mental state of patients. European Archives patients with Cushing's syndrome. Biological Psychiatry 32: 756-
of Psychiatry and Clinical Neuroscience 241: 357-364. 765.
Schmidt, P. J., Nieman, L. K., Danaceau, M. A., Adams, L. F., and Starkman, M. N., and Schteingart, D. E. (1981). Neuropsychiatric mani-
Rubinow, D. R. (1998). Differential behavioral effects of gonadal festations of patients with Cushing's syndrome. Archives of Internal
steroids in women with and in those without premenstrual syn- Medicine 141: 215-219.
drome. New England Journal of Medicine 338: 209-216. Starkman, M. N., Schteingart, D. E., and Schork, M. A. (1986).
Schmidt, R., Fazekas, F., Reinart, B., et al. (1996). Estrogen replace- Depressed mood and other psychiatric manifestations of Cush-
ment therapy in older women: A neuropsychological and brain MRI ing's syndrome: Relationship to hormone levels. Psychosomatic
study. Journal of the American Geriatrics Society 44: 1307-1313. Medicine 43:3-18.
Schneider, L. S., Farlow, M. R., Henderson, V. W., et al. (1996). Effects Stern, R. A., Nevels, C. T, Shelhorse, M. E., Prohaska, M. L., Mason,
of estrogen replacement therapy on response to tacrine in patients G. A., and Prange, A. J., Jr. (1991). Antidepressant and memory
with Alzheimer's disease. Neurology 46: 1580-1584. effects of combined thyroid hormone treatment and electroconvul-
Schneider, M. A., Brotherton, P. L., and Hailes, J. (1977). The effect of sive therapy: Preliminary findings. Biological Psychiatry 30: 623-
exogenous oestrogens on depression in menopausal women. Med- 627.
ical Journal of Australia 2: 162-163. Stern, R. A., Robinson, B., Thorner, A. R., Arruda, J. E., Prohaska,
Schon, M., Sutherland, A., and Rawson, R. (1961). Hormones and M. L., and Pranges, A. J., Jr. (1996). A survey of neuropsychiatric
neuroses—the psychological effects of thyroid deficiency. In Pro- complaints in patients with Graves' disease. Journal of'Neuropsy-
ceedings of the Third World Congress of Psychiatry, McGill Uni- chiatry and Clinical Neurosciences 8: 181-185.
versity Press and University of Toronto Press, Montreal, Canada, Stone, C. P., Girdner, J., and Albrecht, R. A. (1946). An alternate form of
pp. 835-839. the Wechsler Memory Scale. Journal of Psychology 22: 199-206.
Sherwin, B. B. (1988). Estrogen and/or androgen replacement therapy Strachan, M., Dear, I., Ewing, M., and Frier, B. (1997). Is Type II diabetes
and cognitive functioning in surgically menopausal women. Psy- associated with an increased risk of cognitive dysfunction? Diabetes
choneuroendocrinology 13: 345-357. Care 20: 438-445.
206 Erlanger, Kutner, and Jacobs

Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Waring, S. C., Rocca, W. A., Petersen, R. C., O'Brien, P. C., Tangalos,
Journal of Experimental Psychology 18: 643-662. E. G., and Kokmen, E. (1999). Postmenopausal estrogen replace-
Swaab, D. E, and Fliers, E. (1985). A sexually dimorphic nucleus in the ment therapy and risk of AD. Neurology 52: 965-970.
human brain. Science 28: 1112-1115. Warren, S. G., Humphreys, A. G., Jraska, J. M., and Greenough, W. T.
Szalados, J. E., and Vukmir, R. B. (1994). Acute adrenal insufficiency (1995). LTP varies across the estrous cycle: Enhanced synaptic
resulting from adrenal hemorrhage as indicated by post-operative plasticity in proestrus rats. Brain Research 703: 26-30.
hypotension. Intensive Care Medicine 20: 216-218. Webster, J. B., and Bell, K. R. (1997). Primary adrenal insufficiency
Szekely, B. (1966). Tolouse-Pieron Concentration Attention Test. In Los following traumatic brain injury: A case report and review of the
Test, Kapelutz, Buenos Aires, Argentina, pp. 703-709. literature. Archives of Physical Medicine and Rehabilitation 78:
Tang, M. X., Jacobs, D., Stern, Y, et al. (1996). Effect of oestrogen 314-318.
during menopause on risk and age at onset of Alzheimer's disease. Wechsler, D. (1945). A standardized memory scale for clinical use. Jour-
Lancet 348: 429-432. nal of Psychology 19: 87-95.
Taylor, E. M. (1959). Psychological Assessment of Children with Cere- Wechsler, D. (1974). Wechsler Memory Scale Manual, The Psychologi-
bral Deficits, Harvard University Press, Cambridge, MA. cal Corporation, San Antonio, TX.
Thoman, M. L., and Weigle, W. O. (1989). The cellular and subcellular Wechsler, D. (1981). WA1S-R Manual, The Psychological Corporation,
bases of immunosenescence. Advances in Immunology 46: 221- New York.
261. Wechsler, D. (1987). Wechsler Memory Scale—Revised Manual, The
Thome, D. R., Genser, S. G., Sing, H. C,, and Hegge, F. W. (1985). The Psychological Corporation, San Antonio, TX.
Walter Reed Performance Assessment Battery. Neurobehavioral Weingartner, H., Kaye, W., Gold, P., Smallberg, S., Peterson, R,, Gillin,
Toxicology and Teratology 7:415-418. J. C., and Ebert, E. (1981). Vasopressin treatment of cognitive
Thurstone, L. L., and Thurstone, T. G. (1963). Primary Mental Abilities, dysfunction in progressive dementia. Life Sciences 29: 2721-
Science Research Associates, Chicago. 2726.
Toran-Allerand, C. D., Miranda, R. S., Bentham, W. D., et al. (1992). Weissman, M. M. (1996). Epidemiology of major depression in women.
Estrogen receptors colocalize with low-affinity nerve growth fac- Paper presented at the American Psychiatric Association Annual
tor receptors in cholinergic neurons of the basal forebrain. Pro- Meeting, May, New York City, New York.
ceedings of the National Academy of Sciences, USA 89: 4668- Whelan, T. B., Schteingart, M. N., Starkman, M. N., and Smith, A.
4672. (1980). Neuropsychological deficits in Cushing's syndrome. Jour-
Trzepacz, P. T., McCue, M., Klein, I., Greenhouse, J., and Levey, G. S. nal of Nervous and Mental Disease 168: 753-757.
(1988). Psychiatric and neuropsychological response to propranolol Whybrow, P. C., Prange, A. J., Jr., and Treadway, C. R. (1969). Mental
in Graves' disease. Biological Psychiatry 23: 678-688. changes accompanying thyroid gland dysfunction. A reappraisal
Trzepacz, P. T., McCue, M., Klein, I., Levey, G. S., and Greenhouse, using objective psychological measurement. Archives of General
J. (1988). A psychiatric and neuropsychological study of patients Psychiatry 20: 48-63.
with untreated Graves' disease. General Hospital Psychiatry 10: Williams, C. L., Barnett, A. M., and Meck, W. H. (1990). Orga-
49-55. nizational effects of early gonadal secretions on sexual differ-
Tun, P. A., Perlmuter, L. C.,Russo, P.,and Nathan, D. M. (1987). Memory entiation in spatial memory. Behavioral Neuroscience 104: 84-
self-assessment and performance in aged diabetics. Experimental 97.
Aging Research 13: 151-157. Wilson, B. A., Cockburn, J., and Baddeley, A. (1985). The River-
Uhlmann, R. W. (1962). Test of Color Recognition, Form De X-27-61, mead Behavioral Memory Test, Thames Valley Test Co., Reading,
Detroit Edison, Detroit, MI. England; National Rehabilitation Services, Gaylord, MI.
U'Ren, R. C., Riddle, M. C., Lezak, M. D., and Bennigton-Davis, M. Winograd, D. L., Sensheima, P., Barrett-Connor, E. L., and McPhillips,
(1990). The mental efficiency of the elderly person with Type II T. B. (1990). Community-based study on the prevalence of NIDDM
diabetes mellitus. Journal of the American Geriatric Society 38: in older adults. Diabetes Care 13 (Suppl. 2): 3-8.
505-510. Wolf, O. T., Koster, B., Bierings, H. G., et al. (1997). A single administra-
Vandenberg, S. G., and Kuse, A. R. (1978). Mental Rotations: A group tion of dehydroepiandrosterone (DHEA) does not enhance memory
test of three-dimensional spatial visualization. Perceptual and Mo- performance in young healthy adults, but immediately reduces cor-
tor Skills 47: 599-601. tisol levels. Biological Psychiatry 2: 845-848.
Van Goozen, S. H. (1994). Male and Female: Effects of Sex Hormones Wolf, 0. T., Neumann, O., Hellhammer, D. H., et al. (1997). Effects
on Aggression, Cognition, and Sexual Motivation, University of of a two-week physiological dehydroepiandrosterone substitution
Amsterdam Press, Amsterdam. on cognitive performance and well-being in healthy elderly women
Van Goozen, S. H., Cohen-Kettenis, P. T., Gooren, L. J., Frijda, N. H., and men. Journal of Clinical Endocrinology and Metabolism 82:
and Van de Poll, N. E. (1995). Gender differences in behavior: Ac- 2363-2367.
tivating effects of cross-sex hormones. Psychoneuroendocrinology Wolkowitz, O. M. (1994). Prospective controlled studies of the be-
20: 343-363. havioral and biological effects of exogenous corticosteroids. Psy-
Vanhulle, R., and Demol, R. (1976). A double-blind study on the in- choneuroendocrinology 19: 233-255.
fluence of estriol on a number of psychological tests in post- Wolkowitz, 0. M., Reus, V. I., Roberts, E., et al. (1994). Antidepressant
menopausal women. In van Keep, P. A., Greenblatt, R. B., and and cognition-enhancing effects of DHEA in major depression. An-
Albeaux-Fernet, M. (eds.), Consensus on Menopausal Research, nals of the New York Academy of Sciences 774: 337-339.
MTP Press, London, pp. 94-99. Wolkowitz, O. M., Reus, V. I., Roberts, E., et al. (1997). Dehy-
Van Wimersma Greidanus, T. B., Burbach, J. P., and Veldhuis, H. D. droepiandrosterone (DHEA) treatment of depression. Biological
(1986). Vasopressin and oxytocin: Their presence in the central Psychiatry 41:311-318.
nervous system and their functional significance in brain processes Wolkowitz, O. M., Reus, V. I., Weingartner, H., et al. (1990). Cogni-
related to behavior and memory. Acta Endocrinologica 276: 85-94. tive effects of corticosteroids. American Journal of Psychiatry 147:
Vitti, P., Aghini-Lombardi, F., Antonangeli, L., et al. (1992). Mild io- 1297-1303.
dine deficiency in fetal/neonatal life and neuropsychological per- Wolters, E. C., Riekkinen, P., Lowenthal, A., Van der Plaats, J. J.,
formances. Acta Medica Austriaca 19 (Suppl. 1): 57-59. Zwart, J. M., and Sennef, C. (1990). DGAVP (Org 5667)
Wallace, J. E., MacCrimmon, E. J., and Goldberg, W. M. (1980). Acute in early Alzheimer's disease patients: An international double-
hyperthyroidism: Cognitive and emotional correlates. Journal of blind, placebo-controlled, multicenter trial. Neurology 40: 1099-
Abnormal Psychology 4: 519-527. 1101.
Hormones and Cognition 207

Wooley, C. S., and McEwen, B. S. (1993). Roles of estradiol and proges- Yaffe, K., Sawaya, G., Lieberburg, I., and Grady, D. (1998). Estrogen
terone in regulation of hippocampal dendritic spine density during therapy in postmenopausal women. Journal of the American Med-
the estrous cycle in the rat. Journal of Comparative Neurology 336: ical Association 279: 688-695.
293-306. Yashino, A., Hovda, D., Kawamato, T., Katayama, Y., and Becker,
Wynn, V. T., and Arendt, J. (1988). Effect of melatonin on the human D. (1991). Dynamic changes in local cerebral glucose utiliza-
electrocardiogram and simple reaction time responses. Journal of tion following cerebral concussion in rats: Evidence of a hyper-
Pineal Research 5: 427-435. and subsequent hypometabolic state. Brain Research 561: 106-
Yaffe, K., Ettinger, B., Pressman, A., et al. (1998). Neuropsychiatric 119.
function and dehydroepiandrosterone sulfate in elderly women: A Young, R. J., Ewing, D. J., and Clarke, B. F. (1983). Nerve function and
prospective study. Biological Society 43: 694-700. metabolic control in teenage diabetics. Diabetes 32: 142-147.