Content uploaded by Scott J Cameron
Author content
All content in this area was uploaded by Scott J Cameron on Aug 27, 2018
Content may be subject to copyright.
ENDOCRINE EPIDEMIOLOGY
Pre-analytic considerations for the proper assessment of hormones
of the hypothalamic-pituitary axis in epidemiological research
Rachel L. Derr
1
, Scott J. Cameron
2
& Sherita Hill Golden
1,3
1
Department of Medicine, Johns Hopkins University, Baltimore, MD, USA;
2
Department of Pathology, Johns Hopkins
University, Baltimore, MD, USA;
3
Department of Epidemiology, Johns Hopkins University, Baltimore, MD, USA
Accepted in revised form 1 February 2006
Abstract. There is growing epidemiological interest in
hormones as predictors of chronic diseases. The
correct handling and analysis of hormones can be
cumbersome, and great care must be taken in these
processes in order to gain the most information
possible. Given differences in sampling, processing,
and stability of the various hormonal assays, we
sought to provide a comprehensive review to aid fu-
ture epidemiological research. We have coupled a
thorough literature search with our own analytical
experience to outline common laboratory problems
one must consider in analyzing the hormones of the
hypothalamic-pituitary axis. In addition, we describe
the benefits and limitations of using alternative media
– including urine, saliva, and blood spots on filter
paper – to measure endocrine hormones in epidemi-
ological studies.
Key words: Biological assay, Storage, Epidemiology, Hormones, Immunoassay, Saliva
Abbreviations: ACTH = adrenocorticotropin hormone; GH = growth hormone; ICMA = immunochemilu-
minometric assay; IGF-1 = insulin-like growth factor; IMA = immunoassay; IRMA = immunoradiometric
assay; NA = no data available; RIA = radioimmunoassay; TSH = thyroid stimulating hormone; T4 =
thyroxine; T3 = triiodothyronine
Introduction
There has been growing interest in endocrine hor-
mones as risk factors for a variety of chronic diseases,
including cardiovascular disease, diabetes mellitus,
and cancer. Many large-scale epidemiological studies
with stored serum, plasma, and urine samples are
well-suited for nested case–control studies in this
area. In order for hormonal exposures to be accu-
rately determined, specimen collection, storage, and
preservation must be appropriate at the study
inception. Inappropriate sample storage and pro-
cessing can result in a lost research opportunity.
The pituitary gland has been described as the
body’s ‘‘master gland,’’ in that the anterior portion of
the gland secretes hormones that stimulate various
end organs – the gonads, adrenal glands, thyroid, and
liver – to secrete hormones that control reproduction,
growth, and metabolism (see Figure 1). The release of
these pituitary hormones is regulated by positive
feedback from hypothalamic hormones and negative
feedback from the end-organ hormones, such that a
state of hormonal equilibrium is maintained. The
hypothalamic-pituitary–gonadal axis has been the
most extensively studied axis; many population-based
studies have examined the association between
endogenous sex hormones and cardiovascular disease
risk factors and outcome [1–9]. In these large cohort
studies, as well as smaller studies [10–15], sex hor-
mones measurements were made on previously
banked frozen serum samples, allowing the exami-
nation of new hypotheses regarding the relationship
between sex hormones and cardiovascular disease
that may not have been a major specific aim of these
studies at their inception.
This example raises the possibility of more com-
prehensively examining the other pituitary-end-organ
axes, including the thyroid, growth hormone, and
adrenal axes, in relation to chronic disease outcomes,
which may result in the identification of novel,
potentially modifiable risk factors. The ability to use
existing datasets and banks of serum and urine to
explore new hypotheses, however, is dependent upon
the proper processing and storage of samples at the
time of collection. Also, if future hormonal mea-
surements are to be considered in cohort studies
currently being designed, then proper collection and
storage considerations are important at the study
inception. The purpose of this review is to summarize
the collection, storage, and stability issues related to
the measurement of hormones of the other pituitary
end-organ axes that might affect their ability to be
measured in epidemiological studies. We also exam-
ine alternative collection methods (i.e. blood spot,
European Journal of Epidemiology (2006) 21: 217–226 ÓSpringer 2006
DOI 10.1007/s10654-006-0011-0
urine, and saliva samples) that may be more amena-
ble to use in large epidemiological field studies. We
will specifically examine measurement issues for the
hormones of the thyroid, adrenal, and growth hor-
mone axes. Because gonadal steroids have been
extensively measured in many large cohort studies
and issues of stability have been examined previously
[16], these will not be reviewed. We will not explore
measurement and storage issues related to prolactin
because it has primarily local effects on breast tissue
and not generalized systemic effects like the other
anterior pituitary hormones.
Assay considerations
The first step in planning for hormone measurement
is deciding the assay that will be used, as many of the
collection considerations depend on the assay choice.
Currently, most endocrine hormones are measured
by immunometric assays. These may be competitive,
where exogenously applied reagents are limited, or
non-competitive (two-site or sandwich immunoas-
says) where exogenously applied reagents are in
excess. One type of immunometric assay, the radio-
immunoassay (RIA), involves a radioisotope-labeled
compound and an antibody directed against it that
can be inhibited by an unlabeled compound [17, 18].
RIA is commonly used to measure cortisol. The im-
munoradiometric assay (IRMA), which uses excess
labeled antibody, achieves greater sensitivity and
specificity for assaying GH, IGF-1, and ACTH
[19–21]. In place of the radioactive isotope, chemilu-
minescence can be used to label antigens and antibodies.
This technique, the immunochemiluminometric assay
(ICMA), has improved the sensitivity capabilities of
the TSH assay, allowing for reliable diagnosis of
hyperthyroidism [22, 23].
Protein binding
The precision of an immunoassay and the accuracy
of measurements of active hormone levels may be
disrupted by hormone binding proteins; as a result,
assay methodologies have been developed to abro-
gate the influence of binding protein interference
[24]. The vast majority of the circulating T4, T3,
GH, IGF-1, and cortisol in plasma exist in a com-
plex with specific proteins, while TSH and ACTH
are not bound to proteins in blood. Circumstances
such as pregnancy, estrogen therapy, oral contra-
ceptive therapy, and obesity increase binding protein
levels, and others states such as nephrotic syndrome,
starvation, and chronic liver disease decrease levels
[25, 26]. Since total hormone levels may not accu-
rately reflect unbound, or active, hormone levels in
these settings, free hormone measurements should
be considered.
Measurement of the free hormone level by immu-
noassay is now commonplace for T4 in the clinical
Figure 1. Summary of hypothalamic-pituitary-end organ axes.
218
laboratory. Although using equilibrium dialysis to
partition bound from free thyroid hormone remains
the somewhat cumbersome gold standard [27], the
free T4 immunoassay is adequate and more feasible
for most epidemiological studies of large populations.
Free levels can also be measured for T3 and cortisol,
but since the process is expensive and time-consum-
ing, it is rarely performed. The most accurate method
for extracting IGF Binding Proteins (IGFBPs) is acid
chromatography; however, because this procedure
is very labor- and time-intensive, acid extraction
methods are frequently used [24].
Collection considerations
Sample tube choice
The assay chosen may dictate whether samples are
collected as plasma, the liquid separated from blood
cells by centrifugation, or serum, the fluid that re-
mains after fibrinogen, prothrombin, and other clot-
ting factors have been removed from plasma. Some
laboratories have optimized particular hormone
assays using only serum while others use only plasma;
hence, plasma and serum should not be used inter-
changeably. Serum is the preferred matrix for anal-
ysis of thyroid hormones, growth hormone, IGF-I,
and cortisol. In contrast, ACTH, which is distinct
because of its very short half-life, is usually measured
in plasma [28].
Proper sample tube selection must also consider the
potential interference from additives such as heparin
and ethylene diamine tetra-acetic acid (EDTA), which
are present in common collection tubes. EDTA–
plasma sample tubes, which prevent coagulation
allowing plasma to be more effectively separated, have
a chelating effect on Europium
3+
(Eu
3+
). Since Eu
3+
is a component in the DELFIAäassay, which has
been used with success to measure samples containing
TSH [29], GH [30], and cortisol [31], the presence of
EDTA can be a major source of interference.
Quantity requirements
From an analytical perspective, the quantity of ser-
um or plasma required for hormone assays is vari-
able and largely depends on the type of assay and
instrumentation being used for the analysis. Mini-
mum aliquot requirements for assay purposes also
differ according to the hormone of interest and
reported minimum assay volumes necessary range
from 0.1 to 0.3 ml; however, this may vary accord-
ing to laboratory (see Table 1) [32]. Even though
most hormone assays do not use volumes greater
than these for the actual assays, a larger volume
may be required to be physically taken up by the
analyzer (‘‘pick-up volume’’); however, the unused
serum (‘‘dead space’’) is returned at the end of the
analysis. Thus, as a general guideline, if a minimum
of 0.8 ml of blood is aliquoted in an epidemiological
study; this will allow one to assay almost all
hormonal analytes. In long-term epidemiological
cohort studies where there is not an opportunity to
collect another sample under the same conditions,
samples should be stored in aliquots so that they are
available as duplicate or back-up samples for hor-
monal measurements.
Diurnal variation
While TSH, triiodothyronine (T3), thyroxine (T4),
and IGF-1 secretion vary minimally throughout the
day permitting random sampling, the secretion of
growth hormone, ACTH, and cortisol exhibits sig-
nificant diurnal variation, making the timing of their
collection critical to the interpretation of the results
(see Table 1). GH levels fluctuate greatly over the
day, ranging from undetectable at certain times of the
day to peaking in the setting of hypoglycemia and
Stage III and IV sleep. Single, random measurements
are therefore difficult to interpret. The diurnal vari-
ation in ACTH and cortisol secretion is somewhat
more predictable, characterized by peaks in the early
morning hours (8 AM) and nadirs in the late evening.
Random measurements of ACTH should be corre-
lated with cortisol measurements from the same
sample [33]. Although IGF-1 is not known to vary
according to a daily pattern, a recent study showed
that IGF-1 values measured two weeks apart in the
same individual varied to an important extent, sug-
gesting the need for more than single sampling when
studying the relationship between IGF-1 and disease
endpoints [34].
Extrinsic factors at the time of collection
In addition to diurnal variation, extrinsic factors such
as smoking, stress, and medications can alter pitui-
tary hormone levels, and thus should be considered in
the decision of when to collect samples. Cigarette
smoking has an acute dose-related effect of increasing
ACTH and cortisol levels, while TSH and GH are not
changed [35, 36]. Withdrawal from cigarettes may
also increase cortisol and ACTH levels. Physical and
psychological stress, in general, increase cortisol,
ACTH, and growth hormone levels, in proportion to
the level of stress [37–39]. It is therefore recom-
mended that blood samples for measurement of these
hormones be collected when the patient is relaxed and
sitting quietly at rest. In one study, TSH increased in
response to the psychological stress of a parachute
jump [39], but in cases of physical stress such as
critical illness, levels of TSH, as well as T4 and T3,
are commonly decreased [40].
In addition to estrogen, which alters hormone levels
through an effect on binding protein levels, other
drugs can affect pituitary hormone measurements,
and this interaction should be recognized for proper
interpretation of results. Thyroid status can be chan-
ged by many diverse medications, through mecha-
nisms including decreasing TSH and altering T3
and T4 secretion, and affecting the transport and
219
Table 1. Hormonal collection, measurement, and storage considerations for thyroid, growth hormone, and cortisol axes
Assay commonly used Binding
proteins present
Blood component
commonly measured
Minimum volume
required (ml)
a
Diurnal variation
important
Proven stability at
room temp
Suggested stability frozen
TSH 3rd generation ICMA No Serum 0.2 No 8 days 6–8 years
T4 IMA Yes Serum 0.3 No 2 weeks 6–8 years
T3 IMA Yes Serum 0.3 No 2 weeks NA
GH 2-site IRMA Yes Serum 0.3 Yes 3 days NA
IGF-1 2-site IRMA Yes Serum 0.2 No 5 days 8 years
ACTH 2-site IRMA No Plasma 0.3 Yes Several hours if succinylated NA
Cortisol RIA Yes Plasma or serum 0.1 Yes 2 weeks 10 years
ACTH=adrenocorticotropin hormone; GH=growth hormone; ICMA=immunochemiluminometric assay; IGF-1=insulin-like growth factor; IRMA=immunoradiometric assay;
IMA=immunoassay; NA=No data available; RIA=radioimmunoassay; TSH=thyroid stimulating hormone; T4=thyroxine; T3=triiodothyronine.
a
These are averages based on literature review and may vary by assay type and laboratory. In general, if a minimum of 0.8 ml of blood is aliquoted in an epidemiological study, this will
allow one to assay almost all hormonal analytes.
Table 2. Storage and stability considerations for filter paper, urine, and saliva assays
Blood spot filter paper Urine Saliva
Hormones Measurable RT Stability Measurable Stability Measurable Stability
TSH Yes, used widely for neonatal screening 30 days No, because of low levels – No, because of low levels –
T4 Yes 2 weeks No, because of low levels – Rarely, since unreliable NA
T3 Yes 30 days No, because of low levels – No, because of low levels –
GH No, because of diurnal variation – Yes, but limited by variability 4 °C: 2 days,
Frozen: 2–12 months
Rarely, because of low levels NA
IGF-1 Yes 30 days Yes, but complicated by
binding proteins
NA Rarely, because inaccurate NA
ACTH No, because of low levels – NA – NA –
Cortisol Yes 30 days Yes 4 °C: 1–2 days if pH 3–7;
Frozen: months
Yes RT: 4–6 weeks
ACTH=adrenocorticotropin hormone; GH=growth hormone; IGF-1=insulin-like growth factor; NA=No data available; RT=room temperature; TSH=thyroid stimulating
hormone; T4=thyroxine; T3=triiodothyronine.
220
metabolism of these hormones. Amiodarone, lithium,
iodine, glucocorticoids, and anticonvulsants are
among the most common medications to affect thyroid
hormone levels [41]. Glucocorticoids also alter corti-
sol, ACTH, growth hormone, and IGF-1 levels [42].
Sample transport and processing
If fresh samples are transported to the laboratory
without delay and processing is timely, samples can
generally remain at room temperature. An exception
is for samples measuring ACTH, which is highly
unstable. These samples should be drawn into a cold
syringe, placed immediately into an EDTA tube, and
taken immediately on ice to the laboratory, where the
plasma attained from centrifugation should be frozen
directly [33].
If laboratory analysis of samples is to be delayed
for days or weeks, in addition to considering storage
temperature, which will be discussed in the next
section, processing steps, such as the addition of
protease inhibitors, may be considered. Variability in
IGF-1 measurements due to post-sampling proteol-
ysis was avoided by adding protease inhibitors to
samples stored at 4 °C for 4 weeks in one study [43].
Storage and stability
Since certain hormones have stabilities and half-lives
which can change with temperature, correlation
studies are often conducted to compare levels mea-
sured immediately to those measured after certain
time periods of storage at various temperatures. The
initial characterization of most hormone assays also
usually includes studying the reproducibility of
measurements when assay aliquots have been sub-
jected to sequential freeze-thaw cycles (at least four).
TSH measured by radioimmunoassay in plasma has
been found to be stable at temperatures as high as
37 °C for up to 8 days as well as stable through five
repeat freeze-thaw cycles [44]. T4 and T3 levels, as
well as cortisol levels, are stable in serum samples
stored at room temperature (22 °C) and below for up
to 2 weeks [45]. GH, alternatively, is less stable at
room temperature, with stability for only 3 days, but
refrigerated GH (4 °C) is stable for at least 8 days
[46]. GH, like TSH, was shown to be stable through
five cycles of freezing and thawing. IGF-1 measure-
ments are subject to significant inter-assay variability
secondary to post-sampling proteolysis, when stored
at 4 °C for 4 weeks, but this effect can be blocked by
processing the sample with protease inhibitors [43].
The least stable hormone is ACTH, which is stable in
plasma for only 18 h when refrigerated [46], but
certain modifications may help. One study showed
that succinylating ACTH can allow it to be stable for
several hours at room temperature and for 1 week at
4°C [47].
If stored frozen and thawed only once, many hor-
mones have been considered stable for years, allow-
ing epidemiologists to determine the association
between hormone levels and clinical outcomes from
samples collected years earlier. Stored at )20 °C,
thyroid hormones are stable for many years [48], and
studies have measured TSH and T4 from frozen
samples collected from cohorts 6 to 8 years earlier
[49, 50]. Numerous recent nested case–control studies
have found significant associations between IGF-1
levels and development of various cancers and
ischemic heart disease [51–58]. However, these studies
were not designed specifically to assess hormonal
stability, rather stability was assumed. A recent study
demonstrated that IGF-1 frozen at )80 °C for
2 years was stable, but measurements from samples
frozen at )20 °C for 8 years were unreliable [43].
Cortisol measured from frozen samples has not
been used in epidemiologic studies to date, but if
samples were carefully collected in terms of time of
day, such studies could be possible since cortisol re-
mains stable at )25 °C for more than 10 years in
storage [59]. Table 1 summarizes the available data
on storage and stability of each hormone.
Alternative methods of measuring endocrine hormones
Blood samples are not always available, for example,
if the stored supply has been already exhausted by
prior studies. Therefore, alternative sources from
which to measure hormones may be useful. Table 2
summarizes the characteristics of several alternative
assays for measurement of thyroid, growth, and
cortisol hormones.
Blood spot samples
The blood spot filter paper technique involves placing
a drop of the subject’s blood on a small disc of filter
paper and later extracting and measuring the hor-
mone concentration. Because only a minimal amount
of blood is needed, which can be obtained from a
finger or heel prick, this technique is less invasive and
less time-consuming than venipuncture. Since blood
spots on filter paper are small and quite stable at
room air, this method circumvents many problems
involved with liquid samples like cost, storage, and
transporting to distant laboratories [60]. TSH, T3
(but not T4), IGF-1, and cortisol are all stable on
filter paper for 1 month or more at room temperature,
making them useful in field studies where refrigerated
storage is limited [61–66]. In fact, measuring TSH on
filter paper has allowed for widespread screening for
hypothyroidism in neonates. Filter paper assays of
GH and ACTH are less likely to be clinically useful
and have not been successfully developed.
Urine
Potential advantages of urine collection include the
ability to measure hormones in their free form
221
without the interference of binding proteins, the
ability to measure integrated hormone secretion in
hormones affected by diurnal variation, and the
ability to perform collection as outpatients. However,
potential limitations include incomplete compliance
with overnight or 24-h collections, poor stability of
some hormones in urine, insensitivity of certain as-
says to low levels of hormones in urine, and inter-
ference by comorbidities such as renal disease. For
example, thyroid hormone levels are extremely low in
urine (TSH being detectable in only 50% of normal
patients’ urine); as a result, they do not reliably cor-
relate with serum levels and are heavily affected by
proteinuria [67]. Moreover, the tendency of T4 in
urine to rapidly hydrolyze makes urinary measure-
ments more cumbersome than serum measurements
[68, 69]. However, highly-sensitive techniques, such as
gas chromatography-mass spectrometry (GC-MS),
allow for accurate detection of hormones including
the androgens [70] and cortisol and its metabolites
[71]. This method, however, necessitates thorough
purification of urine to avoid assay interference,
which is very labor-intensive.
Accurate measurement of urinary GH is possible
with sensitive immunometric assays, and urinary GH
is stable for 2 days at room temperature and for
2–12 months when frozen between )20 °Cand
)70 °C [72, 73]. However, the clinical utility of uri-
nary GH measurement is small. While the correlation
between serum GH and urinary GH is strong in
hypopituitarism and acromegaly [74], it is weaker in
more subtle clinical scenarios, such as short stature
and partial GH deficiency, where it would be of
particular use [75–77]. Because only minute quantities
of GH are excreted in the urine, interindividual var-
iation and day-to-day variability in GH excretion
impact its measurement, and limit the diagnostic
potential of this approach. For similar reasons, it
would likely have limited utility in assessing GH
status in population-based studies.
As an alternative to long collections of urine for
GH assays, spot urine IGF-1 is easier to collect and
correlates well with growth hormone status and ser-
um IGF-1 levels [78, 79], except in patients with
chronic renal failure, type 2 diabetes [80], or who are
over 60 years of age [81]. Unlike most of the other
hormones, IGF-1 found in the urine is not predom-
inantly in the free form; only 30% exists unbound to
binding proteins [78]. In order to assess urinary IGFs
adequately, assays must separate it from IGFBPs
[81], a difficulty that limits its utility.
Cortisol is useful to measure in the urine as a 24-h
collection or as a timed spot sample. The 24-h col-
lections of urinary cortisol have long been accepted as
a reliable measure of adrenocortical secretion [82].
The stability of cortisol in urine is suitable for out-
patient sampling since it is usually stable if refriger-
ated for 1–2 days before the assay, though freezing
immediately after collection is advised [83].
Saliva
Like urine, collection of saliva for assessment of
hormone levels provides a convenient method of
measuring the unbound fraction of a hormone, which
can be collected multiple times at the subject’s own
home.Cortisol is an example of a hormone that is
currently measured in saliva. Saliva is collected using
a device called a Salivette
Ó
(Sarstedt, Newton, NC),
which contains a cotton swab that is chewed for
2–3 min. Cortisol in saliva, which is about 5% of the
total concentration of plasma cortisol, is in equilib-
rium with plasma cortisol. Because of the lack of
albumin and CBG in saliva, an increase in plasma
cortisol is reflected by a change in salivary cortisol
concentration within a few minutes that is more
pronounced than in the plasma, making salivary
cortisol measures especially useful in dynamic tests of
the hypothalamic pituitary adrenal (HPA) axis (e.g.
insulin tolerance test, corticotropin releasing hor-
mone stimulation test) [84]. Salivary cortisol is
unaffected by salivary flow rate and displays the same
diurnal variation as plasma cortisol. Salivary mea-
sures of cortisol correlate well with plasma cortisol
levels for normal people of all ages and in states of
abnormal adrenal function [85–87]. Applications of
the salivary cortisol assay have also extended to the
field of psychoneuroendocrinology, being used to test
subjects with psychiatric disorders and in stressed
states [88]. Cortisol is fairly stable in saliva for up to
6 weeks at room temperature when preserved in citric
acid [89], and for extended periods of time when
frozen [90].
Attempts at measuring salivary T3, T4, IGF-1, and
GH have been complicated by their very low con-
centrations in saliva and have yielded conflicting re-
sults. While one report of T4 in saliva found no
correlation with serum levels [91], another study
showed a significant correlation between salivary and
serum T4 [92]. Salivary growth hormone was shown
in one study to generally correlate with serum levels,
but since its concentrations are 1000 fold lower than
serum values, testing saliva for GH is difficult [93].
Salivary IGF-1 has been found to generally reflect
growth hormone levels, being low in GH-deficient
subjects and high in acromegalics when compared to
controls [94]; however, its sensitivity for diagnosing
acromegaly is very poor, with nearly half of acro-
megalic patients in one study having salivary IGF-1
concentrations within the normal range [95]. For
these reasons, none of these hormones are routinely
measured in the saliva today.
Discussion
Levels of certain endocrine hormones can be mea-
sured from stored samples of plasma or serum,
thereby allowing the opportunity to easily and rap-
idly conduct epidemiologic studies that examine the
222
association between hormone levels and clinical out-
comes of interest. When choosing an assay for hor-
mone measurement, one must consider the sample
matrix and quantity of the sample available, as well
as the potential for interference by binding proteins if
they are present. Precise collection times are neces-
sary for diurnally varying hormones like GH, ACTH,
and cortisol.
Stability of hormones is influenced by the length of
time since sample collection, the temperature at which
the samples have been stored, and the number of
freeze-thaw cycles they have undergone. However,
long-term stability data are significantly lacking.
There is some evidence that IGF-1, TSH, and T4 may
be stable for many years if stored frozen, based on the
fact that several studies have found statistical signifi-
cance when correlating these hormone levels with
clinical outcomes years later. However, the only
studies with a primary aim of comparing the stability
of a hypothalamic-pituitary axis hormone before and
after storage were done by Kley et al. for cortisol al-
most two decades ago [59] and by Khosravi et al. for
IGF-1 recently [43]. Further studies to validate the
frozen stability of the other hormones, using the most
common assays available today and experimenting
with different storage temperatures and numbers of
freeze-thaw cycles, are important so that currently
available stored samples throughout the world can be
confidently used in epidemiological studies to examine
new hypotheses. In addition, most data regarding
stored thyroid hormone includes data on total T4;
however, with the current widespread use of free T4,
which eliminates the confounding effects of thyroid
binding globulin, future studies should focus on the
stability of free T4. To date, only the hypothalamic-
pituitary–gonadal and growth hormone axes have
been appreciably studied with respect to their impact
on chronic diseases; however, there is likely much to
be learned by investigating the thyroid and adrenal
axes in relation to chronic diseases as well.
Methods of measuring levels of some hormones
through alternative means have been successful in
circumventing some of the problems inherent with
blood collections. For example, the dilemma of
diurnal variation and binding proteins can be avoided
with urine and saliva sampling. Urine and saliva as-
says are now popular for cortisol. Measurement of
TSH for the purpose of diagnosing hypothyroidism
in field studies has been greatly enhanced by the use
of blood spots on filter paper, which decreases cost
and the requirement for extensive storage space.
These methods can provide alternatives to traditional
blood drawing techniques that are more burdensome
to the participants, and importantly, they may also
allow for the preservation of serum banks for mea-
surements of analytes that can only be accomplished
with blood samples.
Pituitary hormones are key regulators of metabo-
lism, growth, and reproduction. With proper atten-
tion to collection and storage considerations,
hormones of the hypothalamic-pituitary axis can be
the targets of case–control and prospective cohort
studies designed to determine the hormonal basis of
major chronic diseases.
Acknowledgements
Dr. Golden was supported by a grant through the
Robert Wood Johnson Minority Medical Faculty
Development Program Award, Princeton, NJ and by
a minority supplement grant through the National
Institute of Diabetes, Digestive, and Kidney Diseases,
Bethesda, MD (3 U01 DK48485–10S1).
References
1. Barrett-Connor E, Goodman-Gruen D. Prospective
study of endogenous sex hormones and fatal cardio-
vascular disease in postmenopausal women. Br Med J
1995; 311: 1193–1196.
2. Goodman-Gruen D, Barrett-Connor E. Sex differences
in the association of endogenous sex hormone levels
and glucose tolerance status in older men and women.
Diabetes Care 2000; 23: 912–918.
3. Haffner SM, Newcomb PA, Marcus PM, Klein BE,
Klein R. Relation of sex hormones and dehydroepi-
androsterone sulfate (DHEA-SO4) to cardiovascular
risk factors in postmenopausal women. Am J Epi-
demiol 1995; 142: 925–934.
4. Gapstur SM, Gann PH, Kopp P, Colangelo L, Long-
cope C, Liu K. Serum androgen concentrations in
young men: a longitudinal analysis of associations with
age, obesity, and race. The CARDIA male hormone
study. Cancer Epidemiol Biomarkers Prev 2002; 11:
1041–1047.
5. Srinivasan SR, Sundaram GS, Williamson GD, Web-
ber LS, Berenson GS. Serum lipoproteins and endog-
enous sex hormones in early life: observations in
children with different lipoprotein profiles. Metabolism
1985; 34: 861–867.
6. Srinivasan SR, Freedman DS, Sundaram GS, Webber
LS, Berenson GS. Racial (black–white) comparisons of
the relationship of levels of endogenous sex hormones
to serum lipoproteins during male adolescence: the
Bogalusa Heart Study. Circulation 1986; 74: 1226–
1234.
7. Golden SH, Maguire A, Ding J, et al. Endogenous
postmenopausal hormones and carotid atherosclerosis:
a case-control study of the atherosclerosis risk in com-
munities cohort. Am J Epidemiol 2002; 155: 437–445.
8. Haffner SM, Moss SE, Klein BE, Klein R. Sex hor-
mones and DHEA-SO4 in relation to ischemic heart
disease mortality in diabetic subjects. The Wisconsin
Epidemiologic Study of Diabetic Retinopathy. Diabe-
tes Care 1996; 19: 1045–1050.
9. Price JF, Lee AJ, Fowkes FG. Steroid sex hormones
and peripheral arterial disease in the Edinburgh Artery
Study. Steroids 1997; 62: 789–794.
223
10. Bernini GP, Sgro M, Moretti A, et al. Endogenous
androgens and carotid intimal-medial thickness in
women. J Clin Endocrinol Metab 1999; 84: 2008–2012.
11. Kumagai S, Kai Y, Sasaki H. Relationship between
insulin resistance, sex hormones and sex hormone-
binding globulin in the serum lipid and lipoprotein
profiles of Japanese postmenopausal women. J Ath-
eroscler Thromb 2001; 8: 14–20.
12. Mantzoros CS, Georgiadis EI, Young R, et al. Relative
androgenicity, blood pressure levels, and cardiovascu-
lar risk factors in young healthy women. Am J Hy-
pertens 1995; 8: 606–614.
13. Phillips GB, Pinkernell BH, Jing TY. Relationship
between serum sex hormones and coronary artery dis-
ease in postmenopausal women. Arterioscler Thromb
Vasc Biol 1997; 17: 695–701.
14. Seidell JC, Cigolini M, Charzewska J, et al. Androge-
nicity in relation to body fat distribution and metabo-
lism in 38-year-old women – the European Fat
Distribution Study. J Clin Epidemiol 1990; 43: 21–34.
15. Soler JT, Folsom AR, Kaye SA, Prineas RJ. Associa-
tions of abdominal adiposity, fasting insulin, sex hor-
mone binding globulin, and estrone with lipids and
lipoproteins in post-menopausal women. Atheroscle-
rosis 1989; 79: 21–27.
16. Cauley JA, Gutai JP, Kuller LH, Powell JG. Reliability
and interrelations among serum sex hormones in
postmenopausal women. Am J Epidemiol 1991; 133:
50–57.
17. Moore WT, Eastman RC. Diagnostic Endocrinology.
Toronto: B.C. Decker, 1990.
18. Burtis CA, Ashwood ER. Tietz Textbook of Clinical
Chemistry. Philadelphia, PA: W.B. Saunders Co, 1999.
19. Ranke MB, Feldt-Rasmussen U, Bang P, et al. How
should insulin-like growth factor I be measured? A
consensus statement. Horm Res 2001; 55(Suppl 2):
106–109.
20. Ranke MB, Orskov H, Bristow AF, Seth J, Baumann
G. Consensus on how to measure growth hormone in
serum. Horm Res 1999; 51(Suppl 1): 27–29.
21. White A, Smith H, Hoadley M, Dobson SH, Ratcliffe
JG. Clinical evaluation of a two-site immunoradio-
metric assay for adrenocorticotrophin in unextracted
human plasma using monoclonal antibodies. Clin
Endocrinol (Oxf) 1987; 26: 41–51.
22. Spencer CA, Takeuchi M, Kazarosyan M. Current
status and performance goals for serum thyrotropin
(TSH) assays. Clin Chem 1996; 42: 140–145.
23. Demers LM, Spencer CA. Laboratory medicine prac-
tice guidelines: laboratory support for the diagnosis
and monitoring of thyroid disease. Clin Endocrinol
(Oxf) 2003; 58: 138–140.
24. Chestnut RE, Quarmby V. Evaluation of total IGF-I
assay methods using samples from Type I and Type
II diabetic patients. J Immunol Methods 2002; 259:
11–24.
25. Dhillo WS, Kong WM, Le Roux CW, et al. Cortisol-
binding globulin is important in the interpretation of
dynamic tests of the hypothalamic–pituitary–adrenal
axis. Eur J Endocrinol 2002; 146: 231–235.
26. Roelen CA, Koppeschaar HP, de Vries WR, et al.
Visceral adipose tissue is associated with circulating
high affinity growth hormone-binding protein. J Clin
Endocrinol Metab 1997; 82: 760–764.
27. Wong TK, Pekary AE, Hoo GS, Bradley ME, Hersh-
man JM. Comparison of methods for measuring free
thyroxin in nonthyroidal illness. Clin Chem 1992; 38:
720–724.
28. Meakin JW. Catabolism of ACTH: the stability of
ACTH in blood, plasma, serum and saline. Endocri-
nology 1960; 66: 59–62.
29. Wang ST, Pizzolato S, Demshar HP. Diagnostic
effectiveness of TSH screening and of T4 with second-
ary TSH screening for newborn congenital hypothy-
roidism. Clin Chim Acta 1998; 274: 151–158.
30. Laursen T, Moller J, Jorgensen JO, Orskov H, Chris-
tiansen JS. Bioavailability and bioactivity of intrave-
nous vs subcutaneous infusion of growth hormone in
GH-deficient patients. Clin Endocrinol (Oxf) 1996; 45:
333–339.
31. Reynolds RM, Bendall HE, Whorwood CB, Wood PJ,
Walker BR, Phillips DI. Reproducibility of the low
dose dexamethasone suppression test: Comparison be-
tween direct plasma and salivary cortisol assays. Clin
Endocrinol (Oxf) 1998; 49: 307–310.
32. Fisher D, Ladenson P. The Corning Endocrine Man-
uel. Capistrano CA: Corning Nichols Institute, 1996.
33. Wilson J. William Textbook of Endocrinology. Phila-
delphia: W.B. Saunders Co, 1998.
34. Milani D, Carmichael JD, Welkowitz J, Ferris S, Reitz
RE, Danoff A, Kleinberg DL. Variability and reli-
ability of single serum IGF-1 measurements: impact
on determining predictability of risk ratios in disease
development. J Clin Endocrinol Metab 2004; 89:
2271–2274.
35. Mendelson JH, Sholar MB, Goletiani N, Siegel AJ,
Mello NK. Effects of low- and high-nicotine cigarette
smokin on mood states and the HPA axis in men.
Neuropsychopharmacology 2005; 30: 1751–1763.
36. Szostak-Wegierek D, Bjorntorp P, Marin P, Lindstedt
G, Andersson B. Influence of smoking on hormone
secretion in obese and lean female smokers. Obes Res
1996; 4: 321–328.
37. Luger A, Deuster PA, Kyle SB, et al. Acute hypotha-
lamic–pituitary–adrenal responses to the stress of
treadmill exercise. Physiologic adaptations to physical
training. N Engl J Med 1987; 316: 1309–1315.
38. Singh A, Petrides JS, Gold PW, Chrousos GP, Deuster
PA. Differential hypothalamic–pituitary–adrenal axis
reactivity to psychological and physical stress. J Clin
Endocrinol Metab 1999; 84: 1944–839.
39. Richter SD, Schurmeyer TH, Schedlowski M, Hadicke
A, Tewes U, Schmidt RE, Wagner TO. Time kinetics of
the endocrine response to acute psychological stress. J
Clin Endocrinol Metab 1996; 81: 1956–1960.
40. McIver B, Gorman CA. Euthyroid sick syndrome: An
overview. Thyroid 1997; 7: 125–132.
41. Surks MI, Sievert R. Drugs and thyroid function. N
Engl J Med 1995; 333: 1688–1694.
42. Prummel MF, Wiersinga WM, Oosting H, Endert E.
The effect of long-term prednisone treatment on
growth hormone and insulin-like growth factor-1.
J Endocrinol Invest 1996; 19: 620–623.
43. Khosravi J, Diamandi A, Bodani Um, Khaja N,
Krishna RG. Pitfalls of immunoassay and sample for
IGF-1: comparison of different assay methodologies
using various fresh and stored serum samples. Clin
Biochem 2005; 38: 659–666.
224
44. Livesey JH, Hodgkinson SC, Roud HR, Donald RA.
Effect of time, temperature and freezing on the stability
of immunoreactive LH, FSH, TSH, growth hormone,
prolactin and insulin in plasma. Clin Biochem 1980; 13:
151–155.
45. Kubasik NP, Ricotta M, Hunter T, Sine HE. Effect of
duration and temperature of storage on serum analyte
stability: examination of 14 selected radioimmunoassay
procedures. Clin Chem 1982; 28: 164–165.
46. Evans MJ, Livesey JH, Ellis MJ, Yandle TG. Effect of
anticoagulants and storage temperatures on stability of
plasma and serum hormones. Clin Biochem 2001; 34:
107–112.
47. Kertesz G, Bourcier B, Cailla H, Jean F. Immunora-
diometric assay of succinylated corticotropin: an im-
proved method for quantification of ACTH. Clin
Chem 1998; 44: 78–85.
48. Kashiwai T, Ichihara K, Tamaki H, et al. The stability
of immunological and biological activity of human
thyrotropin in buffer: its temperature-dependent dis-
sociation into subunits during freezing. Scand J Clin
Lab Invest 1991; 51: 417–423.
49. Oomen HA, Schipperijn AJ, Drexhage HA. The prev-
alence of affective disorder and in particular of a rapid
cycling of bipolar disorder in patients with abnormal
thyroid function tests. Clin Endocrinol (Oxf) 1996; 45:
215–223.
50. Opravil M, Pechere M, Speich R, et al. HIV-associated
primary pulmonary hypertension. A case control study.
Swiss HIV Cohort Study. Am J Respir Crit Care Med
1997; 155: 990–995.
51. Juul A, Scheike T, Davidsen M, Gyllenborg J,
Jorgensen T. Low serum insulin-like growth factor I is
associated with increased risk of ischemic heart disease:
a population-based case–control study. Circulation
2002; 106: 939–944.
52. Palmqvist R, Hallmans G, Rinaldi S, et al. Plasma
insulin-like growth factor 1, insulin-like growth factor
binding protein 3, and risk of colorectal cancer: a
prospective study in northern Sweden. Gut 2002; 50:
642–646.
53. Ma J, Pollak MN, Giovannucci E, et al. Prospective
study of colorectal cancer risk in men and plasma levels
of insulin-like growth factor (IGF)-I and IGF-binding
protein-3. J Natl Cancer Inst 1999; 91: 620–625.
54. Nomura AM, Stemmermann GN, Lee J, Pollak MN.
Serum insulin-like growth factor I and subsequent risk
of colorectal cancer among Japanese–American men.
Am J Epidemiol 2003; 158: 424–431.
55. Chan JM, Stampfer MJ, Giovannucci E, et al. Plasma
insulin-like growth factor-I and prostate cancer risk: A
prospective study. Science 1998; 279: 563–566.
56. Woodson K, Tangrea JA, Pollak M, et al. Serum
insulin-like growth factor I: tumor marker or etiologic
factor? A prospective study of prostate cancer among
Finnish men. Cancer Res 2003; 63: 3991–3994.
57. Hankinson SE, Willett WC, Colditz GA, et al. Circu-
lating concentrations of insulin-like growth factor-I
and risk of breast cancer. Lancet 1998; 351: 1393–1396.
58. Spitz MR, Barnett MJ, Goodman GE, Thornquist
MD, Wu X, Pollak M. Serum insulin-like growth
factor (IGF) and IGF-binding protein levels
and risk of lung cancer: a case-control study
nested in the beta-Carotene and Retinol Efficacy
Trial Cohort. Cancer Epidemiol Biomarkers Prev
2002; 11: 1413–1418.
59. Kley HK, Schlaghecke R, Kruskemper HL. Stability of
steroids in plasma over a 10-year period. J Clin Chem
Clin Biochem 1985; 23: 875–878.
60. Worthman CM, Stallings JF. Hormone measures in
finger-prick blood spot samples: new field methods for
reproductive endocrinology. Am J Phys Anthropol
1997; 104: 1–21.
61. Bourdoux PP, Van Thi HV, Courtois PA, Ermans AM.
Superiority of thyrotropin to thyroxine as a tool in the
screening for congenital hypothyroidism by the filter
paper spot technique. Clin Chim Acta 1991; 195: 97–
105.
62. Waite KV, Maberly GF, Eastman CJ. Storage condi-
tions and stability of thyrotropin and thyroid hor-
mones on filter paper. Clin Chem 1987; 33: 853–855.
63. Diamandi A, Khosravi MJ, Mistry J, Martinez V,
Guevara-Aguirre J. Filter paper blood spot assay of
human insulin-like growth factor I (IGF-I) and IGF-
binding protein-3 and preliminary application in the
evaluation of growth hormone status. J Clin Endocri-
nol Metab 1998; 83: 2296–2301.
64. Jones JS. Insulin-like growth factor I measurement on
filter paper blood spots. Horm Res 2001; 55(Suppl 2):
80–83.
65. Mitchell ML, Hermos RJ. Cortisol in dried blood
screening specimens from newborns with raised 17-
hydroxyprogesterone and congenital adrenal hyper-
plasia. Clin Endocrinol (Oxf) 1998; 48: 757–760.
66. Kraiem Z, Kahana L, Elias V, Ghersin S, Sheinfeld M.
Radioimmunoassay of cortisol in blood collected on
filter paper. Clin Chem 1980; 26: 1916–1917.
67. Yoshida K, Sakurada T, Kaise K, et al. Measurement
of thyroid stimulating hormone (TSH) in human urine.
Endocrinol Jpn 1988; 35: 733–739.
68. Moss BW, Jordan JW, McAllister A. The measurement
of thyroxine in urine using a competitive protein
binding technique. Clin Chim Acta 1979; 92: 387–398.
69. Shakespear RA, Burke CW. Triiodothyronine and
thyroxine in urine. I. Measurement and application.
J Clin Endocrinol Metab 1976; 42: 494–503.
70. Johansson A, Andrew R, Forsberg H, Cederquist K,
Walker BR, Olsson T. Glucocorticoid metabolism and
adrenocortical reactivity to ACTH in myotonic dys-
trophy. J Clin Endocrinol Metab 2001; 86: 4276–4283.
71. Reynolds RM, Walker BR, Syddall HE, et al. Altered
control of cortisol secretion in adult men with low birth
weight and cardiovascular risk factors. J Clin Endo-
crinol Metab 2001; 86: 245–250.
72. Weissberger AJ, Ho KY, Stuart MC. Quantification of
urinary growth hormone (GH) excretion by centrifugal
ultrafiltration and radioimmunoassay: Appraisal of the
relationship between 24 h urinary GH and mean 24 h
serum GH levels in normal and abnormal states of GH
secretion. Clin Endocrinol (Oxf) 1989; 30: 687–698.
73. Hourd P, Edwards R. Current methods for the mea-
surement of growth hormone in urine. Clin Endocrinol
(Oxf) 1994; 40: 155–170.
74. Bates AS, Evans AJ, Jones P, Clayton RN. Assessment
of GH status in acromegaly using serum growth hor-
mone, serum insulin-like growth factor-1 and urinary
growth hormone excretion. Clin Endocrinol (Oxf)
1995; 42: 417–423.
225
75. Leger J, Reverchon C, Porquet D, Noel M, Czernichow
P. The wide variation in urinary excretion of human
growth hormone in normal growing and growth hor-
mone-deficient children limits its clinical usefulness.
Horm Res 1995; 44: 57–63.
76. Gill MS, Toogood AA, O’Neill PA, Thorner MO,
Shalet SM, Clayton PE. Urinary growth hormone
(GH), insulin-like growth factor I (IGF-1), and IGF-
binding protein-3 measurements in the diagnosis of
adult GH deficiency. J Clin Endocrinol Metab 1998; 83:
2562–2565.
77. Popii V, Baumann G. Laboratory measurement of
growth hormone. Clin Chim Acta 2004; 350: 1–16.
78. Hizuka N, Takano K, Tanaka I, et al. Demonstration
of insulin-like growth factor I in human urine. J Clin
Endocrinol Metab 1987; 64: 1309–1312.
79. De Palo EF, Gatti R, Lancerin F, et al. Urinary insu-
lin-like growth factor I in athletes, before and after
physical exercise, and in sedentary subjects. Clin Chim
Acta 2002; 322: 51–57.
80. Fujihara M, Uemasu J, Kawasaki H. Serum and uri-
nary levels of insulin-like growth factor I in patients
with chronic renal disease and diabetes mellitus: its
clinical implication. Clin Nephrol 1996; 45: 372–378.
81. Hasegawa Y, Cohen P, Yorgin P, Rosenfeld RG.
Characterization of urinary insulin-like growth factor
binding proteins. J Clin Endocrinol Metab 1992; 74:
830–835.
82. Baum CK, Davison MJ, Landon J. Proceedings:
Urinary free cortisol excretion by normal subjects.
J Endocrinol 1974; 63: 47P–48P.
83. Miki K, Sudo A. Effect of urine pH, storage time, and
temperature on stability of catecholamines, cortisol,
and creatinine. Clin Chem 1998; 44: 1759–1762.
84. Aardal-Eriksson E, Karlberg BE, Holm AC. Salivary
cortisol – an alternative to serum cortisol determina-
tions in dynamic function tests. Clin Chem Lab Med
1998; 36: 215–222.
85. Laudat MH, Cerdas S, Fournier C, Guiban D, Guil-
haume B, Luton JP. Salivary cortisol measurement: a
practical approach to assess pituitary-adrenal function.
J Clin Endocrinol Metab 1988; 66: 343–348.
86. Bolufer P, Gandia A, Rodriguez A, Antonio P. Sali-
vary corticosteroids in the study of adrenal function.
Clin Chim Acta 1989; 183: 217–225.
87. Reid JD, Intrieri RC, Susman EJ, Beard JL. The
relationship of serum and salivary cortisol in a sample
of healthy elderly. J Gerontol 1992; 47: P176–P179.
88. Kirschbaum C, Hellhammer DH. Salivary cortisol in
psychoneuroendocrine research: recent developments
and applications. Psychoneuroendocrinology 1994; 19:
313–333.
89. Chen YM, Cintron NM, Whitson PA. Long-term
storage of salivary cortisol samples at room tempera-
ture. Clin Chem 1992; 38: 304.
90. Garde AH, Hansen AM. Long-term stability of sali-
vary cortisol. Scand J Clin Lab Invest 2005; 65: 433–
436.
91. Vining RF, McGinley RA. The measurement of hor-
mones in saliva: possibilities and pitfalls. J Steroid
Biochem 1987; 27: 81–94.
92. Putz Z, Vanuga A, Veleminsky J. Radioimmunoassay
of thyroxine in saliva. Exp Clin Endocrinol 1985; 85:
199–203.
93. Rantonen PJ, Penttila I, Meurman JH, Savolainen K,
Narvanen S, Helenius T. Growth hormone and cortisol
in serum and saliva. Acta Odontol Scand 2000; 58:
299–303.
94. Costigan DC, Guyda HJ, Posner BI. Free insulin-like
growth factor I (IGF-I) and IGF-II in human saliva.
J Clin Endocrinol Metab 1988; 66: 1014–1018.
95. Halimi S, Tepavcevic D, Suchanek D, Giljevic Z,
Palvsic V, Korsic M. Measurement of salivary insulin-
like growth factor-I in acromegaly: comparison with
serum insulin-like growth factor-I and growth hormone
concentrations. Eur J Clin Chem Clin Biochem 1994;
32: 705–707.
Address for correspondence: Sherita Hill Golden, Division
of Endocrinology and Metabolism, Johns Hopkins Uni-
versity School of Medicine, 2024 E. Monument Street, Suite
2-600, Baltimore, MD 21205, USA
Phone: +1-410-614-1135; Fax: +1-410-955-0476
E-mail: sahill@jhmi.edu
226
